BIS Papers

No 30
Asian bond markets: issues
and prospects

Monetary and Economic Department

November 2006
The papers in this volume were presented at a BIS/Korea University conference of central
bankers, scholars and market participants held in Seoul on 21-23 March 2004. The papers
should be read as reflecting market conditions at that time. The views expressed are those of
the authors and do not necessarily reflect the views of the BIS or the central banks
represented at the meeting. Individual papers (or excerpts thereof) may be reproduced or
translated with the authorisation of the authors concerned.

Requests for copies of publications, or for additions/changes to the mailing list, should be
sent to:
Bank for International Settlements
Press & Communications
CH 4002 Basel, Switzerland

E-mail: [email protected]
Fax: +41 61 280 9100 and +41 61 280 8100

© Bank for International Settlements 2006. All rights reserved. Brief excerpts may be
reproduced or translated provided the source is stated.

ISSN 1609 0381 (print)

ISBN 92-9131-726-8 (print)
ISSN 1682 7651 (online)
ISBN 92-9197-726-8 (online)
 List of participants and authors from Korea

Bank of Korea
Seongtae Lee Deputy Governor1
Keun-Man Yook Deputy Director General, International Relations Office
Junggon Oh Head, International Finance Research Team,
International Department

Ministry of Finance and

Economy
Tae-Shin Kwon Deputy Minister of Finance and Economy

Financial Supervisory Service

Kap-Soo Oh Vice President

Korea Fixed Income Research

Institute
Gyutaeg Oh Chief Executive Officer

Korea Institute of Finance

Jae-Ha Park Director, Macro Finance Team

Korea Securities Depository

Huhn-Gunn Ro Chairman and CEO

Korea Securities Research

Institute
Sang Yong Park President

Hanyang University
Daekeun Park Professor, Department of Economics

Korea University
Yoon Dae Euh President
Young Sup Yun Professor of Finance/Director, Institute of Northeast
Asian Business and Economics
Hasung Jang Professor of Finance/Director, Asian Institute of
Corporate Governance
Yung-Chul Park Professor, Department of Economics
Kee-Hong Bae Associate Professor of Finance, College of Business

Seoul National University

Dong Hyun Ahn Professor of Finance
Changyong Rhee Professor, Department of Economics

Sogang University
Woon Ryul Choi Dean, Graduate School of Business/Professor of
Finance

BIS Papers No 30 iii

The Bank of New York, Seoul
Branch
Ho Yang Managing Director and General Manager

Maekyung Business News

Jong Won Nam Securities Department Chief 1

1
Now Governor. All affiliations and titles as of conference date.

iv BIS Papers No 30
 List of participants and authors from outside Korea

Reserve Bank of Australia

Ric Deverell Chief Manager, International Department

People’s Bank of China

Jing Chen Section Chief
Shaofeng Niu Section Chief

Université de la Mediterranée
Éric Girardin Professor, Centre de Recherche sur les Dynamiques et
Politiques Économiques et l’Économie des Ressources

Hong Kong Monetary Authority

Grace Lau Head, Market Research

Deutsche Bank
Kyungjik Lee Associate, Asia Fixed Income Research

The Bank of New York, Hong

Kong Branch
Francis Braeckevelt Product Management

Nomura International (Hong

Kong) Ltd
Toshiyasu Iiyama Business Development Manager, Fixed Income -
Asia Pacific

Reserve Bank of India

Jasbir Singh Chief General Manager, Internal Debt Management
Department
Rosemary Sebastian General Manager, Internal Debt Management
Department

Bank Indonesia
Treesna Suparyono Deputy Director, Directorate of Monetary Management
Evy Berliana Deputy Manager, Directorate of Monetary Management

Bank of Japan
Atsushi Takeuchi Deputy Director, Planning and Coordination Division,
International Department

Ministry of Finance, Japan

Masahiro Danno Deputy Director, Legal Office, Research Division,
International Bureau

BIS Papers No 30 v
Nomura Securities Co Ltd
Fumiaki Nishi Managing Director, Investment Banking Products
Division
Waseda University
Toshiharu Kitamura Professor

Bank Negara Malaysia

Arlina Ariff Senior Manager, Monetary & Financial Policy
Department

Bangko Sentral ng Pilipinas

Carlos O Santos Bank Officer VI, Treasury Department
Marissa C De Vera Bank Officer III, Department of Economic Research

Monetary Authority of Singapore

Celeste Cheo Associate

Deutsche Bank
Martin Hohensee Head of Fixed Income and Credit Research, Asia

JPMorgan Chase Bank

Bernhard Eschweiler Managing Director, Central Bank and Government
Institutions Group Asia/Pacific
David G Fernandez Economist
Simon Klassen Vice President, Asian Rates Research

PIMCO
Aaron Low Senior Portfolio Manager

Central Bank of China, Taipei

Pohann Wu Junior Specialist

Bank of Thailand
Atchana Waiquamdee Assistant Governor

International Primary Market

Association (London)
Cliff Dammers Secretary General

Moody’s Investors Service

Tom Byrne Vice President/Senior Credit Officer, Sovereign Risk
Unit

University of California-Berkeley
Barry Eichengreen Professor
Pipat Luengnaruemitchai Student

vi BIS Papers No 30
Citigroup
Kate Kisselev Vice President, Global Country Risk Management

Cornell University
Warren Bailey Professor

Bank for International

Settlements
Gunter Baer Secretary General
Shinichi Yoshikuni Chief Representative, Representative Office for Asia
and the Pacific
Robert N McCauley Deputy Chief Representative, Representative Office for
Asia and the Pacific
Frank Packer Senior Economist
Guorong Jiang Senior Economist, Representative Office for Asia and
the Pacific
Patrick McGuire Economist
Theo Kaitis Senior Risk Control Officer, Representative Office for
Asia and the Pacific

BIS Papers No 30 vii

 Contents

List of participants and authors from Korea....................................................................... iii

List of participants and authors from outside Korea ..........................................................v

Introduction
The development of Asian bond markets
Barry Eichengreen ....................................................................................................................1

Opening addresses
For the advent of a promising and sound Asian bond market
Tae-Shin Kwon .......................................................................................................................13
Asian financial cooperation as seen from Europe
Gunter Baer ............................................................................................................................16

Overview
Developing the bond market(s) of East Asia: global, regional or national?
Robert N McCauley and Yung-Chul Park ...............................................................................19
Why doesn’t Asia have bigger bond markets?
Barry Eichengreen and Pipat Luengnaruemitchai ..................................................................40
Comments by Ric Deverell .....................................................................................................78
Comments by Junggon Oh .....................................................................................................80
Consolidating the public debt markets of Asia
Robert N McCauley ................................................................................................................82
Comments by Junggon Oh .....................................................................................................99

Lunch address
Huhn-Gunn Ro......................................................................................................................101

History, structure and prospects of East Asian bond markets

Determinants of bond holdings by foreign investors
Kee-Hong Bae, Young Sup Yun and Warren Bailey ............................................................102
Choice of currency by East Asia bond issuers
David G Fernandez and Simon Klassen...............................................................................129
Asian bond issues in Tokyo: history, structure and prospects
Fumiaki Nishi and Alexander Vergus....................................................................................143
Comments by Toshiharu Kitamura .......................................................................................168

Dinner address
Seongtae Lee .......................................................................................................................171

BIS Papers No 30 ix
Credit risk management
Minding the gap in Asia: foreign and local currency ratings
Kate Kisselev and Frank Packer.......................................................................................... 174
Comments by Tom Byrne .................................................................................................... 200
Building infrastructure for Asian bond markets: settlement and credit rating
Daekeun Park and Changyong Rhee .................................................................................. 202
Comments by Tom Byrne .................................................................................................... 222
Creation of a regional credit guarantee mechanism in Asia
Gyutaeg Oh and Jae-Ha Park.............................................................................................. 224
Comments by Guorong Jiang .............................................................................................. 241

Lunch adrress
Kap-Soo Oh ......................................................................................................................... 243

Impediments, hedging supervision and clearing

Identifying impediments to cross-border bond investment and issuance
in Asian countries
Atsushi Takeuchi.................................................................................................................. 246
A survey on hedging markets in Asia: a description of Asian derivatives
markets from a practical perspective
Martin Hohensee and Kyungjik Lee ..................................................................................... 261
Comments by Aaron Low..................................................................................................... 282
Clearing, settlement and depository issues
Francis Braeckevelt ............................................................................................................. 284
Comments by Aaron Low..................................................................................................... 333
Bond market regulation and supervision in Asia
Bernhard Eschweiler............................................................................................................ 335

Concluding panel discussion

Yung-Chul Park.................................................................................................................... 353
Tom Byrne ........................................................................................................................... 354
Aaron Low............................................................................................................................ 355
Robert N McCauley.............................................................................................................. 357

x BIS Papers No 30
 The development of Asian bond markets 1

Barry Eichengreen

1. The problem
The 1997-98 crisis in Asia prompted considerable rethinking of the role of financial markets
in the region’s economic development. Banks had long been at the centre of Asian financial
systems. For a set of late-developing economies with urgent needs for financial
intermediation, banking systems were easier to get up and running. Governments could
supply the equity capital and in some cases the managerial talent. Close cooperation
between banks and governments allowed the authorities to influence the flow of funds -
ideally, to ensure that finance flowed towards sectors that were the locus of productivity
spillovers and generators of export revenues. Large corporations in need of funding for
expensive investment projects that might require a lengthy incubation period could be
confident of a stable source of external finance.
Up to the mid-1990s this bank-centred financial system was one of the foundation stones of
East Asian economic growth. The crisis that followed then revealed that this form of financial
organisation also had serious weaknesses. The short maturity of bank loans meant that
when confidence was disturbed, as happened in 1997-98, what had once been a set of
patient lenders might not be so patient any more. Seeing their funding decline, banks might
call in their loans, subjecting their borrowers to a painful credit crunch. Moreover, with the
opening of capital accounts, banks might be in a favoured position to access foreign funds,
not least because of the perception that their obligations were guaranteed by the public
sector. They aggressively extended their intermediation role by borrowing offshore and
onlending the proceeds to domestic customers. Generally, the tenor of these foreign credits
was even shorter than that of the banks’ own loans, exposing them to a maturity mismatch
that might cause serious problems if confidence was shaken. Since most foreign funds were
denominated in dollars, euros or yen, the banks were exposed to either a dangerous
currency risk if they onlent in local currency or an equally serious credit risk if they onlent in
those same foreign currencies. Meanwhile, deregulation allowed banks to take on additional
risks using techniques with which supervisors found it difficult to keep pace. And insofar as
the banks had allowed themselves to be utilised as instrumentalities of the government’s
industrial policies, they anticipated help from the official sector in the event of difficulties.
Thus, the moral hazard inevitably associated with the existence of a financial safety net
appears to have been particularly pervasive in the Asian case.
This episode of financial turmoil led to the restructuring of banking systems and to efforts at
upgrading their supervision and regulation. But it also created an awareness of the need for
better diversified debt markets and specifically for bond markets to supplement the
availability of bank finance. Bank and bond finance have different advantages. Bonds and
securitised finance generally are thought to have better risk-sharing characteristics. Risks
can be more efficiently diversified when they are spread across a large number of individual
security holders. This spreading of risks and the existence of liquid secondary markets in
standardised securities encourages creditors to make long-term commitments and allows
debtors to borrow for extended periods of time.

1
Revised, November 2004.

BIS Papers No 30 1
Banks, in contrast, have a comparative advantage in the information-impacted segment of
the economy. They invest in building dedicated monitoring technologies. (This is one way of
thinking about what distinguishes banks from other financial market participants.)
Consequently they are well placed to identify and lend to small, recently established
enterprises about which public information is scarce. In addition, by pooling the deposits of
households and firms with non-synchronised demands for liquidity, they are able to provide
maturity transformation services for small savers reluctant to lock up their funds for extended
periods. As concentrated stakeholders, they contribute to effective corporate governance and
are prepared to incur the costs of litigation when legal recourse is required.
The point is not that banks or bond markets are better; there is little systematic evidence of
the unconditional superiority of one financial form over the other. Rather, there is a growing
body of evidence that countries benefit from well diversified financial systems with a role for
both well regulated banks and well functioning securities markets. 2 Banks have a
comparative advantage in providing external finance to smaller, younger firms operating in
information-impacted segments of the economy, while securities markets, including debt
markets, do the job more efficiently for large, well established companies. Similarly, banks
and securities markets are subject to different risks. Hence, in financial structure, as in other
areas, diversification may help an economy attain a superior position on the frontier of
feasible risk-return trade-offs. That is, the existence of a well diversified financial system, with
a role for both banks and securities markets, should be conducive both to an efficient
allocation of resources compatible with sustainable medium-term economic growth and to
financial stability - and specifically to minimisation of the risk of late 1990s-style financial
crises.

2. The policy response

It is in this context that recent efforts to foster the development of Asian bond markets should
be understood. These efforts have focused on the development of a more robust and
efficient market infrastructure at the national and regional levels. Among the prominent
initiatives in this area is the Asian Bond Market Initiative (ABMI) of the ASEAN+3 countries. 3
As endorsed by ASEAN+3 finance ministers at their August 2003 meeting in Manila, the
ABMI takes as its goal the development of more robust and efficient primary and secondary
markets. To this end ASEAN+3 has established working groups concerned with the creation
of standardised debt instruments, the establishment of rating agencies, the provision of
technical assistance, foreign exchange transactions and settlement issues, credit guarantee
mechanisms, and the role of multilateral development banks, foreign government agencies
and Asian multinational corporations in issuing in local markets and local currencies.

2
See Demirgüç-Kunt and Levine (1996, 2001).
3
The members of ASEAN are Brunei, Cambodia, Indonesia, Laos, Malaysia, Myanmar, the Philippines,
Singapore, Thailand and Vietnam; the “plus 3” countries are Japan, Korea and China. Another initiative
deserving of mention is the APEC Regional Bond Market Initiative agreed to by the APEC Finance Ministers’
Process (FMP). The FMP was established following the 1997 financial crisis to provide a forum for the
exchange of views and information on regional financial developments and to cooperatively pursue
programmes for the promotion of financial sector development and liberalisation. In 2002 APEC finance
ministers then agreed to a second policy initiative on the development of securitisation and credit guarantee
markets, which aims at using high-level policy dialogues and expert panels to identify impediments to the
development of these markets. For details see www.apec.org, and in particular www.apec.org/
apec/ministerial_statements/sector_ministerial/finance/2003_finance/annex.html.

2 BIS Papers No 30
These working groups can be seen as mechanisms for sharing information and providing
technical assistance about best practice in fostering and regulating bond markets. They can
be seen as working towards the establishment of benchmarks for the development of market
infrastructure against which national policy and practice can be assessed. Private sector
practice has shown such benchmarks to be an effective focal point for reform. 4 The working
groups may thus function as a source of peer pressure for governments to move more
quickly in the direction of creating active and liquid bond markets than they might otherwise,
something that is desirable insofar as the official sector often enjoys privileged access to
bank finance and therefore faces a moral hazard of its own.
Other initiatives seek to remove the obstacles to the development of a pan-Asian bond
market. They seek to encourage Asian investors to build regional bond portfolios by
removing obstacles to cross-border capital flows and by harmonising the regulations,
withholding tax provisions, accounting practices, rating conventions and clearing and
settlement systems that pose obstacles to foreign participation in regional bond markets.
These initiatives respond to the perception that the small size of Asian bond markets is part
of what limits their liquidity, efficiency and growth. To be attractive for investors, a bond
market must operate at a certain minimum efficient scale. Otherwise market participants will
not be able to acquire or dispose of their holdings without moving prices. 5 Small markets with
a limited number of participants may also create scope for strategic behaviour by competitors
and counterparties to deter entry and participation by other investors. 6 There may exist
significant scale efficiency effects in clearing and settlement, payment system data
processing, trading operations, firm-specific information processing activities (such as
listing), and even regulation. 7 In addition, a small market may not be able to develop liquidity
in the full range of marketable instruments, including the derivative instruments needed by
investors to hedge market risk, which in turn may deter participation. 8 For all these reasons,
small countries may find it difficult to develop deep and liquid bond markets. Securing foreign
participation through the removal of impediments to cross-border issuance and investment is
in turn a potential way around this problem.
The most prominent initiative in this area the Asian Bond Fund created by the Executives’
Meeting of East Asia-Pacific Central Banks (EMEAP). 9 Launched in June 2003, the Asian
Bond Fund (ABF) had an initial size of US$1 billion. It invests in a basket of US dollar-
denominated bonds issued by Asian sovereign and quasi-sovereign issuers in EMEAP
countries other than Japan, Australia and New Zealand. It is managed by the Bank for
International Settlements and supervised by an EMEAP Oversight Committee. A second
Asian Bond Fund, under discussion at the time of writing, is to include investments in bond
denominated in regional currencies issued by sovereigns, quasi-sovereigns and creditworthy

4
The use of benchmarking to generate peer pressure for reform is a widespread private sector practice. It is
also used by the European Union as part of its method of “open cooperation”. See Wyplosz (2004).
5
McCauley and Remolona (2000) provide evidence on the relationship between market size and liquidity, as
measured by inter alia bid-ask spreads and market turnover.
6
Mohanty (2001) cites a number of real-world instances where such behaviour has been evident in small and
even medium-sized markets.
7
For evidence on this see Hancock et al (1999), Saloner and Shepard (1995), Malkamaki (1999) and Bossone
et al (2001).
8
See Turner and Van’t dack (1996).
9
EMEAP is a forum of central banks and monetary authorities in the East Asia-Pacific region with 11 members:
Australia, China, Hong Kong SAR, Indonesia, Japan, Korea, Malaysia, New Zealand, the Philippines,
Singapore and Thailand.

BIS Papers No 30 3
companies. 10 By encouraging the reinvestment of central bank reserves in the qualifying
bond markets and securities of the region, the ABF initiative can be seen as helping to
augment the installed base of local securities holdings and thus overcome the problem of
inadequate scale. More generally this initiative can be seen as one of a set of measures
designed to foster the development of a deep and liquid bond market at the regional level.

3. Dilemmas
There is an almost instinctual tendency on the part of economists to applaud such efforts,
given the compelling nature of the arguments for developing more active bond markets to
round out Asia’s bank-dominated financial systems. But there is also a dilemma. In reality
what we are talking about is capital account convertibility, and capital account convertibility in
advance of the development of regional financial markets. This, of course, is the opposite of
what most of us thought that we had learned from the Asian crisis about the right time at
which to liberalise the capital account. One of the key lessons of the Asian crisis is that it is
important to have strong, diversified and well developed domestic financial markets,
including by implication bond markets, before liberalising the capital account. If financial
markets are underdeveloped, market discipline will be weak, and banks and firms will be
prone to overborrow. Capital account liberalisation will then cause funds to flow in through
the banking system. Cheap funding will encourage the banks to expand their loan portfolios,
resulting in a decline in the average quality of loan projects. Maturity mismatches will be
accentuated if banks use this short-term finance to fund long-term loans, and currency
mismatches will result either for the banks (if they lend in local currency) or their customers
(if their loans are denominated in foreign currency but the borrowers are active in the
production of non-traded goods - as in the case of construction firms). If the flow of foreign
capital then turns around, the whole financial edifice can come crashing down. The Asian
crisis is a stark reminder of the havoc that can be wreaked by this combination of
circumstances.
Thus, macroeconomists will insist that governments should not proceed with capital account
liberalisation unless they have first made progress in developing local bond markets. And
market participants will insist that countries cannot have local bond market development
unless they first have open capital accounts. Lee Hsien Loong, Deputy Prime Minister of
Singapore and head of that country’s Monetary Authority, put the point well in an address
given in 2002: “There is a trade-off between tightening up the capital account, and
developing the bond markets. Measures to restrict offshore foreign currency trading have
been effective, in so far as reducing or eliminating offshore markets is concerned. But these
safeguards come at a cost - they also hinder the development of capital markets, especially
the bond markets. Size and liquidity are essential attributes for a market to attract
international interest. Already in size and liquidity, we clearly lag behind our counterparts in
the West. If Asian markets are fragmented and unable to grow, they risk being ignored by
global investors.” 11
Thus, Asia would seem to be in a classic Catch-22 situation. Without removing capital
controls, attempting to foster domestic bond markets can be an uphill fight. Yet trying to win it
by removing restrictions on the ability of residents and foreigners to invest across borders

10
The stated purpose of ABF2 is to encourage the development of index bond funds in regional markets and
to act as a catalyst for the improvement of domestic bond markets and for greater harmonisation of bond
market infrastructure and legal, regulatory and tax arrangements across the region. For details, see
www.emeap.org/press/15apr04.htm.
11
I owe this quote to Dwor-Frecaut (2003).

4 BIS Papers No 30
could be a risky strategy. It is widely recognised that these trade-offs are implicit in efforts to
build domestic bond markets by removing capital controls. 12 What is less well understood is
that even seemingly benign steps like harmonising regulations and taxation, or creating an
Asian rating agency (or a common standard for national rating agencies), or using central
bank reserves to jump-start private cross-border investment are the equivalent of capital
account liberalisation in the sense that they too would work to encourage cross-border
capital flows. This is their intent, and it would certainly be their effect. And these measures
create risks - as well as conflicting with the conventional wisdom regarding sequencing -
insofar as they encourage capital mobility first and only produce stronger markets later.
The positive message is that governments should proceed with all due speed to strengthen
market infrastructure at the national level: more efficient clearing and settlement systems,
more efficient information provision and assessment (through inter alia the establishment of
disclosure requirements for issuers and the creation of rating agencies), stronger creditor
rights and the development of benchmark assets and yield curves. Even small countries can
make progress in this direction, although they may have to forgo some of the cost savings
associated with the scale efficiency effects enumerated above. They can also overcome
some of the disadvantages of small market capitalisation by consolidating the public debt
and overfunding their fiscal needs. 13 One reading of European experience from the 1950s to
the 1980s is that, through the dedicated pursuit of such measures, reasonably robust and
liquid markets in debt securities can be created. 14 At that point it becomes safe to remove
residual capital controls, as Europe did in the 1990s, and to encourage market participants to
build pan-regional portfolios.
This perspective suggests that Asian countries, especially lower-income Asian countries with
a less developed financial infrastructure, should proceed cautiously with capital account
liberalisation. It suggests that a relatively small Asian Bond Fund (recall that an ABF-I funded
to the tune of US$1 billion compares with regional bond markets with a market capitalisation
of some US$1.5 trillion) is appropriate in that it does not put the cart before the horse. That
is, it does not commit Asian central banks to large amounts of cross-border portfolio
investment before a stronger market infrastructure is in place. It suggests that efforts to foster
the development of bond markets should focus, in the first instance, on measures to
strengthen the market infrastructure at the national level and not on measures to harmonise
and integrate those market structures, thereby encouraging cross-border capital flows, per
se. Measures to harmonise and integrate institutions and regulations should come later, once
those stronger market structures are in place.
The other issue raised by Europe’s experience in creating a regional bond market is the role
of the exchange rate. In Europe, the elimination of currency risk by the creation of the euro
strongly stimulated the development of regional bond markets. This is evident in the dramatic
increase in corporate bond issuance, speculative grade issuance in particular, following the
irrevocable locking of exchange rates in 1999, and in the adoption of the 10-year German
government bond as the benchmark issue for the region. 15 This experience suggests that an

12
The econometric results in Eichengreen and Luengnaruemitchai (in this volume) are consistent with this
emphasis, in that they find a number of alternative measures of capital controls to be negatively associated
with domestic bond market capitalisation in a panel of data for 41 countries.
13
See McCauley (2003).
14
Wyplosz (2001) advances this position.
15
In the first year of the euro, the value of euro-denominated corporate bond issues more than tripled, and the
share of corporate bond issues accounted for by speculative (sub-investment grade) issues rose from 4% to
15%. Corporations were able to place unpredecentedly large issues on European markets; see Detken and
Hartmann (2000). These extraordinary early growth rates have now tailed off a bit, but the rate of growth of
issuance of debt securities by non-financial corporations continues to outrun the growth of their other sources

BIS Papers No 30 5
exchange rate regime that minimises currency risk can lend strong stimulus to the
development of regional bond markets by encouraging investors to build pan-regional
portfolios, thereby enhancing market liquidity and in turn inducing additional issuance and
investment. The paper by Barry Eichengreen and Pipat Luengnaruemitchai in this volume
provides further support for this association between exchange rate stability and bond market
capitalisation.
For Asia, these facts again create something of a dilemma. Another widely drawn lesson of
the Asian crisis is that countries should gravitate towards more flexible exchange rate
regimes in order to limit crisis risk and be able to better tailor domestic money and credit
conditions to local needs. Moreover, the presumption that Asian countries will continue to
move down the road towards capital account convertibility reinforces the argument for
greater exchange rate flexibility, insofar as moving to managed flexibility is an essential
precondition for full capital account liberalisation. 16 Hence, the exchange rate regime
consistent with financial stability in the short run may not be conducive to bond market
development in the longer run.
The severity of this problem is not entirely clear. The observation that countries with more
stable exchange rates have better capitalised bond markets is based on an all-else-equal
comparison. In practice, everything else may not be equal. In particular, macroeconomic
policies that minimise currency risk by holding exchange rates stable may heighten credit risk
by encouraging banks, firms and governments to borrow more freely, thereby exposing them
to financial distress when cyclical conditions deteriorate. Robert McCauley and Guorong
Jiang (2004) find a closer conformance of bond yields across countries with flexible
exchange rates. One interpretation is that credit risk is even more important than currency
risk in driving a wedge between national markets and that in countries where the bulk of debt
is domestic currency denominated these two forms of risk are negatively correlated. If this is
right, then greater exchange rate flexibility may not in the end be an impediment to bond
market development.
The other solution, also suggested by European experience, is monetary unification to
reconcile the desire for exchange rate stability with the reality of capital account convertibility,
along with stronger financial market institutions and regulation to prevent overborrowing and
avoid unnecessary credit risk. From this point of view the Chiang Mai Initiative for swap lines
and other financial supports, ongoing discussions of a collective currency peg and initiatives
to foster the development of bond markets are of a piece. That is to say, these various efforts
to further economic and financial development and cooperation at the regional level are
complementary to one another. The problem is that the time horizon relevant to these
different initiatives is not the same. While furthering the development of bond markets is an
urgent task that should be advanced now in order to foster growth and buttress financial
stability, monetary unification is a long-run objective that presupposes a significantly more
extensive political commitment. 17
The other question in this context is whether Asia is the right level at which to pursue these
objectives. A positive answer is generally presupposed in discussions of exchange rate
stabilisation and monetary unification. There is both the European precedent and the fact of
rapidly growing intraregional trade and foreign investment linkages, heavily centred on

of finance. This enhanced access of euro-denominated corporate debt markets helped to finance a wave of
mergers and acquisitions which in turn promises to strengthen Europe’s corporate sector.
16
See for example Fischer (2003).
17
This is something that is acknowledged even by the proponents. Thus Mallet (2004), in describing discussions
at the 2004 Asian Development Bank meetings for achieving currency union in Asia, reports that “economists
and bankers say a common east Asian currency would take two or three decades.”

6 BIS Papers No 30
China. But it is not clear that a positive answer is appropriate in discussions of bond market
development. There already exist well developed global securities markets into which Asian
countries can link, as emphasised by Robert McCauley and Yung Chul Park in their
contribution to this volume. Many of the large issuers and large investors - multilateral
institutions, foreign government agencies and multinational corporations alike - whose
participation in local markets is desired are headquartered outside Asia. Harmonising
institutions and policies across Asian countries is not the most obvious way of encouraging
their participation; better would be to harmonise institutions and regulations with those
prevailing in global markets. Even if the answer to the question of whether Asian countries
should attempt to integrate into global or regional bond markets is not obvious, that question
should at least be asked.

4. The papers that follow

The papers that follow shed additional light on a number of these issues. In keeping with the
thematic structure of this introduction, I review these papers in a somewhat different order
than they appear below.
Atsushi Takeuchi’s paper sets the stage by describing the rationale for the development of
deeper and more liquid bond markets, the progress that has been made to date and the
obstacles going forward. As the author notes, the size of local bond markets in Asia, as
measured by the volume of issuance, has more than doubled since 1998. However,
secondary markets remain relatively illiquid, and foreign participation, in particular, is
disappointing. Takeuchi highlights capital controls, taxation, the difficulty of cross-border
clearing and settlement, and the limited availability of hedging instruments as obstacles to
greater participation by nonresidents.
Barry Eichengreen and Pipat Luengnaruemitchai further set the stage by using multiple
regression and cross-country comparisons to analyse the obstacles to the development of
Asian bond markets. While a variety of alternative explanations have been offered in the
past, Eichengreen and Luengnaruemitchai show that bond market undercapitalisation in Asia
is in fact a phenomenon with multiple causes. To some extent the problem is one of
minimum efficient scale: smaller countries find it more difficult to develop well capitalised
bond markets (even when capitalisation is measured relative to GDP). But market size is not
the entire problem. In addition, corruption and low bureaucratic quality, which are signs of
unreliable securities market regulation, and the failure of countries to follow internationally
recognised accounting and disclosure standards have slowed the development of debt
markets. Macroeconomic policy, for its part, appears to have played both a supporting and
impeding role. On the one hand, Asia’s strong fiscal balances, while admirable on other
grounds, have not been conducive to the growth of government bond markets. At the same
time, there is little evidence that the small size of public debt markets is a serious obstacle to
corporate bond market development. And the stability of exchange rates in the region
appears, if anything, to have encouraged bond market development.
Robert McCauley, in his paper, builds on the observation that small countries find it difficult to
achieve the minimum efficient scale required for a deep and efficient bond market. He notes
that the sterilisation operations engaged in by Asian central banks in the process of
accumulating international reserves have provided an opportunity to get a larger installed
base of public debt securities into the market. The problem is that the market has been
segmented into government debt and central bank debt. McCauley therefore recommends
consolidating this debt into a uniform set of securities by “overfunding” the fiscal deficit
(issuing more government debt securities than needed to fund the deficit, and purchasing
central bank bills in return). Finance ministries may be reluctant to permit the de facto
issuance of additional government debt as a device for mopping up excess liquidity; among

BIS Papers No 30 7
other things, doing so is likely to undermine their control of the public debt market. But they
still may wish to consider this alternative if they attach priority to the creation of a liquid
domestic market in public debt.
Kee-Hong Bae, Warren Bailey and Young-Sup Yun look more closely at the issue of foreign
participation, analysing data gathered by the IMF for 165 countries on the holdings of local
bonds by foreign investors. They find that measures of property rights protection similar to
those analysed by Eichengreen and Luengnaruemitchai - corruption, risk of expropriation of
private property and the risk that contracts may be repudiated - are the most influential
determinants of foreign participation. By comparison, they find little evidence of a role for
macroeconomic variables like inflation, interest rates and the volatility of growth. This
reinforces the message that countries seeking to develop their bond markets, and specifically
to encourage foreign participation, should focus on building investment-friendly institutions.
Atsushi Taneuchi, in a companion paper, examines these same issues and in addition
characterises the obstacles to non-resident issuance. Compared to Bae, Bailey and Yun, he
puts more emphasis on statutory restrictions such as capital controls, the opacity and lack of
uniformity of withholding tax regimes, and the absence of adequate instruments for hedging
interest rate and currency exposures.
Martin Hohensee and Kyungjik Lee pursue the problem of hedging instruments, both those
traded on futures exchanges and over-the-counter interest rate derivatives such as interest
rate swaps and options. They show that Hong Kong and Singapore, two of the leading bond
markets in the region, also have the most advanced derivatives markets - a fact that is surely
not coincidental. It is impossible to imagine the development of the relevant hedging markets
absent the growth of the underlying bond market, for without trading in the underlying bonds
there would be nothing on which to base the swaps and options in question. The growth of
hedging markets in Hong Kong and Singapore thus reflects the success of these centres in
growing their local bond markets. But it also reflects a transparent and flexible regulatory
regime, which provides market participants the opportunity and the incentive to engage in
derivative transactions. While other Asian countries have launched their own derivatives
markets, these remain relatively illiquid. This suggests that markets in the relevant hedging
instruments tend to develop as a natural by-product of bond market development, although
their growth can also be fostered by putting in place a transparent, market-friendly regulatory
regime.
The case for developing local bond markets is strongest to the extent that the resulting
issues are long in tenor and denominated in local currency, thereby helping to relieve the
double mismatch problem. David Fernandez and Simon Klassen focus on the currency
mismatch aspect, analysing the choice of currency by East Asian bond issuers. In contrast to
Eichengreen and Luengnaruemitchai, they argue for the existence of strong spillovers
between the sovereign and corporate segments of the market. They conclude that sovereign
issuance has played a catalytic role in the genesis of regional bond markets, this despite the
constraints resulting from the traditionally strong fiscal stance of Asian governments. Since
the Asian crisis, however, sovereign issuance has soared, and corporate issues have
followed in its train. Indeed, as the authors emphasise, corporate issuance has been the
most rapidly growing segment of Asian bond markets in the last five years. The constraint on
the further growth of the corporate market, they suggest, is not so much inadequate supply
as inadequate demand - or at least a mismatch in the structure of supply and demand. High-
grade corporates have either ample retained earnings or easy access to equity finance;
hence much of the potential supply of corporate bonds is from sub-investment grade credits.
The demand, from institutional investors in particular, is however for investment grade debt
securities (given restrictive covenants, regulations, etc). One potential solution to this
problem is the development of structured products that allow investors to unpack credit risk
from other characteristics of debt securities. While the market in structured products is
growing as well, Fernandez and Klassen suggest that Asian financial institutions, which are

8 BIS Papers No 30
potential suppliers of such products, need to develop further their expertise and involvement
in this area.
Another potential way of addressing this supply-demand imbalance is the provision of credit
guarantees. Gyutaeg Oh and Jaeha Park argue that the most important constraint on the
development of local currency bond markets is not weakness of creditor rights, imperfections
in the rating function or the absence of a pan-Asian clearing and settlement system, but the
absence of credit guarantees. The underlying constraint to bond market development in the
region, they argue, is the mismatch between the credit quality of potential issuers (which is
often speculative grade) and the credit quality required by provident funds, insurance
companies and other institutions (which, as noted, are often required or prefer to limit their
holdings to investment grade securities). Guarantees against credit and political risk could
bridge this gap, Oh and Park argue: they would guarantee a high rating for issuers, facilitate
the securitisation of outstanding debts, broaden the investor base and improve marketability
by limiting the danger of downgrades. To this end the authors propose the creation of either
public-private partnerships or multilateral institutions to provide guarantees for qualifying
Asian issuers. The question here is why, if there is a demand for such insurance, private
agencies have not sprung up to screen potential customers and provide this service at a
price. And if the answer is that there exist distortions preventing the market from doing this
job, then there is still the danger that the public provision of guarantees will only reintroduce
in another guise the moral hazard problem that arose in the 1990s as a result of the
commercial banks’ implicit guarantees.
One model for Asian countries seeking to develop their bond markets is Japan, which has a
large and highly liquid debt market. Fumiaki Nishi and Alexander Vergus consider the history,
structure and prospects of this market. They show that government debt dominates the
Japanese market, not surprisingly given the large budget deficits run by the government in
the 1990s in the effort to jump-start a deflation-ridden economy. 18 Their conclusion is that the
Japanese corporate bond market has developed relatively slowly due to the long-standing
dominance of Japanese banks over the country’s corporate finance. In this sense Japan is
not a reassuring precedent for other Asian countries, which similarly inherit financial systems
heavily dominated by commercial banks.
Japan also provides lessons, as Nishi and Vergus shows, for Asian countries seeking to
encourage foreign participation in their local bond markets. The first yen-denominated bond
publicly offered by a non-resident issuer in the domestic market was issued by the Asian
Development Bank in 1970. This was followed by sovereign issues by Singapore in 1976 and
the Korea Development Bank in 1978. As controls on non-resident issuance in yen were
gradually relaxed, a variety of foreign corporate issuers followed, creating the so-called
“Samurai market”. 19 However, the development of the Samurai market has been relatively
slow, a disappointing record that the authors attribute the onerous regulations and
registration requirements imposed by the Japanese authorities - and that might be best
avoided by other Asian governments.
The remaining papers consider challenges for the development of market infrastructure, a
task that emerges from these analyses as a key step for countries seeking to foster local
bond markets. Kate Kisselev and Frank Packer consider the rating function, focusing on the

18
In fact, large-scale government bond issuance started already in the second half of the 1980s, and Nishi and
Vergus trace the development of the Tokyo market back to this period.
19
There may be a more general lesson here for Asia’s less developed countries - and for the Asian
Development Bank. Non-resident issuance often starts with the international financial institutions. In addition
to creating a local currency benchmark asset and stimulating liquidity, such issuance provides an instrument
that local issuers tapping foreign currency bond markets can use to swap out of their foreign currency
exposures, limiting the currency mismatch problem.

BIS Papers No 30 9
rating of local currency bonds. Transparent and efficient ratings are essential to creating a
broad and diversified customer base for local currency bonds. But, as Kisselev and Packer
show, local currency sovereign bonds often receive very different ratings than the foreign
currency issues of the same governments. Rating agencies tend to give higher ratings to
local currency issues on the grounds that the sovereign may find it easier to raise domestic
currency denominated resources in times of stress. (In extremis they can always print
money.) Corruption appears to be important for explaining these rating gaps: the greater
perceived corruption, the smaller the rating advantage to local currency bonds. In addition,
countries with higher investment rates, and therefore, presumably, superior growth
prospects, receive more favourable ratings for their domestic currency bonds, relative to their
foreign currency counterparts. There are also important differences between S&P and
Moody’s in how they calibrate this rating gap, suggesting that the market still has some way
to go in arriving at a standard methodology for rating local currency bonds.
One response to dissatisfaction with the global rating agencies is to develop local
counterparts. The performance of local rating agencies, one or more of which now exists in
most Asian countries, is analysed by Daekeun Park and Changyong Rhee. Park and Rhee
argue for standardising the rating systems used by these agencies and creating a pan-Asian
settlement system as a way of fostering the development of a pan-regional bond market.
Their case for local agencies is based on local knowledge and on the observation that global
agencies like S&P, Moody’s and Fitch do not find it worthwhile to provide ratings for the
multitude of small local issuers that comprise the most rapidly growing segment of Asian
borrowing. Unfortunately, local agencies in different countries follow incompatible practices,
assigning government bonds the highest credit ratings and rating other entities’ ratings below
that sovereign ceiling. Since sovereign creditworthiness differs across countries, this renders
ratings of corporate creditworthiness incomparable. Harmonising practices and abandoning
the sovereign ceiling would help, but this is easier said than done. 20
Park and Rhee also consider whether the underprovision of clearing, settlement and
depository services is an obstacle to the development of regional bond markets. They
compare the advantages of clearing and settling cross-border transactions using local
agents, global custodians and cross-border settlement systems operated by international
central securities depositories (ICSDs) like Euroclear and Clearstream. Local custodians
must be hired in each relevant market, and the quality of their services varies. Global
custodians essentially do little more than arrange local custodians for their clients. ICSDs
avoid this duplication and quality variation but have limited coverage in Asia, partly because
of regulatory restrictions on financial transactions in various Asian countries. In addition, time
zone differences mean that Euroclear and Clearstream do not provide real-time clearing for
many Asian transactions. Park and Rhee argue for the creation of an Asian clearing and
settlement system to rectify these problems. The question is whether creating a new system
is really necessary or whether existing networks like Euroclear would provide an expanding
range of services if Asian countries simply relaxed regulatory restrictions on financial
transactions and the volume of bond market turnover increased.
Frank Braeckevelt also finds much to criticise in these areas, describing Asia’s clearing and
settlement infrastructure as opaque and fragmented. But he does not find that infrastructure
is the principal barrier to the development of efficient bond markets in Asia; despite their
fragmentation, existing clearing and settlement systems operate relatively well. Rather, the
principal barriers to the development of regional bond markets, Braeckevelt concludes, are
capital controls that limit the participation of foreign investors, along with factors limiting
market liquidity such as the absence of incentives for Asian institutions to actively manage

20
In any case, it is not clear that local rating agencies provide a meaningful alternative to global agencies, since
in practice they are often affiliated with those global agencies, which are also among their major shareholders.

10 BIS Papers No 30
their portfolios. This is consistent with the view that clearing and settlement issues, rather
than requiring the development of an Asian clearing system, could be resolved by relaxing
regulatory restrictions and encouraging additional market liquidity and turnover.
Finally, the paper by Bernhard Eschweiler analyses the role played by supervision and
regulation in the development of Asian bond markets. Eschweiler finds that supervision and
regulation in the region increasingly resemble global practices, in both structure and content.
At the same time, there is considerable variation within Asia in the quality of regulation, Hong
Kong and Singapore being the only countries that are fully compliant with global standards
and best practices. Where other Asian countries tend to fall down is not so much in the
design of regulation as in its enforcement, reflecting weaknesses in legal systems and
creditor rights. Eschweiler concludes that the fastest route to developing bond markets in
Asia is not through efforts to harmonise market rules and regulations but rather through the
adoption and implementation of global best practices at the national level.
Given this variety of viewpoints and conclusions, it will be evident that there is less than
complete consensus on the priority actions that should be taken to effectively foster the
development of bond markets in Asia. If there is one thing on which observers agree, it is
Eichengreen and Luengnaruemitchai’s point that the slow growth of Asian bond markets is a
problem with multiple dimensions whose solution requires multiple interventions:
strengthening creditor rights, building stronger market infrastructure, improving regulatory
design and enforcement, and removing the capital controls and tax measures that limit
foreign issuance and investor participation - while adapting macroeconomic policies,
including the exchange rate regime, to the reality of financial integration. The small scale of
many Asian economies and financial markets also remains a barrier to the development of
deep and liquid local markets at the national level; on this too there is agreement, although
there is no consensus, at least yet, on whether this means that priority should be attached to
harmonising bond market rules, integrating clearing and settlement systems, and creating
pan-Asian standards for rating agencies so that market growth can proceed at the regional
rather than the national level. Clearly, there is no shortage of positive steps that can be taken
to promote the development of Asian bond markets. The key task going forward is to identify
which such measures should be priorities.

References
Bossone, B, P Honohan and M Long (2001): “Policy for small financial systems”, Financial
Sector Discussion Paper no 6, Washington, DC: The World Bank.
Demirgüç-Kunt, Asli and Ross Levine (1996): “Stock market development and financial
intermediary growth: stylized facts”, World Bank Economic Review 10, pp 291-321.
Demirgüç-Kunt, Asli and Ross Levine, eds (2001): Financial structure and economic growth,
Cambridge, Mass: MIT Press.
Detken, C and P Hartmannn (2000): “The euro and international capital markets”, ECB
Working Paper 19, Frankfurt: European Central Bank.
Dwor-Frecaut, Dominique (2003): “The Asian bond market: from fragmentation to
aggregation”, Asian Rates and Credit Research, Barclays Capital (12 September).
Fischer, Stanley (2003): “The importance of financial markets for economic growth”, paper
presented to International Derivatives and Financial Markets Conference of the Brazilian
Mercantile & Futures Exchange, Campos do Jordao, Brazil, 21 August.
Hancock, D, D B Humphrey and J A Wilcox (1999): “Cost reductions in electronic payments:
the roles of consolidation, economies of scale, and technical change”, Journal of Banking
and Finance 23, pp 391-421.

BIS Papers No 30 11
Malkamaki, M (1999): “Are there economies of scale in stock exchange activities?”
Discussion Paper 4/99, Helsinki: Bank of Finland.
McCauley, Robert (2003): “Unifying public debt markets in East Asia”, BIS Quarterly Review
(December), pp 89-98.
McCauley, Robert and Guorong Jiang (2004): “Diversifying with Asian local currency bonds”
BIS Quarterly Review (September), pp 51-66.
McCauley, Robert and Eli Remolona (2000): “Size and liquidity of government bond
markets”, BIS Quarterly Review (November), pp 52-60.
Mallet, Victor (2004): “East Asia’s financial activities grow closer: regional bankers’ meeting
says inexorable moves towards integration could lead to a single currency”, Financial Times
(17 May), p 1.
Mohanty, M S (2001): “Improving liquidity in government bond markets: what can be done?”
BIS Papers, no 11, pp 49-80.
Saloner, G and A Shepard (1995): “Adoption of technologies with network effects: an
empirical examination of the adoption of automated teller machines”, Rand Journal of
Economics 25, pp 479-591.
Turner, Philip and J Van’t dack (1995): “Changing financial systems in small open
economies: an overview”, BIS Papers, no 1 (December).
Wyplosz, Charles (2001): “A monetary union in Asia? Some European lessons”, unpublished
manuscript, Graduate Institute for International Studies, Geneva.
——— (2004): “Economic integration and structural reforms: the European experience”, in
World Economic Outlook: Advancing Structural Reforms, Washington, DC: International
Monetary Fund, April, pp 130-131.

12 BIS Papers No 30
 For the advent of a promising and sound
Asian bond market

Tae-Shin Kwon

Introduction
Good morning, distinguished guests, ladies and gentlemen! Let me first thank the President
of Korea University, Dr Yoon Dae Euh, for the invitation to participate in this very important
conference, and also the organisers for their efforts in hosting this occasion. It is my great
honour to have this opportunity to address such a distinguished group of experts and
practitioners from government, academia, financial markets and international organisations.
What policies should we adopt to further develop the Asian bond market? All Asian countries
already share a common understanding of the significance of developing a regional bond
market and have embarked on efforts towards this end. Now it is appropriate for us to
evaluate past performance and elaborate future direction. In this context, I believe strongly
that this conference is very timely and meaningful.

Why is a regional bond market so important?

There has been wide recognition that the Asian financial crisis broke out due to two major
reasons: first, excessive reliance on bank-intermediated financing, and second, foreign short-
term financing leading to mismatches in currency and maturity. Regrettably, many borrowers
in this region still have only limited access to long-term local currency financing. For this
reason, how to develop the local currency bond markets has recently become one of the
hottest policy issues in the region.
Another reason for developing the Asian bond market is the need to recycle accumulated
capital directly back into the region. Although Asian countries have replenished their foreign
exchange reserves substantially since the financial crisis, most of the funds have been
invested in the developed markets, including the United States and Europe, and only later
flowed back into the region in the form of foreign currency denominated loans, purchases of
equities, and foreign direct investments. This situation needs to change.
It is sad to note that there is hardly any demand for local currency bonds from cross-border
investors. In this regard, we have to create investment products that are of interest to
regional investors. All things considered, regional borrowers will remain dependent on the
international financial markets to a large extent until the Asian bond market is well
established.
Therefore, the policymakers and economists in the region have reached a consensus on the
need to promote the regional bond market in order to obtain the following benefits. The first
and most important is that an advanced bond market would allow Asian countries to prevent
the recurrence of financial crises as well as facilitating the recycling of regional savings. The
Asian financial crisis taught this region a hard lesson: that the development of a strong and
deep Asian bond market is an essential element in ensuring that long-term, local currency
funds are available for investment in the region. In addition, bond market development would
help to narrow the gap between capital markets in advanced and emerging market
economies, and also to achieve better balanced capital markets within the region. The more
advanced emerging economy bond markets become, the more efficient regional capital

BIS Papers No 30 13
markets will be as a whole. And this, in turn, will raise the efficiency and growth potential of
the region’s economies.
I’d like to make clear that deeper and better balanced capital markets provide a shared
benefit to all economies, serving as more reliable and stable sources of financing. Moreover,
they bring additional benefits for both issuers and investors, such as transparent accounting
practices, well educated financial experts and various types of financial instruments. Such
improvements would undoubtedly bring about a virtuous circle of prosperity all over the
continent.

What has been done so far?

Against this background, each of the Asian economies has made tremendous efforts with a
view to promoting bond markets at both national and regional level. For instance, the Korean
government has undertaken significant financial sector reforms at unprecedented speed
since the Asian crisis, and introduced a series of important measures for the advancement of
the country’s capital market. Notably, Korea has successfully developed its securitisation and
credit guarantee market through asset-backed security (ABS) schemes.
We have also launched a regional financial cooperation project - the Asian Bond Market
Initiative - aimed at developing the bond market through various forums such as APEC,
ASEAN+3 and EMEAP. Under these regional groups, a number of concrete steps are being
taken to complement individual country efforts to develop the bond market. For example,
APEC and ASEAN+3 are focusing on the supply side of the bond market. That is to say, they
are mainly studying ways to provide high grade bonds for investors. In the meantime,
EMEAP is attending to the demand side by establishing an Asian Bond Fund with members’
foreign reserves for the purpose of securing regional demand for bonds.
These regional initiatives are historic and unprecedented. Through these forums, the region’s
policymakers have maintained regular dialogue and discussion for the last couple of years. In
addition, a number of seminars and symposiums like today’s conference have been
organised to foster the exchange of views and knowledge to be shared among academics,
think tanks and the private sector.
Over the course of studies and discussions, Asian countries have reached an agreement that
establishing bond market infrastructures, such as a credit guarantee mechanism and
settlement and credit rating systems, is essential for regional bond markets to prosper. As a
result, six working groups have been established and are currently addressing these issues
on a voluntary basis under the Asian Bond Market Initiative. There has already been
substantial progress in the drive to set up this bond market infrastructure. I believe that we
will see tangible outcomes in the near future.

Future challenges
Despite much endeavour and many achievements to date, we have yet to overcome many
challenges. We have stored up a vast number of ideas and suggestions from numerous
studies and meetings on this issue. It is now time for individual countries to take practical
steps in a more comprehensive and well organised manner. At this moment, we need to pay
more attention to how to create an attractive financial market for domestic and overseas
investors alike.
To this end, first of all, financial reform and liberalisation initiatives need to maintain their
momentum in each economy. In addition, we must create highly competitive financial
institutions within the region to prepare for any unexpected financial market vulnerability. By

14 BIS Papers No 30
doing so, we can strengthen our financial markets and also afford regional market
participants more opportunities to invest in fully qualified bonds equivalent to those issued by
western countries.
Second, we should seek advancement and harmony simultaneously in the regulatory and
supervisory systems of Asian economies by developing advanced transaction regulations
and a more transparent accounting and disclosure system. This, in turn, would enable
investors to minimise liquidity and credit risk.
It is true that one size does not fit all. However, there are many common denominators that
exist in well functioning regulatory and supervisory systems. In this context, we should
endeavour to harmonise regulations for the protection of investors’ interests, which will lead
to a more active regional bond market.
Third, we should accelerate the already extensive efforts to develop market infrastructure.
Solid market infrastructure is the basis for a promising and sound bond market. Therefore,
we should focus more on setting up efficient settlement and credit rating systems,
strengthening institutional investors and improving risk management techniques.
Meanwhile, the so-called credit quality gap between advanced and emerging market
economies is also a major impediment to the development of regional bond markets. To
bridge this gap, we should make better use of securitisation and credit guarantees, which I
know is one of this conference’s primary subjects. From our experience, I am sure that the
combination of securitisation and credit enhancement mechanisms will serve as a model for
the Asian bond market and reduce the gap between borrowers’ credit standing and investors’
requirements.
Korea is keen to contribute to the further development of local currency bond markets in the
region by drawing on the experience gained in developing the country’s bond market and
ABS market after the financial crisis. Finally, we should offer further technical assistance and
advice to emerging market economies in the region. Although many technical assistance
programmes are available, our hard-earned experience can be put at the service of these
economies more efficiently.
It would be in every regional country’s interest to lessen the trials and errors of emerging
market economies in the process of strengthening their bond markets, and at the same time
the developed economies would be able to secure new markets.

Conclusion
The question of whether bond market development will be in vain or will become a stepping
stone for sustainable economic growth wholly depends on our future attitude. Building on the
efforts and progress made so far, member economies have to continuously cooperate with
each other to secure our common goal of developing sound and efficient bond markets in the
region. I can assure you that the Korean government will spare no effort in achieving our
goal, the advent of a promising and sound Asian bond market.

BIS Papers No 30 15
 Asian financial cooperation as seen from Europe

Gunter D Baer

It is a great pleasure for me to attend this conference, jointly organised by Korea University
and the BIS. We at the BIS have organised conferences with monetary institutions in the
region, including the State Administration of Foreign Exchange in Beijing and the Bank of
Thailand in Bangkok, but this is the first with an academic institution in the region. We take
this opportunity to salute Korea University on its 100th anniversary and to wish it well in its
next century. As Seoul aspires to become a hub for Northeast Asia, Korea’s world-class
centres of learning will come into their own.
We meet in this splendid academic environment to discuss a practical and topical subject.
The work to be reported and commented on here is certainly of great relevance to a matter
that is enjoying high priority on the agenda of Asian policymakers, namely promoting the
development of Asian bond markets.
In fact, I am convinced that steps to promote the Asian bond market have the potential to
make a contribution to monetary and financial cooperation in Asia that goes beyond simply
deepening and enhancing the efficiency of today’s bond markets. Let me explain what I have
in mind by looking at Asian monetary cooperation through the European rear-view mirror. In
doing so, I will first make some broad-brush comparisons between Asian and European
developments and then present some observations on the forces that, in my mind, have
driven the process of cooperation in Europe - leaving it to you to decide whether a similar
development could be expected in Asia.
At the risk of oversimplifying, there are at least two developments that are prompting closer
monetary cooperation in Asia even as they did in Europe. The first is a marked increase in
trade integration across the region, accompanied by the emergence of poles of economic
growth independent of US demand and by a growing awareness of vulnerability to exchange
rate changes.
In Europe, this process was partly the result and partly the cause of closer monetary
cooperation. And I am proud to say that the BIS, in offering technical and meeting-related
support, played an important role in this process. Without going into detail, let me illustrate
this with a couple of examples. Between 1950 and 1958 the European Payments Union used
the BIS as an agent to permit the multilateralisation of bilateral surpluses and deficits in
Europe, thus preparing the ground for a return to current account convertibility. In the Treaty
of Rome of 1958 the European Community established the Monetary Committee, composed
of very senior central bank and finance ministry officials. And in 1964 the European central
banks set up the Committee of Governors as their central forum for cooperation. This
Committee met for 30 years at the BIS - until a new European institution was established in
Frankfurt.
In Asia the response to growing regional integration broadly paralleled the European
development. Regional central bank forums were launched, such as SEANZA (in 1957) and
SEACEN (formally in 1982), to promote joint training and research. EMEAP started in 1991,
operating first at Deputy Governor and later at Governor level, and soon established working
groups in three financial areas of particular concern to central banks. In fact, these groups
complemented those which had been in existence at the BIS for many years.
The second common trend in the development of monetary cooperation in Asia and Europe
was the impact of crises on the building of institutional arrangements. In Europe, the shock of
the collapse of the Bretton Woods system and the beginning of generalised floating
prompted the creation of the so-called narrow margins arrangement, better known as the

16 BIS Papers No 30
“snake”, supported by the European Monetary Cooperation Fund (EMCF), which settled
intervention balances and provided short-term balance of payments support. Incidentally, this
Fund existed more on paper - all operations were performed by the BIS acting as agent. Still,
the record of these early efforts at monetary and exchange rate stability was pretty
chequered, as inflation differentials necessitated parity changes and/or forced countries to
opt out of the system.
The shock of the foreign exchange and banking crisis in Asia in 1997-98 also initiated steps
towards building firmer institutional structures - though not necessarily of the kind set up in
Europe. The main strategy for building defences against currency crises was to increase the
availability of reserves, either through swap lines as agreed by ASEAN+3 under the Chiang
Mai Initiative, or simply by bolstering reserves as a kind of self-insurance. A different and, in
its multilateral character, potentially far-reaching institutional measure was EMEAP’s
launching last June of the Asian Bond Fund (ABF) in dollar-denominated instruments, aimed
at promoting the development of a regional bond market. It was recognised that a dollar-
denominated fund was the art of the possible rather than the desideratum. Accordingly,
active discussions are now under way to add a second ABF in domestic currency. An
extremely significant and multipronged approach to improving the underlying financial
structure is being pursued in parallel by ASEAN+3.
These developments in monetary cooperation pertain mainly to the past, with Asia lagging
behind Europe, where such cooperation led in 1999 to the creation of a single currency and a
common central bank. Naturally, this gives rise to the speculative question of whether
developments in Asia will take the same turn, that is, follow the European path. Since we at
the BIS do not speculate, I can neatly sidestep this question. However, having been closely
involved for many years in the process of European monetary cooperation, I can highlight
some of the forces that were instrumental in reaching the goal of monetary unification. Then I
shall leave it to you to infer what this could mean for future monetary cooperation in Asia.
Let me start with two more principal observations. The first is that the move towards
monetary union in Europe was foremost a political process. Without the political will and
impetus, monetary and financial integration at today’s level could not have happened.
Having said this, my second principal observation is that the political objective of monetary
union would also not have been achieved without the active part played by the financial
authorities and, in particular, the central banks. Just to remind you, the breakthrough in
moving to monetary union was based on a blueprint of how to realise monetary union (the
Delors Report of 1989) presented by a group of central bank Governors.
But, in addition to these points, I would identify at least three lessons to be drawn from the
process of monetary cooperation in Europe.
First, financing arrangements such as short-term swaps or medium-term balance of
payments loans (as granted by the EMCF and the European Community during the first
phase of the exchange rate mechanism) have frequently been cited as an important
prerequisite for the success of the exchange rate arrangements. I doubt that their
contribution in terms of providing financial resources was really decisive, but I recognise that
these mechanisms were sometimes of enormous psychological and tactical importance for
coming to an agreement in the negotiations.
Second, institutional aspects matter. European experience, however, suggests that big
institutional structures are not necessary for success - at least not until the moment that
responsibility for policy is transferred to a new, common institution. In fact, the process of
European central bank cooperation relied for decades on a very small permanent secretariat,
working independently under the roof of the BIS.
Third, there is nothing better than an operational framework to promote and focus monetary
cooperation. Such a framework could be a swap arrangement requiring accounting and
settlement services, or it could be an ABF or some form of exchange rate mechanism. What

BIS Papers No 30 17
matters is that any such arrangement makes it necessary to meet, to exchange views and to
take decisions in common. This in turn builds knowledge and mutual trust, which provide the
basis for getting through difficult moments in more ambitious cooperative undertakings.
Let me conclude with these observations and, as I said earlier, I leave it to you to judge to
what extent they are relevant for the process of Asian monetary cooperation. In one respect,
however, I am certain the discussions held here between researchers and policymakers form
part of the grand tradition of promoting monetary cooperation. On that note, I wish you all a
challenging and fruitful exchange of views at this conference.

18 BIS Papers No 30
 Developing the bond market(s) of
East Asia: global, regional or national?

Robert N McCauley and Yung-Chul Park 1

The various initiatives to develop Asian bond markets tend to draw on a shared analysis of
the Asian crisis of 1997-98. It is generally agreed that the mismatch between foreign
currency debt and domestic currency cash flows, on the one hand, and short-term debt and
long-term investments, on the other, left Asian firms and banks vulnerable to changing
evaluations of creditworthiness and to exchange rate depreciation. More controversial is a
related argument, which gained force as East Asia, excluding Japan, moved into a
substantial current account surplus after the crisis. Asia is thought to be missing an
opportunity if its savings flow into global capital markets only to be reinvested in some
measure in the region at higher yields and at the discretion of global investors. 2 The
development of a regional bond market or domestic bond markets is promoted to make
financial structures more resilient, to diversify sources of financing and to increase the asset
menu for investment in Asia.
Discussion of means to promote bond market development in East Asia can lose clarity
owing to the very different images of the desired outcome held by the participants. In
particular, some participants envision the creation of a regional market in which borrowers
from around the region obtain funding in regional currencies from regional investors. Others
envision improvement to the markets in which predominantly domestic borrowers meet
predominantly domestic investors. For the sake of completeness and of drawing distinctions
as boldly as possible, it is also worthwhile to consider a third image, namely that of
globalised financial markets in which East Asian borrowers and investors participate as
relatively small players.
This paper first defines terms and proceeds to sketch out these three alternative paths:
global, regional and national. It then considers where markets currently stand, recognising
that reality cuts across the neat ideal types sketched. Next, policies proposed to develop
bond markets are lined up with the different images. Finally, we conclude with our own views
on the preferred image.

1
The authors are grateful to Clifford Dammers, Guorong Jiang, Malcolm Knight, Francis Lau, Bob Rankin and
Philip Wooldridge for drawing various points to our attention and to Christian Dembierment, Denis Pêtre and
Swapan-Kumar Pradhan for statistical assistance. Any errors remain those of the authors. Views expressed
are those of the authors and not the Bank for International Settlements.
2
This is not the place to analyse these widely shared presumptions in depth. Suffice it to say that Korea’s
sizeable bond market before the crisis did not prevent a crisis. Moreover, it is not clear, at least at the
aggregate level, that Korea suffered a currency mismatch problem. Bond market development can only keep
East Asian savings in East Asia on a net basis if it increases domestic investment or consumption, leading to
higher absorption and narrow current account surpluses. Gross flows are another matter. For development of
the currency mismatch question, the prospects for narrower current accounts and two-way capital flows,
respectively, see Cho and McCauley (2003), Park (2004) and McCauley (2003a).

BIS Papers No 30 19
1. Defining terms
Asian bonds are defined by residence of issuer. They are interest bearing obligations of
Asian governments, financial institutions or corporations, wherever marketed and in whatever
currency of denomination.
Bond markets can be classified according to residence of issuer, targeted investors and
currency of denomination. For instance, the BIS international securities data cover everything
but issues by residents targeted at resident investors denominated in domestic currency
(Table 1). Issues by non-residents targeted at resident investors and denominated in
domestic currency are part of the foreign bond markets, which go by various colourful names
(yankee for United States; samurai for Japan; bulldog for the United Kingdom). Offshore (or
“euro” in the old sense) markets involve targeting investors with bonds not denominated in
their domestic currency.

Table 1
Classification of BIS securities statistics
Issues by residents Issues by non-residents
In domestic currency
Targeted at resident Domestic International (foreign: yankee,
investors samurai, bulldog)
Targeted at non-resident International (offshore or International (offshore or
investors euromarket) euromarket)
In foreign currency International International
Source: BIS (2003a), p 14.

Our approach to defining global, regional and national or domestic markets relies primarily on
the “who’s who” of issuers and investors and to a lesser extent on currency of denomination.
Thus, global markets require broad international participation on the sell and the buy side,
but can, conceptually at least, operate in as few as one or as many as all of the world’s
currencies. A regional bond market would be defined primarily as one that brings together
issuers and investors from a region, and secondarily as one that uses the currencies of the
region. Finally, domestic bond markets feature mostly domestic issuers and investors,
although foreign investors may play a more or less important role, while the currency of
choice is the local currency.
We fine-tune our definitions of global, regional and national or domestic markets to East Asia and
play down the distinction between onshore and offshore markets. Global bond markets would
mostly feature dollar or euro bonds underwritten in London, placed in Asia and Europe and
housed in Euroclear or Clearstream, as well as truly global bonds, which are also
SEC-registered, offered simultaneously offshore and in the United States and housed in both the
offshore depositories and the US Depository Trust Company. 3 Yen issues by Asian borrowers
are taken to be examples of regional bonds whether they are legally sold offshore (relative to
Japan) as euroyen bonds, onshore as samurai bonds or onshore as private placements. Issues
by Asian borrowers non-resident in Hong Kong SAR or Singapore denominated in Hong Kong or
Singapore dollars (foreign bonds rather than offshore bonds) are also termed regional bonds. It
should be clear, therefore, that we consider that there are potentially several regional bond
markets in East Asia. Further, one can imagine domestic regulations or withholding taxes

3
As well as the near-global bonds, underwritten offshore and placed in the United States under the SEC Rule
144a.

20 BIS Papers No 30
favouring Thai or Korean borrowers selling Thai baht or Korean won bonds in Tokyo or Hong
Kong. Such an offshore market would also count as a regional bond market, owing to issuers
and investors sharing East Asian residence and the use of a regional currency.

2. Global bond markets

East Asian issuers and investors could integrate themselves into global bond markets. Global
markets require cosmopolitan participation of issuers and investors, but could function with one
or many currencies. At one extreme, global markets might operate with only one currency, or
only with major currencies like dollar, euro and yen. At the other extreme, global markets might
be very accepting of different currencies.

Global bond market integration with few currencies

The extreme of one or few currencies would maximise the liquidity of bonds issued by East
Asian issuers and the liquidity of bonds bought by East Asian investors. With well developed
derivatives markets, issuers could swap the proceeds of their bond issues back into domestic
currencies; similarly, investors could contract asset swaps to transform bonds denominated in
major currencies into domestic currency assets.
The example of Canada is more to the point than that of any emerging market. When the
issuers from an emerging market borrow in major currencies, the presumption is often made
that they have no choice. That is, it is assumed that international investors, and perhaps
domestic investors as well, will not buy bonds denominated in the home currency. Controversy
attaches to whether this is a result, as it were, of natal curse (“original sin”) or is rather the
reaction of investors to a history of variable inflation and less subtle violations of creditors’
rights. With Canada, by contrast, there is obviously a choice, although that choice has
increasingly favoured the greenback rather than the loony.
Canadian corporations’ bond issues show an evolution to this first version of globalised
markets. In the mid-1970s, Canadian firms denominated 80% of their bonds in the home
currency. This proportion declined over the next 10 years, but recovered as Canadian firms
sold Canadian dollar bonds in the eurobond market (moving from the domestic bond market to
the global bond market in the second sense). This offshore demand for Canadian dollar bonds
was associated with the higher interest rates on Canadian dollar bonds than on US dollar
bonds at the time. As Canadian interest rates converged to US interest rate levels in the
1990s, the offshore demand dropped off and the Canadian dollar share started to fall again. By
2001, Canadian firms used the US dollar as much as the Canadian dollar to denominate their
bonds.
It might be thought that the Canadian corporate sector’s use of the US dollar to denominate its
bonds merely reflects the general level of the Canadian economy’s integration with the US
economy, and thus the general use of the US dollar by Canadian households and firms. It is
important to recognise, however, that the predominant role of the US dollar in the bond market
stands out as an exception (Murray and Powell (2002)). Canadian holdings of dollars amount
to no more than 10% of Canadian holdings of cash or money, and no more than 20% of bank
loans or institutional portfolios. Canadian companies denominate their bonds, but not their
bank loans, in US dollars because that way their bonds can find a wider market and thus fetch
higher prices. To some extent, this wider market depends on differences on the buy and sell
side in the readiness to use the derivatives markets. Thus, some managers of US dollar
portfolios are not prepared to swap Canadian dollars into US dollars, 4 while the large Canadian

4
In the bank loan market, the major investors, namely banks, are in the swap business, while the smaller firms
that rely on bank loans are less ready to manage a swap book.

BIS Papers No 30 21
firms that sell US dollar bonds could all be presumed to be prepared to swap out of the US
dollar liabilities into Canadian dollar liabilities to achieve a desired currency mix of liabilities. In
addition, the tilt towards the US dollar bond market by Canadian firms would enable their
bonds to gain value from the greater depth, breadth and liquidity of US dollar markets,
including the superior interest rate hedging and dealer financing capacities in the US dollar.

Graph 1
Currency denomination of bonds
issued by Canadian corporations
100%

90%
Other
80%

70%
Euro-US$
60%

50% US$

40%

30% Euro-CA$

20%
CA$
10%

0%
1975 1980 1985 1990 1995 2000 2001

Source: Bank of Canada.

Graph 2
US dollar share of Canadian assets/liabilities
Per cent
50
45
40
35
30
25
20
15
10
5
0
Cash M3 Funds Business loans Corporate bonds
Source: Murray and Powell (2002).

22 BIS Papers No 30
The Canadian corporate sector’s integration into the global dollar bond market is matched in
East Asia by firms in Hong Kong and Singapore. In part, this reflects the importance of
multinational firms headquartered in the two city economies, such as Hutchison Whampoa,
which in 2003 built up a single dollar bond to the size of $4 billion. Relying on a different
source of data, Fernandez and Klassen (2004) find that Philippine firms denominate the bulk
of their bonds in the dollar.

Graph 3
International share of corporate debt securities
Per cent
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
CA AU CN HK JP KR MY SG TH

Source: BIS Quarterly Review, Tables 12C and 16B, data for September 2003.

One can see aspects of this global vision at work in the market for East Asian dollar bonds.
Thus it is widely thought that one of the largest holders of the Republic of Korea’s 2013 dollar
bond is an insurance company connected to one of the largest chaebol in Korea. It is thought
to have bought the dollar paper and to have swapped the dollar cash flows for Korean won
cash flows, thereby matching its liabilities to its policyholders.
Proponents of the global image of bond market development for East Asia would readily
acknowledge that cross-currency hedging markets need to develop further. Only then could it
be assumed that firms can sell dollar bonds and hedge into domestic currency liabilities, and
institutional investors can buy dollar bonds and hedge into domestic currency assets. In the
face of capital controls, non-resident equity investors and multinational firms have
contributed to the development of non-deliverable forward foreign exchange markets in the
region (Ma et al (2004)). More to the point are the longer-term cross-currency swap markets,
which allow the hedging of whole streams of cash flows stretching over years. These tend to
have been small at the time of the last comprehensive measurement, in April 2001, although
they have generally grown since (Table 2).

Table 2
Cross-currency swap markets: daily
turnover in millions of US dollars

AU CN HK IN ID JP KR MY NZ PH SG TW TH CA EU

2001 510 0 285 1 13 1,969 46 0 101 2 18 21 11 361 2,190

2003 na 0 30-50 250-300 0 na 100 0 na “Volatile” “Volatile” 20-30 15-20 na na
Sources: Hohensee and Lee (2006) for 2003; BIS (2002), pp 78-81 for 2001.

BIS Papers No 30 23
Global bond market integration with many currencies
There is another, more inclusive image of global bond markets. Instead of a duopoly or
oligopoly of currencies, the international bond market can be conceived of as an open
competition among currencies. Currencies from East Asia and the Pacific could be, and to
some extent are, integrated into this global bond marketplace. The euro has surpassed the
dollar in this market and, taken together, the two currencies represent about 85% of
outstanding international bonds - close to a duopoly in practice (Graph 4). Sterling represents
the next biggest currency sector, with 7% of outstanding bonds. Taken together, currencies
of East Asia and the Pacific (broken out on the right-hand side) amount to $650 billion in
international bonds, about 6% of the total of over $11 trillion. Of these, the Japanese yen
represents the largest part (about three quarters of the regional total), with 4% of outstanding
international bonds. The Australian dollar, Hong Kong dollar, Singapore dollar and New
Zealand dollar bonds follow. There is a small New Taiwan dollar segment as well, while a
few equity-linked capital issues for Thai banks were denominated in Thai baht and sold to
international investors. All in all, five or six of the EMEAP economies have a presence in the
international bond market.
The international bond markets have shown a willingness to accept peripheral or “exotic”
currencies, especially when these offer higher yields to compensate for lack of familiarity,
greater perceived exchange rate risk and often lower liquidity. Thus, higher coupon
payments have characterised the so-called dollar bloc currencies (the Canadian, Australian
and New Zealand dollars) when these have sold well. The process of monetary unification in
Europe led to “convergence plays” on the Finnish markka, Irish pound, Portuguese escudo,
Spanish peseta, Italian lira and, most recently, the Greek drachma. This same thinking now
warms investors to Polish zloty and Czech koruna bonds (Table 3). In contrast, investors
interested in Hungarian forint bonds have had to enter the domestic market and buy
government bonds.

Table 3
Minor currency bonds outstanding in the
international bond market, end-2003
Billions of US dollars

Argentine peso 0.9 Singapore dollar 9.4

Czech koruna 8.0 South African rand 9.3
Hong Kong dollar 45.6 New Taiwan dollar 4.0
Polish zloty 5.2 Thai baht 1.7
Source: BIS.

Outside Europe, foreign investors have had their choice between buying South African rand
bonds in the international bond market and buying the government bonds in the domestic
market. Chile and Mexico have not allowed their bonds denominated in their respective
pesos to be sold in international markets.
A common element in the dollar bloc, peripheral European and other issues is a higher
coupon than that available on bonds denominated in major currencies. One could argue, in
fact, that, although all of the currency sectors listed in Table 3 satisfy the BIS definition of
international bonds, the relatively low-coupon Hong Kong dollar, Singapore dollar and New
Taiwan dollar bonds have not been widely marketed outside the three economies. If wider
demand in the international bond market does indeed depend on attractive coupons, then the
higher-coupon, moderate-inflation currencies of East Asia may have the best shot at
international portfolios. In particular, the Korean won, Philippine peso and Indonesian rupiah

24 BIS Papers No 30
in East Asia, and Indian rupee bonds in South Asia, could meet with the greatest demand.
The acceptance of such bonds to investors in global offshore markets has not been tested to
date owing to the unwillingness of domestic authorities to permit them.

Graph 4
Currency composition of the international bond market

Dollar Euro Pound Other Yen Aust dollar

HK dollar Sing dollar NZ dollar NT dollar Baht

3. Regional bond market

In a regional bond market, governments, banks and companies in the region would tap
institutional investors, banks, mutual funds and individuals in the region. There is a strong
feeling in East Asia that the region has never achieved what Europe had before the euro. As
Donald Tsang, Hong Kong’s then Financial Secretary, asked rhetorically, “How is that we in
Asia have never been able to replicate the eurobond market success in this part of the world?”
(Tsang (1998)). We interpret the reference to have been to European issuers’ selling bonds
denominated in European currencies largely to European buyers. This section first gives an
example of an East Asian issuer tapping regional portfolios in a regional currency. Then the
truth of Tsang’s observation is demonstrated in terms of currency sectors. This section then
considers whether there are important regional elements in the international dollar bond market
and, more narrowly, in the international market for Australian dollar bonds.

KAL bond issue

An example of an Asian firm tapping a regional bond market is provided by the Korean Air
Lines issue in 2003. Given Korea’s proximity to Japan and the flow of tourists from Japan to
Korea, Korean Air has a regular flow of yen receipts from travel agents in Japan. By pooling
these cash flows, and adding a credit enhancement from the Korea Development Bank, a
yen-denominated bond could be issued that met the quality demands of Japanese investors.

Regional issuers in regional currencies: Europe versus Asia

It is well known that the introduction of the euro has helped to encourage European issuance
in the new currency, and led to rapid growth of the euro bond market. Part of this growth has
been in the international bond segment, and has led the euro sector to overtake the dollar
sector (Graph 4 above). The relevance of all these observations for East Asia is at best
distant, since few foresee the introduction of a common currency in East Asia for a
generation.

BIS Papers No 30 25
Of greater relevance is the record of the international bond market before the introduction of
the euro. Scrolling back 10 years, what role did the Deutsche mark, its surrogates (like the
Dutch guilder, Danish krone or ECU), and its immediate competitors like the French franc
and others, play in the meeting the international financing needs of European governments,
banks and corporations?
Looking at the left-hand panel of Graph 5, it is clear that the euro’s predecessor currencies
played a predominant role in the international bond offerings of European (defined here as
current EU) borrowers. To be sure, dollar issues figure importantly, but since at least five years
before the euro, its predecessor currencies have accounted for more issuance than the dollar.
The regional element is even larger when the share of sterling bonds is taken into account.
The right-hand panel tells a very different story. Issuance by East Asian governments, banks
and corporations in the international bond market is overwhelmingly dollar-denominated.
Regulation, buy-side characteristics and exchange rate management have all played roles in
preventing Asia’s currencies from posing tougher competition to the dollar.
To begin with, the authorities in important Asian countries have not been prepared to accept
non-resident issues targeted at resident investors or offshore issues in their currencies. In
the case of Singapore, foreign issuers have been allowed to sell Singapore dollar bonds, but
only to swap the proceeds into foreign currency. Thus, a multinational company with
operations in Indonesia, the Philippines and Thailand that found the Singapore dollar an
attractive currency to borrow in, both because of its movement with regional currencies and
because of the low interest rate, might be able to access the Singapore dollar bond market,
but not hold liabilities in the Singapore dollar at the end of the day. Regulation in the form of
Japanese-language registration requirements has also made especially opportunistic
issuance in the yen difficult.

Graph 5
Announced international bonds and
notes issuance by nationality and currency
In billions of US dollars
(semi-logarithmic scales)

European Union1 East Asia and Pacific2

1
European Union refers to the current membership. Euro is the euro or its predecessor
currencies. 2 Comprises Australia, China, Hong Kong SAR, India, Indonesia, Japan, Korea, Malaysia,
New Zealand, the Philippines, Singapore, Taiwan (China) and Thailand.
Sources: Dealogic; Euroclear; ISMA; Thomson Financial Securities Data; BIS.

26 BIS Papers No 30
Nishi and Vergus (2004) emphasise the risk appetite of Japanese investors, and this factor,
unlike regulation which has only eased over time, can explain why East Asian issuance in
yen has not regained the levels reached before the Asian crisis (while that in the dollar has).
Japanese investors are increasingly prepared to run currency risk, otherwise there would not
be such an active market for Australian dollar paper there. When it comes to credit risk,
however, Japanese investors’ own recent domestic financial history has not well prepared
them to accept it. Moreover, they feel burned by their experience with domestic high-yield
issuers like Mycal and more so with foreign high-yield, high-risk issuers like Argentina. Thus,
while a below investment grade issuer like the Republic of the Philippines, or even
investment grade Federation of Malaysia, have tapped the dollar and euro segments of the
international bond market, they have not sold much in the way of yen bonds. The yen bond
market is also limited by the risk appetites of Japanese institutional investors. With rare
exceptions like the leasing group Orix, Japanese institutional investors, like Japanese
households, have been more willing to take on currency risk than credit risk in external
investments. To be fair, as argued by Remolona and Schrijvers (2003), starting with a
low-risk portfolio, yield enhancement through acceptance of credit risk is inherently a trickier
proposition than yield enhancement through acceptance of foreign exchange risk. Because
of the fat left tail of the distribution of returns on risky bonds, diversification requires very
broad portfolios (which are particularly hard for a bond-picking household to assemble).
Exchange rate management may also play a role in the limited development of an Asian
regional bond market. Regional currencies have tracked the yen only to a limited extent, at
least until recently. This means that from the perspective of issuers worried about the
possibility of their liabilities blowing up, selling yen bonds may have posed greater exchange
rate risk than selling dollar bonds. The contrast with Europe would be that after the onset of
generalised floating in 1973, a number of European currencies shared much of the Deutsche
mark’s movements against the dollar, thereby reducing the risk of governments’ and
corporations’ mark borrowing. Nevertheless, Schmidt (2004) has argued that the yen
markets have missed an opportunity in recent years insofar as regional currencies, especially
the Korean won, have shared much of the yen’s movements against the dollar.
Exchange rate management also bears on the attractiveness of the Hong Kong dollar as a
currency to denominate bonds. Typically, there is a premium of long-term Hong Kong dollar
yields over US dollar yields, in part reflecting currency risk and in part reflecting liquidity. To
pay more for a Hong Kong dollar bond than a US dollar bond thus requires a view on the
Hong Kong dollar. Thus, most international issuance of Hong Kong dollar bonds has been
either opportunistic (that is, driven by profitable opportunities to swap the proceeds) or for
funding assets in Hong Kong.

Regional elements in international bond markets

There may be important regional elements in the global bond markets as they concern East
Asia and the Pacific. 5 One regional element in the international bond market is the placement
of international Australian dollar bonds in Japan among retail investors. Another, broader
regional element may be a regional bias in the investor base for US dollar bonds sold by
East Asian borrowers.

5
Above, it was argued that the Canadian corporate sector had integrated its bonds into global capital markets,
progressively eschewing the Canadian dollar in favour of issuing bonds into the broader, deeper and more
liquid US dollar market. This was taken to be a case of embracing the global bond market. From another
perspective, this is a case of regionalism in bond markets, since many US buy-side investors’ portfolio
guidelines or restrictions would treat Canadian issuers the same as US issuers. What is on one view a strong
case of globalisation, might therefore on another view seem to be a case of regionalism within global markets.

BIS Papers No 30 27
Uridashi market
As explained in the paper in this conference by Nishi and Vergus (2004), Japanese securites
houses market Australian dollar bonds formally issued as international bonds to Japanese
households. Since the household buyers are averse to credit risk, if not currency risk, the
issuers of the bonds are mostly very high-quality governments or agencies from outside
Australia. They are opportunistic issuers looking only for cheap funding when measured
against US dollar Libor or Euribor. Through the swap market, their liabilities ultimately are
taken on by Australian banks or firms financing assets in Australia. While complicated, the
essence of the transactions is the willingness of Japanese households to take on the
currency risk of the Australian dollar in exchange for a decent coupon. And what is clear is
that these bonds require an ongoing investment in providing information to Japanese
households by the Japanese securities firms.
Is it possible to imagine this same marketing being applied to the sale of, say, Korean won
bonds to Japanese households? While Korea does not possess Australia’s aura of a
vacation and honeymoon destination, it has had another advantage in recent years. As noted
above, the won has shared a considerable, albeit varying, part of the yen’s movements
against the dollar. A Japanese investor in a won bond would have experienced less volatility
in the value of their holdings compared to an investment in a US dollar bond. Were such a
co-movement to persist it would favour the development of Japanese demand for won
bonds. Indeed, the relative stability of the won in terms of yen led Korean companies,
reportedly small and medium-sized enterprises with little in the way of yen cash flows, to
build up $7 billion in yen debt from Korean banks in 2002. 6

A regional bias in investment in US dollar bonds issued by Asian borrowers?

A question has arisen whether Asian buyers figure disproportionately among investors in US
dollar bonds issued by East Asian governments and corporations. Market participants have
coined the term “Asian bid” to describe this asserted clientele, which is used to explain the
spreads of such bonds or their stability (Fernandez and Li (2002), Woods (2002), Schmidt
(2004)). McCauley et al (2002) consulted the trade press for reports of the placement of
dollar- and euro-denominated bonds of East Asian issuers and reported that almost half were
placed in Asia. Eichengreen and Park (2003) question whether there is anything more to the
Asian bid than home country investors’ buying dollar bonds.
There is agreement that there is home bias. For instance, Korean banks and institutional
investors are reported to be important holders of the Republic’s and Korean Development
Bank’s dollar bonds; Philippine banks match their US dollar deposits with the shorter-dated
Philippine government dollar bonds; and Chinese banks are reported to be important holders
of Chinese dollar bonds.
The question is whether, in addition to home bias, there is substantial regional bias.
Eichengreen and Park show that during the period covered by McCauley et al, Japanese
holdings of Asian bonds (presumably mostly dollar-denominated) actually fell in dollar
amounts. They ask, if not Japanese investors, then who are the Asian investors with a
regional taste in bond buying? Moreover, they argue that Asian investors do not plausibly
have any informational advantage in buying Asian bonds and that, given the similarity of
economic structure and business cycles, Asian investors cannot sensibly diversify by buying
Asian dollar bonds.
The issue will not be resolved here. The Box reviews the evidence from the BIS banking data
and from the IMF’s portfolio capital survey in 2002. Substantial holdings of Asian bonds in

6
See Financial Supervisory Service (2002).

28 BIS Papers No 30
Hong Kong or Singapore leave open the question of the ultimate beneficial owner (eg an
Indonesian bank branch there holding an Indonesian bond or a French-owned insurer there
holding a Korean bond). There remains room for diverging interpretations of the data on
cross-border holdings of Asian bonds in Asia.

Regional bond market: a summary

To conclude, although Asia enjoys more than one financial centre featuring the issuance of
bonds by non-resident borrowers, the yen, the Hong Kong dollar and the Singapore dollar
have not to date attracted a large share of the offshore issuance from the region. Unlike the
EU countries in the years before the introduction of the euro, the international dollar market
still captures most of the offshore issuance of bonds by regional borrowers.
A variety of factors seem to be responsible for the relatively small role of existing regional
bond markets in the international fund-raising of regional borrowers. Regulation used to limit
access to the yen market by lower-rated borrowers and still imposes some costs. At this
point, however, it is probably the aversion of Japanese investors to credit risk which poses
the larger hurdle to regional issuance. Regulation does limit the ability of all foreign issuers to
arrange Singapore dollar liabilities, while the currency board system in Hong Kong makes
Hong Kong dollar funding generally unattractive to international borrowers.
While there may be important regional currents in the flow of funds in the US and Australian
dollar global bond markets, our overall result is that the regional bond markets to date have
only offered limited funding options. Thus, we turn to the national markets, without having
found a very solid alternative to them in the existing regional markets.

Box
Cross-border holdings of Asian bonds: banks and all investors
Robert N McCauley and Patrick McGuire

While there is broad agreement among policymakers in East Asia that further financial integration in
the region would be desirable, no such consensus has emerged regarding the proper understanding
of the current extent of such integration. Market-based analysts highlight the importance of the
“Asian bid” - that is, a disproportionate representation of regional buyers - in the primary and
secondary market for dollar bonds issued by East Asian governments, banks and firms.1 This view
has been challenged, however, by reference to official Japanese data on holdings of bonds by
Japanese residents, which suggest low and declining holdings of the obligations of Asian issuers.
This box consults two sources of evidence to shed light on the extent of the regional bias in holdings
of international bonds issued by East Asian borrowers. First, the BIS international banking data
report banks’ cross-border claims that take the form of bonds, providing country detail and a time
series perspective. Banks are natural buyers of bonds, especially those of relatively short maturity
or those bearing floating interest rates, but represent just one investor segment. Second, the IMF
portfolio survey of securities holdings provides broader coverage of the investor base, capturing
institutional investors as well as banks, but represents only a snapshot at end-2002. The IMF data
are in principle universal, while the BIS reporting area does not include all the important Asian
economies.

BIS international banking data

Even as a means to profile a single segment of investor demand, the BIS data are limited by the
reporting area, which does not include some important economies in East Asia (BIS (2003b)). In

____________________________
1
See Schmidt, 2004.

BIS Papers No 30 29
 Box (cont)
Cross-border holdings of Asian bonds: banks and all investors

particular, while Hong Kong SAR, Japan and Singapore are long-time reporters, Australia and
Taiwan (China) have joined only recently. Yet to participate are China, Korea, Malaysia and
Thailand. Thus, Asian holdings of Asian bonds as measured by the BIS data will be smaller than the
actual amount insofar as banks in these latter countries hold bonds issued by their neighbours. The
data include both international bonds and domestic securities held offshore, for instance a Hong
Kong bank’s holdings of a Korean treasury bond (which are, judging by Korean flow of funds data,
very small).2
The BIS banking data do suggest a regional bias in holdings of Asian bonds by Asian banks. This
conclusion emerges from two findings. First, as of the fourth quarter of 2003 BIS area banks held an
estimated $66 billion in bonds issued by borrowers from Asia excluding Japan.3 In terms of country
composition, the largest holdings are vis-à-vis Singapore and Korea (as suggested by the BIS data
on international bonds issued by Asia excluding Japan). Second, an estimated two thirds of these
bonds are held in Asia, including Hong Kong, Japan, Singapore and Taiwan (see the graph below).
About half the rest are held by banks in the United Kingdom.4 Holdings of Asian bonds by reporting
banks in Asia were squeezed by the combination of regional banks’ loss of access to international
interbank markets during the period of the Japan premium and the Asian crisis, but have risen since
late 1999.

Estimated holdings of Asian bonds by BIS area banks

In billions of US dollars

Asia
United Kingdom
Rest of the world
40

20

0
1996 1997 1998 1999 2000 2001 2002 2003

IMF portfolio survey

The IMF survey of cross-border portfolio holdings of bonds provides a matrix of holdings for East
Asia and allows these holdings to be put into a global context (Bae et al (2006)). It shows holdings
of long-term debt securities at the end of 2002 and includes both foreign currency and local
currency bonds. These data need to be interpreted with some care because the decomposition by
country is often not complete.

____________________________
2
The data also include some holdings of short-term paper, such as certificates of deposit, that are not relevant
to the question under discussion.
3
Asia excluding Japan includes Hong Kong, Singapore and Macao, typically classified as offshore centres by
the BIS.
4
The country composition of Hong Kong banks’ bond holdings is estimated using the composition of loans,
and bond holdings are estimated for Japan and Singapore from country by country data on non-loan claims.

30 BIS Papers No 30
 Box (cont)
Cross-border holdings of Asian bonds: banks and all investors
The data indicate an uneven but in aggregate high degree of regional bias in bond holdings across
Asia excluding Japan. Asia excluding Japan holds over half (51.3%) the bonds issued by borrowers
in that area (last row of the table). In the first column of the table, for instance, investors in Hong
Kong put 12.8% of their international bond portfolio into Asian bonds, and, given the size of their
aggregate portfolio, they account for a high share (7%) of international holdings of such bonds.
Excluding Japanese bonds, Hong Kong holds 16% of global holdings of Asian bonds. Singapore puts
a higher fraction of its overall international bond portfolio in Asian bonds, but, given its portfolio size,
accounts for a smaller share (13.9%) of global holdings of Asia excluding Japan’s bonds. These
portfolio data support the hypothesis of a regional bias.
It turns out that the largest foreign investor in the region, Japan, does not show an Asian bias. While
Japan’s holdings of Asian bonds amount to more than Hong Kong’s or Singapore’s holdings (last row
of the table), they are very small from the Japanese perspective. Of the grand total of $7.7 trillion in
cross-border bond investment captured by the survey, Asian bonds amount to about $225 billion
(about 3%), of which Japanese bonds account for around two thirds ($160 billion). Global holdings of
bonds from Asia excluding Japan thus amount to approximately 1% of holdings. Japan’s holdings of
bonds from Asia excluding Japan are also around 1%, which is about par. Despite the scale of the
Japanese portfolio and the country’s proximity, therefore, Japan has no disproportionate holdings of
Asian bonds. In contrast, with double digit percentage weights on Asian bonds, investors in Hong
Kong, Indonesia, Korea, Macao (where the currency board vis-à-vis the Hong Kong dollar must play
a role), Malaysia and Singapore do favour regional bonds. Given the scale of holdings, the regional
bias derives mostly from the behaviour of portfolio managers in Hong Kong and Singapore. The
result of a neutral Japanese weight, on the one hand, and regional bias elsewhere in the region, on
the other, is the high fractions of internationally held bonds of Asia excluding Japan to be found in
Asia (Table A, last column).

Table A
Cross-border investment in bonds, end-2002
In millions of US dollars

Investment from: Total

Total Asia
Invested in: from
HK ID JP KR MO MY PH SG TH in Asia share1
Asia
China 1,232 ... 578 38 15 ... 2 416 – 2,281 3,430 67
Hong Kong SAR ... 57 1,137 455 521 40 58 1,653 20 3,941 7,208 55
India ... ... 159 47 8 1 ... 241 – 456 788 58
Indonesia ... ... 49 78 ... 1 4 869 – 1,001 2,462 41
Japan 5,351 ... – 29 21 ... 5 3,828 – 9,234 159,937 6
Korea 4,202 ... 5,348 ... 23 51 15 2,586 – 12,225 25,015 49
Macao SAR ... ... – ... ... ... ... ... – 0 1 0
Malaysia 2,085 3 1,823 332 3 ... 9 1,830 – 6,085 8,844 69
Philippines ... 5 1,389 81 ... 4 ... 595 – 2,074 7,805 27
Singapore 1,842 23 680 144 31 41 23 ... – 2,784 6,451 43
Taiwan, China 674 ... 46 ... 13 ... 7 333 – 1,073 1,372 78
Thailand 447 ... 550 24 ... 1 ... 542 – 1,564 1,895 83
Total in Asia 15,833 88 11,759 1,228 635 139 123 12,893 20 42,718 225,208 19
Total investment 123,528 703 1,135,519 9,608 2,637 471 1,553 52,830 1,344 1,328,193 7,733,214 17
1
Asia share 12.8 12.5 1.0 12.8 24.1 29.5 7.9 24.4 1.5 3.2 2.9 .
Share of inv in
1
Asia 7.0 0.0 5.2 0.5 0.3 0.1 0.1 5.7 0.0 19.0 . .
Share of inv in
1
Asia excl JP 16.1 0.1 18.0 1.8 0.9 0.2 0.2 13.9 0.0 51.3 . .
1
In per cent.
Source: IMF.

BIS Papers No 30 31
 Box (cont)
Cross-border holdings of Asian bonds: banks and all investors

It can still be asked: who are the beneficial owners of the bonds held in the financial centres of Hong
Kong and Singapore? To the extent that they are held at branches of banks headquartered outside
the region, one could question whether there really is a regional bias. Whether institutional investors
like insurance companies and pension funds would hold bonds in these centres to fund liabilities to
retirees and policyholders outside the region is another issue.
Based on the data reviewed, it can be said that a disproportionate share of cross-border holdings of
bonds issued by East Asian borrowers is held in bank and institutional portfolios located in East
Asia. Whether the ultimate beneficial ownership of these securities, in some sense, is likewise
concentrated in Asia remains an open question.

4. National bond markets

A third image for bond market development is the improvement of the working of the existing
national bond markets. This image calls for many markets, not one global or regional market,
to be developed.
In the wake of the Asian crisis, and given deliberate attempts in places to increase issuance,
these markets have reached substantial size, aggregating $1.2 trillion across East Asia
excluding Japan (Jiang and McCauley (2004)). Even if one accepts HSBC’s definition of an
investable universe of bonds, one still is confronted with an aggregate size of $270 billion,
and this does not (yet) include China and Indonesia. This is considerably larger than the
stock of dollar-denominated Asian bonds and even a larger multiple of outstanding
yen-denominated Asian bonds. While these markets could no doubt be larger (Eichengreen
and Luengnaruemitchai (2004)), the size of the local bond markets alone should give one
pause when considering proposals that would ignore the development of national markets in
favour of regional markets.
This hesitation only increases when one considers that these national markets suffer to
varying degrees from a lack of liquidity and a lack of investor diversity. One finding that holds
across G10 government bond markets is that size matters for liquidity (McCauley and
Remolona (2000)). That is, the larger the outstanding bonds, the higher the transactions
volume and the narrower the bid-ask spread. This result holds across the local economies as
well, although it appears to be weaker partly because of the developmental efforts of Hong
Kong and Singapore (Graph 6). To be sure, other factors, such as the concentration of
issuance in particular issues and the breadth of financing markets, make a difference. The
implication of the importance of size for liquidity, however, is that global or regional issuance,
particularly by the benchmark issuer, the government, comes at an opportunity cost. Every
billion dollars of bonds sold abroad are bonds that will not contribute to the liquidity of the
domestic market.
The lack of investor diversity is also related to liquidity. Lack of a diverse investor base tends
to make a bond market one-sided, with all the players at times attempting to adjust their
portfolios in the same direction. In particular, a predominance of buy-and-hold investors can
leave the secondary markets quite inactive. Even if the market has more active accounts,
they may, like the Korean investment trust companies or the Thai bond mutual funds, be hit
simultaneously with liquidity pressure, leading liquidity to dry up and prices to gap. It appears
that a lack of diversity, as measured by the Herfindahl index of the concentration of bond
holding, is related to the bid-ask measure of liquidity (Graph 6).

32 BIS Papers No 30
 Graph 6
Liquidity in East Asian bond markets
Size, trading, issue size and concentration

Size (x-axis) and Trading volume (x-axis)1

trading volume (y-axis)1 and bid-ask spread (y-axis)2

y = 0.81x + 0.96 y = –2.79x + 8.84 7

3.5 ID
R = 0.24 R = 0.76
TW KR 6
3
5
SG IN 2.5 TH 4
HK 3
HK PH 2
MY MY
2
TH 1.5 KR
SG 1
ID
IN TW
1 0
0 0.5 1 1.5 2 2.5 3 1 2 3 4

Issue size (x-axis)1 and Concentration of bond holdings (x-axis)3

bid-ask spread (y-axis)2 and bid-ask spreads (y-axis)2

TH
3.5 y = 8.67x – 1.75 ID 7
MY R = 0.60
HK 3 6

y = –0.98x + 3.31 2.5 5

R = 0.41
2 TH 4
MY HK
1.5 3
KR
1 2
SG IN SG
0.5 KR 1
TW TW
IN
0 0
0 1 2 3 4 0 0.2 0.4 0.6 0.8 1
1 2 3
In billions of US dollars; in logs. In basis points. Herfindahl-Hirschman index.
Sources: Barclays Capital; Bloomberg; Deutsche Bank; HSBC; BIS calculations.

One way of diversifying the investor bases in national bond markets is to open them up to
foreign investors, but efforts to develop a regional market could actually hold back such an
opening. In Korea, for instance, while about 40% of the equity market is foreign-owned, only
0.4% of the government bond market is foreign-held. Why this is so is not clear: Takeuchi
(2004) considers the impediments to foreign investment in national bond markets and
McCauley (2004) considers the costs and benefits of doing so. It is sometimes proposed that
an easy way to get around these impediments might be to issue bonds offshore in a regional
bond market. But this would not really bring foreign investors into the domestic bond market.
The next time that investment trust companies or bond mutual funds suffered heavy
withdrawals, there would still be no bid from foreign investors who could see a buying
opportunity in the temporary liquidity pressure on selected institutions.

BIS Papers No 30 33
5. Images of bond market development and policies
This section considers the mapping between images of bond market development and
policies that have been proposed to accelerate bond market development in Asia. Different
intentions imply different policies.

Policies for the global bond market

Recall that we sketched out two versions of global bond market development, one with a
narrow range of currencies and another with a wide range. Somewhat different policy
considerations apply for each.
The first version of global bond market development, which emphasises the network
externalities of a currency oligopoly of dollar and euro, would at first blush seem
unacceptable to many observers in the region, because it would apparently not respond to a
central lesson of the Asian crisis. That is, while reliance on dollar or euro bond funding would
perhaps minimise liquidity risks, it would involve firms in the region mismatching projects
generating local currency cash flows with debt requiring foreign currency payments. In short,
integration into the dollar and euro bond markets might address the maturity mismatch
problem but would seem to leave the currency mismatch problem dangerously unaddressed.
Proponents of this version of the global bond market development path might respond,
however, that derivatives markets could transform dollar assets and liabilities into local
currency exposures. After all, the Canadian example is not really a case of mismatch of
currency obligations and receipts. Well developed currency swap markets allow Canadian
firms to transform their US dollar obligations back into Canadian dollars. Against this, it might
be argued that such hedging markets are developed to varying extents in the region (BIS
(2002), Hohensee and Lee (2004)). Those who embrace this image of global bond market
development, therefore, would need to consider the means for governments to encourage, or
at least allow, the development of cross-currency swaps.
Some would argue that even this global approach needs healthy national bond markets.
Recently, the Australian Treasury considered whether to repay all of the Commonwealth’s
bonds denominated in Australian dollars. As noted above, Australian firms and banks enjoy
access to Australian dollar fixed and floating rate finance through the global (and regional)
bond market, both directly and through currency swaps. Should the central price discovery
mechanism in the Australian dollar bond market, the nexus of cash government bonds,
repurchase markets and 10-year futures be allowed to wither? The overwhelming answer
during the government’s consultation was no. To market participants, it was not clear that the
currency and interest rate swap markets could function successfully, in both normal and
stressed markets, as the central price discovery mechanism, that is, in the absence of a base
of pricing of government bonds. In the end, Australia (2003) decided to retain its domestic
government bond market, even if there were no funding need. 7
Regarding the many-currency version of global bond market development, its implied policy
agenda would be regulatory change that would permit the issuance of bonds denominated in
Asian currencies in London, New York or Tokyo. As discussed in Eschweiler (2004), this
would be quite a programme. Like the first image of global bond market development, this
second version, it might be argued, requires national bond market development as a base for
the pricing of bonds denominated in local currency, even if they are to be sold abroad.

7
See McCauley (2002) for a discussion (now moot) of whether global fixed income markets could function
without US Treasury debt.

34 BIS Papers No 30
Policies for a regional bond market
Policies to promote regional bond market development include those on the sell side, those
on the buy side and infrastructure. Consider each in turn.
Just as integration of new, Asian currency sectors into the global bond market would require
the authorities to permit offshore use of their currencies, so, too, any widening of regional
bond markets from the status quo of the yen would require regulatory change. The European
experience suggests that the regional development occurred on a wider base than the
Deutsche mark alone, over which the German authorities continued to exercise control.
Hybrid currencies like the ECU served as ways around that control.
On the buy side as well, a genuinely regional market would require investment from a
number of countries. Many large portfolios in the region, national pension or provident funds,
for instance, have barely started their external diversification. When they are permitted to
diversify externally, the global dollar and euro markets are among the natural first steps. In
Thailand recently, the Bank of Thailand has authorised external investment of selected
portfolios. In these authorisations, it is reported, some amounts are earmarked for investment
in regional bonds. Thus, there are opportunities in the process of opening up fixed income
portfolios to external investment for channelling funds into regional markets. Analytically, the
question is whether such regional allocations come at the expense of global investment, or
are in addition to them, as policymakers accelerate the opening in the pursuit of regional
bond market development.
Many observers take three policy initiatives for infrastructure to advance regional bond
market development as a package. In particular, they see a regional credit guarantee
agency, a regional bond rating facility and a regional clearing and settlement capacity all as
pieces of infrastructure needed for regional bond market development (Oh and Park (2006),
Park and Rhee (2006)).
From our perspective, the regional credit guarantee agency is less specific to a particular
image of bond market development. A regional credit guarantee agency could support the
credit of borrowers from the region in accessing global, regional or domestic markets. For
example, the Electricity Generating Authority of Thailand obtained World Bank guarantees
for the principal and the next interest payment of a 10-year bond (Schmidt (2004, pp 49-50)).
The bond was denominated in dollars and sold in the global market. To take another
example, the Korean Air Lines deal described above used a Korean Development Bank
credit enhancement to access the regional market in Tokyo. To take still another example,
the Hong Kong Mortgage Corporation (a wholly owned subsidiary of the Hong Kong
Monetary Authority), promotes mortgage securitisation in the territory with guarantees.
Substitute a regional credit rating agency for the World Bank, the official Korean guarantor or
the Hong Kong Mortgage Corporation, and it is apparent that a regional credit guarantee
agency could serve any of the three images of bond market development.
Proposals for a regional rating agency or a regional clearing capacity, by contrast, strike us
as specific to the image of regional bond market development. Global rating agencies
already exist, and are increasingly targeting domestic bond markets with ratings specific to
them (Packer (2003), Kisselev and Packer (2006)). National rating agencies also already
exist. Similarly, a regional clearing capacity would sit between the global clearing capacity of
Euroclear and Clearstream, on the one hand, and national clearing operations, on the other
(Braeckevelt (2006)).

Policies for domestic bond markets

Policies for domestic bond market development in general, or for domestic government bond
market development in particular, have been reviewed in a number of forums (APEC (1999),
CGFS (1999)).

BIS Papers No 30 35
One aspect of developing domestic bond markets is opening them up to foreign investment.
As argued above, foreign investment makes for a more diverse base of investors, even if the
inward capital flow is not needed given current balance of payments surpluses in the region.
The reason for the low levels of inward investment in local bond markets is not clear, and
Takeuchi (2006) surveys market writings to identify the most important impediments to
foreign investment.

6. Conclusions
We recommend that emphasis be placed on the third image of bond market development for
Asia. That is, national bond markets should be developed with a view to integrating them with
global markets at some stage. Even if one embraces the image of a global bond market,
development of the national markets would probably be necessary under current
circumstances.
The impulse to regional development can contribute to national bond market development by
bringing politically acceptable peer pressure to bear. The process of discussing the
circumstances under which there could be more regional investment in domestic bond
markets may raise the political salience of policy changes that will make domestic bond
markets more friendly to foreign investors. The announcement by Thailand of an intended
waiver of withholding tax on coupon interest paid to foreigners, which took place at the time
of an international conference on Asian bond market development in Bangkok in October
2003, may be a case in point.
One might ask why peer pressure could accomplish what market pressure has failed to do.
One answer is that market pressure has not been very strong because underlying balance of
payments positions mean that most countries do not need the foreign capital. When
countries in the region were running deficits, eg in the pre-crisis period in Thailand, there
were a large current account deficit and a real funding need, but no government bonds. Now
there are government bonds, but no particular need for additional capital inflows. It might be
noted in this connection that the United States repealed its withholding tax on non-resident
holdings of bonds only once a large current account deficit opened up in the mid-1980s.
Moreover, market pressure is subject to the interpretation that market participants are
arguing their narrow interests - to gain access to new revenue sources in domestic bond
markets - rather than the national interest. This interpretation of self-interested advice in part
reflects the memory of 1997-98. In contrast, peer pressure is less subject to the interpretation
that advice to make markets more investor-friendly is self-interested. Finally, peer pressure
may be more effective when it is collective.
Care must be taken that measures taken to develop regional bond markets do not slow the
development of domestic bond markets: the Korean Air Line deal provides an example of a
very sensible regional issue using both securitisation of cash flows and guarantees to meet
the high demand for credit quality of the Japanese investor base. A less sensible example
would be further duplication of ABF1 - of course, not in the works - which could encourage
dollar issuance by borrowers in the region at the expense of domestic market growth.
Another untoward example would be the sale of government bonds offshore as part of an
effort to develop regional markets. In particular, Kingdom of Thailand baht bonds might be
underwritten and sold in Tokyo. As argued above, however, liquidity divided is liquidity lost.
Every baht bond not traded in Bangkok would be one less bond that could be repurchased
there or that could form part of a benchmark bond there, making the domestic market that
much smaller and less liquid. In addition, in political economy terms, the easy option of
offshore issuance may militate against removing domestic impediments.
Similarly, care must be taken that infrastructure development for the region proves both
consistent with eventual global integration and financially self-sustainable. We have argued

36 BIS Papers No 30
above that a regional credit guarantee agency could serve the goal of domestic bond market
development as well as regional bond market development. The ambition to bring small and
medium-sized enterprise liabilities to the bond market must be informed by an analysis of
losses on such programmes in recent years in Japan, Korea and Hong Kong (Jiang (2004)).
Otherwise such an effort cannot be sustained. Similarly, it is easier to extend guarantees to
highly leveraged firms not enjoying investment grade ratings than it is to ensure the revolving
nature of the guarantees and capital supporting them. The view that Asia is stuck with a
mismatch between the credit ratings that investors desire and the credit ratings that its
companies are assigned is underpinned by a very partial view of corporate finances. The
example of PCCW in Hong Kong, which started as a leveraged buyout of the local telephone
company but is now managing its finances to achieve an A rating, reminds us that, within
limits, corporate credit ratings are choice variables of corporate management.
Regional initiatives for a rating agency and clearing system are structurally more risky in their
dependence on an image of regional bond market development. A regional rating agency will
ultimately have to pass the test of being at least a point of reference for investors from
outside the region. In other words, its establishment needs to anticipate the integration of
bond markets in the region into global markets. By its nature, a regional clearing system
must ultimately be hooked up with national systems on one side and global ones like
Euroclear and Clearstream on the other.
Since this conference is being held in Korea to mark the hundredth anniversary of Korea
University, perhaps we could end with a success criterion for national bond market
development for Korea: instead of two orders of magnitude difference between foreign
ownership of bonds and stocks in Korea, just one order of magnitude. That is, the Korean
bond market might better have 4% foreign ownership than its present level of 0.4%.

References
Asia-Pacific Economic Cooperation (1999): Compendium of sound practices: guidelines to
facilitate the development of domestic bond markets in APEC member economies,
September.
Australia, Commonwealth of (2003): “Statement 7: budget funding”, Budget for 2003-03,
Canberra, May.
Bae, Kee-Hong, W Bailey and Young-Sup Yun (2006): “Determinants of bond holdings by
foreign investors”, this volume.
Bank for International Settlements (2002): Triennial Central Bank Survey: foreign exchange
and derivatives market activity in 2001, March.
——— (2003a): Guide to the international financial statistics, BIS Papers, no 14, February.
——— (2003b): Guide to the international banking statistics, BIS Papers, no 16, April.
Braeckevelt, Francis (2006): “Clearing, settlement and depository issues in Asia”, this
volume.
Cho, Yoon Je and Robert McCauley (2003): “Liberalising the capital account without losing
balance: lessons from Korea”, in China’s capital account liberalisation: international
perspectives, BIS Papers, no 15, April 2003, pp 75-92.
Committee on the Global Financial System of the G-10 Central Banks (1999): “How should
we design deep and liquid markets: the case of government securities”, CGFS publication,
no 13, October.
Eichengreen, Barry and Yung-Chul Park (2003): “Financial liberalization and capital market
integration in East Asia”.

BIS Papers No 30 37
Eichengreen, Barry and Pipat Luengnaruemitchai (2006): “Why doesn’t Asia have bigger
bond markets?”, this volume.
Eschweiler, B (2006): “Bond market regulation and supervision in Asia”, this volume.
Fernandez, David G and Amy Li (2002): “Asian dollar bonds: in a class by themselves”,
JPMorgan Chase, Asian Credit Markets Research: Asia Markets Outlook and Strategy,
19 September, pp 3-6.
Fernandez, David G and S Klassen (2006): “Choice of currency by Asian issuers”, this
volume.
Financial Supervisory Service (2002): “FSC/FSS plans to improve the risk management of
foreign currency-denominated bank lending”, Financial Supervisory Service, Weekly
Newsletter, 14 December (www.fss.or.kr).
Hohensee, M and K Lee (2006): “A survey of hedging markets in Asia”, this volume.
Jiang, Guorong (2006): “Comments on Professor Oh and Mr Park’s paper ‘ASEAN+3
Regional Guarantee Mechanism’”, this volume.
Jiang, Guorong and Robert N McCauley (2004): “Asian Local Currency Bond Markets”, BIS
Quarterly Review, June, pp 67-79.
Kisselev, Kate and Frank Packer (2006): “Minding the gap in Asia: foreign and local-currency
ratings”, this volume.
Ma, Guonan, Corrinne Ho and Robert N McCauley (2004): “The markets for non-deliverable
forwards in Asia”, BIS Quarterly Review, June, pp 81-94.
McCauley, Robert N (2002): “International market implications of declining Treasury debt”,
Journal of Money, Credit and Banking, vol 34, no 3, part 2, August, pp 952-66.
——— (2003a): “Capital flows in East Asia since the 1997 crisis”, BIS Quarterly Review,
June, pp 45-59.
——— (2003b): “Unifying the government bond markets of East Asia”, BIS Quarterly Review,
December, pp 89-98.
——— (2004): “Striking the balance: benefits and costs of foreign investment in Asian bond
markets”, paper presented to Korea Institute of Finance Korea Day seminar, “Financial
cooperation in East Asia”, Asian Development Bank, Jeju, Korea, 14 May 2004.
McCauley, Robert N and Eli Remolona (2000): “Size and liquidity of government bond
markets”, BIS, International Banking and Financial Market Developments, November,
pp 52-8.
Murray, J and J Powell (2002): “Dollarization in Canada: the buck stops there”, Bank of
Canada Technical Report, 90.
McCauley, Robert N, San-Sau Fung and Blaise Gadanesz (2002): “Integrating the finances
of East Asia”, BIS Quarterly Review, December, pp 83-95.
Nishi, Fumiaki and Alexander Vergus (2006): “Asian bond issues in Tokyo: history, structure
and prospects”, this volume.
Oh, Gyutaeg and Jae Ha Park (2006): “Survey of securitization and credit guarantee
proposals”, this volume.
Packer, F (2003): “Mind the gap: domestic versus foreign currency sovereign ratings”, BIS
Quarterly Review, September, pp 55-63.
Park, Yung-Chul (2004): “The trans-Pacific imbalance: what can be done about it?”,
processed.

38 BIS Papers No 30
Park, Daekeun and Changyong Rhee (2006): “Building infrastructures for the Asian bond
markets: settlement and credit rating”, this volume.
Remolona, E and M Schrijvers (2003): “Reaching for yield: tips for reserve managers”, BIS
Quarterly Review, December, pp 65-73.
Schmidt, Florian (2004): Asia’s credit markets: from high yield to high grade, New York,
Wiley.
Takeuchi, Atsushi (2006): “Identifying impediments to cross-border bond investment and
issuance in Asian countries”, this volume.
Tsang, Donald (1998): Speech to Asian Debt Conference, 6 July,
www.info.gov.hk/gia/general/ 199807/06/0706078.htm.
Woods, J (2002): “Overview: Asian bonds - a class apart?”, HSBC Fixed Income Research,
August.

BIS Papers No 30 39
 Why doesn’t Asia have bigger bond markets?

Barry Eichengreen and Pipat Luengnaruemitchai 1

1. Introduction
The 1997-98 financial crisis highlighted the problem of bond market underdevelopment in
Asia. The small size and slow growth of regional bond markets, many observers noted, left
corporate borrowers excessively dependent on bank finance. Given the short tenor of bank
loans, a shock to confidence left Asian economies vulnerable to a disruptive credit crunch.
Since banks denominated many of their loans in foreign currency, exchange rate
depreciation resulted in serious balance sheet damage and thrust highly leveraged
corporations into bankruptcy.
Analysts argued further that Asia’s heavy dependence on banks increased the weight of
political and economic connections in resource allocation. Banks and the companies to which
they lent were linked by family control. Banks were used by the authorities to extend
preferential credit to firms favoured on political or developmental grounds. Financial
institutions carrying out these tasks came to be seen as too big and politically important to
fail, and the guarantees they consequently enjoyed weakened market discipline over their
lending.
The lesson drawn was that Asian countries need better diversified financial systems, and
specifically deep and liquid bond markets, to supplement their banking systems. Better
diversified financial markets would reduce financial fragility and enhance the efficiency of
capital allocation. The development of bond markets would lengthen the tenor of debt and
facilitate the placement of domestic currency bonds, limiting maturity mismatches on
corporate balance sheets. Corporations would be encouraged to disclose more information
and follow internationally recognised accounting practices, strengthening corporate
governance. Borrowers would be distanced from lenders, anonymous and decentralised
bond markets being hard to influence, and markets would be better insulated from
governments, limiting moral hazard and political interference.
The problem of Asia’s underdeveloped bond markets was known to close observers, of
course, even before the 1997-98 crisis. In some cases the absence of bond markets
complicated efforts to finance large infrastructure projects, and enterprises with a high
minimum efficient scale found it hard to meet their financial needs. 2 In principle they could
borrow from a syndicate of banks which could securitise their loans, but securitisation was
costly and difficult in the absence of a bond market. Banks therefore found it hard to diversify
risk created by their acquisition of concentrated stakes in the large enterprises that were their
leading customers. And the development of other financial instruments was limited by the
absence of bond markets on which to base forwards, futures and more exotic derivatives

1
We thank Nancy Brune and Geoffrey Garrett for help with data and Robert McCauley and Ric Deverell for
helpful comments.
2
Infrastructure finance was a particular problem with regard to the privatisation of electricity supply,
telecommunications and transportation services in Asian countries. More generally, securing adequate finance
often required diluting corporate control by issuing equity or giving banks representation on corporate boards.
Since owners saw the dilution of control as unattractive, dynamic enterprises sometimes found it difficult to
access external finance.

40 BIS Papers No 30
contracts. 3 These problems were not specific to Asia, to be sure, but they seemed to
manifest themselves there in particularly dramatic ways.
Coincident with the Asian crisis, contributions to the theoretical literature explained how
countries benefit from well diversified financial systems (see eg Boot and Thakor (1997)).
Equity finance encourages risk-taking, since holders of equity stakes share in supernormal
returns while their losses are truncated on the downside, whereas debt holders, who do not
share in exceptional profits, encourage risk aversion; a well diversified financial system
therefore facilitates risk management. Banks have a comparative advantage in providing
external finance to smaller, younger firms which typically operate in information-impacted
segments of the economy, while securities markets do the job most efficiently for large, well
established companies.
Thus, as early as 1995, before the Asian crisis, the World Bank had issued studies
recommending that Asian countries accelerate bond market development (see eg Dalla et al
(1995)). The crisis then directed additional attention to the problem. The 17 Asian
governments participating in the Asia Cooperation Dialogue at that time set up a Working
Group on Financial Cooperation to establish guidelines for the development of Asian bond
markets. APEC finance ministers agreed on a comprehensive approach to developing sound
and sustainable regional bond markets, including credit guarantees and markets in a variety
of new products (bonds denominated in a basket of Asian currencies being the most
attractive candidate). ASEAN+3 established a Study Group on Capital Market Development
and Cooperation under the leadership of Thailand, Japan, Korea and Singapore.
The most prominent of these responses was a proposal to use the international reserves of
Asian central banks to encourage the development of regional bond markets. The Asian
Bond Fund (ABF) was launched by EMEAP in June 2003, and its members committed to
investing USD 1 billion of the region’s international reserves in Asian sovereign and
quasi-sovereign dollar bonds. 4
The question is whether this use of central bank reserves will have the desired effect.
Perhaps, but some critics of this use of central bank reserves will object that other factors
- improved regulation, enhanced transparency, stronger investor protection and stable
macroeconomic policies - are more important for the development of deep and liquid bond
markets. 5 In their view these fundamentals, and not the allocation of some small fraction of
the reserves of regional central banks to local debt securities, should be the focus of efforts
to develop Asian bond markets.
This uncertainty about what initiatives are most urgently needed to promote Asian bond
markets reflects our incomplete understanding of why those markets are underdeveloped in
the first place. 6 This paper therefore considers the historical, structural, institutional and

3
Herring and Chatusripitak (2000) observe that it may still be possible, despite the absence of these markets,
to tailor forwards, futures and derivatives contracts to the needs of individual customers, but doing so can be
costly, limiting the use of such instruments.
4
Some of the proposal’s initiators had envisaged utilising 1% of the international reserves of Asian central
banks, which would have amounted to USD 12 billion, purchasing domestic currency as well as dollar bonds,
and investing in corporate as well as government securities. At the time of writing, EMEAP is discussing a
second Asian bond fund that might be larger in size and would invest in high-grade domestic currency issues.
5
See Fernandez and Klassen (2003).
6
While earlier studies touched on the issue, none of them, so far as we know, has analysed it systematically.
Burger and Warnock (2003, 2004) are the studies closest in spirit to our own, but they consider only long-term
bonds (not also the short-term bonds considered here) and a subset of the potential determinants of local
market issuance. Claessens et al (2003) consider both domestic and foreign currency denominated issues,
but they limit their analysis to government bonds, putting aside the determinants of corporate bond market
growth. Eichengreen et al (2002) consider corporate as well as government issues, but they too are

BIS Papers No 30 41
macroeconomic determinants of bond market development in a cross section of developing
and developed economies. Section 2 presents an overview of bond markets in emerging
Asia with comparisons to other regions. Sections 3 and 4 enumerate the hypotheses that
have been described to explain bond market underdevelopment. Sections 5 and 6 present
our regression analysis. Section 7 draws out the implications for the development of Asian
bond markets.
The results confirm that small size and fragmentation are part of the explanation for the
underdevelopment of Asia’s bond markets, but only part. In addition, corruption, poor
regulatory quality and failure to compel firms to follow internationally recognised accounting
standards have slowed the development of private debt markets. Countries with competitive,
well capitalised banking systems also have larger bond markets (both public and private),
suggesting the existence of complementarities between banking and bond market
development.
This suggests that, in order to promote the development of bond markets, governments need
to encourage adherence to internationally recognised accounting standards and enhance the
reliability of regulation and contract enforcement. They should distance themselves from the
lending operations of banks in order to accentuate the complementarities between banking
and bond market development. Through this combination of policies, our results suggest,
Asian countries could acquire bond markets as liquid and well capitalised as those of other
regions.

2. Overview
Table 1 describes the stock of external finance in various economies at the end of 2001. For
emerging Asia, bond market capitalisation (the sum of corporate, financial institution and
public sector issues) was 45% of GDP; this was actually higher than the average for all
emerging markets, at 39%, if lower than that for developed economies, at 139%. Note that
we include here only domestic currency bonds issued by residents and targeted to local
investors. 7 At this level of aggregation, Asia is not behind Latin America or emerging central
Europe in terms of bond market development, although it is considerably behind the
developed economies, and in particular the United States. 8
These regional aggregates disguise considerable variation across countries. Corporate bond
market capitalisation is 50% of GDP in Malaysia and 28% in Korea but only 5% in Thailand. 9
Financial institutions are important for bond issuance in Hong Kong SAR, Korea and
Singapore, but less so in China and Malaysia. They figure hardly at all in external finance in
Thailand.

concerned with currency denomination, not market capitalisation. Domowitz et al (2000) and Hale (2003)
analyse the choice between bank and bond finance, but they analyse international bonds and bank loan
syndications, not their domestic counterparts.
7
For more discussion of our measure of bond market capitalisation, see Section 4 below.
8
The picture is not much different when we distinguish between bond issues by non-financial corporations,
financial institutions and governments. Public issues are slightly less important in emerging Asia than in other
emerging markets, reflecting the traditionally strong fiscal position of Asian governments, while issues by
corporations and financial institutions are slightly more important in emerging Asia than elsewhere.
9
These aggregates need to be interpreted cautiously; in some cases they may tell us less about the scale and
health of the bond market than might be naively supposed. Thus, in the Korean case, a considerable fraction
of bond market capitalisation is in the form of asset-backed securities in which the government and its
agencies have absorbed the risky junior tranche that accounts for the majority of the outstanding stock.

42 BIS Papers No 30
Table 2 compares the relative importance of bonds, bank loans and equity markets in
domestic external finance outstanding at the end of 2001. 10 In terms of the composition of
external finance, Asia relies less on bond markets than other emerging market regions; the
share of bonds is a bit more than half that of Latin America and emerging central Europe.
Again, these generalisations disguise considerable variation across countries. For well
known historical reasons, the banking sector is particularly important for external finance in
China, Korea and Thailand. The stock market is important only in Hong Kong, Malaysia and
Singapore, where the authorities have aggressively promoted it. The bond market is the least
important of the three sources of finance in virtually every country (the exception being
Thailand, where it is approximately the same size as the stock market). Bonds are least
important in total external finance in Hong Kong and most important in Malaysia and Korea.
The preceding data are for stocks; flows may offer a clearer picture of recent trends.
According to Table 3, new domestic bank loans were 10% of GDP in emerging Asia in 2001
but only 4% of GDP for the emerging markets as a whole. Domestic bond flotations, in
contrast, amounted to 12% of GDP in 2001 for emerging markets as a whole but only 8% in
Asia.
In sum, this overview confirms that emerging Asia relies less on bonds and more on banks
than other emerging markets, and very much less on bonds and very much more on banks
than developed economies. Recent data suggest that these distinctive characteristics of
Asian financial systems are not growing noticeably less pronounced; in some cases the
opposite may be true.

3. Hypotheses
Five broad hypotheses have been advanced to explain the underdevelopment of Asian bond
markets. One is the region’s history. Banks have dominated Asian financial markets for many
years. Once upon a time there may have been good reasons for their dominance.
Imperfections in the information and contracting environment gave a strong comparative
advantage to bank intermediation, while governments found banks to be convenient vehicles
for advancing their industrial policies. But although these circumstances have now changed,
banks retain their “first mover” advantage. Markets, institutions and social conventions have
adapted to the dominance of bank intermediation. Examples of that adaptation include the
importance of family connections and state involvement in financial relationships. As a result,
bonds may face an uphill battle when seeking to acquire market share.
A second hypothesis emphasises structural characteristics of the region’s economies. Small
countries presumably find it more difficult to develop bond markets insofar as liquid securities
markets have a certain minimum efficient scale. Endowment theories suggest that the
geographical environment shapes the long-standing institutions that influence financial
development. The strength of bondholder protections may depend on a country’s legal
tradition (see La Porta (1998)). Not all of these structural characteristics are impervious to
change, but even the most malleable of them may be difficult to change quickly.
A third hypothesis focuses on the developmental stage of the region’s economies. Compared
to the economies of western Europe and North America, most Asian countries have

10
Strictly speaking, total external finance (that is, financing from outside the corporation, excluding retained
earnings and depreciation) would also include credit provided by foreign sources, for which we lack
information. To avoid double-counting, we exclude bonds issued by financial institutions from this comparison.
Including them makes little difference for the comparisons with which we are concerned in this paper. The
main effect is to further increase the value of bond market capitalisation in the advanced economies.

BIS Papers No 30 43
undergone the transition to modern economic growth relatively recently. Some are still poor.
At the core of this situation is the underdevelopment of market-supporting institutions,
including the institutions needed to support financial markets. From this perspective, Asian
financial markets are underdeveloped because of the unreliability of contract enforcement
and uncertainty of investor rights that are characteristic of less developed economies. These
are problems that economies presumably grow out of, though how quickly they do so
depends on country-specific circumstances.
A fourth hypothesis focuses on the structure and management of the financial system. This
explanation considers, inter alia, the intensity of competition among financial institutions, the
quality of prudential supervision and regulation, the existence of a well defined yield curve,
the absence of institutional investors and rating agencies, and the adequacy of trading,
settlement and clearing systems. 11
Fifth and finally are macroeconomic policies. The currency risk created by flexible exchange
rates may limit the market for domestic currency denominated securities. Domestic interest
rate volatility may make it unattractive to hold long-term debt securities. Such instability may
be a serious impediment to bond market development. Finally, controls on capital flows, such
as those limiting the ability of foreigners to purchase domestic capital and money market
securities or to repatriate their interest earnings and principal, may discourage foreign
participation in domestic markets and rob those markets of liquidity.

4. Empirical implications
We now turn from broad hypotheses to empirical implications, illustrating our points with
information for 41 economies. The data are for all economies for which the BIS reports
estimates of domestic bond market capitalisation. 12 Hence, the sample is not limited to Asia.
But neither are questions about the development of bond markets limited to Asia. In
analysing the determinants of bond market development we seek to take advantage of the
information content of a wide cross section of economies. The variables that we use to
operationalise our five hypotheses are shown in Table 4. 13
Economic size. Small countries may lack the minimum efficient scale needed for deep and
liquid bond markets. 14 The amount of money that can be raised by issuing on the local

11
Independent agencies that rate corporate issuers provide information that should help to attract a large base
of active investors into the bond market. While some Asian countries have independent rating agencies
(Malaysia, for example, has two), others do not. A large population of institutional investors is important for
creating a demand for domestic bonds (Schinasi and Smith (1998)). Conversely, heavy regulation of mutual
funds may prevent fund managers from actively participating in corporate bond markets. Finally, it has been
argued that the absence of well developed clearing, settlement and trading systems have rendered some
Asian bond markets illiquid and unattractive (Trairatvorakul (2001)).
12
The BIS compiles these data from national sources, and attempts to eliminate international debt securities
from its estimates of domestic bond market capitalisation. Capitalisation is only one measure of bond market
development, of course; turnover is another obviously relevant dimension. But only capitalisation is available
for a broad cross section of countries. Previous studies (eg McCauley and Remolona (2000)) suggest that
capitalisation and turnover on domestic bond markets are strongly if imperfectly correlated.
13
Information on data sources can be found in the appendix.
14
Eichengreen et al (2002) provide evidence that small size is similarly the most robust determinant of the
inability of emerging markets to borrow abroad in their own currencies (“original sin”). Here the obvious
explanation is that countries whose debt issuance is small have trouble getting international investors to add
securities denominated in “exotic” currencies to their investment portfolios. This will be the case when the
increase in management costs is constant but the diversification benefits decline with each additional
currency. This is probably an appropriate point at which to discuss how domestic bond market development

44 BIS Papers No 30
market may be too small to attract multinational corporations and other potential foreign
issuers. The market may be too small to justify inclusion in the global bond market indices
constructed by the leading investment banks, in which case there will be no demand to hold
local securities in order to track the index. Markets in small issues may be characterised by
price volatility as buyers and sellers enter and exit. Similarly, it may be difficult to put on and
take off positions without being noticed. There being fixed costs of learning about the
performance characteristics of an issue, investing in small issues may not be attractive for
portfolio managers, who will consequently demand a yield premium in order to do so. 15 And if
adverse selection is present, it may be that no premium will create a demand. A bivariate
scatter plot of bond market development (measured as domestic bond market capitalisation
as a share of GDP, averaged over the 1990s) and country size (GDP at purchasing power
parity (PPP), also averaged over the 1990s) shows a weakly positive relationship between
the two variables (Figure 1). 16
Natural openness. Entrenched interests will seek to prevent their advantaged position from
being undermined by market competition. Banks, for example, will attempt to prevent their
dominant market share from being eroded by competition from securities markets. But
entrenched interests will be less able to insist on policies that suppress competing sources of
supply when the economy is exposed to international competition. This is Rajan and
Zingales’ (2001) explanation for why more open economies do less to suppress securities
markets. That said, Figure 2 does not suggest a particularly strong relationship between
openness, measured as the ratio of exports to GDP, and bond market development.
Legal system. Legal traditions differ in the priority they attach to protecting minority
investors. La Porta et al (1998) predict that common law systems in the British tradition,
which offer stronger investor protection than systems in the French civil law tradition, should
promote the development of financial markets. But the same legal traditions may not affect
all aspects of financial development equally. Where investor rights are weak, savers may
prefer investing through banks rather than bonds since politically well connected banks are
better able to enforce their claims (Sharma (2000)). Systems with weak investor rights may
also encourage creditors to demand assets with seniority (bonds rather than stocks). 17
Geographical/disease endowments. Endowment theories suggest that environmental
factors shape long-standing institutions influencing financial development. Authors like Beck
et al (2002) argue that countries with less favourable geographical and disease environments
should have less developed financial markets. They present evidence that endowments
(measured by settler mortality or distance from the equator) are correlated with financial

relates to original sin. In principle, domestic bond market development is a route to solving this problem. As
domestic markets gain scale and liquidity, foreign participation will be easier to attract, both because those
local currency markets will become easier to enter and exit (transaction costs will decline) and because they
will constitute a greater share of the global portfolio (diversification benefits may increase). In practice,
however, this route to “redemption” appears to work only very slowly. Data in Burger and Warnock (2004)
indicate that as of 2001 US residents held only USD 2.5 billion bonds issued by emerging markets, whereas
emerging markets had more than USD 1.6 trillion of local currency bonds outstanding (and more than
USD 2.2 trillion of total bonds outstanding). In other words, while foreign participation in local bond markets
has attracted much comment, as a quantitative phenomenon it remains inconsequential.
15
This phenomenon is familiar in the context of foreign bond issues; see Eichengreen and Mody (2000).
16
All variables are similarly measured as averages for the 1990s in the scatter plots that follow, except where
expressly noted otherwise.
17
La Porta et al (1998), when reporting a positive correlation between debt/GNP and common law legal tradition,
define debt as the sum of corporate bonds and bank loans. Beck et al (2002) consider financial intermediary
credits to the private sector divided by GDP. Thus, neither set of authors addresses the impact on bond
markets that is our concern here.

BIS Papers No 30 45
intermediary and stock market development. Figure 3 suggests the existence of a positive
relationship between distance from the equator and bond market development.
Riskiness of the investment environment. Bonds are a way for investors to limit risk. It
follows that entities issuing bonds are generally of higher credit quality than those issuing
equity claims (Harwood (2000)). In some countries, however, there may be a dearth of high-
quality issuers with proven business models and records of financial probity. Consistent with
this idea, Figure 4 suggests that bond market capitalisation rises as investment risk declines.
Law and order. Countries with more reliable law enforcement are more attractive to
investors. Figure 5 confirms the existence of a positive relationship between the size of bond
markets and the International Country Risk Guide (ICRG) measure of law and order. To the
extent that corruption undermines law enforcement, corruption and bond market
development should be negatively correlated. Figure 6 is consistent with this hypothesis
(since, on the ICRG scale utilised here, a higher score indicates a lower level of corruption).
Weak corporate governance and transparency. If corporate governance is weak,
managers will be able to enrich themselves at the expense of holders of debt and equity
claims. If banks enjoying long-term relationships with borrowers have a comparative
advantage in detecting and correcting insider abuses, savers may prefer to invest via banks
rather than securities markets. Lenders will also prefer banks to bond markets where
transparency is low, since banks have a comparative advantage in information-impacted
markets (Diamond (1991), Hale (2003)). In support of this hypothesis, Figure 7 shows that
the quality of accounting standards is positively associated with bond market development.
Developmental stage of the economy. There are a number of reasons why economic
development and bond market development go hand in hand. Less developed countries
have volatile investment environments and heavy government involvement in commercial
activity. Often they have weak creditor rights, inadequate transparency and poor corporate
governance. GDP per capita can be thought of as capturing these aspects of
underdevelopment insofar as they are not already absorbed by our other explanatory
variables. Figure 8 is consistent with the notion that economic development and bond market
development are positively associated. 18
Size of the banking system. Banks and bond markets compete in providing external
finance; in some circumstances, well developed banking systems may succeed in depriving
bonds of market share. At the same time, banks serve as dealers and market-makers, whose
presence is needed for the development of a liquid and well functioning bond market. 19
Figure 9 suggests that the complementarities dominate - that, on balance, banking systems
and bond markets develop together.
Banking concentration. Benston (1994), Schinasi and Smith (1998), Smith (1998) and
Rajan and Zingales (2003) suggest that banks with market power may attempt to stifle the
development of securities markets by setting loan and deposit rates strategically or use
moral suasion to discourage public placements by firms with which they have relationships.
That said, Figure 10, which juxtaposes banking sector concentration against bond market
development, does not show a particularly strong relationship between the two variables.

18
It suggests that bond markets are less developed than levels of per capita GDP and a broader sample of
national experiences would predict in, inter alia, Hong Kong SAR, Singapore and Japan, while they are rather
better developed in Malaysia.
19
See Harwood (2000) and Hawkins (2002). In many countries, regulators require that bond business be done
in a separately capitalised subsidiary, although such firewalls may be more apparent than real. At the same
time, dealers need a diversified and active investor base with which to buy and sell; they cannot simply trade
among themselves. Without such a base, dealing will not be profitable. One suspects, therefore, that dealers
are not so much a precondition for bond market development as a corollary.

46 BIS Papers No 30
Absence of public sector funding needs. The development of a government securities
market “helps promote a class of dynamic, profitable fixed-income dealers” (Harwood
(2000)). In addition, an active and liquid corporate bond market requires a benchmark yield
curve on the basis of which risky credits can be conveniently priced. 20 That yield curve is
typically constructed from a suite of outstanding treasury securities, requiring governments to
issue a range of maturities on a regular schedule. If a government has modest funding
requirements, there may be little need to develop an active and liquid bond market and little
regular issuance to maintain a well defined yield curve. 21 Figure 11 is consistent with the
existence of a positive relationship between private and public sector bond market
capitalisation. 22
Poor regulatory enforcement. Investors will be reluctant to take positions in markets
characterised by opportunistic participants and delivery risk, problems that regulation is
designed to mitigate. Elements of an adequate regulatory framework include disclosure
standards, penalties for accountants and auditors providing false information, and sanctions
for insider trading and market manipulation. Equally important is the clear and consistent
implementation of regulations. Figure 12 shows that bureaucratic quality is positively
correlated with bond market development. 23
Interest rate variability. Where interest rates are variable, investors will have little appetite
for long-term fixed rate notes, since there is significant risk that the purchasing power of such
assets will be eroded. Investors’ limited appetite for long-term bonds may thus limit the
demand for securitised debt. In addition, high levels of interest rate volatility may be an
indication of inadequate market liquidity, insofar as returns are affected by the entrance or
exit of a few buyers and sellers from the market. Figure 13 illustrates the negative
relationship between nominal interest rate volatility and bond market development.
Level of interest rates. Since few firms can service debts when interest rates are high, high
rates tend to have a depressing impact on issuance. It follows, as shown in Figure 14, that
countries with high interest rates show signs of having poorly capitalised bond markets.
Exchange rate regime. It is argued (by eg Goldstein (1998)) that pegged exchange rates
encourage foreign investors to underestimate the risks of lending to banks and corporations,

20
Schinasi and Smith (1998) note other advantages of the existence of a benchmark issue: since they are liquid,
benchmark assets are widely used in repo markets and are typically usable as collateral for a wide range of
other financial contracts.
21
It is in principle possible for governments without ongoing funding needs to circumvent this constraint by
overfunding the fiscal deficit (issuing more debt than strictly necessary, rolling it over as it matures, and
depositing the resulting cash surplus with the central bank, which allows the central bank to retire its
sterilisation bonds, thereby unifying the public sector bond market); see McCauley (2003). Thus, despite not
running current budget deficits, the Hong Kong Monetary Authority has been able to create a liquid market in
Exchange Fund Paper (EFP), with a 10-year yield curve, even in the absence of current government budget
deficits. EFP was introduced in 1990 with the issuance of 91-day bills, followed by 182- and 364-day bills in
1990 and 1991, two- and three-year notes in 1993, five-year notes in 1994, seven-year notes in 1995 and
finally 10-year notes in 1996. The outstanding stock of EFP is more than HKD 100 billion, or more than 8% of
GDP, and more than 20% of total debt instruments. It is issued through competitive tender bids, was listed on
the stock exchange in 1999 to enhance liquidity, and can be used as collateral for trading stock options and
futures. Taiwan, China financed a significant part of its National Development Plan starting in 1991 through
bond issuance, using a US Treasury bond-type auction system (Lynch (2001)). Similarly, despite limited public
funding needs, the government of Singapore decided in 1998 to increase the issuance of government
securities, especially longer-term bonds of benchmark size, and in May 2000 it introduced a repo facility for
primary dealers.
22
Note that the variable on the vertical axis, private market capitalisation, is different from that in the other
figures.
23
Using simple tabulations, Domowitz et al (2000) similarly find that the share of domestic finance accounted for
by bonds in emerging markets rises with the quality of accounting standards.

BIS Papers No 30 47
and that the resulting foreign competition may slow the development of domestic
intermediation. From this point of view, greater exchange rate flexibility should encourage the
development of domestic bond markets (as argued by, inter alia, World Bank (2003)). Of
course, to the extent that foreign participation is valuable for the growth and development of
domestic markets, discouraging the participation of international investors by introducing
additional risk into the market may not produce the desired result. 24 In fact, countries with
fixed exchange rate regimes do not appear to have bigger bond markets (Figure 15). Figure
16, however, is consistent with the view that stable exchange rates are conducive to bond
market development.

5. Multivariate analysis
We now test the importance of these factors using a multivariate regression analysis of
annual data from 1990 to 2001. The dependent variable, as in the scatter plots, is bond
market capitalisation as a share of GDP. Recall that this measure includes only domestic
currency bonds issued by residents and targeted to local investors. 25
All equations are estimated using panel generalised least squares (GLS) with corrections for
heteroskedasticity and panel-specific autocorrelation. We start in Table 5 with preliminary
regressions exploring the importance of, alternately, historical, structural, financial,
developmental and macroeconomic factors. Definitive hypothesis tests, of course, require
considering all five categories of explanation simultaneously. We do so in the final column of
the table.
The first three columns show the effects of structural characteristics of countries. Consistent
with earlier arguments, country size and openness are positively related to bond market
development. Distance from the equator, a proxy for endowment theories, similarly enters
with its expected positive sign. 26 But where previous studies have shown that English
common-law legal tradition favours equity market development and bank intermediation, the
same does not appear to be true of bond markets. It may be that stronger investor rights
encourage investors to attach less importance to seniority and to substitute equity for debt
securities. 27 Overall, these results lend support to structural explanations for bond market
development.

24
There is also the possibility that the correlation reflects causality running in the other direction, from the
existence of a large domestic financial market to the willingness of countries to countenance additional
exchange rate variability (Calvo and Reinhart (2002)).
25
Thus, a limitation of our analysis is that we do not have information on foreign currency denominated issues or
issues by non-residents denominated in local currency targeted to resident investors. We also do not know
what share of the domestic currency issues we include are interest rate or exchange rate indexed. Note that
our measure excludes issues denominated in foreign currency, issues by non-residents, and issues by
residents targeted to non-residents, all of which are counted as international securities, as they presumably
should be.
26
It is not possible to use settler mortality rates in an analysis of Asian bond markets, since relatively few Asian
countries were colonised by the European powers, and settler mortality estimates (and logic) are based on
data for and the experience of one-time colonies.
27
However, the coefficient on this variable is significantly different from zero in only one of the two equations in
which it is included. Adding dummy variables for other legal origins does not alter these findings. For example,
when we add French legal origin, the new variable enters positively (and significantly), while English legal
origin continues to enter negatively and significantly. Since the French civil law tradition is associated with
relatively weak investor rights, the opposite signs on the two variables are consistent with the explanation in
the text.

48 BIS Papers No 30
These regressions also include a dummy variable for Asia, which we interpret as reflecting
aspects of the region’s history not captured by other variables. The negative coefficient on
this variable supports historical explanations for the undercapitalisation of the region’s bond
markets. 28
The specification in column 4 considers proxies for the developmental stage of the economy:
the safety of the investment environment (predictability of contract enforcement, danger of
expropriation), an index of the reliability of law enforcement, and per capita GDP as a
summary measure of development. While per capita GDP has its expected positive
coefficient, investment risk and law and order (which are scaled so that higher values
indicate a more stable investment environment) enter with negative signs. We will return to
these variables below.
Columns 5 and 6 consider governance and regulation of the corporate and financial sectors.
Column 5 shows that countries which have better rankings on the ICRG’s measure of
corruption and which adhere to international accounting standards (which is likely to enhance
the effectiveness of corporate governance) have larger bond markets. 29 Column 6 shows that
countries ranking higher in terms of bureaucratic quality have larger bond markets, which we
interpret in terms of the efficiency and reliability of regulation. Similarly, countries with better
developed banking sectors have better developed bond markets - bank and bond market
intermediation appear to be complements rather than substitutes. On the other hand,
countries with more concentrated banking systems appear to have smaller bond markets,
consistent with arguments suggesting that banks with market power may use it to discourage
bond flotations. Again, we will return to these findings below.
Column 7 considers macroeconomic factors. While the volatility of interest rates is not
significant, their level, as measured by the interbank rate minus Libor, suggests that higher
interest rates are associated with smaller bond markets. 30 The coefficient on the volatility of
changes in exchange rates is marginally significant, although it is, surprisingly, positive.
Finally, the capital controls dummy (where a value of one indicates an open capital account)
suggests that controls slow bond market development. 31 As we show below, this last result is
the one that turns out to be robust.
Column 8 considers the entire range of hypotheses. 32 It suggests that no single class of
factors is wholly responsible for the underdevelopment of Asian bond markets; rather, the
present state of affairs reflects a confluence of influences. Structure and inheritance matter:
the size of the economy, its openness, its location, and the origin of its legal system all

28
Note that the coefficient predictably becomes smaller in absolute value the more other independent variables
are included in the specification.
29
This is consistent with results in Burger and Warnock (2004) suggesting that countries with stronger
institutions have larger domestic bond markets.
30
Domowitz et al (2000) similarly provide evidence that countries with higher rates of inflation issue less
domestic debt and more equity.
31
A variety of alternative measures of capital controls point in the same direction. Thus, in addition to the binary
(“IMF-style”) open or closed measure, we experimented with Brune et al’s (2003) measure, which ranges from
zero to nine depending on how many of the nine categories of capital account restrictions a country has in
place. We looked separately at capital account openness for inflows and outflows. We also looked separately
at controls on inflows and outflows pertaining to capital and money market securities. In virtually all cases we
obtained the same positive and statistically significant coefficient on controls when using the specification in
column 7.
32
Adding all of the explanatory variables substantially reduces the number of observations (from 475 in the full
sample to 284 in column 8). However, the observations from countries in Asia remain well represented. While
accounting for 22% of the observations in the full sample, they account for 25% of the observations in
column 8.

BIS Papers No 30 49
influence bond market capitalisation. Factors like these may be difficult to change, although
some of them, such as the handicap of small size, may be overcome through initiatives like
the Asian Bond Fund. In addition, adherence to internationally recognised accounting
standards and the size and concentration of the banking sector are important for bond
market capitalisation. These are policy variables; our results thus suggest that countries can
accelerate the development of their bond markets by improving the quality and reliability of
regulation, requiring corporations to adhere to internationally recognised accounting
standards, and encouraging competition in financial intermediation. In addition, there is a role
for macroeconomic policy: both the level of interest rates and the presence or absence of
capital controls matter in the consolidated specification.
At first blush, a number of the results are anomalous or at least counter-intuitive. Thus, we
appear to find that interest rate volatility is good for bond market development. At the same
time, there is little evidence of a relationship between exchange rate volatility and bond
market development. We will have more to say about these counter-intuitive results below.
Note also that when we add direct measures of institutions - such as bureaucratic quality,
corruption, law and order, and the investment profile - the effect of per capita GDP washes
out. This is not inconsistent with explanations for bond market growth emphasising the
developmental stage of the economy, but it suggests that the effects of economic
development and underdevelopment operate through the aforementioned institutional
channels.
We looked further at the robustness of the positive association of bank and bond market
development, which runs contrary to some popular arguments, and which is likely to be
controversial. We also regressed non-public bond market capitalisation on bank credit to the
private sector as a share of GDP, adding the entire vector of controls. 33 Excluding public
sector bonds and considering only bank credit to the private sector avoids the possibility that
the positive association between the two variables is simply picking up liquidity requirements
and other policies forcing the banking sector to hold government bonds - and the greater
ability of the government to compel such behaviour in countries where the banking system is
relatively large. In this alternative specification the coefficient on bank credit continues to
enter with a positive coefficient and differs from zero at the 99% confidence level.
Finally, note that the dummy variable for Asia continues to matter statistically and
economically. Its effect is large: the coefficient of –17 suggests that Asian bond markets are
17% smaller as a share of GDP than their counterparts in countries with comparable
characteristics in other parts of the world. One interpretation of this is that the development of
bond markets continues to be held back by Asia’s history and current circumstances in ways
that are not fully captured by the other explanatory variables. We will revisit this finding
below.
An eclectic set of policy implications would seem to flow from these findings. The Asian Bond
Fund and the removal of capital account restrictions may help domestic bond market
development by relaxing the constraint of small market size, although such policies may be a
mixed blessing insofar as capital account liberalisation prior to domestic market development
poses risks as well as promising rewards. But market size is far from the entire problem. In
addition, governments seeking to promote domestic bond markets must require adherence to
international accounting standards by security issuing firms and encourage growth and
competition in banking so as to maximise the complementarities between banking system
and bond market development. They should adopt stable macroeconomic policies to make it
attractive to hold domestic currency denominated debt instruments.

33
In further regressions not reported here.

50 BIS Papers No 30
Even if they take these steps, the results of this section suggest, Asian governments still
should not expect to succeed in developing bond markets with the depth and liquidity
characteristic of continental Europe and the English-speaking economies, due to the extent
to which the region’s markets, institutions and social conventions have adapted to the
dominance of bank intermediation. This is undoubtedly the most controversial conclusion
seeming to emerge from the present section. It is important, therefore, to subject it to further
analysis.

6. Adding fiscal policy

We subjected our results to a variety of robustness checks. We dropped influential
observations. The results were robust to these changes. We limited the sample to the period
before the Asian crisis to test for structural breaks. The results were again very similar. 34
The one sensitivity test that did make an important difference was adding fiscal policy. 35 We
measured this in three ways: as the public debt/GDP ratio, as the past year’s budget balance
as a percentage of GDP, and as a three-year moving average of past budget balances. The
last of these alternatives is probably preferable, since the budget balance in a single year will
tend to be dominated by transient factors, while public indebtedness is likely to have a
spuriously strong coefficient given that the public debt is itself a major component of bond
market capitalisation.
The results in the first three columns of Table 6 confirm that fiscal policy is important for
overall bond market development. 36 Stronger fiscal balances are negatively associated with
bond market capitalisation. The coefficient in the third column reinforces our trepidation
about using the public debt ratio in that the coefficient is almost exactly unity. We do not
consider this measure further in what follows.
Adding past budget balances has a number of other effects. We now obtain a significantly
negative coefficient on exchange rate volatility. Higher interest rates continue to be obstacles
to more rapid bond market development. An earlier anomaly, that greater interest rate
volatility is associated with faster bond market development, is now evident in only one of the
three specifications. 37
Adding past budget deficits also eliminates previously significant coefficients on the
investment profile, accounting standards and bureaucratic quality, while strengthening (at
least in some cases) the effects of corruption and rule of law. 38 In the case of the investment
profile, this is reassuring, since the previous result anomalously suggested that safer
investment environments are associated with less well developed bond markets. 39 The now

34
The main differences were that the corruption and law and order variables became significant (lower levels of
corruption and more reliable law enforcement were associated with larger bond markets), while distance from
the equator and domestic credit provided by the banking sector lost their significance.
35
Asian governments have tended to run surpluses, with a few prominent exceptions, and this otherwise
admirable behaviour may have stymied the development of bond markets (for reasons explained above).
36
The observations here are only about half the number in the full sample. However, Asian economies are still
well represented: they account for 21% of the reduced sample.
37
As we will see shortly, disaggregating public and private debt makes this anomaly disappear entirely.
38
To be precise, accounting standards are significant in column 1, where the three-year average of the fiscal
balance is included, but not in column 2, where an alternative measure of fiscal policy is used.
39
Table 4 shows that there is a positive correlation between the strength of fiscal policy and the quality/safety of
the investment environment, which may explain this result.

BIS Papers No 30 51
greater importance of corruption and rule of law is also reassuring. However, the loss of
significance of bureaucratic quality and accounting standards is less reassuring; at face
value this suggests that financial transparency and the quality and reliability of regulation are
not so important after all. At a minimum, it suggests that it is hard to distinguish the effects of
transparency, regulation and fiscal policy.
But when one distinguishes public debt from private debt (debt issued by both non-financial
corporations and financial institutions), one finds that budget deficits are a significant
determinant of public debt market capitalisation (columns 6 and 7) but not private debt
market capitalisation (columns 4 and 5). In other words, while governments that run deficits
have significantly more public debt (as a matter of definition), public sector deficits do not
appear to encourage private debt issuance. That there is no net effect is unsurprising given
arguments that a history of strong fiscal policies is both good and bad for private debt
markets. (It creates a more stable investment environment, but complicates the creation of a
well defined yield curve and slows the development of a class of dynamic fixed income
dealers.) 40
Note, further, that in the regressions for private debt the coefficients on accounting standards
regain their significance even through fiscal policy is still included. 41 In contrast, they are
insignificant in the equations for public debt. The same is true for corruption and bureaucratic
quality. Thus, while institutional characteristics and regulatory practices like accounting
standards, corruption and bureaucratic quality matter for private debt market capitalisation,
they evidently matter less for public debt market capitalisation. 42
Another difference introduced by disaggregating public and private debt has to do with the
relationship between banking systems and bond markets. Earlier, when considering total
debt, we found evidence that both the size and concentration of the banking sector matter
(positively and negatively respectively). 43 Disaggregating reveals that the size of the banking
system matters mainly for the capitalisation of private debt markets - in other words, there is
evidence of complementarities between the development of banking and the development of
private debt markets. In contrast, banking system concentration is negatively associated with
public debt. Readers familiar with Asia’s economic and financial history will conjecture that in
countries with concentrated banking systems the government was able to use the banks as
agents for its industrial policy, channelling private savings towards favoured industries and
activities, whereas in countries with atomistic banking systems less subject to manipulation,
direct government expenditures were required for these purposes.
We also find, upon disaggregating public and private debt, that the earlier evidence of a
positive relationship between interest rate volatility and bond market development
disappears. In contrast, the level of interest rates and the stability of the exchange rate
continue to matter, as before, for both private and public debt.
Finally, analysing public and private debt separately reveals that the significance of capital
controls derives from their impact on the volume of public debt. Evidently, governments that

40
To put the point the other way, chronic deficits create an ample supply of sovereign securities from which to
construct a benchmark yield curve but at the same time crowd out private debt issues (McCauley and
Remolona (2000)). Our results suggest that these two effects roughly cancel each other out.
41
Corruption and bureaucratic quality are not significant in the regression for public debt, except in one case
where the coefficient on bureaucratic quality is marginally significant at the 90% level, and there it counter-
intuitively enters with a negative sign.
42
This explains the unstable pattern of coefficients when total debt is considered and measures of fiscal policy
are added or dropped.
43
However, the evidence that the size of the banking system is important was much weaker when we included
measures of fiscal policy.

52 BIS Papers No 30
open the capital account are better able to fund themselves, whether by selling debt to
foreigners or owing to credibility effects. Of course, we know from the Asian crisis that to fund
government deficits in this way before putting the other prerequisites for capital account
liberalisation in place can be a risky business. And the insignificance of both capital account
openness and past deficits for private bond market capitalisation suggests that any benefits
for corporate bond market development are at best indirect.
Note that adding a measure of past fiscal policies eliminates the previously negative
coefficient on the dummy variable for Asia in the equations for total debt. This is true whether
fiscal policy is measured as the past year’s deficit or as a moving average of past deficits.
Moreover, the coefficient on Asia is now positive, not negative, including in column 3, where
past fiscal policy is measured by the public debt and the coefficient is significantly different
from zero (columns 1-3). Once we control for the traditionally strong fiscal stance of Asian
countries, in other words, there is no longer support for the notion that their bond markets are
smaller than can be explained by their economic characteristics and policies. 44

7. Conclusions
Asia’s underdeveloped bond markets and dependence on bank finance have attracted
concern since before the crisis of 1997-98. The result has been a host of official responses,
ranging from reports by the multilateral financial institutions on the importance of reliable
contract enforcement, strengthened prudential regulation and improved market infrastructure
to the Asian Bond Fund funded by EMEAP central banks. But it remains uncertain whether
these initiatives will succeed in surmounting the fundamental obstacles to bond market
development in the region, since there has been little systematic analysis of the nature of
those obstacles. This is a gap that the present paper seeks to fill.
We find that the slow development of local bond markets is a phenomenon with multiple
dimensions. To some extent the problem is one of minimum efficient scale: larger countries
have better capitalised bond markets when capitalisation is measured relative to GDP. 45 But
market size is not the entire problem. In addition, the failure of countries to adhere to
internationally recognised accounting standards has slowed the development of private debt
markets. Corruption and low bureaucratic quality, which are signs of unreliable securities
market regulation, work in the same direction. Countries with competitive, well capitalised
banking systems, on the other hand, have larger bond markets.
Macroeconomic policy appears to have played both a supporting and impeding role. On the
one hand, Asia’s strong fiscal balances, while admirable on other grounds, have not been
conducive to the growth of government bond markets. Fortunately, there is little evidence
that the small size of public debt markets is an insurmountable obstacle to corporate bond
market development. On the other hand, the stability of exchange rates in the region
appears, if anything, to have encouraged bond market development.
Over time, markets, institutions and social conventions have adapted to the status quo,
which in the case of Asia is the dominance of bank finance. Some may worry that, as a result

44
Indeed, when we limit our attention to private debt (columns 4 and 5), both estimates of the Asia dummy are
significantly greater than zero. For public debt, the sign of the coefficient on the Asia dummy is sensitive to
how fiscal policy is measured, and it is never significant at the 95% confidence level.
45
In addition to being supported by our empirical results, this fact is evident in Europe’s experience, where the
advent of the euro has relaxed the constraint of market size at the national level and greatly enhanced the
liquidity of the bond markets, the corporate bond market in particular.

BIS Papers No 30 53
of this inheritance, Asian countries will not be able to develop bond markets as efficient and
well capitalised as those of the advanced industrial countries. In this respect our results are
reassuring: they suggest that the region’s structural characteristics and macroeconomic and
financial policies account fully for differences in bond market development between Asia and
the rest of the world. Once one controls for these characteristics and policies, in other words,
there is no residual “Asia effect”.
One obstacle that the region must overcome in order to accelerate this process is the legacy
of capital controls. The evidence is strong that capital controls discourage foreign
participation in domestic bond markets and that they discourage bond market development
more generally. But we also know, not least from the Asian crisis, that capital account
liberalisation is only prudent when domestic financial markets are already deep, liquid and
robust. Here, obviously, is a dilemma. Capital account liberalisation makes sense only when
domestic market development is sufficiently advanced, but developing domestic financial
markets is harder when the capital account remains fully or partly closed. There is no easy
way of finessing this problem. 46 The only solution is to work harder at strengthening market
regulation, market infrastructure and the other domestic preconditions for the development of
local bond markets before giving that process a further push by finally opening the capital
account.

Data appendix
The data set covers the period 1990-2001 at an annual frequency. Sample economies are
Argentina, Australia, Austria, Belgium, Brazil, Canada, Chile, China, the Czech Republic,
Denmark, Finland, France, Germany, Greece, Hong Kong SAR, Hungary, Iceland, India,
Ireland, Italy, Korea, Japan, Malaysia, Mexico, the Netherlands, New Zealand, Norway, Peru,
the Philippines, Poland, Portugal, Russia, Singapore, South Africa, Spain, Sweden,
Switzerland, Thailand, Turkey, the United Kingdom and the United States.

Domestic debt securities

Domestic debt securities are taken from Table 16A of the BIS domestic and international
securities statistics, which are regularly published in the annex tables of the BIS Quarterly
Review. The series are accessible at http://www.bis.org/statistics/secstats.htm.

Interbank interest rates

Interbank rates are taken from online provider Global Financial Data. Twelve-month
interbank rates are used wherever they are available. When 12-month rates are not
available, shorter rates are used, and where shorter rates are not available, we use the
monthly average of daily overnight interbank rates.

46
In particular, harmonising market regulations and withholding tax regimes or creating a pan-Asian payment
and settlement system with the goal of encouraging more cross-border investment in the region and thereby
producing deeper and more liquid markets would be tantamount to encouraging more capital flows and thus
equivalent to early capital account liberalisation. In other words, doing so would promote market development
but also heighten crisis risk, which is the very dilemma referred to in the text (Eichengreen (2004)).

54 BIS Papers No 30
Exchange rates
Exchange rates are end-month (local currency per US dollar) from line AE in International
Financial Statistics.

Institutional variables
Measures of government stability, investment profile, law and order, corruption and
bureaucratic quality are from the International Country Risk Guide:
Investment profile is an assessment of factors affecting the risk to direct investment. The
risk rating assigned is the sum of three subcomponents, each with a maximum score of four
points and a minimum score of zero points. A score of four points equates to very low risk
and a score of zero points to very high risk. The subcomponents are:
• Contract viability/expropriation
• Profit repatriation
• Payment delays
Law and order are assessed separately, with each subcomponent comprising zero to three
points. The law subcomponent is an assessment of the strength and impartiality of the legal
system, while the order subcomponent is an assessment of popular observance of the law. A
higher score indicates better law and order.
Corruption is an assessment of corruption within the political system. The index ranges from
zero to six, where a higher score means a lower degree of corruption.
Bureaucratic quality is the measure of institutional strength and quality of the bureaucracy.
High points are given to countries where the bureaucracy has the strength and expertise to
govern without drastic changes in policy or interruptions in government services. In the
low-risk countries, the bureaucracy tends to be somewhat autonomous from political
pressure and to have an established mechanism for recruitment and training.

Capital control variable

Coded from the IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions
by Nancy Brune and Geoffrey Garrett; see Brune (2003).

Other variables
The following series are from the World Bank’s World Development Indicators:
GDP (constant 1995 US dollars)
GDP (current US dollars)
GDP per capita (constant 1995 US dollars)
GDP per capita, PPP (current international US dollars)
GDP, PPP (current international US dollars)
Interest rate spread (lending rate minus deposit rate)
Interest rate spread (lending rate minus Libor)
Lending interest rate (in per cent)
Credit to private sector (as a percentage of GDP)
Deposit interest rate (in per cent)

BIS Papers No 30 55
Domestic credit provided by banking sector (as a percentage of GDP)
Market capitalisation of listed companies (as a percentage of GDP)
Market capitalisation of listed companies (current US dollars)
Overall budget deficit, including grants (as a percentage of GDP)
Real effective exchange rate index (1995 = 100)
S&P/IFC investable index (annual change, in per cent)
Stocks traded, total value (as a percentage of GDP)
Stocks traded, turnover ratio (in per cent)

56 BIS Papers No 30
 Table 1
Total outstanding external finance (end-2001)
As a percentage of GDP

Domestic Outstanding domestic debt securities

credit Stock
provided by market
Issued by Issued by Issued by
banking capitalisation
corporate public financial
sector
issuers sector institutions

Emerging markets 90.21 56.87 5.76 24.96 8.28

Asia 131.91 75.56 9.27 23.52 12.00
China 140.59 45.21 0.90 25.04 8.80
Hong Kong SAR 141.98 310.81 3.07 11.78 11.90
Korea 110.37 54.97 27.84 18.32 23.20
Malaysia 156.23 135.92 50.40 36.57 7.54
Singapore 102.95 138.25 6.71 34.16 20.61
Thailand 112.03 31.67 4.96 26.17 0.35
Latin America 41.21 38.70 1.73 26.12 4.53
Argentina 37.13 71.62 2.71 9.11 2.04
Brazil 59.19 37.06 0.56 51.99 9.40
Chile 76.74 89.28 8.82 29.76 13.86
Mexico 24.69 20.49 1.52 12.10 0.68
Central Europe 42.48 16.04 1.23 29.32 0.92
Czech Republic 51.84 16.22 4.79 36.35 4.61
Hungary 49.54 19.80 1.53 35.31 0.00
Poland 37.34 14.85 0.00 25.26 0.00

Developed
economies 194.13 122.92 20.55 85.18 33.64
Australia 93.99 101.55 12.25 17.13 16.75
Canada 93.18 90.86 10.50 59.60 14.33
Japan 308.67 92.10 16.48 104.45 15.94
New Zealand 120.00 36.82 0.00 28.36 0.00
United States 160.56 137.48 23.90 83.53 43.32
Europe 121.30 156.25 8.05 44.12 28.44
Sources: World Bank, World Development Indicators (WDI); BIS.

BIS Papers No 30 57
 Table 2
Composition of external finance (end-2001)
As a percentage of total

Outstanding
Domestic credit domestic debt
Stock market
provided by banking securities (corporate
capitalisation
sector issuers and public
sector)

Emerging markets 50.74 31.98 17.28

Asia 54.90 31.45 13.65
China 66.40 21.35 12.25
Hong Kong SAR 30.36 66.46 3.17
Korea 52.18 25.99 21.82
Malaysia 41.21 35.85 22.94
Singapore 36.50 49.01 14.49
Thailand 64.08 18.11 17.81
Latin America 38.24 35.91 25.85
Argentina 30.79 59.40 9.80
Brazil 39.78 24.91 35.31
Chile 37.51 43.64 18.86
Mexico 41.99 34.85 23.17
Central Europe 47.69 18.01 34.30
Czech Republic 47.48 14.85 37.67
Hungary 46.66 18.65 34.69
Poland 48.21 19.18 32.61

Developed
economies 45.92 29.08 25.01
Australia 41.79 45.15 13.06
Canada 36.66 35.75 27.58
Japan 59.17 17.65 23.18
New Zealand 64.80 19.88 15.32
United States 39.60 33.91 26.50
Europe 42.32 38.72 18.96

Sources: WDI; BIS.

58 BIS Papers No 30
 Table 3
New external finance in emerging markets
As a percentage of GDP

1997 1998 1999 2000 2001

Emerging
markets 22.47 27.03 18.69 23.20 20.28
Domestic 18.05 24.47 15.77 19.56 17.33
Equities 1.00 0.92 1.26 0.67 0.54
Bonds
Private 0.30 0.33 0.30 2.59 3.25
Public 10.45 17.73 11.50 10.25 9.09
Bank loans
Private 4.55 4.49 2.25 5.29 2.72
Public 1.74 1.00 0.46 0.76 1.72
International 4.42 2.56 2.93 3.64 2.95
Equities 0.50 0.16 0.46 0.84 0.25
Bonds
Private 1.12 0.56 0.64 0.58 0.88
Public 1.12 0.81 1.01 0.86 0.83
Bank loans
Private 1.14 0.61 0.65 1.04 0.79
Public 0.54 0.42 0.18 0.32 0.20

Asia 12.63 15.88 16.77 19.72 22.20

Domestic 8.46 14.38 14.57 16.21 19.03
Equities 1.49 0.99 1.93 1.03 0.60
Bonds
Private 0.00 0.00 0.05 2.12 3.00
Public 0.36 2.52 2.49 2.78 5.28
Bank loans
Private 6.70 7.56 8.54 9.08 7.53
Public –0.09 3.31 1.56 1.21 2.61
International 4.16 1.50 2.20 3.51 3.16
Equities 0.55 0.24 0.76 1.32 0.43
Bonds
Private 1.04 0.20 0.45 0.62 1.21
Public 0.74 0.35 0.54 0.39 0.57

BIS Papers No 30 59
 Table 3 (cont)
New external finance in emerging markets
As a percentage of GDP

1997 1998 1999 2000 2001

Bank loans
Private 1.18 0.27 0.26 0.81 0.80
Public 0.65 0.43 0.21 0.37 0.16

Central Europe 20.88 30.19 24.52 24.30 33.47

Domestic 17.52 26.81 21.49 21.67 31.08
Equities 0.54 2.52 1.30 0.51 0.34
Bonds
Private 0.21 0.11 0.13 0.07 0.11
Public 17.69 18.03 23.07 22.35 22.85

Bank loans
Private 1.46 4.50 –1.48 0.45 4.24
Public –2.38 1.66 –1.53 –1.71 3.54
International 3.36 3.38 3.03 2.63 2.39
Equities 1.07 0.56 0.45 0.15 0.00
Bonds
Private 0.52 0.82 0.69 0.33 0.66
Public 0.52 0.96 1.05 0.50 0.78
Bank loans
Private 0.73 0.55 0.49 1.49 0.43
Public 0.52 0.50 0.35 0.17 0.52

Latin America 34.52 38.45 20.20 27.59 15.35

Domestic 29.63 34.89 16.31 23.60 12.55
Equities 0.50 0.60 0.33 0.23 0.50
Bonds
Private 0.67 0.72 0.68 3.61 4.17
Public 21.44 33.98 21.50 18.05 11.08
Bank loans
Private 2.46 1.19 –5.56 1.12 –3.47
Public 4.56 –1.59 –0.65 0.59 0.27
International 4.89 3.55 3.89 3.99 2.80
Equities 0.34 0.00 0.05 0.33 0.08
Bonds
Private 1.32 0.91 0.89 0.56 0.52
Public 1.67 1.27 1.64 1.54 1.17

60 BIS Papers No 30
 Table 3 (cont)
New external finance in emerging markets
As a percentage of GDP

1997 1998 1999 2000 2001

Bank loans
Private 1.15 0.98 1.21 1.27 0.84
Public 0.41 0.39 0.11 0.28 0.19

Note: Dollar amounts are from Tables 4.2 and 4.3 in the IMF’s Global Financial Stability Report: Market
Developments and Issues (March 2003). GDP data are from the World Bank’s World Development Indicators.
Emerging markets include China, Hong Kong SAR, Korea, Malaysia, Singapore, Thailand, Argentina, Brazil,
Chile, Mexico, the Czech Republic, Hungary and Poland.

BIS Papers No 30 61
62

Table 4
Correlations of explanatory variables
Budget
Dummy Dis- Concen- balance
GDP Interest Exchange
Exports Dummy for tance Invest- Law Account- tration Bureau- Interest (% of
GDP, per Corrup- Bank rate rate
to GDP for English from ment and ing in cratic rate GDP)
PPP capita, tion credits vola- vola-
(%) Asia legal equa- profile order standards banking quality spread 3-year
PPP tility tility
origin tor sector moving
average
GDP, PPP 1
Exports to
GDP (%) –0.2511 1
Dummy for
Asia –0.1045 0.517 1
Dummy for
English legal
origin 0.2960 0.3069 0.2452 1
Distance
from equator 0.0131 –0.4452 –0.6842 –0.3749 1
Investment
profile 0.1807 0.1415 –0.1082 0.0953 –0.0496 1
Law and
order 0.1658 0.0917 –0.3004 0.2509 0.5557 0.0247 1
GDP per
capita, PPP 0.4217 –0.0032 –0.4491 0.1586 0.6141 0.171 0.7602 1
Corruption –0.0273 –0.0552 –0.3841 –0.0111 0.6711 –0.0405 0.6468 0.6298 1
Accounting
standards 0.1402 0.3326 0.1509 0.4985 0.1489 –0.0344 0.4411 0.4720 0.4028 1
Bank credit 0.3339 0.0322 0.1145 0.1976 0.1700 –0.0498 0.4414 0.4301 0.2420 0.3269 1
BIS Papers No 30

Concent-
ration in
banking
sector –0.5322 0.2670 –0.1784 –0.1083 0.2408 –0.0719 0.1574 0.1041 0.4159 0.2559 –0.1242 1
BIS Papers No 30

Table 4 (cont)
Correlations of explanatory variables
Budget
Dummy Dis- Concen- balance
GDP Interest Exchange
Exports Dummy for tance Invest- Law Account- tration Bureau- Interest (% of
GDP, per Corrup- Bank rate rate
to GDP for English from ment and ing in cratic rate GDP)
PPP capita, tion credits vola- vola-
(%) Asia legal equa- profile order standards banking quality spread 3-year
PPP tility tility
origin tor sector moving
average
Bureaucratic
quality 0.2337 0.0745 –0.3017 0.2666 0.5625 0.1557 0.7582 0.8176 0.7016 0.5669 0.4373 0.1927 1
Interest rate
volatility –0.1436 –0.1417 0.0729 –0.2013 –0.3703 –0.0713 –0.6644 –0.5824 –0.4733 –0.3767 –0.3942 –0.0222 –0.5799 1
Interest rate
spread –0.1516 –0.2699 –0.0111 –0.2893 –0.2219 –0.2787 –0.6057 –0.5791 –0.295 –0.3633 –0.389 –0.0602 –0.5275 0.7165 1
Exchange
rate volatility –0.2516 –0.0326 –0.0347 –0.2457 0.1406 –0.2046 –0.0866 –0.0696 0.0998 0.0258 0.1191 0.1883 –0.0383 0.1795 0.2693 1
Budget
balance (%
of GDP)
3-year
moving
average –0.0459 0.5713 0.4628 0.3817 –0.549 0.3514 –0.087 –0.1121 –0.2364 0.1223 –0.0909 –0.0481 –0.0294 0.0907 –0.1719 –0.1969 1
63
64

Table 5
Multivariate analysis

(1) (2) (3) (4) (5) (6) (7) (8)

GDP, PPP (current international billions of US 0.010 0.012 0.012 0.012

dollars) (7.97) (11.72)*** (11.98)*** (19.04)***
Exports to GDP (in per cent) 0.209 0.237 0.351 0.265
(3.53)*** (4.19)*** (8.25)*** (5.02)***
Dummy for Asia –32.702 –29.669 –10.674 –16.899
(6.33)*** (7.23)*** (3.81)*** (4.85)***
Dummy for English legal origin –9.492 –3.587 –18.426
(4.96)*** (1.17) (3.74)***
Distance from equator 110.339 65.177
(11.79)*** (4.10)***
Investment profile –0.371 –0.542
(1.97)** (1.71)*
Law and order –0.674 0.808
(1.47) (0.82)
GDP per capita, PPP (current international 3.179 –0.291
thousands of US dollars) (30.53)*** (1.26)
Corruption 3.383 0.023
(4.90)*** (0.03)
Accounting standards (La Porta et al (1998)) 0.630 0.775
(6.66)*** (3.47)***
Domestic credit provided by banking sector (as a 0.213 0.090
percentage of GDP) (9.17)*** (2.84)***
Concentration in banking sector –9.031 –18.909
BIS Papers No 30

(3.47)*** (4.60)***
Bureaucratic quality 12.327 1.554
(11.10)*** (1.17)
Standard deviation of interbank interest rates –0.222 0.605
(0.83) (2.19)**
BIS Papers No 30

Table 5 (cont)
Multivariate analysis

(1) (2) (3) (4) (5) (6) (7) (8)

Interest rate spread (interbank rate minus Libor) –0.713 –0.484

(5.65)*** (3.47)***

Standard deviation of change in log of exchange 39.393 –5.936

rates (1.87)* (0.34)
IMF capital controls dummy variable 3.226 2.641
(2.87)*** (1.85)*
Constant 45.368 46.363 –3.903 13.413 7.538 2.535 52.352 –25.739
(23.25)*** (26.75)*** (0.86) (5.24)*** (1.26) (0.63) (42.46)*** (2.75)***

Observations 475 475 421 469 395 405 400 284

Number of id 41 41 36 41 34 41 38 30

Note: Absolute value of z statistics in parentheses; * = significant at 10%; ** = significant at 5%; *** = significant at 1%.
65
66

Table 6
Sensitivity analysis

Total Private Public

(1) (2) (3) (4) (5) (6) (7)

GDP, PPP (current international billions of US 0.011 0.011 0.005 0.006 0.005 0.006 0.010
dollars) (13.78)*** (12.75)*** (3.95)*** (13.00)*** (13.43)*** (5.99)*** (9.09)***
Exports to GDP (in per cent) 0.176 0.159 0.113 0.130 0.106 0.131 0.236
(2.45)** (1.87)* (3.58)*** (3.59)*** (3.10)*** (3.03)*** (6.05)***
Dummy for Asia 7.484 5.217 13.259 8.805 7.743 8.673 –8.124
(1.26) (0.80) (4.94)*** (3.90)*** (3.07)*** (1.57) (1.71)*
Dummy for English legal origin –4.718 –7.270 –20.759 –15.939 –14.394 9.714 4.957
(1.16) (1.54) (8.35)*** (7.59)*** (7.20)*** (2.77)*** (1.55)
Distance from equator 111.762 79.264 69.202 58.224 48.715 73.631 63.661
(6.02)*** (4.08)*** (8.02)*** (6.91)*** (6.18)*** (6.03)*** (6.03)***
Investment profile 0.357 0.111 0.187 –0.149 –0.087 0.260 –0.028
(1.06) (0.33) (1.31) (0.97) (0.59) (1.14) (0.11)
Law and order 2.066 –0.097 –0.387 0.217 0.288 1.452 1.021
(2.08)** (0.09) (1.07) (0.49) (0.65) (1.87)* (1.18)
GDP per capita, PPP (current international –0.035 0.662 –0.372 –0.143 –0.203 –0.712 –0.745
thousands of US dollars) (0.09) (1.65)* (2.21)** (0.70) (1.09) (2.58)*** (2.95)***
Corruption 2.500 2.552 0.201 1.208 1.353 0.456 0.978
(2.99)*** (2.73)*** (0.57) (3.03)*** (3.58)*** (0.72) (1.34)
Accounting standards (La Porta et al (1998)) 0.330 –0.095 0.351 0.480 0.446 –0.134 0.102
BIS Papers No 30

(1.72)* (0.32) (4.01)*** (5.83)*** (5.56)*** (0.69) (0.60)

Domestic credit provided by banking sector (as a 0.039 0.004 0.071 0.070 0.103 0.040 0.092
percentage of GDP) (0.89) (0.08) (4.25)*** (3.52)*** (4.89)*** (1.29) (3.05)***
Concentration in banking sector –11.878 –20.028 –4.101 –2.739 –1.279 –11.415 –11.823
(2.63)*** (3.87)*** (1.68)* (1.12) (0.51) (3.30)*** (3.35)***
BIS Papers No 30

Table 6 (cont)
Sensitivity analysis

Total Private Public

(1) (2) (3) (4) (5) (6) (7)

Bureaucratic quality –1.397 –0.554 1.114 1.544 1.754 –1.665 –1.571

(1.06) (0.32) (2.04)** (2.68)*** (3.17)*** (1.67)* (1.28)
Standard deviation of interbank interest rates 0.509 0.295 0.159 0.084 0.110 0.218 0.191
(2.02)** (0.87) (1.55) (0.78) (1.07) (1.00) (0.81)
Interest rate spread (interbank rate minus Libor) –0.391 –0.289 –0.104 –0.116 –0.166 –0.285 –0.295
(3.25)*** (1.85)* (1.80)* (1.91)* (2.81)*** (2.81)*** (2.69)***

Standard deviation of change in log of exchange –57.932 –95.382 –17.051 –21.608 –25.153 –31.280 –38.613
rates (3.07)*** (4.21)*** (2.14)** (2.24)** (2.51)** (2.23)** (2.63)***
IMF capital controls dummy variable 5.740 4.859 –0.218 1.336 1.056 4.385 5.667
(1= if capital account is open) (3.24)*** (2.81)*** (0.29) (1.44) (1.20) (3.00)*** (4.07)***
Fiscal balance (as a percentage of GDP) three-year –1.357 0.165 –1.204
moving average (5.91)*** (1.58) (6.83)***
Lagged overall budget balance (as a percentage of –0.871 0.078 –0.348
GDP) (5.09)*** (1.21) (2.52)**
Outstanding domestic debt securities issued by 1.094
public sector (as a percentage of GDP) (45.88)***
Constant –41.317 9.632 –36.844 –49.076 –45.686 7.940 –8.686
(3.93)*** (0.63) (7.52)*** (9.93)*** (9.84)*** (0.71) (0.91)

Observations 231 235 284 231 235 231 235

Number of id 28 29 30 28 29 28 29

Note: Absolute value of z statistics in parentheses; * = significant at 10%; ** = significant at 5%; *** = significant at 1%.
67
 Figure 1
Bond markets and country size

y = 47.0216 + 0.0132 x
t-stat: 7.72) (3.25) r-squared: 0.21
Outstanding domestic debt securities (% of GDP)

150

Japan
100

Malaysia

50 Korea
Singapore
Philippines
India
China
Hong Kong SAR
Thailand
0
0 2,000 4,000 6,000 8,000
GDP (constant 1995 billions of US dollars)

Source: World Bank, World Development Indicators (WDI).

Figure 2
Bond markets and exports to GDP

y = 56.6124 + −0.0343 x
t-stat: (5.98) (0.15) r-squared: 0.00
Outstanding domestic debt securities (% of GDP)

150

Japan
100

Malaysia

50 Korea

Philippines Singapore

India China
Thailand Hong Kong SAR

0
0 50 100 150

Exports to GDP (%)

Source: WDI.

68 BIS Papers No 30
 Figure 3
Bond markets and distance from equator

y = 16.2301 + 106.4054 x
t-stat: (1.08) (3.11) r-squared: 0.22

Outstanding domestic debt securities (% of GDP) 150

Japan
100

Malaysia

50 Korea

Singapore
Philippines
India
China
Thailand Hong Kong SAR
0
0 .2 .4 .6 .8
Distance from equator

Note: Measured by absolute value of the latitude of a country, scaled between zero and one.
Source: La Porta et al (1999).

Figure 4
Bond markets and investment profile

y = −83.1888 + 19.4645 x
t-stat: (2.16) (3.64) r-squared: 0.25
Outstanding domestic debt securities (% of GDP)

150

Japan
100

Malaysia

50 Korea

Philippines Singapore

India China
Hong Kong SAR
Thailand
0

4 5 6 7 8 9
Investment profile

Note: See data appendix.

Source: The PRS Group, International Country Risk Guide (ICRG).

BIS Papers No 30 69
 Figure 5
Bond markets and law and order

y = −28.4493 + 17.4018 x
t-stat: (1.29) (3.91) r-squared: 0.28
Outstanding domestic debt securities (% of GDP)

150

Japan
100

Malaysia

50 Korea

Singapore
Philippines
India China
Thailand Hong Kong SAR

0
2 3 4 5 6
Law and order

Note: See data appendix.

Source: ICRG.

Figure 6
Bond markets and corruption

y = −22.9929 + 17.8387 x
t-stat: (1.02) (3.58) r-squared: 0.25
Outstanding domestic debt securities (% of GDP)

150

Japan
100

Malaysia

50 Korea

Singapore
Philippines
India China Hong Kong SAR
Thailand
0
2 3 4 5 6

Corruption

Note: See data appendix.

Source: ICRG.

70 BIS Papers No 30
 Figure 7
Bond markets and accounting standards

y = 0.7577 + 0.9480 x
t-stat: (0.02) (1.53) r-squared: 0.07
Outstanding domestic debt securities (% of GDP)
150

Japan
100

Malaysia

50 Korea

Philippines Singapore

India
Hong Kong SAR
Thailand
0
40 50 60 70 80
Accounting standards (La Porta et al (1998))

Note: A higher score means a better accounting standard.

Source: La Porta et al (1998).

Figure 8
Bond markets and GDP per capita

y = 22.6556 + 2.0626 x
t-stat: (2.92) (5.33) r-squared: 0.42
Outstanding domestic debt securities (% of GDP)

150

Japan
100

Malaysia

50 Korea

Singapore
Philippines
India
China
Hong Kong SAR
Thailand
0
0 10 20 30 40 50
GDP per capita (constant 1995 thousands of US dollars)

Source: WDI.

BIS Papers No 30 71
 Figure 9
Bond markets and banking sector development

y = 17.5699 + 0.4158 x
t-stat: (1.43) (3.43) r-squared: 0.23
Outstanding domestic debt securities (% of GDP)

150

Japan
100

Malaysia

50 Korea
Singapore
Philippines
India China Hong Kong SAR
Thailand
0
0 100 200 300
Domestic credit provided by banking sector (% of GDP)

Source: WDI.

Figure 10
Bond markets and bank concentration

y = 73.9533 + −30.3529 x
t-stat: (3.49) (0.91) r-squared: 0.02
Outstanding domestic debt securities (% of GDP)

150

Japan
100

Malaysia

50 Korea

Singapore
Philippines
India Thailand China
Hong Kong SAR

0
.2 .4 .6 .8 1
Concentration in banking sector

Source: Beck et al (1999).

72 BIS Papers No 30
 Figure 11
Public and private sector bond market development

y = 6.8623 + 0.3966 x

Domestic debt securities by non-public sector (% of GDP)

t-stat: (1.28) (3.14) r-squared: 0.20
100

80

60

40 Korea
Malaysia
Japan

20
Singapore
Hong Kong SAR
China
Thailand
0 India Philippines

0 20 40 60 80 100
Outstanding domestic debt securities issued by public sector (% of GDP)

Source: BIS.

Figure 12
Bond markets and bureaucratic quality

y = −36.1273 + 29.1189 x
t-stat: (1.85) (4.85) r-squared: 0.38
Outstanding domestic debt securities (% of GDP)

150

Japan
100

Malaysia

50 Korea

Singapore
Philippines
China India
Hong Kong SAR
Thailand
0
1.5 2 2.5 3 3.5 4
Bureaucratic quality
Note: See data appendix.
Source: ICRG.

BIS Papers No 30 73
 Figure 13
Bond markets and interest rate volatility

y = 68.9338 + −5.1011 x
t-stat: (10.18) (3.41) r-squared: 0.24
Outstanding domestic debt securities (% of GDP)

150

Japan
100

Malaysia

50 Korea

Singapore Philippines
China
Thailand
Hong Kong SAR
0

0 5 10 15 20
Standard deviation of interbank interest rates

Sources: Global Financial Database (GFD); authors’ calculations.

Figure 14
Bond markets and the level of interest rates

y = 66.8278 + −1.1315 x
t-stat: (9.03) (2.30) r-squared: 0.12
Outstanding domestic debt securities (% of GDP)

150

Japan
100

Malaysia

50 Korea
Singapore
Philippines
China India
Hong Kong SAR
Thailand
0

0 20 40 60 80
Interest rate spread (lending rate minus Libor)
Sources: GFD; WDI.

74 BIS Papers No 30
 Figure 15
Bond markets and fixed exchange rate regime

y = 52.6573 + 14.9173 x
t-stat: (6.70) (0.59) r-squared: 0.01

Outstanding domestic debt securities (% of GDP) 150

Japan
100

Malaysia

50 Korea

Singapore
Philippines
India China
Thailand Hong Kong SAR

0
0 .2 .4 .6 .8
Fixed exchange rate regime dummy (Reinhart-Rogoff)

Source: Reinhart and Rogoff (2002).

Figure 16
Bond markets and exchange rate volatility

y = 71.9603 + −647.6625 x
t-stat: (5.34) (1.37) r-squared: 0.05
Outstanding domestic debt securities (% of GDP)

150

Japan
100

Malaysia

50 Korea

Singapore Philippines
India
Hong Kong SAR ChinaThailand

0
0 .02 .04 .06 .08
Standard deviation of change in log of exchange rates

Sources: IMF, International Financial Statistics (IFS); author’s calculations.

BIS Papers No 30 75
References
Asia-Pacific Economic Cooperation (1999): Compendium of sound practices, guidelines to
facilitate the development of domestic bond markets in APEC member economies, report of
the collaborative initiative on the development of domestic bond markets, APEC,
Washington, DC.
Beck, T, A Demirguc-Kunt and R Levine (2002): “Law, endowments, and finance”, NBER
Working Papers, no 9089, August.
Benston, G J (1994): “Universal banking”, Journal of Economic Perspectives, vol 8,
pp 121-43.
Boot, A and A Thakor (1997): “Financial system architecture”, Review of Financial Studies,
vol 10, pp 693-733.
Brune, N, A Guisinger, J Sorens and G Garrett (2003): The political economy of capital
account liberalization in developing countries, UCLA, unpublished manuscript.
Burger, J D and F E Warnock (2003): “Diversification, original sin, and international bond
portfolios”, International Finance Discussion Papers, no 755, Board of Governors of the
Federal Reserve System, Washington, DC, January.
——— (2004): “Foreign participation in local-currency bond markets”, International Finance
Discussion Papers, no 794, Board of Governors of the Federal Reserve System,
Washington, DC, February.
Calvo, G and C Reinhart (2002): “Fear of floating”, Quarterly Journal of Economics, no 117,
pp 379-408.
Claessens, S, D Klingebiel and S Schmukler (2003): “Government bonds in domestic and
foreign currency: the role of macroeconomic and institutional factors”, CEPR Discussion
Papers, no 3789, February.
Dalla, I, D Khatdhate, D R Rao, K Kondury, L Jun and T Chuppe (1995): The emerging Asian
bond market, World Bank, Washington, DC.
Diamond, D (1991): “Monitoring and reputation: the choice between bank loans and directly
placed debt”, Journal of Political Economy, vol 99, pp 689-721.
Domowitz, I, J Glen and A Madhavan (2000): International evidence on aggregate corporate
financing decisions, unpublished manuscript, Pennsylvania State University, January.
Eichengreen, B (2004): The unintended consequences of the Asian bond fund, University of
California, Berkeley, unpublished manuscript, February.
Eichengreen, B, R Hausmann and U Panizza (2002): Original sin: the pain, the mystery, and
the road to redemption, UC Berkeley, Harvard University and IDB, unpublished manuscript.
Eichengreen, B and A Mody (2000): “What explains the changing spreads on emerging
market debt? Fundamentals or market sentiment?”, in Sebastian Edwards (ed), Capital flows
and the emerging economies, University of Chicago Press, Chicago, pp 107-36.
Fernandez, D and S Klassen (2003): Asian bond fund crawls, but should it ever walk?,
JP Morgan Chase, mimeograph, 12 June.
Goldstein, M (1998): The Asian financial crisis, Institute of International Economics,
Washington, DC.
Government of Japan (2003): “ASEAN+3 high-level seminar on fostering bond markets in
Asia”, www.mof.go.jp/english/if/hls20030301a.htm.
Hale, G (2003): “Bonds or loans? The effects of macroeconomic fundamentals”, Cowles
Foundation Discussion Papers, no 1403, February.

76 BIS Papers No 30
Harwood, A, ed (2000): Building local bond markets: an Asian perspective, International
Finance Corporation, Washington, DC.
Hawkins, J (2002): “Bond markets and banks in emerging economies”, BIS Papers, no 11,
June-July.
Herring, R J and N Chatusripitak (2000): “The case of the missing market: the bond market
and why it matters for financial development”, ADB Institute Working Papers, no 11, July.
Ho, T (2003): The Singapore bond market, Deutsche Bank Global Markets Research, June.
Kim, Y (2000): “Korea’s bond market following the onset of the financial crisis”, in Alison
Harwood (ed), Building local bond markets: an Asian perspective, International Finance
Corporation, Washington, DC, pp 63-83.
Kim, Y-H (2001): Developing treasury securities markets in Asia, Asian Development Bank,
mimeo, June.
La Porta, R l, F Lopez-de-Silanes, A Shleifer and R W Vishny (1998): “Law and finance”,
Journal of Political Economy, vol 106, no 6, pp 1113-55.
——— (1999): “The quality of government”, Journal of Law, Economics, and Organization,
vol 15, pp 222-79.
Lynch, D (2001): Asian bond markets, International Banks and Securities Association of
Australia, mimeo.
McCauley, R (2003): “Unifying government bond markets in East Asia”, BIS Quarterly
Review, pp 89-98, December.
McCauley, R and E Remolona (2000): “Size and liquidity in government bond markets”, BIS
Quarterly Review, November, pp 52-8.
Rajan, R and L Zingales (2001): “The great reversals: the politics of financial development in
the 20th century”, NBER Working Papers, no 8178, March.
——— (2003): Saving capitalism from the capitalists, Crown Books, New York.
Reinhart, C M and K S Rogoff (2002): “The modern history of exchange rate arrangements: a
reinterpretation”, NBER Working Papers, no 8963, May.
Schinasi, G J and R T Smith (1998): “Fixed-income markets in the United States, Europe and
Japan: some lessons for emerging markets”, IMF Working Papers, no 173, December.
Sharma, K (2000): “The underlying constraints on corporate bond market development in
southeast Asia”, United Nations Department of Economic and Social Affairs Discussion
Papers, no 14, September.
Smith, R T (1998): Banking competition and the development of securities markets,
University of Alberta, unpublished manuscript.
Trairatvorakul, P (2001): Improvement of clearing and settlement for sound development of
bond markets, Securities and Exchange Commission of Thailand, mimeo.
World Bank (2003): Global development finance, World Bank, Washington, DC.

BIS Papers No 30 77
 Comments on Barry Eichengreen and
Pipat Luengnaruemitchai’s paper
“Why doesn’t Asia have bigger bond markets?”

Ric Deverell

I thank the Korea University and the BIS for organising this highly topical conference, and for
inviting me to participate. Thanks also to the authors for their thought-provoking presentation
and paper, which touch on many of the issues which will be discussed over the coming two
days. My comments can be loosely categorised into three sections. First, I will elaborate on
the genesis of the Asian Bond Fund, and the motivations underpinning its introduction.
Second, I will comment on the methodology adopted in the paper, and suggest some
avenues for further work. And third, I will discuss conclusions.

1. ABF origin and objectives

The Asian Bond Fund had its genesis within the EMEAP grouping of regional central banks
in mid-2002 - independent of other initiatives. The first stage was launched in the middle of
2003, and involved investment by regional central banks of USD 1 billion in a basket of US
dollar-denominated bonds issued by government and quasi-government issuers in eight of
the EMEAP economies. While this initiative demonstrates the EMEAP grouping’s
commitment to the continued development of regional bond markets, its impact on the
breadth and liquidity of domestic markets is likely to be small. The second stage, which is not
mentioned explicitly in Barry and Pipat’s paper, is likely to be far more significant. While the
final details are still being finalised, this stage will initially involve investment by central banks
in bonds denominated in domestic currencies. However, more significantly, it will put in place
the structures to allow access by private investors.
In their paper, Barry and Pipat question whether the use of central bank reserves in the
Asian Bond Fund will have the desired effect - which they suggest is to jump-start supply and
demand by increasing the size of the installed base through government and central bank
purchases of bonds.
This characterisation does not, in my view, fully capture the objectives underlying the ABF -
and ignores the ABF2 extension, which was always part of the scoped model. The idea of
ABF2 is not so much to “jump-start” demand with our own demand, but to blaze a trail that
others could follow. By establishing ABF2, we intend to show that it can be done (identifying
and removing regulatory or other impediments along the way), and provide the infrastructure
for private sector investors to utilise. This is very clear in the focus on private sector
involvement following on from the central bank investments, once we have resolved
infrastructure issues such as price dissemination services and the setting-up of indices. In
fact, we were always conscious not to make ABF so big as to risk locking up securities and
depressing turnover as a result! Something that larger investments by the government sector
might have done.

78 BIS Papers No 30
2. The results
The paper attempts to shed light on what initiatives are “most urgently needed to promote
Asian bond markets”. In order to achieve this, it sets out five hypotheses (historical,
structural, developmental stage, structure of financial system, and macroeconomic) and
proxies each with several variables. Using these variables, a broad-based cross-country
econometric analysis is undertaken.
While this broad-brush approach is useful as part of the initial sorting process, in general the
discussion of the empirical results is presented as being more certain than the regressions
seem to support. As many of the implications drawn depend on model specification, it would
be useful to make a judgment as to the preferred model and then to discuss the results.
Alternatively, variables that are robust across model specifications could be singled out as
the most likely determinants.
Consistent with this, several of the conclusions warrant further analysis. First, the paper
sends mixed messages about the effect of exchange rate stability/volatility. On page 53, last
full paragraph, it is suggested that “On the other hand, the stability of exchange rates in the
region appears, if anything, to have encouraged bond market development”. However, on
page 50, first full paragraph, the paper suggests “...there is little evidence of a relationship
between exchange rate volatility and bond market development”. One of the things we have
been trying to achieve in emerging markets over recent years is better management of
currency mismatch. Part of the solution in our view has to be demand-oriented. That is, there
will be little interest in managing currency mismatch if everyone thinks that the exchange rate
will be stable. We should not sell currency stability as a means of promoting financial
stability.
Second, the paper does not adequately distinguish the type of bond markets that we are
seeking to promote, and as such does not explore the important issue of currency
denomination of bond markets. Presumably one of our ultimate goals is for bond markets to
facilitate borrowing in domestic currency in order to reduce currency mismatch (the so-called
problem of original sin). Understanding what allows countries to do this is one of the more
important questions the international community faces at the moment.
A third point relates to the type of bond market desired. In the paper, bond market
development is defined primarily as aggregate capitalisation. While this may be a useful
general proxy, it is not clear that bigger is always better - particularly if a large part of
capitalisation is due to large levels of government debt. It is interesting to note that, on this
metric, the Australian bond market is almost identical as a share of GDP to that seen in
developing Asia. Development may be better defined relative to some metric of secondary
market liquidity - for example, bid-offer spread or turnover (ratio of outstandings).
A fourth point relates to the relatively controversial finding that bank financing is
complementary to bond market development. Given that market-based financing dominates
in some countries, while bank-based financing dominates in others, it may also be useful to
split the samples into bank financing and market-based financing countries and examine the
relationship between banking sector and bond market development in the subsamples.

3. Conclusions
While the paper says that its intention is to help prioritise potential initiatives, its conclusions
are very general, with many factors found to be relevant. Given the general nature of the
analysis, no convincing argument is put forward that there is any particular initiative that will
be more crucial than others. This suggests that more work needs to be done to test which
reforms are more urgent.

BIS Papers No 30 79
 On the Asian bond market: comments on
Barry Eichengreen and Pipat Luengnaruemitchai’s paper
“Why doesn’t Asia have bigger bond markets?”

Junggon Oh 1

The main arguments of the above paper are as follows: causes of the underdevelopment of
the Asian bond markets include small economic size, poor legal systems, risky investment
environments, weak corporate governance and insufficient transparency, low levels of
economic development, less developed banking systems, inappropriate interest rate and
exchange rate policies, and capital controls. Results of empirical tests using panel
generalised least squares (GLS) on data for 41 countries for the period of 1990-2001 support
this general hypothesis.
An important finding in this paper is to show empirically the existence of complementarities
between banking sector and bond market development. It is necessary to stress again the
importance of the banking sector in the period of transition from a bank-centred financial
system to a market-centred one after the East Asian crisis, because the sector contributes to
a decrease in information asymmetry and transaction costs through long-run close
relationships with customers, etc.
There is a hierarchy of external finance in the financial system. The share of bank credit is
the largest among external financing sources in most countries mainly because transaction
costs and information asymmetry are relatively low in the banking sector (Table 1).

Table 1
Composition of external finance
2001, in per cent

Stock market Outstanding

Bank credit
capitalisation1 domestic bonds

Emerging markets 50.74 31.98 17.28

Asia 54.90 31.45 13.65
Latin America 38.24 35.91 25.85
Central Europe 47.69 18.01 34.30
Developed Countries 45.92 29.08 25.01
United States 39.60 33.91 26.50
Japan 59.17 17.65 23.18
1
Stock market capitalisation is not exactly equivalent to the amount of finance obtained through the stock
market.
Source: Eichengreen and Luengnaruemitchai (2006).

1
The views expressed herein are those of the author and do not necessarily reflect those of the Bank of Korea.

80 BIS Papers No 30
In order to develop the bond market, infrastructure development and the introduction of legal
requirements in areas such as information disclosure, accounting standards, credit ratings,
etc, are necessary in order to reduce transaction costs and information asymmetry. A
competitive banking system can also reduce information asymmetry and transaction costs
through long-run close relationships with customers and can thereby contribute to the
development of the bond market.
However, there are some points to be discussed regarding this paper: Are conclusions
specific to Asia or general to all regions? Can empirical tests for Asia only, rather than for
41 countries, obtain the same results?

BIS Papers No 30 81
 Consolidating the public debt markets of Asia

Robert N McCauley 1

The large reserves of East Asian central banks have received a great deal of attention
(Aizenmann and Marion (2002)). Some observers consider that these have made regional
finances more robust and better able to weather sudden withdrawals of capital. Others have
criticised the reserves as low-yielding external assets that are accumulated at the expense of
higher-yielding domestic investment. Others worry that exchange rate management that
gives rise to the reserves might result in investment in the traded goods sector that will prove
wasteful if exchange rates subsequently appreciate.
Less attention has been paid to the financing of the reserve build-up. The financing, or
sterilisation, of the foreign exchange reserve build-up has presented an opportunity for bond
market development, but policy has not made the most of this opportunity. While the interest
bearing debts issued by central banks to finance the reserve build-up have added to the sum
of public debts outstanding, they have generally also segmented that market into government
debt per se and central bank debt. While from a macroeconomic standpoint this choice
seems innocent, from a market development standpoint it has serious drawbacks.
This paper starts by considering the alternatives faced by a central bank in financing large
holdings of foreign exchange reserves. These choices are ranked, with use of government
securities in the first position. Then the transactions needed to use government securities to
finance reserves when these are held by the central bank are outlined. Then the benefits of
this approach are adduced and the issues that must be faced are discussed. These include
the attitude of the government and the rating agencies above all, and the practical questions
of the return to be paid to the government for its deposit at the central bank, the duration of
the extra government debt and consistency with the government budgetary process.
This proposal was originally made with reference to East Asia, especially Indonesia, Korea,
Malaysia, Taiwan (China) 2 and Thailand (McCauley (2003)). 3 Recently, the People’s Bank of
China (PBOC) and the Reserve Bank of India (RBI) both reached a crossroads as they ran
out of government securities to sell to sterilise purchases of foreign exchange. While the
PBOC opted for central bank bills, the RBI persuaded the government to issue new
government debt to sterilise. The contrast between these two cases illustrates that the
greatest impediment to the use of government securities is the natural reluctance of finance
ministers to issue, and parliaments to authorise, the needed expansion of recognised
government debt.

1
The author is grateful to Claudio Borio, Brian Coulton, Jeong-Ho Hahm, Corrinne Ho, Hak-Ryul Kim, Kyunjik
Lee, Guonan Ma, Madhu Mohanty, Ramon Moreno, T K Ogawa, Junggun Oh and William White and
participants in seminars at the Reserve Bank of Australia, the Hong Kong Institute for Monetary Research, the
Reserve Bank of India, the Bank of Korea and the Bank of Thailand for helpful discussions. Any errors remain
those of the author. Views expressed are those of the author and not the Bank for International Settlements.
2
Hereinafter Taiwan.
3
Until very recently the government of Hong Kong SAR had not issued any government debt, and the argument
of this paper did not apply there. Similarly, it would not apply to Chile, where the central bank is the only issuer
of public debt. With the 2004-05 budget, however, the Hong Kong government will become an issuer in Hong
Kong dollars, and the opportunity for a bond market development through consolidating the Exchange Fund
paper and government debt per se will arise.

82 BIS Papers No 30
1. The choice of sterilisation instrument
Most Asian central banks have seen their foreign exchange reserves (or foreign assets more
generally) outgrow their monetary liabilities (base money). This observation points to the
practice of sterilising foreign asset growth, typically implemented initially by selling domestic
assets like government paper. At some point, the central bank runs out of government paper
and must then mop up any additional excess liquidity by issuing its own liabilities.
Major Asian central banks reached this crossroads some time ago. Central banks in small
open economies like Singapore and Malaysia have had foreign exchange reserves well in
excess of base money for many years, and have managed money market liquidity at least in
part via the liability side of their balance sheets. The Indonesian, Korean and Taiwanese
central banks all reached similar crossroads in the 1980s. The PBOC and the RBI only just
reached it in 2003 and 2004, respectively. 4
The crossroads between sale of domestic assets and issuing a central bank liability is the
first branching on the diagram below (Figure 1). While it is possible to increase the demand
for central bank liabilities by increasing reserve requirements, this alternative goes against
the international trend towards lower reserve requirements and is not considered here. 5 The
key choice is then between “non-market” and “market” liabilities. In the first case, the central
bank accepts a deposit from the government or quasi-government body; in the second case,
the central bank sells an interest bearing liability to market participants. This distinction is
drawn based on the issuance mechanism and the identity of the immediate claimant on the
central bank.
The burden of this paper is that taking government deposits (the non-market approach) is the
best choice because it is most conducive to the development of money and bond markets
when the government has debt outstanding. In this case, the central bank’s issuance of its
own liability to market participants creates in effect two sovereign issuers in the domestic
bond market. 6 This segments the bond market in a manner that works against liquidity. Thus,
the best advised approach is to use a combination of government securities and a

4
This stylised progression towards a need to issue other liabilities for liquidity management purposes can be
accelerated if the central bank holds a substantial sum of unmarketable domestic assets. Thus the Bank of
Korea, burdened with loans to particular sectors at below market interest rates, started to sell its own interest
bearing liabilities long before foreign exchange reserves reached the level of the monetary base. See Oh
(2004) in this volume for a current proposal to fiscalise directed central bank credit as a first step towards
consolidating public bond markets. Similarly, quasi-fiscal burdens caused the PBOC to reach the crossroads
well before its foreign exchange reserves attained the level of the monetary base. In particular, the PBOC’s
claims on the asset management companies and other domestic assets of questionable market value meant
that it ran out of tradable domestic assets at a relatively early stage (Ma and Fung (2002)).
5
The RBI (2004b) notes: “In case CRR [cash reserve ratio] is not remunerated, it has the distortionary impact of
a ‘tax’ on the banking system. CRR is also discriminatory in that it has an in-built bias in favour of financial
intermediaries that are not required to maintain balances with the Reserve Bank… It is also to be noted that
the medium term objective of monetary policy is to bring down the CRR to its statutory minimum level of
3.0 per cent of NDTL [net demand and time liabilities]… Nevertheless, use of CRR as an instrument of
sterilisation, under extreme conditions of excess liquidity and when other options are exhausted, should not be
ruled out altogether by a prudent monetary authority ready to meet all eventualities.”
6
The RBI (2004b) reasons: “Issuance of central bank securities can fragment the debt market due to the
availability of two competing sovereign issues, one of the Central Government and the other of the Reserve
Bank. Normally, central banks issue securities at the short end of the maturity spectrum, on the premise that
the capital inflows are transient and may reverse over a short period; in the event of reversal, liquidity could be
matched by the maturing central bank paper. However, the Group felt that in the Indian context, issuance of
government securities at the short end, particularly for the cash management needs, would also be quite
significant and, therefore, market fragmentation remains a key issue.”

BIS Papers No 30 83
government deposit at the central bank as the sterilisation instrument once the central bank
has run out of domestic assets to sell or to repo into the market.

Figure 1
Instrument choices for absorbing liquidity

(a) Sell (or reverse repo)

domestic assets CN, IN,
KR, PH, SG, TH

CN (social security?)
Non-market IN (government deposit 2004)
SG (government deposit)
TW (postal saving)

CN (PBOC bills)
Public ID (SBIs)
securities KR (MSBs)
MY (BNM bills)
TW (NCDs)
(b) Issue liabilities TH (BOT bonds)

Take
deposits ID (“FASBI”)
MY (tender)
Market
SG

Private
transaction
IN (2003)
SG
TH
Swap dollars for
domestic currency

The other alternative sterilisation instruments are ordered by their desirability in terms of
developing the domestic money and bond markets. Thus, if the central bank must issue its
own liabilities, these can contribute to market development best if they are tradable
securities. Thus public securities are to be preferred to private transactions. Among private
transactions, ones that involve only the domestic currency are probably better suited to
market development than ones that involve foreign exchange. The latter tend to channel the
development of the local money market into the foreign exchange swap market. 7 Indeed,
before the crisis, such swap markets were the best developed money markets in East Asia.
Thus deposit-taking from banks can be ranked above short-term foreign exchange swaps.
The issuance of any central bank liability to market participants is seen as inferior to
“overfunding”, which the next section outlines.

7
The RBI (2004b) put forward another argument against the use of foreign exchange swaps, namely that “forex
sold by the Reserve Bank through swaps has been used by the market for extending forex loans to customers
for meeting rupee expenditure. To the extent that such loans are not hedged, the forex finds its way back into
the reserves of the Reserve Bank attenuating the efficacy of swaps as a sterilisation instrument”.

84 BIS Papers No 30
2. Overfunding the fiscal deficit to transform central bank debt
To unify the domestic public bond market, the government can “overfund” its own fiscal
needs in order to replace debt issued by the central bank to market participants. First, the
government sells more debt than it needs to finance any deficit and to roll over maturing
issues (overfunding). This produces a cash surplus that the government places on deposit
with the central bank, thereby draining bank reserves. The central bank is then in a position
to pay off its maturing obligations to market participants, thereby reinjecting bank reserves.
From the standpoint of the private sector, this would essentially mean a swap of claims on
the central bank for claims on the government. The case shown in Table 1 entails an
overfunding of sufficient scale to permit the central bank to buy some government securities
outright for further use in monetary operations.

Table 1
Mechanics of overfunding and refunding

Government overfunds its deficits and places the proceeds on deposit with the central bank

Assets Liabilities

+ deposits due from central bank + government securities

Central bank shifts its liabilities from market participants to the government

Assets Liabilities

Foreign exchange reserves Monetary base

+ government securities + deposits due to government
− central bank debt to market

Singapore has recently engaged in such an operation. In order to develop its bond market,
the Singapore government more than doubled its outstanding government securities, thereby
raising the outstanding stock to 39% of GDP at end-2001, despite fiscal surpluses (see Lian
(2002, p 184)). In fiscal 2001/02 and 2002/03, deposits placed by the government with the
MAS grew by SGD 21.7 billion, mainly reflecting “the proceeds from the larger issuance of
Singapore Government Securities through the [Monetary] Authority to the public and the
Central Provident Fund Board”. 8 This allowed “provisions and other liabilities” to fall by
SGD 10.9 billion over the two years, “due largely to the reduction in the Authority’s
borrowings from banks as part of its money market operations”. At the same time, holdings of
Singapore government securities (SGSs) by the MAS rose by SGD 118 million. “The
increase was in line with the Authority’s policy to build up its portfolio of SGSs for more active
use in repurchase transactions as part of its money market operations.” These transactions
implied the changes shown in Table 2.

8
This and the following citations are from MAS (2002, 2003, p 62 and p 84, respectively).

BIS Papers No 30 85
 Table 2
Selected changes to the Monetary Authority
of Singapore’s balance sheet, 2001/02-2002/03
In millions of Singapore dollars

Assets Liabilities

Foreign assets1 +23,967 Provisions and other liabilities −10,866

(“largely … borrowings from banks”)
Singapore government securities +118 Deposits of Singapore government +21,699
1
Includes SGD 15,986 million from the merger of the Currency Fund on 1 October 2002.

The authorities in India decided not to issue a central bank security - which would have
required a change in the RBI’s legislation - in favour of overfunding from the outset.
Reflecting its assessment of the balance of the arguments laid out below, the RBI persuaded
the government and parliament to accept selling more government paper than needed to
satisfy the public sector borrowing requirement and to place the proceeds in a non-interest
bearing blocked account at the RBI. This decision took effect in April 2004, more or less just
as the RBI ran out of government securities available for draining operations.
Despite the call in RBI (2004a) for an inframarginal instrument to sterilise surplus liquidity of
an “enduring” nature, it was easy to imagine that the new Market Stabilisation Scheme would
operate marginally. That is, as the RBI intervened and acquired further foreign exchange,
additions would be made to auctions of government bills or bonds, with the proceeds placed
in a blocked account at the RBI. The contrast between the top and bottom panels of Table 3
covering selected changes in the RBI’s balance sheet in the first and second quarters of
calendar 2004, respectively, is consistent with this interpretation. Foreign asset growth
slowed, but remained substantial. Whereas, in the first quarter of calendar 2004, reverse
repos and outright sales of Indian government securities did the heavy lifting, in the second
quarter the deposits under the Scheme took over. Instead of selling government securities
outright or on a reversed basis, the RBI received deposits from the government, which in turn
was funding the deposits with additional sales of its securities.
Events during the second quarter, however, showed that the Scheme was operating
increasingly inframarginally. As capital flows reversed starting in May, the RBI began net
sales of dollars that continued to July (RBI (2004d, p 79)). As a result, there was no need for
sterilisation at the margin. Still, additions to the Scheme put the RBI in a position to allow
reverse repos to run off. In effect, the Scheme came to be used to rebuild the RBI’s stock of
government securities that can be used to absorb liquidity in the future. Issuance of
government securities continued under the Scheme into the third quarter, and by the middle
of the quarter (14 August 2004) INR 464.8 billion had been raised. 9

9
Because government deposits are invested in government securities held in the RBI’s portfolio, the reduction
in regular deposits by the Indian government shown in the lower panel of Table 3 also released government
securities. Thus, the government deposits under the Scheme have released government securities for use for
future absorption through two channels.

86 BIS Papers No 30
 Table 3
Selected changes to the Reserve Bank of India’s balance sheet
In billions of Indian rupees

27 December 2003-27 March 2004

Assets Liabilities

Foreign assets +461.71 Reverse repos +319.10

Indian government securities –53.32 Deposits of Indian government –66.85

27 March-25 June 2004

Assets Liabilities

Foreign assets +349.71 Reverse repos –0.35

Indian government securities –4.29 Deposits of Indian government –185.77
Deposits of Indian government
under Market Stabilisation
Scheme +378.12
Source: RBI (2004e), p 128.

3. Benefits to the bond market and monetary operations

Significant benefits could be gained from the overfunding operation described in the previous
section. The main benefit arises from the increased liquidity in the secondary market that
could be fostered by consolidating all the public debt. In some Asian economies, the increase
in the size of the government bond market could be significant, representing growth of
anything from 137 to 222%. In aggregate, the five markets considered could be $220 billion
larger. In the next section, the general advantage that size provides for liquidity is elaborated.
Measures are then offered for how much larger regional bond markets could be were central
bank debt to be transformed into government debt.

3.1 Size and liquidity in government bond markets 10

The relationship between the size and liquidity of government bond markets is complicated
by the fact that size has several dimensions. In dealer markets, liquidity is generally supplied
by market-makers, who not only provide quotes but also take positions. How far size matters
for liquidity thus hinges on the various economies of scale in market-making. The size of
individual issues matters and debt managers can attain larger sizes by concentrating
issuance in fewer maturities, holding auctions less frequently or reopening issues, and
buying back illiquid issues. In addition, the overall size of the market matters. One economy
of scale arises from market-makers’ assembling information about the future path of interest
rates. The cost of this in a $500 billion government bond market is not likely to be 10 times its

10
This section draws on McCauley and Remolona (2000) and Jiang and McCauley (2004).

BIS Papers No 30 87
cost in a $50 billion bond market. Similarly, if the extraction of information from order flows
entails economies of scale, then overall trading activity may also matter.
The evidence from G10 bond markets suggests that size does make a difference to the
liquidity of government bond markets (Graph 1), though it is not the only factor of
importance. 11 The larger the outstanding stock of publicly issued central government debt,
the higher the turnover in cash and futures trading. And the higher the turnover, the better
the liquidity as measured by the tightness of the bid-ask spread. 12 Nevertheless, other factors
also play a role. These include: holdings by government accounts and other “buy and hold”
investors; the concentration of outstanding debt in benchmark issues; the industrial
organisation of the dealers and construction of trading platforms; taxes; arrangements for
sale and repurchase; and the efficiency of clearing and settlement systems (CGFS (1999b)).

Graph 1
Size and liquidity

Log of turnover value Bid-ask spreads (in basis points)

6 SE 15

US 5 12
DE JP
FR CH FR
CA IT 4 9
SE GB
JP
BE 3 CA 6
BE IT
CH GB DE
2 US 3

1
0
1.0 1.5 2.0 2.5 3.0 3.5 4.0
1 2 3 4 5 6

Source: H Inoue, The structure of government securities markets in G10 countries: summary of
questionnaire results, in CGFS (1999a).

Size matters for liquidity in Asia (Graph 2, upper panels). A larger market tends to be
associated with higher trading volumes (both variables are in logs), which are in turn
associated with tighter bid-ask spreads. This is similar to (although somewhat weaker than)
the relationship between size, turnover and liquidity observed in G10 government bond
markets and ascribed to economies of scale in market-making.
Using the existence of an active government bond futures market as well as bid-ask spreads
in G10 markets, McCauley and Remolona (2000) suggest that the critical size for a liquid
market is around $100-200 billion. In Asia, China and India have crossed this threshold, and
Korea and Taiwan are approaching it. Australia’s experience, however, suggests that, under
the right circumstances, liquid government bond cash and futures markets can both be
sustained at a much smaller size (Australia (2003)). Equally, though, the $100-200 billion
threshold may be too low under less favourable circumstances.

11
See CGFS (1999a).
12
The bid-ask spread measures only one dimension of liquidity, since it does not capture market depth or
resilience in respect of absorbing large orders. See CGFS (1999a,b) for a discussion.

88 BIS Papers No 30
 Graph 2
Liquidity in East Asian bond markets
Size, trading, issue size and concentration

Size (x-axis) and Trading volume (x-axis)1

trading volume (y-axis)1 and bid-ask spread (y-axis)2

y = 0.807x + 0.956 y = −2.7866x + 8.8401 7

ID
R = 0.2372 3.5 R = 0.7567
TW KR 6
3
5
SG IN 2.5 TH 4

HK HK 3
PH 2
MY MY
2
TH 1.5
SG KR
ID 1
IN TW
1 0
0 0.5 1 1.5 2 2.5 3 1 2 3 4

Issue size (x-axis)1 and Concentration of bond holdings (x-axis)3

bid-ask spread (y-axis)2 and bid-ask spreads (y-axis)2

TH
3.5 y = 8.6963x − 1.7479 ID 7
MY R = 0.6001
HK 3 6

y = −0.9816x + 3.3138 2.5 5

R = 0.4099
2
TH 4
1.5 MY HK
3
KR
1
2
SG IN
0.5 SG
TW KR 1
TW
0 IN
0
0 1 2 3 4
0 0.2 0.4 0.6 0.8 1
1 2 3
In billions of US dollars; in logs. In basis points. Herfindahl-Hirschman index.
Sources: Barclays Capital; Bloomberg; Deutsche Bank; HSBC; BIS calculations.

As noted by RBI (2004c, p 416), “issuance of central bank securities can fragment the debt
market due to availability of two competing sovereign issues, one of the Central Government
and the other of the Reserve Bank”. The cost of such fragmentation can be illustrated by the
case of Korea. The Bank of Korea sells monetary stabilisation bonds of a maturity of up to
two years, while the government’s treasury bonds extend out to five or 10 years. Where the
two debt programmes overlap, for instance at the one-year or two-year maturity, the yields
are generally identical. 13 In this case, at least, the fear expressed in RBI (2004c) that two
sovereign issuers could produce two separate yield curves does not seem justified. Another

13
I am indebted to Kyunjik Lee for pointing out to me that the data from the Korea Money Broker Corp and
Korea Securities Dealers Association quoted by Reuters are inconsistent or otherwise erroneously suggest
minor differences in yields on the two public bonds.

BIS Papers No 30 89
observation, however, does suggest a possible loss of liquidity from two sovereign issuers.
The yield on the very liquid three-year bond, which is served by a successful futures contract,
is often below or about the same as the yield on the two-year monetary stabilisation bond
(Graph 3). Confronted by such a strong demand for a benchmark issue, a single debt
manager might well issue more three-year bonds and fewer two-year bonds. If the single
debt manager did not want to extend the duration of the debt by selling more three-year
bonds, then a “barbell” of issuance - more one- and three-year paper and less two-year
paper - would better satisfy market demand and thereby reduce financing costs, given the
yields shown in Graph 3.

Graph 3
Yields on public obligations in Korea
Selected dates in 2003; in percentages

Monetary stabilisation bonds: 5 June

Treasury bonds: 5 June 5.0
Monetary stabilisation bonds: 6 November
Treasury bonds: 6 November
4.8

4.6

4.4

4.2

4.0
One-year Two-year Three-year Four-year Five-year

Sources: Korea Money Broker Corp; Korea Securities Dealers Association.

The case of Korea suggests that transforming central bank debt into debt of longer maturity
might be particularly advantageous in that it would allow greater issuance at longer
benchmark maturities. But it also suggests that market functioning would be improved even if
government debt simply replaced central bank debt at the shorter maturities characteristic of
the latter.

3.2 Prospective increase in the size of government bond markets in East Asia
How much of a difference would the transformation of central bank debt into government
debt make to the government bond markets in East Asia? The answer varies across the
region. The potential stock of government debt would be a third as high again as its current
level in Indonesia, half as high again as its current level in Malaysia and Thailand (Graph 4),
and more than twice its current level in Korea and Taiwan (Table 4 and Graph 5). This could
make a substantial difference to liquidity. For instance, Malaysia’s bond market is dominated
by such buy and hold investors as the provident fund (see Harun (2002)). Were the level of
government debt to rise by 50%, a significant amount of this debt might be available for
trading by more active accounts.

90 BIS Papers No 30
 Graph 4
Outstanding public debt in three Southeast Asian economies
In billions of domestic currency1

Indonesia Malaysia Thailand

Bank Indonesia certificates Deposits in central bank Bank of Thailand bonds

750 BOT net forex forward² 1,750
Government bonds less statutory reserves
200 BOT net repo
Central bank bills/bonds Government bonds
Government securities 1,500
600
160
1,250
450
120 1,000

300 750
80
500
150 40
250

0 0 0
Oct 99 Oct 00 Oct 01 Oct 02 92 94 96 98 00 02 99 00 01 02 03
1 2
For Indonesia, in trillions. Truncated at zero between February 2001 and December 2002.
Sources: CEIC; national data.

Table 4
Potential increase in size of government bond markets

Government bond market

Central bank debt to market Memo: Size
size
of combined
market as a
Domestic Domestic
$ billion $ billion % of current
currency currency

Indonesia1 398.2 trillion 47.1 147.3 trillion 17.3 137

2
Korea 81.5 trillion 69.0 98.9 trillion 84.0 222
3
Malaysia 109.6 billion 28.8 77.9 billion 20.5 171
Taiwan, China1 2.5 trillion 73.5 2.8 trillion 82.4 212
2
Thailand 1,269.5 billion 30.5 648.0 billion 15.6 151

Total 248.9 219.8 188

Note: Central bank debt: for Indonesia, Bank Indonesia certificates, August 2003; for Korea, monetary
stabilisation bonds (MSBs), August 2003; for Malaysia, Bank Negara Malaysia bills/bonds and net deposits of
banks, finance companies and merchant banks with Bank Negara Malaysia other than statutory reserves,
September 2003; for Taiwan (China), negotiable certificates of deposit (NCDs), September 2003; for Thailand,
net borrowing under repo from banks and other financial institutions, Bank of Thailand net forward sales of
baht, and Bank of Thailand bonds, August 2003.
1 2 3
End-September 2003. End-July 2003. End-December 2002.
Sources: CEIC; national data.

BIS Papers No 30 91
 Graph 5
Outstanding public debt in two Northeast Asian economies
In billions of domestic currency1
Korea Taiwan, China

BOK monetary stabilisation bonds CBC negotiable CDs 5,000

Treasury forex stabilisation bonds Treasury bills
Treasury bonds Treasury bonds
150 4,000

3,000
100

2,000

50
1,000

0 0
Jul 97 Jul 99 Jul 01 Jul 03 Sep 94 Sep 97 Sep 00 Sep 03
1
For Korea, in trillions.
Sources: Central Bank of China (CBC); CEIC; national data.

3.3 Monetary policy operations and the repo market

Three related advantages pertaining to monetary operations would arise from the
transformation of central bank liabilities into explicit government debt. Such a step would help
rebalance monetary operations, would allow the central bank to engage in reversed
transactions against government bonds and would thereby help to further develop the bond
market.
First, the central bank could have a firmer influence over short-term rates if the structural
balance in the money market could be shifted from structural surplus to deficit. At present,
redemptions of maturing central bank debt and interest payments on such debt represent
predetermined injections of bank reserves that must be offset by active draining operations.
Sufficient overfunding, and the stability of the government’s deposit with the central bank,
could leave the money market structurally short of funds and therefore dependent on regular
injections of reserves by the central bank. While it is not technically necessary for effective
policy implementation, most central bankers instinctively prefer a situation where market
participants need to come to the central bank for funding.
Second, sufficient overfunding could allow the central bank to hold a substantial stock of
government paper. This would permit it to carry out reversed transactions (repos and reverse
repos) against government securities, either to drain or to inject bank reserves. Moreover, to
the extent that the central bank can encourage the development of a repo market, not only
for its own operations but also among market participants themselves, it would lead the
banking system away from outright and towards collateralised interbank transactions. This
can enhance counterparty risk management.
Third, development of a deep and liquid repo market that benefits from central bank
operations is conducive to the increased depth and liquidity of the government bond market
more generally. Short positions become easier to fund and smaller securities firms find it
easier to finance themselves. This would contribute to a broadening of the dealer market and
more active trading.

92 BIS Papers No 30
4. Issues to be resolved
Five practical issues need to be resolved before central bank debt can be transformed into
government debt. The first two relate to whether the policy is advisable or politically feasible,
and the other three relate to aspects of the implementation of the policy. The first challenge
is to overcome the natural reluctance of finance ministers to increase outstanding debt for
which they are explicitly responsible. Second is the question of whether the rating agencies
would take a dimmer view of the fiscal position. If these deal-breaker issues can be resolved,
then three implementation issues must be faced. What yield should the central bank pay on
the government deposit? What is the maturity profile and duration of the government
securities to be issued? Finally, how can the uncertain scale of sterilisation needs in a year
be reconciled with the budgetary process?

4.1 The reluctance of the ministry of finance

Of all the practical issues, probably the greatest impediment to the consolidation of central
bank and government debt is the unwillingness of finance ministers to increase outstanding
debt for which they are explicitly responsible. 14 It is a rare finance minister who leaves office
bragging, as did Premier Zhu Ronji, of the government debt he has left as a legacy. A
finance minister may fear that a proposal to consolidate public debts would seem to make a
virtue out of more government debt, and thereby open the door to further spending or tax
cutting.
The legislature, for its part, may distrust the argument that the increase in public debt will
have as its counterpart a deposit at the central bank. This may seem an unstable bargain,
with the government then being able to draw down the deposit at will to meet some
unanticipated need without having to go to the legislature to authorise an increase in debt. 15
The discussion below of rating agencies suggests, however, that market discipline
substitutes in some measure for the legal discipline of setting debt ceilings.
In India, comfort is taken from the fact that it would require an act of parliament to authorise
the government to spend the funds placed in the blocked account at the RBI. Furthermore,
the experience of provincial governments placing funds in blocked accounts at the RBI also
supports the expectation that the bargain will prove a stable one.
The different choices of sterilisation instrument in China and India reflect the different
attitudes of the two ministries of finance. The Chinese ministry of finance has given evidence
of a strong aversion to issuing debt beyond the needs of its immediate deficits: after an initial
recapitalisation of the state-owned commercial banks with an addition to explicit government

14
Note that the strictures that have evolved against central banks’ making advances to governments (as
opposed to buying government debt in the market) do not apply to the reverse case of governments making
deposits in the central bank. These strictures attempt to keep monetary policy from becoming subservient to
the needs of the government. No such issue is raised by the government placing deposits with the central
bank. Were a government displeased with a monetary policy choice, it might threaten to withdraw its deposits.
But so long as the central bank had other tools to drain the resultant increase in bank reserves held in reserve,
this threat would not impinge on monetary policy or compromise central bank independence.
15
Adam Smith (1937), in his chapter on the public debt, observed that, in the happy case in which taxes
earmarked to service a debt proved excessive (generally owing to the reduction in interest on the debt), they
were often paid into a sinking fund intended to pay off debt. Such a “fund is almost always applied to other
purposes”, however: “During the most profound peace, various events occur which require an extraordinary
expense, and government finds it always more convenient to defray this expense by misapplying the sinking
fund than by imposing a new tax. Every new tax is immediately felt more or less by the people. It occasions
always some murmur and meets with some opposition… To borrow from the sinking fund is always an
obvious and easy expedient for getting out of the present difficulty.”

BIS Papers No 30 93
debt, subsequent moves to recapitalise the Chinese banks have taken place off-budget. 16
The irony in the contrast between the unwillingness of the Chinese ministry of finance and
the willingness of the Indian ministry of finance is that the reported debt position of the
central Chinese government is among the healthier in Asia, whereas that of the Indian
government is among the least healthy; at the same time, the Chinese government enjoys an
investment grade rating, whereas the Indian government labours under a speculative rating.
As long as Singapore was the outstanding example of overfunding the public sector
borrowing requirement in order to sterilise foreign exchange holdings, policymakers could
conclude that this option is open only to governments with the strongest debt positions and
ratings. Contrary to this conclusion, the government of the large country with the weaker
fiscal accounts has agreed to overfund.
Perhaps another reason for the difference between the Chinese and Indian cases is the
central bank’s relationship to the market, and market participants’ involvement in the larger
political process. The RBI’s consultative process allowed market participants an opportunity
to weigh in and to steer policy in a direction conducive to the development of a broad, deep
and liquid government bond market. 17

4.2 The reaction of the rating agencies

A second issue is whether rating agencies might see the larger gross stock of government
debt as a negative for the sovereign rating. This would happen if the rating agencies looked
strictly at the reported gross debt of the government. 18 In contrast, they would be well
advised to put more emphasis on a net concept, recognising that the government’s deposits
with the central bank (and ultimately the foreign exchange reserves) are assets to be
accounted for. 19

16
This includes the purchase of non-performing loans at par by the asset management companies (Ma and
Fung (2002)), the injection of foreign exchange reserves by the PBOC into the Bank of China and China
Construction Bank, the recent capital injection into the Bank of Commerce, and the use of PBOC bills to buy
non-performing loans from the big banks. Also consistent with the ministry of finance’s aversion to additions to
its explicit debt was resistance to PBOC proposals to turn its claims on distressed financial institutions into
government securities.
17
The RBI formed an internal review group to study the choice of sterilisation instruments. Drawn from the
departments responsible for internal debt, government accounts, monetary policy, economic research, foreign
exchange reserve management and legal affairs, it considered the various options and reviewed the
experience in a number of countries, including China, Korea, Malaysia and Thailand. The group held
discussions with market participants during the review and received written comments after the report was
posted on the RBI website in December 2003 (RBI (2004c,d)). At the time, the Indian financial press featured
well informed commentary on the issue and alternatives. With respect to the PBOC issuance of bills, by
contrast, after-the-fact commentary by market economists rarely addressed the merits of the sale of PBOC
bills or other feasible alternatives. Rather, the focus was put on whether this sterilisation tool would work and
thus whether the pegged exchange rate would hold.
18
The ratios that, say, Standard & Poor’s examines suggest that the operation described above would not have
a significant implication for the assessment. Three out of four fiscal debt concepts in Standard & Poor’s
glossary would not seem to be affected by a change in the locus of financing of the foreign exchange
reserves. The general government debt is a broad aggregate across the public sector that would include the
central bank’s debt. The two net debt aggregates vary in netting out cash, deposits, loans and equity holdings
or, more restrictively, only cash and deposits. Either one should be unaffected both because of the breadth of
the concept of the government and because of the netting. Finally, the central government’s gross debt is
included in the gross debt concept, and this one could well be increased by the overfunding proposed. It
should be noted, however, that this narrow gross concept is last on the list, no doubt because it is the least
comparable or the most manipulable because of its non-inclusion of “non-commerical off-budget and quasi-
fiscal activities” included in the general government concept.
19
If one believes that the rating agencies take no heed of what lies behind government debt, one is led to an
absurd result in the case of Japan. Consider the case in which the ministry of finance sold the bulk of Japan’s

94 BIS Papers No 30
Asked whether the use of government securities to sterilise foreign exchange holdings would
result in a downgrading, representatives of both Moody’s and Fitch pointed to the case of
Singapore. The additions to its government debt described above in Section 2 did not result
in downgradings. Indeed, in the discussion of this paper at the BIS/Korea University
conference, Tom Byrne of Moody’s noted that Singapore received an upgrade even as its
government debt increased. At another conference a month earlier in Seoul, Brian Coulton,
Senior Director of Fitch Ratings in Hong Kong, held that the rating agencies would not
mechanically react to overfunding. 20

4.3 The return on the government deposit

The third issue is what yield the central bank should pay on the government deposit:
available models include profit-sharing, fixed returns and a zero return. The Hong Kong
Exchange Fund shares its profits with the government in proportion to most of the
government’s direct claim, although one small deposit is still serviced at interbank rates
(HKMA (2003)). 21 By contrast, the Reserve Bank of Australia pays interest on the
government deposit with it in line with yields on Australian government paper.
One of the most surprising aspects of the Indian Market Stabilisation Scheme is the absence
of any direct return paid by the RBI to the government on its blocked account. The RBI
(2004c) argued against issuance of interest-paying central bank paper to finance foreign
exchange holdings precisely because such paper could result in substantial central bank
losses. Given the scale of the Indian government’s debt, the RBI reasoned, a recapitalisation
could not be presumed. If no fixed return were paid to the government, losses would be less
likely. At the same time, the RBI argued, the government owns the RBI and has a claim to all
its profits. 22 In a sense, the existing profit-sharing norm was successfully appealed to.

4.4 The duration of the additional government securities

A fourth issue is the choice of duration of the government securities used to finance foreign
exchange reserves. This choice should be considered along with the choice of duration of
the international foreign exchange reserve holdings. A central bank that considers the inflows
that have built up the foreign exchange reserves as temporary and thus invests them in
short-term securities might sensibly finance the reserves with short-term instruments as well.
Conversely, a central bank that regards the reserves, or a portion of them, as stable parts of
the national balance sheet might finance with longer-term instruments.
Central banks in Asia hold down the carrying costs of foreign exchange holdings by financing
an increasingly medium-term portfolio of reserve assets with short-term liabilities. Only in the
cases of Korea and Taiwan does the sterilisation debt extend out to two years. Meanwhile

foreign exchange reserves to the Bank of Japan. In this case, government debt would fall by 10-20% of GDP.
Would the rating agencies upgrade the Japanese government under these circumstances?
20
His text (Coulton (2004, p 3)) reads: “While Korea’s government debt - including guaranteed bonds issued by
KAMCO and KDIC - is in line with its ‘A’ rating peers at 40% of GDP, the prospect of continued fiscal prudence
bodes well for a declining public debt ratio in the next few years, notwithstanding increased issuance of foreign
exchange stabilisation bonds by the Ministry of Finance and Economy (MOFE) to finance foreign exchange
market intervention” [emphasis added].
21
The Bank of Thailand would require legislation to enable it to remunerate government deposits. For practice
across industrial countries, see Borio (1997, pp 60-2).
22
Notwithstanding this, it is said that a disagreement between the Korean ministry of finance and the Bank of
Korea in the early 1990s over the proper rate of return ultimately undid an arrangement whereby the Bank of
Korea issued government debt as a sterilisation instrument.

BIS Papers No 30 95
the duration of reserve portfolios has moved out to the two- to five-year range (McCauley and
Fung (2003)). If the duration of the domestic currency financing portfolio were lengthened
and it were desired to maintain the longer duration of the reserve portfolio, the latter could be
lengthened.
The Indian authorities have tended to view the inflows that have led to the growth of foreign
exchange reserves as not very stable and have thus tended to finance the reserves with
short-term securities. 23 In principle, the Indian government can sell either bills or coupon
securities to fund its blocked account at the RBI. In practice, the majority of the issuance has
taken the form of treasury bills. 24 In the event, this issuance has been very well received by
the market owing to the previous scarcity of such paper. In the past, the Indian government
had prudently avoided selling much short-term paper out of concern for the rollover risk,
given the large government debt. Given the blocked account, however, such a concern for
the rollover of maturing paper seemed no longer relevant. 25

4.5 Reconciliation with the budgetary policy

A final issue is how to reconcile the uncertain extent of sterilisation needs in a year with the
budgetary process or how to resolve the related tension between the time variation of
sterilisation needs and the predictable issuance aimed for by many government debt
managers. The problem calls for flexibility, the solution for predictability. Joseph Yam, Chief
Executive of the Hong Kong Monetary Authority (2003), drew attention to this tension and
offered a proposal:
Flexibility in deciding not until, say, a week before what maturity of paper is to be
issued and what amount is of course helpful, particularly when sterilisation is the
more pressing objective. But the market would like as much information on the
central bank bills programme as possible in order to plan ahead, in terms for
example of managing maturity mismatches. The thing to do may be to fix the
weekly programme and to fine-tune monetary conditions through rather more
frequent money market operations of the type carried out daily by other central
banks…
The difficulty of reconciling the budgetary cycle and unpredictable intervention was recently
illustrated in Japan, where foreign exchange reserves are financed at the margin by
government debt issues. At the turn of this year, the ministry of finance ran short of
authorised debt to finance its massive intervention. In the case of India, a certain amount of
overfunding was authorised by parliament for the current fiscal year, with an understanding
that it might not prove sufficient.
In practice, both the Singaporean and Indian policies to use government debt for sterilisation
had an inframarginal rather than marginal character. That is, in practice, government debt
issues and related deposits by the government in the central bank took weight off other
means of sterilisation. In this spirit, one can imagine using overfunding in an after-the-fact
manner. That is, other instruments, including perhaps central bank debt, might be used to
sterilise in the first instance, and subsequently government debt might be substituted the

23
RBI (2004b, p 291-2) shows that two of the biggest sources of the reserve build-up were non-resident Indian
deposits and foreign institutional investor purchases of Indian equities.
24
RBI (2004e, p 131) reports that “the total amount raised under the MSS [Market Stabilisation Scheme]
amounted to Rs 46,480 crore [464.8 billion rupees] by 14 August 2004, inclusive of Rs 20,000 crore raised
through dated securities of residual maturity of up to 2.5 years”.
25
If the government repaid maturing short-term debt with funds drawn from the blocked account, the RBI might
need to drain bank reserves, so that the liquidity risk is not so much absent as transferred.

96 BIS Papers No 30
next budget year. Such an arrangement would leave the central bank able to intervene and
to sterilise without immediate assistance from the fiscal authorities, an important matter
where the central bank controls intervention policy. At the same time, the eventual
involvement of the finance ministry and parliament in the financing of the foreign exchange
reserves could help ensure that the public sector at large, and not just the central bank,
knowingly takes on the foreign exchange risk.

5. Conclusions
If these issues can be resolved, then the central bank debt that has financed large holdings
of foreign exchange reserves could be consolidated with government debt. In particular,
issuing government debt beyond the need of the public sector borrowing requirement could
finance a government deposit with the central bank. This would allow a run-off of central
bank liabilities.
The benefits from lumping central bank liabilities into government debt are likely to be
substantial. Government bond markets could increase in size to anywhere from 137 to 222%
of their current size in East Asia. Properly handled, such an increase would make these bond
markets more liquid and thereby more attractive to investors.

References
Aizenmann, Joshua and Nancy Marion (2002): “The high demand for international reserves
in the Far East: what’s going on?”, NBER Working Papers, no 9266, October.
Asian-Pacific Economic Cooperation (1999): Compendium of sound practices: guidelines to
facilitate the development of domestic bond markets in APEC member economies,
September.
Australia, Commonwealth of (2003): Statement 7: budget funding, Budget for 2003-04,
Canberra, May.
Borio, Claudio E V (1997): “The implementation of monetary policy in industrial countries: a
survey”, BIS Economic Papers, no 47, July.
Committee on the Global Financial System (1999a): “Market liquidity: research findings and
selected policy implications”, CGFS Working Group Reports, no 11, May.
——— (1999b): “How should we design deep and liquid markets? The case of government
securities”, CGFS publications, no 13, October.
Coulton, Brian (2004): “Global perspectives on the Korean economy”, paper presented to the
International conference on a new vision and strategy under changing leadership in
Northeast Asia, sponsored by the Ministry of Finance and the Economy and the Korean
Development Institute, 27-28 February (www.newvision.go.kr/data/s1-2_coulton.doc).
Eichengreen, Barry and Pipat Luengnaruemitchai (2006): “Why doesn’t Asia have bigger
bond markets?”, this volume.
Harun, S (2002): “The development of debt markets in Malaysia”, in The development of
bond markets in emerging economies, BIS Papers, no 11, June, pp 147-50.
Hong Kong Monetary Authority (2003): Annual report 2002, Hong Kong: HKMA.
Jiang, Guorong and Robert N McCauley (2004): “Asian local currency bond markets”, BIS
Quarterly Review, June, pp 67-79.

BIS Papers No 30 97
Lian, T S (2002): “Debt market development in Singapore”, in The development of bond
markets in bond emerging economies, BIS Papers, no 11, June, pp 183-89.
Ma, Guonan and Ben S C Fung (2002): “China’s asset management companies”, BIS
Working Papers, no 115, August.
McCauley, Robert N (2002): “International market implications of declining Treasury debt”,
Journal of Money, Credit and Banking, vol 34, no 3, part 2, August, pp 952-66.
——— (2003): “Unifying the government bond markets of East Asia”, BIS Quarterly Review,
December, pp 89-98.
——— (2004): “Striking the balance: benefits and costs of foreign investment in Asian bond
markets”, paper presented to Korea Institute of Finance Korea Day seminar, Financial
cooperation in East Asia, Asian Development Bank, Jeju, Korea, 14 May 2004.
McCauley, Robert N and Ben S C Fung (2003): “Choosing instruments in managing dollar
foreign exchange reserves”, BIS Quarterly Review, March, pp 39-46.
McCauley, Robert N and Guorong Jiang (2004): “Diversifying with Asian local currency
bonds”, BIS Quarterly Review, September, pp 51-66.
McCauley, Robert N and Eli Remolona (2000): “Size and liquidity of government bond
markets”, BIS, International Banking and Financial Market Developments, November,
pp 52-8.
Monetary Authority of Singapore (2002): Annual Report 2001/2002, August.
——— (2003): Annual Report 2002/2003, August.
Oh, Junggun (2006): “Comments on McCauley’s paper ‘Consolidating the public debt
markets of Asia’”, this volume.
Reserve Bank of India (2004a): “Report of the Internal Group on Liquidity Adjustment
Facility”, Reserve Bank of India Bulletin, January, pp 23-68.
——— (2004b): “Report on foreign exchange reserves”, Reserve Bank of India Bulletin,
March, pp 289-99.
——— (2004c): “Report of the Working Group on Instruments of Sterilisation”, Reserve Bank
of India Bulletin, April, pp 387-421.
——— (2004d): “Report of the Working Group on Instruments of Sterilisation: suggestions
from market participants/experts and comments of the Reserve Bank of India”, Reserve Bank
of India Bulletin, April, pp 423-8.
——— (2004e): Annual Report, 2003-04.
Smith, Adam (1937): “Of public debts”, in The wealth of nations, New York, Modern Library,
book 5, chapter 3.
Yam, Joseph (2003): “Viewpoint: Monetary management developments on the mainland”,
15 May, www.info.gov.hk/hkma/eng/viewpt/index.htm.

98 BIS Papers No 30
 Comments on McCauley’s paper
“Consolidating the public debt markets of Asia”

Junggon Oh 1

The main arguments of the paper are as follows: dual mismatches of foreign borrowings,
ie currency and maturity mismatches, were important causes of the East Asian financial
crisis. Accordingly, it is necessary to develop the region’s bond markets. In particular,
unifying government bond markets and central bank debt markets may contribute to the
development of bond markets.
The benefits of unifying government bond markets and central bank debt markets are as
follows: an increase in liquidity in the secondary bond market through the development of the
repo market and thereby the development of the government bond market; and advantages
for monetary operations through greater influence on short-term rates with the help of a
reduction in the burden of redemptions of maturing debt and interest payments.

Outstanding amounts of government,

public and corporate bonds in Korea
End of period, in trillions of won

1995 1996 1997 1998 1999 2000 2001 2002 2003

Monetary
Stabilisation
Bonds 25.8 25.0 23.5 45.7 51.5 66.4 79.1 84.3 105.5
Government
bonds 23.3 25.7 28.6 41.6 61.2 71.2 82.4 98.3 135.8
(Treasury
bonds) 3.0 4.9 6.3 18.8 34.2 42.6 50.9 55.6 81.4
Corporate
bonds 61.0 76.0 90.1 122.7 119.7 133.6 154.4 180.0 187.4
Total 164.6 190.0 234.2 343.9 376.2 429.3 501.6 544.8 574.1

Looking at the table, these suggestions seem acceptable, but there are some practical
problems, which include: the difficulty of creating synergy effects from the consolidation of
two markets to reduce the liquidity premium; and the possibility of reducing the efficiency and
independence of monetary policy.
First, as relates to the difficulty of creating synergy effects from the consolidation, it should be
taken into account that the two debts have different characteristics. In particular, government
bonds have a relatively easy periodical issuance and are readily fungible, while central bank
debt is issued to offset changes in bank reserves due to autonomous factors such as flows of
government funds or changes in foreign exchange holdings.

1
Views expressed herein are those of the author and do not necessarily reflect those of the Bank of Korea.

BIS Papers No 30 99
Second, as relates to the possibility of reducing the efficiency and independence of monetary
policy, it should be pointed out that, in fact, it is difficult to reach an agreement on the issue of
government bonds for monetary stabilisation, and the practical procedures for reaching an
agreement on the issue from the parliament are complicated.
As an alternative, it may be suggested that bonds be issued by the central bank and interest
on them be paid by the government, as in Germany, New Zealand and Israel, etc. In
Germany, an issue of three-, six- and nine-month government bonds up to DM 25 billion was
decided by the Bundesbank, and in New Zealand, government bonds and central bank debt
of three-month maturities for monetary operations are used together for monetary
stabilisation, with the interest on the central bank debt paid by the government. More central
bank independence is a prerequisite to the implementation of these policies.
As another alternative in Korea, a more feasible step-by-step approach may be considered,
taking the current situation into account. As a first step, quasi-fiscal burdens of the Bank of
Korea due to aggregate credit ceiling loans, and the underwriting of non-performing asset
management fund bonds etc would be transferred to the government to reduce the issuance
of Monetary Stabilisation Bonds (MSBs). As a second step, interest on MSBs would be paid
by the government. And as a third step, Monetary Stabilisation Government Bonds (MSGBs)
would be substituted for MSBs. Amounts, time and conditions of the issue of MSGBs up to a
certain amount agreed on by the parliament would be decided by the BOK.
In order to introduce these approaches, it is important to have independence and
coordination between government bonds and MSBs, and monetary, fiscal and foreign
exchange rate policies. Step-by-step substitution of the government bonds for MSBs seems
more feasible and desirable, in line with the improvement of circumstances for the
independence of the central bank and of understanding about the use of government bonds
for monetary policy operation.

100 BIS Papers No 30

 Lunch address

Huhn-Gunn Ro

As the head of the central securities depository of Korea, I accept as a great honour and
pleasure the task of conveying a congratulatory message to this international conference on
Asian bond markets, which is being held under the joint auspices of the Institute of Northeast
Asian Business and Economics of Korea University and the Bank for International
Settlements.
In addition, I would like to thank all the prominent scholars and experts in the securities and
financial fields who are gracing this conference with their attendance. It really is a privilege
for the Korea Securities Depository (KSD) to support this wonderful conference, where
outstanding papers are being presented and heated discussions are being held on the
subject of our common interest, the nurture of Asian bond markets.
Also, I would like to take this opportunity to give my special thanks to Professors Yung-Chul
Park and Young-Sup Yun of Korea University and all the researchers at the Institute of
Northeast Asian Business and Economics for having spared no effort to successfully
organise this conference.
As you are well aware, one of the convincing arguments for the root cause of the 1997 East
Asia foreign exchange crisis is that East Asia unwittingly invited the crisis. It did so by
financing long-term investments with short-term liabilities and financing projects producing
domestic currency cash flows with foreign currency debt.
I believe it is encouraging that after the financial crisis we are giving serious consideration to
developing Asian bond markets and discussing this topic in earnest within the framework of
ASEAN+3 and APEC.
The further development of Asian bond markets is expected to bring about the following
three effects. First, it will retain Asia’s savings within the region, helping Asian corporations
easily finance long-term projects with long-term liabilities. Second, it will provide more
diverse investment vehicles to Asian investors. Finally, it will lay the groundwork for
enhanced economic cooperation among Asian countries and ultimately perhaps encourage
the evolution of an Asian key currency.
However, the prospects are not necessarily rosy. There are a lot of problems which need to
be addressed in order to nurture the Asian bond market. Decisions need to be made on
areas such as the issuance methods for Asian bonds, credit enhancement measures, and
development of the depository and settlement systems within the region, to name but a few.
Under the current circumstances, it is very significant that Asian countries have reached a
consensus on the importance of tackling the recognised problems and formed a working
group comprising experts from each Asian country.
I firmly believe that more developed Asian bond markets would make a strong contribution to
bringing each Asian country’s capital market to an international level. Increased financial
cooperation among Asian countries would be a welcome by-product of this effort.
This international conference provides a golden opportunity for us to consider the prospects
for and the tasks relating to the development of Asian bond markets and to seek the joint
prosperity of Asian countries on the firm foundation of stable financial systems.

BIS Papers No 30 101

 Determinants of bond holdings
by foreign investors

Kee-Hong Bae, Young Sup Yun 1 and Warren Bailey

1. Introduction
A key theme in restructuring economies in the developing world is opening local capital
markets to foreign portfolio investments. This can be accomplished by permitting foreign
investors to enter the local capital markets directly or by allowing local assets to trade in
overseas markets. In theory, this permits firms in developing economies to draw from the
global pool of capital to undertake useful investments that generate profits and employment.
Furthermore, the scrutiny of foreign investors, foreign analysts and foreign stock listing
standards can help resolve agency problems, effectively transmitting higher quality reporting
and governance standards to firms in developing countries (Obstfeld (1998), Stulz (1999)).
There is now much theory and empirical evidence to support the notion that foreign equity
capital flows are beneficial. One way foreign equity capital flows benefit local capital markets
is by causing a fall in the cost of equity capital because of increased risk sharing between
domestic and foreign agents. 2 This increased risk sharing reduces systematic risk, which in
turn reduces the cost of equity capital. 3 There is also increasing evidence that openness to
foreign portfolio investment enhances the governance of local corporations. Doidge (2003),
for example, reports evidence that cross-listing in the United States affords greater protection
to minority shareholders. More generally, the evidence in Kaminsky and Schmukler (2002)
suggests that equity market liberalisation tends to spur the process of institutional reform, not
the other way around.
In this paper, we examine what attracts foreign investors to the local bond markets. While we
have much evidence on the dynamics of foreign equity investment, there is little evidence on
foreign bond investment. 4 The issue is important in that liquidity is essential in order to build
up mature bond markets and foreign investors are crucial in building liquidity. Foreign
investors hold at least 20% of government bonds in markets as diverse as Canada, Sweden
and the United States. 5 For emerging markets that largely lack domestic institutional
investors such as mutual funds, pension organisations, insurance companies, etc, foreign
investors are likely to be even more important. They will not only provide demand but also
bring more varied investment objectives and thus provide liquidity.

1
We thank the discussant and conference participants for their comments.
2
There is also greater liquidity following increased capital inflows. Amihud and Mendelson (1986) and Amihud
et al (1997) discuss the effect of liquidity on equity risk premiums.
3
The effect of increased risk sharing on equity premiums is discussed in Stapleton and Subrahmanyan (1977),
Errunza and Losq (1985), Eun and Janakiramanan (1986), Alexander et al (1987), and Stulz (1999a,b).
Empirical evidence consistent with the risk sharing view of stock market liberalisation is provided in Henry
(2000) and Chari and Henry (2002).
4
See Bekaert and Harvey (2003) and the references therein.
5
See Exhibit 2 in Beckert and Pitsilis (2000).

102 BIS Papers No 30

In this paper, we specifically ask whether differences in property rights protection matter for
foreign bond holdings, after controlling for cross-country differences in macroeconomic
variables including GNP per capita, lending rates and/or inflation rates and exchange rates.
Countries differ considerably in terms of the efficiency with which they respect property rights
and enforce contract laws. For example, Sweden provides strong protection for private
property rights, but Argentina only weak protection. Recent research indicates that the extent
to which property rights are secure among countries shows several important differences in
financial systems. More secure property rights are associated with higher values of stock
markets (La Porta et al (1997)); a higher number of listed firms (La Porta et al (1997)); higher
valuation of listed firms relative to their assets (Claessens et al (2002), La Porta et al (2002));
greater use of external finance (La Porta et al (1997, 1998, 2000)); and greater investments
from external funds (Rajan and Zingales (1998), Demirgüç-Kunt and Maksimovic (1998)).
If property rights are weakly protected in emerging markets, which is indeed the case, foreign
investors will charge higher risk premiums to compensate for the additional risk of contract
repudiation, shorten the duration of bond maturity, or will shy away from these markets
completely. Therefore, in countries with poor property rights protection, foreign investors are
likely to make smaller investments.
To examine these predictions, we use data on bond portfolio holdings in 45 countries (out of
165 countries in our total sample) for which the rule of law and property rights protection
quality can be identified. We estimate cross-country regressions in which the dependent
variables are local currency bonds held by foreign investors. The key variable of interest on
the right-hand side is an index of property rights protection. We associate countries with high
levels of corruption, higher risk of expropriation and greater uncertainty about contract
enforcement as countries with poor property rights protection. The regressions control for
other possible macroeconomic variables that may also affect the foreign bond holdings.
Our results show that in countries with better property rights protection, foreign investors buy
and hold more local bonds. Our results suggest that more secure property rights are
important in developing liquid and mature bond markets. To the extent that foreign investors
play a vital role in providing liquidity, our evidence shows that improving property rights
protection is a matter of prime importance.
The rest of the paper is organised as follows. Section 2 presents our data and variables.
Section 3 gives the summary statistics. Section 4 discusses the main empirical results, and
Section 5 concludes.

2. Data and variables

2.1 Basic bond holdings data

We obtain our basic data from the International Monetary Fund website. The most recent
data show that at end-2001 or end-2002, 67 countries made portfolio investments in foreign
equity and debt securities and 236 countries received the investments. 6 Since our objective

6
The data are from the Coordinated Portfolio Investment Survey (CPIS) and includes 67 investing countries
(and bonds held by international organisations and as reserve assets) and 236 countries receiving investment
(other countries classified as “confidential” and “unallocated” and international organisations). For the
purposes of the survey, long-term debt securities include bonds, debentures and notes with an original
maturity of more than one year. Short-term debt securities cover treasury bills, commercial paper and bankers’
acceptances with an original maturity of one year or less.

BIS Papers No 30 103

is to investigate the determinants of bond holdings by foreign investors, we only examine the
long-term and short-term debt securities data. From the IMF sample of 236 countries, we
exclude 71 countries that had investments of less than US$ 1 million in either 2001 or 2002.
This restriction results in a sample of 67 investor countries investing in 165 recipient
countries (in a matrix form), both in long- and short-term debt securities. For the regression
analysis, we use data on bond portfolio holdings of 45 countries for which property rights
protection quality can be identified, out of 165 countries.
There is one point worth mentioning about our dataset. Our data source mixes bonds of
different currencies, especially local currency and US dollar. This might have the effect of
making our property rights variable less relevant in that the governing law of the bonds per
se will often be London or New York law rather than local country law. However, in a recent
paper, Siegel (2005) tests the functional convergence hypothesis, which states that foreign
firms can leapfrog their countries’ weak legal institutions by listing equities in New York and
voluntarily abiding by US securities law, and he shows that the US Securities and Exchange
Commission has rarely been effective in enforcing the law against any US-listed foreign firm.
In other words, the governing law of the country where the securities are issued or listed may
not be effective. Ideally, for our analysis we should use the local currency-denominated
bonds issued in a local country, but the data are not readily available.

2.2 Measuring property rights protection

To measure the extent to which a country respects private property rights, we focus on three
indexes from La Porta et al (1998). These three indexes measure corruption, the risk of
expropriation of private property and the risk that contracts may be repudiated.
La Porta et al (1998) describe these three indexes as follows: The “corruption index” is an
assessment of corruption in government by the International Country Risk Guide (ICRG). Low
scores indicate that “high government officials are likely to demand special payments” and
“illegal payments are generally expected throughout lower levels of government” in the form of
“bribes connected with import and export licenses, exchange controls, tax assessment, policy
protection, or loans”. The “risk of expropriation index” is the ICRG’s assessment of the risk of
“outright confiscation” or “forced nationalization”. The “index for the repudiation of contracts” is
the ICRG’s assessment of the “risk of modification in a contract taking the form of repudiation,
postponement or scaling down” due to “budget cutbacks, indigenization pressure, a change in
government or a change in government economic and social priorities”.
The range for each index is between zero and 10 with low values indicating less respect for
private property. All three ICRG indexes are averages from 1982 to 1995. Following Morck et
al (2000), we combine these three indexes into an additive index of property rights
protection. The index measures the extent to which a country’s legal systems and institutions
enforce all contracts, including government contracts.
We include the natural logarithm of GNP per capita and the level of lending rates in the
country. We include the latter in order to control for the differences in inflation rates and
sovereign risks that might also explain cross-country differences. Countries with high inflation
rates are subject to greater political risks, as there is a greater likelihood that their
governments will introduce wage and price controls or tamper with indexes. Higher inflation
rates can raise contracting costs for firms and their bank lenders. According to
Demirgüç-Kunt and Maksimovic (1999), high and/or variable rates of inflation make it costly
for investors and firms to contract. The cross-country macroeconomic data are from a
database compiled by the World Bank (and available on its website).

104 BIS Papers No 30

3. Descriptive statistics of the basic bond holdings data

3.1 Geographical breakdown

Tables 3.1 and 3.2 show geographical breakdowns of long-term and short-term paper
investments at end-2002. 7 The 67 investing countries in the sample are grouped into six
regions: East Asia, Europe, North America, Latin America, Africa/Middle East/Southeast Asia
and Tax Havens. 8 The 165 recipient countries are grouped into the six regions above plus
Other Nations.
The total amount of long-term bond investment is more than seven times larger than that of
short-term paper investment ($6.6 trillion vs $0.88 trillion). Europe is the region that makes
the largest investment in both long- and short-term paper (48.7% and 43.4%), followed by
Tax Havens (28.6% and 21.6%), East Asia (8.0% and 12.7%) and North America (7.8% and
18.1%). Europe is also the largest recipient of bond investment (57.2% and 58.2%), followed
by North America (23.0% and 27.8%), East Asia (5.4% and 6.1%) and Tax Havens (9.6%
and 5.7%).
Focusing on East Asia, we find that as an investor, it makes only a small proportion of total
bond investment in its own regional bonds (14.3% and 27.8%). On the other hand, East Asia
receives a much larger or the largest proportion of bond investment from the region (21.6%
and 58.0%). In particular, more than half of East Asian short-term paper investment comes
from the region. This might suggest that due to the weak property rights enforced in East
Asia, foreign investors are likely to shorten the duration of bond maturity to minimise the risk
of contract repudiation. This might also suggest that there is a regional bias in bond holdings
by foreign investors, similar to the much-documented home bias found in stock investment. 9

3.2 East Asian country breakdown

Tables 3.3 and 3.4 show East Asian country breakdowns of long-term and short-term paper
investments. There are three dominant countries that invest in East Asian bonds: Hong Kong
SAR, Japan and Singapore. Together they make up more than 90% of the total investment
East Asia makes in its own region. However, the profile of countries receiving bond
investment is quite different. Australia, Japan, Korea, Singapore, Malaysia and New Zealand
receive the majority of the investment. However, it is interesting to note that Australia, which
attracts the largest investment (42.7% and 53.6%) from East Asia, makes an insignificant or
zero amount of bond investment in other East Asian countries. 10

3.3 Top 10 country breakdown

Tables 3.5 and 3.6 show the top 10 countries’ bond investments in seven different regions.
Japan, the United Kingdom, the United States, France and Luxembourg appear to be the
most important buyers. As the group of largest investors, the top 10 countries invest mostly

7
For summary statistics, we only report results for end-2002, because those for end-2001 are similar. We use
the terms long-term and short-term bonds or paper for long-term and short-term debt securities.
8
We use the list of tax havens compiled by EscapeArtist Inc which can be obtained from the website
EscapeArtist.com.
9
For more on the regional bias issue, please also refer to McCauley and Park (2006).
10
This pattern appeared for New Zealand for end-2001 which is not reported here. Considering Australia and
New Zealand are both common law countries, one might maintain that the rule of law argument is at work
here, but it is premature to do so without further analysis.

BIS Papers No 30 105

in Europe and North America, and East Asia receives about 4% of their total bond
investment. Japan’s investment in East Asian bonds also appears to be minimal, taking only
3-7% of its total investment.

3.4 Rule of law breakdown of bond investment

Tables 3.7 and 3.8 show rule of law breakdowns. We follow Stulz and Williamson (2003) who
classify 51 countries into four classes of rule of law: common law, civil/French, civil/German
and civil/Scandinavian. 11 The results show that the largest proportion of bond investment
goes to common law countries (38% and 63%) where investor protection is deemed
strongest. The second largest proportion goes to civil/French law countries (33% and 19%)
that are generally regarded to provide weaker investor protection.
One might speculate that in the common law countries, investors would invest relatively more
in long-term bonds than in short-term paper because investor protection tends to be strong in
these countries, whereas they would invest relatively more in short-term paper in civil/French
countries where investor protection tends to be weaker. However, no such evidence is found
from the descriptive statistics. It appears that more investment is made in common law
countries in both long-term and short-term paper. In the next section, we conduct a formal
analysis on how investor rights protection affects foreign bond holdings.

11
See Table 1 of Stulz and Williamson (2003).

106 BIS Papers No 30

 Table 3.1
Geographical breakdown of long-term
bond investment, end-2002
In millions of US dollars

Africa,
From East North Latin Tax
Europe Middle
Asia America America Havens Subtotal
(28) East and
In (10)1 (2) (7) (17)
Asia (3)

East Asia (14) 76,376 109,154 16,997 12,816 399 138,430 354,172
2
[21.56%] [30.82%] [4.80%] [3.62%] [0.12%] [39.09%] [100%]
{14.43%}3 {3.39%} {3.28%} {2.86%} {9.18%} {7.33%} {5.36%}

Europe (42) 266,466 1,839,011 391,025 327,960 934 958,945 3,784,341

[7.04%] [48.60%] [10.33%] [8.67%] [0.02%] [25.34%] [100%]
{50.35%} {57.16%} {75.44%} {73.08%} {21.51%} {50.75%} {57.28%}

North 141,327 789,129 83,098 92,070 2,061 411,900 1,519,585

America (2)
[9.30%] [51.93%] [5.47%] [6.06%] [0.14%] [27.11%] [100%]
{26.70%} {24.53%} {16.03%} {20.52%} {47.44%} {21.80%} {23.00%}

Latin 7,172 40,733 4,246 3,150 256 47,881 103,438

America (21)
[6.93%] [39.38%] [4.10%] [3.05%] [0.25%] [46.29%] [100%]
{1.35%} {1.27%} {0.82%} {0.70%} {5.88%} {2.53%} {1.57%}

Africa, Middle 1,085 12,266 1,051 1,666 5 8,226 24,299

East and
[4.47%] [50.48%] [4.33%] [6.85%] [0.02%] [33.85%] [100%]
Southeast
Asia (42) {0.21%} {0.38%} {0.20%} {0.37%} {0.11%} {0.44%} {0.37%}

Tax 31,201 384,443 21,728 9,553 689 187,129 634,743

Havens (44)
[4.92%] [60.57%] [3.42%] [1.51%] [0.11%] [29.48%] [100%]
{5.89%} {11.95%} {4.19%} {2.13%} {15.87%} {9.90%} {9.61%}

Other 5,616 42,357 148 1,526 – 136,945 186,592

Nations (73)
[3.01%] [22.70%] [0.08%] [0.82%] [0.00%] [73.39%] [100%]
{1.06%} {0.03%} {0.03%} {0.34%} {0.00%} {7.25%} {2.82%}

Subtotal 529,243 3,217,093 518,293 448,741 4,344 1,889,456 6,607,170

[8.01%] [48.69%] [7.84%] [6.79%] [0.07%] [28.60%] [100%]
{100%} {100%} {100%} {100%} {100%} {100%} {100%}
1
In ( ) next to the region’s name is the number of countries. 2 In [ ] is the percentage when the IN country
subtotal (far right) is 100%. 3 In { } is the percentage when the FROM country subtotal (bottom) is 100%.

BIS Papers No 30 107

 Table 3.2
Geographic breakdown of short-term
paper investment, end-2002
In millions of US dollars

Africa,
From East North Latin Tax
Europe Middle
Asia America America Havens Subtotal
(28) East and
In (10)1 (2) (7) (17)
Asia (3)

East Asia (14) 31,024 10,285 3,336 – 18 8,849 53512

[57.98%]2 [19.22%] [6.23%] [0.00%] [0.03%] [16.54%] [100.00%]
3
{27.78%} {2.46%} {2.09%} {0.00%} {2.29%} {4.65%} {6.07%}

Europe (42) 36,477 243,419 131,240 279 169 101,359 512,943

[7.11%] [47.46%] [25.59%] [0.05%] [0.03%] [19.76%] [100.00%]
{32.66%} {58.25%} {82.19%} {15.20%} {21.50%} {53.28%} {58.15%}

North 21,595 133,897 19,475 1,101 244 69,186 245,498

America (2)
[8.80%] [54.54%] [7.93%] [0.45%] [0.10%] [28.18%] [100.00%]
{19.33%} {32.04%} {12.20%} {60.00%} {31.04%} {36.37%} {27.83%}

Latin 6 959 357 19 2 2,621 3,964

America (21)
[0.15%] [24.19%] [9.01%] [0.48%] [0.05%] [66.12%] [100.00%]
{0.01%} {0.23%} {0.22%} {1.04%} {0.25%} {1.38%} {0.45%}

Africa, Middle 126 833 – 18 34 129 1,140

East and
[11.05%] [73.07%] [0.00%] [1.58%] [2.98%] [11.32%] [100.00%]
Southeast
Asia (42) {0.11%} {0.20%} {0.00%} {0.98%} {4.33%} {0.07%} {0.13%}

Tax Havens (44) 18,267 18,027 5,272 398 319 7,884 50,167
[36.41%] [35.93%] [10.51%] [0.79%] [0.64%] [15.72%] [100.00%]
{16.35%} {4.31%} {3.30%} {21.69%} {40.59%} {4.14%} {5.69%}

Other 4,200 10,488 5 20 – 216 14,929

Nations (73)
[28.13%] [70.25%] [0.03%] [0.13%] [0.00%] [1.45%] [100.00%]
{3.76%} {2.51%} {0.00%} {1.09%} {0.00%} {0.11%} {1.69%}

Subtotal 111,695 417,908 159,685 1,835 786 190,244 882,153

[12.66%] [43.37%] [18.10%] [0.21%] [0.09%] [21.57%] [100.00%]
{100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%}
1
In ( ) next to the region’s name is the number of countries. 2 In [ ] is the percentage when the IN country
subtotal (far right) is 100%. 3 In { } is the percentage when the FROM country subtotal (bottom) is 100%.

108 BIS Papers No 30

BIS Papers No 30

Table 3.3
East Asian country breakdown of
long-term debt securities, year-end 20021
In millions of US dollars

From Hong
New
Australia Kong Indonesia Japan Korea Malaysia Philippines Singapore Thailand Subtotal
In Zealand
SAR
Australia – 11,333 1 17,092 20 26 358 10 3,761 – 32,601
2
[0.00%] [34.76%] [0.00%] [52.43%] [0.06%] [0.08%] [1.10%] [0.03%] [11.54%] [0.00%] [100.00%]
3
{0.00%} {41.72%} {1.14%} {57.01%} {1.66%} {15.78%} {50.42%} {7.52%} {22.53%} {0.00%} {42.67%}
China – 1,232 – 578 38 – – 2 416 – 2,266
[0.00%] [54.37%] [0.00%] [25.50%] [1.69%] [0.00%] [0.00%] [0.09%] [18.37%] [0.00%] [100.00%]
{0.00%} {4.54%} {0.00%} {1.93%} {3.18%} {0.00%} {0.00%} {1.51%} {2.49%} {0.00%} {2.97%}
Hong Kong – – 57 1,137 455 40 – 58 1,653 20 3,421
SAR [0.00%] [0.00%] [1.67%] [33.23%] [13.30%] [1.17%] [0.00%] [1.70%] [48.33%] [0.58%] [100.00%]
{0.00%} {0.00%} {64.80%} {3.79%} {37.86%} {24.45%} {0.00%} {43.82%} {9.91%} {100.00%} {4.48%}
Indonesia – – – 49 78 1 – 4 869 – 1,000
[0.00%] [0.00%] [0.00%] [4.86%] [7.77%] [0.09%] [0.00%] [0.40%] [86.90%] [0.00%] [100.00%]
{0.00%} {0.00%} {0.00%} {0.16%} {6.46%} {0.55%} {0.00%} {3.01%} {5.21%} {0.00%} {1.31%}
Japan – 5,351 – – 29 – 282 5 3,828 – 9,495
[0.00%] [56.36%] [0.00%] [0.00%] [0.31%] [0.00%] [2.97%] [0.05%] [40.31%] [0.00%] [100.00%]
{0.00%} {19.70%} {0.00%} {0.00%} {2.41%} {0.00%} {39.75%} {3.74%} {22.93%} {0.00%} {12.43%}
Korea – 4,202 – 5,348 – 51 69 15 2,586 – 12,271
[0.00%] [34.24%] [0.00%] [43.58%] [0.00%] [0.42%] [0.56%] [0.12%] [21.07%] [0.00%] [100.00%]
{0.00%} {15.47%} {0.00%} {17.84%} {0.00%} {31.38%} {9.77%} {11.12%} {15.49%} {0.00%} {16.06%}
Malaysia – 2,085 3 1,823 332 – – 9 1,830 – 6,083
[0.00%] [34.28%] [0.05%] [29.98%] [5.46%] [0.00%] [0.00%] [0.15%] [30.08%] [0.00%] [100.00%]
{0.00%} {7.68%} {3.41%} {6.08%} {27.65%} {0.00%} {0.00%} {6.85%} {10.96%} {0.00%} {7.96%}
109
110

Table 3.3 (cont)

East Asian country breakdown of
long-term debt securities, year-end 2002
In millions of US dollars

From Hong
New
Australia Kong Indonesia Japan Korea Malaysia Philippines Singapore Thailand Subtotal
In Zealand
SAR
New 251 – – 1,258 – – – – 279 – 1,788
Zealand [14.03%] [0.00%] [0.00%] [70.36%] [0.00%] [0.00%] [0.00%] [0.00%] [15.63%] [0.00%] [100.00%]
{99.93%} {0.00%} {0.00%} {4.20%} {0.00%} {0.02%} {0.00%} {0.00%} {1.67%} {0.00%} {2.34%}
Philippines – – 5 1,389 81 4 – – 595 – 2,074
[0.00%] [0.00%] [0.22%] [66.96%] [3.92%] [0.22%] [0.00%] [0.00%] [28.68%] [0.00%] [100.00%]
{0.00%} {0.00%} {5.12%} {4.63%} [6.76%} {2.73%} {0.00%} {0.00%} {3.56%} {0.00%} {2.71%}
Singapore – 1,842 23 680 144 41 – 23 – – 2,753
[0.00%] [66.91%] [0.82%] [24.69%] [5.23%] [1.49%] [0.00%] [0.84%] [0.00%] [0.00%] [100.00%]
{0.00%} {6.78%} {25.69%} {2.27%} {11.98%} {25.05%} {0.00%} {17.43%} {0.00%} {0.00%} {3.60%}
Taiwan, – 674 – 46 – – – 7 333 – 1,060
China [0.00%] [63.58%] [0.00%] [4.32%] [0.00%] [0.00%] [0.00%] [0.66%] [31.40%] [0.00%] [100.00%]
{0.00%} {2.48%} {0.00%} {0.15%} {0.00%} {0.00%} {0.00%} {5.27%} {1.99%} {0.00%} {1.39%}
Thailand – 447 – 550 24 1 – – 542 – 1,564
[0.00%] [28.58%] [0.00%] [35.20%] [1.53%] [0.04%] [0.00%] [0.00%] [34.64%] [0.00%] [100.00%]
{0.00%} {1.65%} {0.00%} {1.84%} {2.00%} {0.34%} {0.00%} {0.00%} {3.25%} {0.00%} {2.05%}
Subtotal 252 27,167 89 29,951 1,203 165 710 134 16,693 21 76,376
[0.33%] [35.55%] [0.12%] [39.24%] [1.57%] [0.22%] [0.93%] [0.17%] [21.85%] [0.03%] [100.00%]
BIS Papers No 30

{100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%}
1 2 3
Data for Vietnam and Cambodia are not reported here. In [ ] is the percentage when the IN country subtotal (far right) is 100%. In { } is the percentage when the FROM
country subtotal (bottom) is 100%.
BIS Papers No 30

Table 3.4
East Asian country breakdown of
short-term debt securities, year-end 20021
In millions of US dollars

From Hong
New
Australia Kong Indonesia Japan Korea Malaysia Philippines Singapore Thailand Subtotal
In Zealand
SAR
Australia – 9,795 – 1,657 – – – – 5,130 50 16,632
[0.00%]2 [58.89%] [0.00%] [9.96%] [0.00%] [0.00%] [0.00%] [0.00%] [30.84%] [0.30%] [100%]
{0.00%}3 {64.39%} {0.00%] {47.65%} {0.00%} {0.00%} {0.00%} {0.00%} {42.11%} {69.44%} {53.61%}
China – 1,569 – – – – – – 53 – 1,622
[0.00%] [94.40%] [0.00%] [0.00%] [0.00%] [0.00%] [0.00%] [0.00%] [3.17%] [0.00%] [100%]
{0.00%} {10.31%} {0.00%} {0.00%} {0.00%} {0.00%} {0.00%} {0.00%} {0.43%} {0.00%} {5.23%}
Hong Kong – – – 12 4 – – 50 111 2 179
SAR [0.00%] [0.00%] [0.00%] [6.52%] [2.40%] [0.00%] [0.00%] [27.93%] [62.018%] [1.12%] [100%]
{0.00%} {0.00%} {0.00%} {0.34%} {18.70%} {0.00%} {0.00%} {100.00%} {0.91%} {2.78%} {0.58%}
Indonesia – – – 2 – – – – 89 – 90
[0.00%] [0.00%] [0.00%] [1.85%] [0.00%] [0.00%] [0.00%] [0.00%] [98.63%] [0.00%] [100%]
{0.00%} {0.00%} {0.00%} {0.05%} {0.00%} {0.00%} {0.00%} {0.00%} {0.73%} {0.00%} {0.29%}
Japan – 1,396 – – – – 1 – 2,789 – 4,186
{0.00%} [33.35%] [0.00%] [0.00%] [0.00%] [0.00%] [0.02%] [0.00%] [66.63%] [0.00%] [100%]
{0.00%} {9.18%} {0.00%} {0.00%} {0.00%} {0.00%} {100.00%} {0.00%} {22.90%} {0.00%} {13.49%}
Korea – 1,761 – 125 – 9 – – 523 10 2,428
[0.00%] [72.53%] [0.00%] [5.15%] [0.00%] [0.35%] [0.00%] [0.00%] [21.56%] [0.41%] [100%]
{0.00%} {11.58%} {0.00%} {3.60%} {0.00%} {94.68%} {0.00%} {0.00%} {4.30%} {13.89%} {7.83%}
Malaysia – 43 – – 19 – – – 126 – 187
[0.00%] [22.99%] [0.00%] [0.00%] [9.89%] [0.00%] [0.00%] [0.00%] [67.20%] [0.00%] [100%]
{0.00%} {0.28%} {0.00%} {0.00%} {80.43%} {0.00%} {0.00%} {0.00%} {1.03%} {0.00%} {0.60%}
111
112

Table 3.4 (cont)

East Asian country breakdown of
short-term debt securities, year-end 2002
In millions of US dollars

From Hong
New
Australia Kong Indonesia Japan Korea Malaysia Philippines Singapore Thailand Subtotal
In Zealand
SAR
New – – – 171 – – – – 2,754 10 2,935
Zealand [0.00%] [0.00%] [0.00%] [5.83%] [0.00%] [0.00%] [0.00%] [0.00%] [93.82%] [0.34%] [100%]
{0.00%} {0.00%} {0.00%} {4.92%} {0.00%} {0.00%} {0.00%} {0.00%} {22.60%} {13.89%} {9.46%}
Philippines – – – – – – – – 104 – 104
[0.00%] [0.00%] [0.00%] [0.00%] [0.00%] [0.00%] [0.00%] [0.00%] [0.00%] [0.00%] [100%]
{0.00%} {0.00%} {0.00%} {0.00%} {0.00%} {0.00%} {0.00%} {0.00%} {0.86%} {0.00%} {0.34%}
Singapore – 303 – 1,510 – – – – – – 1,814
[0.00%] [16.70%] [0.02%] [83.26%] [0.00%] [0.03%] [0.00%] [0.00%] [0.00%] [0.00%] [100%]
{0.00%} [1.99%} {0.00%} {43.44%} {0.00%} {5.54%} {0.00%} {0.00%} {0.00%} {0.00%} {5.85%}
Taiwan, – 131 – – – – – – 224 – 355
China [0.00%] [36.90%] [0.00%] [0.00%] [0.00%] [0.00%] [0.00%] [0.00%] [62.97%] [0.00%] [100%]
{0.00%} {0.86%} {0.00%} {0.00%} {0.00%} {0.00%} {0.00%} {0.00%} {1.84%} {0.00%} {1.14%}
Thailand – 213 – – – – – – 279 – 492
[0.00%] [43.29%] [0.00%] [0.00%] [0.00%] [0.00%] [0.00%] [0.00%] [56.81%] [0.00%] [100%]
{0.00%} {1.40%} {0.00%} {0.00%} {0.00%} {0.00%} {0.00%} {0.00%} {2.29%} {0.00%} {1.59%}
Subtotal – 15,211 – 3,477 23 9 1 50 12,182 72 31,024
[0.00%] [49.03%] [0.00%] [11.21%] [0.07%] [0.03%] [0.00%] [0.16%] [39.27%] [0.23%] [100%]
BIS Papers No 30

{100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%}
1 2 3
Data for Vietnam and Cambodia are not reported here. In [ ] is the percentage when the IN country subtotal (far right) is 100%. In { } is the percentage when the
FROM country subtotal (bottom) is 100%.
BIS Papers No 30

Table 3.5
Top 10 country breakdown of long-term
debt securities investment, year-end 2002
In millions of US dollars

From
United United Cayman
Germany Luxembourg Italy Netherlands Japan France Canada Subtotal
States Kingdom Islands
In

East 60,227 50,087 13,750 16,113 3,821 3,769 29,951 15,733 627 899 194,977
Asia (14)1 [30.89%]2 [25.69%] [7.05%] [8.26%] [1.96%] [1.93%] [15.36%] [8.07%] [0.32%] [0.46%] [100.00%]
{12.84%}3 {6.42%} {2.52%} {3.12%} {1.25%} {1.10%} {2.79%} {2.63%} {4.10%} {2.13%} {4.16%}
Europe (42) 230,012 345,738 426,727 390,163 190,146 265,737 413,010 450,439 2,605 3,190 2,717,766
[8.46%] [12.72%] [15.70%] [14.36%] [7.00%] [9.78%] [15.20%] [16.57%] [0.10%] [0.12%] [100.00%]
{49.04%} {44.31%} {78.15%} {75.64%} {62.32%} {77.87%} {38.47%} {75.37%} {17.01%} {7.55%} {57.99%}

North 106,024 206,101 57,071 82,675 43,198 60,612 395,281 78,321 7,988 31,770 1,069,044
America (2) [9.92%] [19.28%] [5.34%] [7.73%] [4.04%] [5.67%] [36.98%] [7.33%] [0.75%] [2.97%] [100.00%]
{22.60%} {26.41%} {10.45%} {16.03%} {14.16%} {17.76%} {36.82%} {13.11%} {52.15%} {75.19%} {22.81%}
Latin 33,863 10,794 7,676 4,236 11,489 2,453 8,472 3,264 1,989 5,404 89,640
America [37.78%] [12.04%] [8.56%] [4.73%] [12.82%] [2.74%] [9.45%] [3.64%] [2.22%] [6.03%] [100.00%]
(21)
{7.22%} {1.38%} {1.41%} {0.82%} {3.77%} {0.72%} {0.79%} {0.55%} {12.99%} {12.79%} {1.91%}
Africa, 4,273 2,643 4,890 1,044 1,515 238 3,737 723 29 178 19,270
Middle East
and [22.17%] [13.72%] [25.38%] [5.42%] [7.86%] [1.24%] [19.39%] [3.75%] [0.15%] [0.93%] [100.00%]
Southeast
Asia (42) {0.91%} {0.34%} {0.90%} {0.20%} {0.50%} {0.07%} {0.35%} {0.12%} {0.19%} {0.42%} {0.41%}

Tax 33,746 60,705 35,948 21,436 52,185 8,467 210,916 48,923 1,869 185 474,411
Havens (44) [7.11%] [12.80%] [7.58%] [4.52%] [11.00%] [1.78%] [44.46%] [10.31%] [0.39%] [0.04%] [100.00%]
{7.19%} {7.78%} {6.58%} {4.16%} {17.10%} {2.48%} {19.65%} {8.19%} {12.20%} {0.44%} {10.12%}
113
114

Table 3.5 (cont)

Top 10 country breakdown of long-term
debt securities investment, year-end 2002
In millions of US dollars

From
United United Cayman
Germany Luxembourg Italy Netherlands Japan France Canada Subtotal
States Kingdom Islands
In

Other 910 104,219 3 148 2,750 3 12,152 221 209 629 121,244
Nations (73)
[0.75%] [85.96%] [0.00%] [0.12%] [2.27%] [0.00%] [10.02%] [0.18%] [0.17%] [0.52%] [100.00%]
{0.19%} {13.36%} {0.00%} {0.03%} {0.90%} {0.00%} {1.13%} {0.04%} {1.36%} {1.49%} {2.59%}

Subtotal 469,055 780,287 546,065 515,815 305,105 341,279 1,073,551 597,623 15,316 42,255 4,686,351
[10.01%] [16.65%] [11.65%] [11.01%] [6.51%] [7.28%] [22.91%] [12.75%] [0.33%] [0.90%] [100.00%]
{100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%}
1 2 3
In ( ) next to the region’s name is the number of countries. In [ ] is the percentage when the IN country subtotal (far right) is 100%. In { } is the percentage when the
FROM country subtotal (bottom) is 100%.
BIS Papers No 30
BIS Papers No 30

Table 3.6
Top 10 country breakdown of short-term debt
securities investment, year-end 2002
In millions of US dollars

From
United United Cayman
Germany Luxembourg Italy Netherlands Japan France Canada Subtotal
States Kingdom Islands
In

East Asia 3,323 5,396 49 4,550 346 486 3,477 858 13 158 18,656
(14)1 [17.81%]2 [28.92%] [0.26%] [24.39%] [1.85%] [2.61%] [18.64%] [4.60%] [0.073%] [0.85%] [100%]
{2.14%}3 {7.04%} {0.36%} {4.97%} {3.77%} {5.32%} {7.23%} {1.22%} {0.30%} {2.01%} {3.84%}
Europe (42) 130,171 41,468 9,434 53,290 7,740 6,178 11,940 50,503 1,069 2,956 314,749
[41.36%] [13.17%] [3.00%] [16.93%] [2.46%] [1.961%] [3.79%] [16.04%] [0.35%] [0.94%] [100%]
{83.81%} {54.07%} {69.22%} {58.20%} {84.36%} {67.68%} {24.83%} {71.80%} {24.48%} {37.68%} {64.74%}

North 16,324 18,690 3,990 28,868 246 1,582 15,153 14,886 3,151 3,710 106,600
America (2) [15.31%] [17.53%] [3.74%] [27.08%] [0.23%] [1.48%] [14.21%] [14.00%] [2.96%] [3.48%] [100%]
{10.51%} {24.37%} {29.28%} {31.53%} {2.68%} {17.33%} {31.52%} {21.16%} {72.15%} {47.30%} {21.93%}
Latin 357 712 8 117 46 15 – 47 – 617 1,919
America [18.60%] [37.10%] [0.42%] [6.10%] [2.40%] [0.78%] [0%] [2.50%] [0%] [32.15%] [100%]
(21) {0.23%} {0.93%} {0.06%} {0.13%} {0.50%} {0.16%} {0.00%} {0.07%} {0.00%} {7.87%} {0.39%}
Africa, – 538 7 26 – 72 – 1 – 7 651
Middle East
and [0.00%] [82.64%] [1.08%] [3.99%] [0.00%] [11.06%] [0.00%] [0.15%] [0.00%] [1.08%] [100%]
Southeast
Asia (42) {0.00%} {0.70%} {0.05%} {0.03%} {0.00%} {0.79%} {0.00%} {0.00%} {0.00%} {0.09%} {0.13%}

Tax Havens 5,143 1,962 141 4,685 797 793 17,318 4,023 129 292 35,283
(44) [14.58%] [5.56%] [0.40%] [13.28%] [2.26%] [2.25%] [49.08%] [11.40%] [0.37%] [0.83%] [100%]
{3.31%} {2.56%} {1.03%} {5.12%} {8.69%} {8.69%} {36.02%} {5.72%} {2.95%} {3.72%} {7.26%}
115
116

Table 3.6 (cont)

Top 10 country breakdown of short-term debt
securities investment, year-end 2002
In millions of US dollars

From
United United Cayman
Germany Luxembourg Italy Netherlands Japan France Canada Subtotal
States Kingdom Islands
In

Other – 7,933 – 26 – 2 192 18 5 104 8,280

Nations (73)
[0.00%] [95.81%] [0.00%] [0.31%] [0.00%] [0.02%] [2.32%] [0.22%] [0.06%] [1.26%] [100%]
{0.00%} {10.34%} {0.00%} {0.03%} {0.00%} {0.02%} {0.40%} {0.03%} {0.11%} {1.33%} {1.70%}

Subtotal 155,318 76,699 13,629 91,562 9,175 9,128 48,080 70,336 4,367 7,844 486,138
[31.95%] [15.78%] [2.80%] [18.83%] [1.89%] [1.88%] [9.90%] [14.47%] [0.90%] [1.61%] [100%]
{100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%} {100.00%}
1 2 3
In ( ) next to the region’s name is the number of countries. In [ ] is the percentage when the IN country subtotal (far right) is 100%. In { } is the percentage when the
FROM country subtotal (bottom) is 100%.
BIS Papers No 30
 Table 3.7
Rule of law breakdown of long-term
bonds, year-end 2002
In millions of US dollars

From Civil/
Common Civil/French Civil/German
Scandinavian Subtotal
In (13)1 (18) (5)
(4)

Common (19) 582,835 235,504 722,734 64,783 1,605,856

[36.29%]2 [14.67%] [45.01%] [4.03%] [100%]
3
{42.66%} {23.01%} {42.35%} {36.06%} {37.56%}

Civil/French (22) 439,889 468,920 460,009 42,478 1,411,296

[31.17%] [33.23%] [32.59%] [3.01%] [100%]
{32.20%} {45.82%} {26.95%} {23.65%} {33.01%}

Civil/German (6) 300,796 287,216 449,774 53,658 1,091,444

[27.56%] [26.32%] [41.21%] [4.92%] [100%]
{22.02%} {28.07%} {26.35%} {29.87%} {25.53%}

Civil/ 42,627 31,740 74,228 18,729 167,323

Scandinavian (4)
[25.48%] [18.97%] [44.36%] [11.19%] [100%]
{3.12%} {3.10%} {4.35%} {10.43%} {3.91%}

Subtotal 1,366,149 1,023,379 1,706,745 179,647 4,275,920

[31.95%] [23.93%] [39.92%] [4.20%] [100%]
{100%} {100%} {100%} {100%} {100%}
1
In ( ) next to the region’s name is the number of countries. 2 In [ ] is the percentage when the IN country
3
subtotal (far right) is 100%. In { } is the percentage when the FROM country subtotal (bottom) is 100%.

BIS Papers No 30 117

 Table 3.8
Rule of law breakdown of short-term
paper, year-end 2002
In millions of US dollars

From Civil/
Common Civil/French Civil/German
Scandinavian Subtotal
In (13)1 (18) (5)
(4)

Common (19) 292,708 62,775 30,318 3,922 389,723

2
[75.11%] [16.11%] [7.77%] [1.01%] [100%]
3
{68.96%} {47.84%} {53.48%} {42.74%} {62.70%}

Civil/French (22) 63,700 36,078 17,351 1,046 118,175

[53.90%] [30.53%] [14.68%] [0.89%] [100%]
{15.01%} {27.50%} {30.60%} {11.40%} {19.01%}

Civil/German (6) 53,424 25,119 6,619 1,571 86,733

[61.60%] [28.96%] [7.63%] [1.811%] [100%]
{12.59%} {19.14%} {11.67%} {17.12%} {13.95%}

Civil/ 14,615 7,243 2,406 2,637 26,901

Scandinavian (4)
[54.33%] [26.92%] [8.94%] [9.80%] [100%]
{3.44%} {5.52%} {4.24%} {28.74%} {4.33%}

Subtotal 424,447 131,215 56,694 9,176 621,532

[68.30%] [21.11%] [9.12%] [1.48%] [100%]
{100.00%} {100.00%} {100.00%} {100.00%} {100.00%}
1
In ( ) next to the region’s name is the number of countries. 2 In [ ] is the percentage when the IN country
subtotal (far right) is 100%. 3 In { } is the percentage when the FROM country subtotal (bottom) is 100%.

4. Regression results
We begin our empirical analysis by estimating cross-country regressions that examine
whether differences in foreign bond investments can be accounted for by cross-country
differences in property rights protection, while controlling for other macro factors that may be
important.

4.1 Summary statistics of the regression sample

In Table 4.1, we present the list of 47 sample countries and some of the variables used in the
regression analysis. We partition the sample into two groups by the median of the property
right index. One clear message delivered by the table is that countries with stronger investor
protection show larger GDP and have more local bonds held by foreign investors. The
percentage of local bonds held by foreign investors scaled by GDP is 7.18% for countries
with low investor protection, whereas the same figure is as much as 35.5% for countries with
high property rights protection. Table 4.2 presents some summary statistics and correlations
used in the analysis.

118 BIS Papers No 30

 Table 4.1
Summary statistics

(2) (3) (4)

(1) (6) (7)
Long- Gross Percent- (5)
Short-term Property Creditor
Country term domestic age of Origin of
debts rights rights
debts product debts law
(in m$) Index Index
(in m$) (in m$) over GDP

Countries with low property rights protection

ARG 606 21,485 268,831 8.22 2 16.84 1

BRA 2,984 43,500 508,994 9.13 2 20.24 1
CHL 80 4,332 66,450 6.64 2 19.60 2
COL 31 5,886 82,411 7.18 2 18.97 0
ECU 14 1,233 21,024 5.93 2 16.93 4
EGY 48 491 98,476 0.55 2 16.22 4
GRC 460 61,302 117,169 52.71 2 21.01 1
IDN 79 1,597 141,255 1.19 2 15.40 4
IND 214 1,766 481,440 0.41 1 18.44 4
JOR 18 176 8,829 2.20 2 16.41 .
KEN 22 18 11,396 0.35 1 16.46 4
KOR 2,018 22,425 427,234 5.72 3 22.20 3
LKA 31 130 15,662 1.02 1 16.30 3
MEX 742 42,802 623,890 6.98 2 18.61 0
MYS 292 9,538 88,050 11.16 1 22.76 4
PAK 22 244 58,648 0.45 1 13.47 4
PER 143 1,792 54,218 3.57 2 14.92 0
PHL 332 8,804 71,382 12.80 2 12.94 0
THA 348 3,607 115,310 3.43 1 20.17 3
TUR 575 11,833 145,244 8.54 2 18.13 2
URY 129 1,629 18,561 9.47 2 18.87 2
VEN 226 8,544 126,197 6.95 2 17.89 .
ZWE 17 39 9,057 0.62 1 16.07 4
Average 410 11,008 154,771 7.18 . 18 2

Countries with high property rights protection

AUS 25,448 73,916 368,726 26.95 1 26.50 1

AUT 6,512 104,400 188,546 58.83 3 27.86 3
BEL 22,536 97,354 229,610 52.22 2 27.93 2
CAN 14,872 196,559 694,475 30.44 1 28.63 1
CHE 3,297 13,828 247,091 6.93 3 29.96 1

BIS Papers No 30 119

 Table 4.1 (cont)
Summary statistics

(2) (3) (4)

(1) (6) (7)
Long- Gross Percent- (5)
Short-term Property Creditor
Country term domestic age of Origin of
debts rights rights
debts product debts law
(in m$) Index Index
(in m$) (in m$) over GDP

Countries with high property rights protection

DEU 84,910 802,257 1,846,069 48.06 3 28.60 3

DNK 6,036 51,119 161,542 35.38 4 28.98 3
ESP 5,724 165,876 581,823 29.49 2 25.30 2
FIN 3,639 37,677 120,855 34.19 4 28.82 1
FRA 56,589 331,120 1,309,807 29.60 2 27.89 0
GBR 174,385 394,595 1,424,094 39.95 1 28.44 4
HKG 505 15,691 162,843 9.95 1 25.63 4
IRL 12,496 67,419 103,298 77.36 1 27.15 1
ISR 91 8,996 . . 1 24.12 4
ITA 31,564 428,540 1,088,754 42.26 2 24.65 2
JPN 36,127 167,520 4,141,431 4.92 3 27.88 2
NLD 39,538 371,334 380,137 108.09 2 29.33 2
NOR 2,584 32,007 166,145 20.82 4 29.59 2
NZL 3,844 9,145 50,425 25.76 1 28.98 3
PRT 3,278 44,617 109,803 43.62 2 24.85 1
SGP 1,233 13,228 84,871 17.04 1 26.38 4
SWE 14,220 84,280 209,814 46.95 4 28.98 2
USA 418,135 1,660,138 10,065,270 20.65 1 27.61 1
ZAF 255 7,085 114,174 6.43 1 23.07 3
Average 40,326 215,779 1,036,939 35.47 . 27 2

This table presents the amount of local bonds held by foreign investors as of the end of 2001, gross
domestic product in 2001, percentage of bonds over GDP, legal origin indicator, property rights
index and creditor rights index for each of 47 sample countries. The property rights index is the sum
of three indexes from La Porta et al (1998). Legal origin indicator “1” is English origin, “2” French,
“3” German, and “4” Scandinavian. Each index ranges from zero to 10. Each index measures
government corruption, the risk of expropriation by the government and the risk of the government
repudiating contracts. High values of property rights indexes indicate better protection of property
rights. The creditor rights index is the sum of four dummy variables, each of which measures “no
automatic stay on assets”, “secured creditors first”, “restrictions for going into reorganization” and
“current management does not stay in the reorganized firm”. High values of creditor rights indicate
better protection of creditor rights.

120 BIS Papers No 30

 Table 4.2
Summary statistics of variables

Panel A: summary statistics

Variable N Mean Std dev Minimum Maximum

Short-term debts/GDP 47 2.33 3.37 0.04 12.25

Long-term debts/GDP 47 18.56 20.61 0.15 97.68
Property rights index 47 22.68 5.33 12.94 29.96
Creditor rights index 45 2.00 1.00 1.00 4.00
GDP per capital 47 15,053 13,686 332 48,160
Inflation rate 46 5.90 12.52 –3.97 64.87
Growth rate of GDP 47 0.62 4.22 –14.36 10.03
Lending rate 43 14.32 14.41 2.16 64.02

Panel B: correlations

(1) (2) (3) (4) (5) (6) (7) (8)

(1) Short-term
debts/GDP 1.00
(2) Long-term 0.72 1.00
debts/GDP (0.00)
(3) Property rights index 0.65 0.61 1.00
(0.00) (0.00)
(4) Creditor rights index –0.09 –0.22 –0.11 1.00
(0.55) (0.15) (0.46)
(5) GDP per capita 0.52 0.53 0.90 –0.18 1.00
(0.00) (0.00) (0.00) (0.25)
(6) Inflation rate –0.24 –0.22 –0.38 0.07 –0.37 1.00
(0.11) (0.14) (0.01) (0.66) (0.01)
(7) Growth rate of GDP 0.33 0.30 0.41 0.01 0.34 –0.67 1.00
(0.02) (0.04) (0.00) (0.96) (0.02) (0.00)
(8) Lending rate –0.40 –0.41 –0.62 0.15 –0.57 0.79 –0.72 1.00
(0.01) (0.01) (0.00) (0.34) (0.00) (0.00) (0.00)

BIS Papers No 30 121

4.2 Baseline regressions
Table 4.3 presents results from regressions on the cross-country data. The dependent
variable is the local bonds held by foreign investors scaled by the size of GDP in the sample
countries. 12 We first look at the regression results reported in Column (1) of Table 4.3. The
explanatory variables are the log of the lending rate level in the country and the log of per
capita GNP. We also include the growth rate of GNP. The coefficient on the lending rate level
is positive and that on the growth rate of GNP is also positive, but both of these coefficients
are not significant at conventional levels. The coefficient on per capita GNP is positive and
significant at the 1% significance level. This implies that richer countries attract more
investment in local bond markets from abroad.
In Column (2) we examine whether the cross-country variation in the creditor rights index
explains the variation in foreign bond investments. The coefficient on the creditor rights index
is again not statistically significant. In the regression in Column (3), we drop the creditor
rights index and replace it with the property rights index. The coefficient on the property
rights index variable is positive and significant at the 1% level. The coefficient on the per
capita GNP loses its significance, which indicates that the institutional variable is more
important than the degree of economic development.
The positive relation between foreign bond investments and the strength of property rights
protection supports the view that property rights protection allows more efficient contracting,
and that foreign investors are willing to take part in countries where property rights protection
is strong. Together with other variables, the property rights index explains about 38.5% of the
variation in cross-country foreign bond investments. Notice also that the explanatory power
significantly increases from the regression in Column (1) by 14.7 percentage points to 38.5%.
The fact that the creditor rights index does not explain the cross-country variation in foreign
bond investments supports the argument that what matters to foreign investors is not the
actual law that provides creditor rights protection, but, instead, how the law is enforced. The
enforcement is a function of property rights protection. Demirgüç-Kunt and Maksimovic
(1999) reach a similar conclusion. These authors argue that a direct statistical relation
between the existence of creditor rights and financial contracts is not expected, because the
“existence of rights may be necessary but not sufficient to make a financial contract
enforceable.” In a different context, Esty (2002) finds that creditor rights and property rights
affect foreign bank participation in project loan syndicates.
When both the property rights index and the creditor rights index variables are included
together with the log of per capita GNP and the log of lending rates (in COLUMN (4)), only
the property rights index variable is statistically significant.
Our results provide additional evidence in the literature underscoring the importance of
property rights protection in the development of debt markets. Miller and Puthenpurackal
(2002) examine the costs, wealth effects and determinants of international capital raising for
a sample of 260 public debt issues made by non-US firms in the yankee bond market. They
find that investors demand economically significant premiums on bonds issued by firms that
are located in countries that do not protect investors’ rights and do not have a prior history of
ongoing disclosure. Their results support the idea that better legal protections and more
detailed information disclosure increases the price investors will pay for financial assets. Bae
and Goyal (2003) examine how property rights affect private contracting in bank loan
markets. They find that when property rights are weaker, banks offer less credit, charge
higher spreads and lend only on a short-term basis.

12
We also scale the foreign investor-held bonds by the size of the bond markets in the sample countries. The
results are similar.

122 BIS Papers No 30

 Table 4.3
Regression of bond holdings by foreign investors on
property rights index and macroeconomic variables

Panel A. Short-term paper

(1) (2) (3) (4)

Constant 0.414 0.503 –9.607*** –9.541***

(0.81) (1.22) (3.14) (3.33)
Lending rate 0.022 0.023 0.015 0.019
(0.04) (0.05) (0.04) (0.04)
Growth rate of GNP 0.179 0.188 0.081 0.096
(0.14) (0.15) (0.13) (0.14)
Per capita GNP 0.114 0.114*** –0.077 –0.080
(0.03) (0.04) (0.06) (0.06)
Creditor right index –0.048 –0.136
(0.36) (0.32)
Property rights index 0.573*** 0.583***
(0.17) (0.18)
Observations 45 43 45 43
Adjusted R-squared 0.238 0.200 0.385 0.356

Panel B. Long-term bonds

(1) (2) (3) (4)

Constant 6.727 12.187 –39.078* –35.015*

(4.98) (7.32) (20.20) (20.97)
Lending rate 0.137 0.154 0.104 0.137
(0.29) (0.31) (0.28) (0.29)
Growth rate of GNP 0.906 1.030 0.460 0.597
(0.86) (0.94) (0.84) (0.90)
Per capita GNP 0.728*** 0.686*** –0.145 –0.227
(0.21) (0.22) (0.42) (0.44)
Creditor right index –2.276 –2.689
(2.15) (2.04)
Property rights index 2.620*** 2.741**
(1.12) (1.15)
Observations 45 43 45 43
Adjusted R-squared 0.236 0.217 0.311 0.303
The table presents results from cross-sectional regression of bond holdings by foreign investors. The sample
includes 45 countries. The dependent variable is the short/long-term bonds held by foreign investors as of the
end of 2001 over GDP. Independent variables include the inflation rate and log of per capita GNP, property
rights index, creditor rights index. Numbers in parentheses are standard errors. ***Significant at the 1% level.
**Significant at the 5% level. *Significant at the 10% level.

BIS Papers No 30 123

While not reported, we also examine whether property right protection affects the trading
activity of foreign investors in local bond markets. We compute the size of bond portfolio
flows between the United States and each country in our sample. US investors represent a
significant fraction of the portfolio capital flows. Furthermore, comparable data are not
available from other countries. The monthly flow of portfolio capital between the United
States and virtually every country in the world is made available online by the US Treasury
Department starting with 1988 data. 13 We use the sum of inward and outward flows of US
investors scaled by the market’s GDP to proxy for the amount of bond portfolio activity
crossing a particular country’s borders in a particular month. The results show that we
observe more active foreign trading activity in countries with strong property rights.

4.3 Robustness tests

This section presents results from sensitivity tests that examine whether there are other
explanations for our results. A concern with cross-country analysis is that the regression may
omit an important explanatory variable that is really driving the result and that is highly
correlated with property rights protection. To rule out several alternative explanations, we
have experimented with other plausible institutional and macroeconomic factors and examine
if including them reduces the significance of the coefficient on the property rights index. In
particular, we focus on the legal origin dummies, country risk and variables that measure the
size and activity of debt and equity markets. These sensitivity tests rule out a large number of
alternative explanations.
Legal origin dummies: La Porta et al (1998) show that the legal origin of a country’s laws
explains the degree of investor protection in that country. English common-law countries
offer creditors stronger legal protection against managers. German civil-law countries are
protective of secured creditors. Scandinavian civil-law countries are the best in law
enforcement. The legal variables are from the La Porta et al (1998) dataset. While the results
are not reported, none of the legal origin dummy variables is significant.
We also examine the effect that including only the French legal origin dummy has on the
coefficient of the property rights index in the baseline regression. French civil-law countries
are considered weak in investor protection. However, these unreported results show that the
conclusions are not sensitive to the set of legal origin dummies included in the regression.
The French legal origin dummy continues to remain insignificant while the property rights
index remains positive and significant at the 1% level.
GNP growth volatility: To address concerns that per capita GNP or the lending rate level
does not adequately capture country risk, we include GNP growth volatility as an additional
measure of country risk. GNP growth volatility is estimated as the standard deviation of the
annual growth rate in GNP. As one would predict, the coefficient on the GNP growth rate
volatility is negative but not significant. The inclusion of GNP growth volatility does not
change our main results on the property rights index.
Size of stock market: We examine the stock market capitalisation to GDP ratio, which equals
the value of listed shares divided by the GDP and the stock market total value traded to
GDP. Our results on the importance of the property rights index are robust to controlling for
the importance of the stock market in the economy. In addition, our results are robust to the
inclusion of variables that measure the size of the bond markets and the size of the primary
equity markets.

13
The data were downloaded from www.treas.gov/tic/country-longterm.html, and are also available in the
Treasury’s monthly bulletin. Across our sample of emerging market countries, only Sri Lanka lacks data from
this source.

124 BIS Papers No 30

 Table 4.4
Robustness tests

Panel A. Short-term paper

(1) (2) (3) (4)

Constant –11.176*** –8.055** –9.662*** –14.606

(3.90) (3.56) (3.22) (15.07)
Lending rate 0.024 0.034 0.009 0.029
(0.04) (0.05) (0.04) (0.05)
Growth rate of GNP 0.133 0.098 0.102 0.109
(0.14) (0.14) (0.14) (0.19)
Per capita GNP 0.018 –0.088 –0.091 –0.107
(0.08) (0.07) (0.07) (0.08)
Creditor right index 0.055 –0.132 –0.001 –0.266
(0.37) (0.32) (0.32) (0.46)
Property rights index 0.477** 0.561*** 0.615*** 0.687***
(0.19) (0.18) (0.18) (0.31)
English origin dummy 2.463
(1.77)
French civil origin 2.862
dummy (1.82)
German civil origin –0.794
dummy (1.84)
GNP growth volatility –0.329
(0.29)
Stock market cap/GDP –0.113*
(0.61)
Country credit rating 0.717
(4.56)
Observations 43 43 43 36
Adjusted R-squared 0.390 0.361 0.396 0.285

Panel B. Long-term bonds

(1) (2) (3) (4)

Constant –60.172** –21.426 –35.309* –159.53*

(23.67) (22.05) (21.13) (93.68)
Lending rate 0.057 0.270 0.113 0.139
(0.27) (0.29) (0.29) (0.35)
Growth rate of GNP 1.165 0.614 0.611 0.284
(0.86) (0.88) (0.91) (1.20)
Per capita GNP –0.067 –0.300 –0.254 –0.410
(0.50) (0.43) (0.44) (0.50)

BIS Papers No 30 125

 Table 4.4 (cont)
Robustness tests

Panel B. Long-term bonds (cont)

(1) (2) (3) (4)

Creditor right index 0.642 –2.518 –2.359 –1.847

(2.23) (2.00) (2.11) (2.87)
Property rights index 3.053*** 2.538** 2.819** 1.151
(1.16) (1.13) (1.16) (1.96)
English origin dummy –0.424
(10.65)

French civil origin 19.163*

dummy (11.02)
German civil origin –2.832
dummy (11.19)
GNP growth volatility –3.014*
(1.80)
Stock market cap/GDP –0.275
(0.40)
Country credit rating 38.769
(28.36)
Observations 43 43 43 36
Adjusted R-squared 0.391 0.335 0.293 0.206

The table presents results from an additional cross-sectional regression of bond holdings by foreign investors
to test the robustness of results reported in Table 3. The sample includes 45 countries. The dependent variable
is the short/long-term bonds held by foreign investors as of the end of 2001 over GDP. This table presents
several additional country level regressions of country median loan spreads to test the robustness of results
reported in Table 4.3. Independent variables include the lending rate and log of per capita GNP, property rights
index, creditor rights index, legal tradition dummies, GNP growth volatility, stock market capitalisation over
GDP and country credit rating. Numbers in parentheses are standard errors. ***Significant at the 1% level.
**Significant at the 5% level. *Significant at the 10% level.

Country credit rating: We examine if our results survive when we include the country credit
rating. We use Standard and Poor’s Foreign Currency Sovereign Credit Rating as a proxy for
country credit rating. While not reported, we also examine country credit rating scores
obtained from IMD survey data as of 2001 and find similar results. The S&P ratings data are
available for 40 countries. For our sample countries, the S&P ratings range from AAA to B−
with a rank of one to 16. We convert these rank values to continuous variables. Intuitively,
bonds of countries with better credit ratings are likely to be preferred by foreign investors and
one would expect that this variable is positively related to our dependent variable. The results
show that the property rights index is significant and positive in explaining both short-term
and long-term bonds held by foreign investors even after controlling for credit rating.
However, for long-term bonds, neither the property rights index nor the country credit rating
is significant, perhaps due to strong correlation between these two variables.
2002 data: To examine if the results are robust across time, we reexamine the data in year
2002. The results are very similar to those using the data in year 2001.

126 BIS Papers No 30

These sensitivity tests show that our results concerning the relation between property rights
protection and foreign bond investments are robust.

5. Summary and conclusions

This paper investigated how the protection of property rights affects foreign bond
investments. Some countries provide stronger protection for private property rights than do
other countries. The rights that lenders have are likely to be better enforced in countries with
stronger protection of property rights. We asked if cross-country differences in property rights
protection affect foreign bond holdings, after controlling for cross-country differences in
macroeconomic variables including GNP per capita, lending rates and/or inflation rates and
exchange rates.
Our findings suggest that differences in property rights protection translate into large
differences in foreign bond investments. In countries that provide weak property rights
protection, foreign investors make smaller investments.
These results imply that by improving property rights protection, a country or region can
expect to attract foreign interest and participation in local or regional bond markets. To the
extent that foreign investors play a vital role in providing liquidity, our evidence shows that
improving property rights protection is a matter of prime importance.

6. References
Alexander, Gordon, Cheol Eun and S Janakiramanan (1987): “Asset pricing and dual listing
on foreign capital markets: a note”, Journal of Finance, 42, 151-58.
Amihud, Yakov and Haim Mendelson (1986): “Asset pricing and the bid-ask spread”, Journal
of Financial Economics, 17, 223-49.
Amihud, Haim Mendelson and Beni Lauterbach (1997): “Market microstructure and securities
values: evidence from the Tel Aviv Stock Exchange”, Journal of Financial Economics, 45,
365-90.
Bae, Kee-Hong and Vidhan Goyal (2003): “Property rights protection and bank loan
contracts”, Journal of Financial Economics, submitted.
Becker, Robert and Emmanuel Pitsilis (2000): “A case for Asian bond markets”, the
McKinsey Quarterly 2000, Number 4.
Bekaert, Geert and Campbell Harvey (2003): “Emerging markets finance”, Journal of
Empirical Finance, 10, 3-55.
Chari, A and Peter Blair Henry (2002): “Risk sharing and asset prices: evidence from a
natural experiment”, Journal of Finance, 59, 1295-324.
Claessens, S, S Djankov, J P H Fan and L H P Lang (2002): “Disentangling the incentive and
entrenchment effects of large shareholdings”, Journal of Finance, 57, 2741-71.
Demirgüç-Kunt, A and V Maksimovic (1998): “Law, finance, and firm growth”, Journal of
Finance, 53, 2107-37.
——— (1999): “Institutions, financial markets, and firm debt maturity”, Journal of Financial
Economics, 54, 295-336.
Doidge, Craig (2003): “U.S. cross-listings and the private benefits of control: evidence from
dual-class firms”, Journal of Financial Economics, forthcoming.

BIS Papers No 30 127

Errunza, Vihang and Etienne Losq (1985): “International asset pricing under mild
segmentation: theory and tests”, Journal of Finance, 40, 105-24.
Esty, B C (2002): “When do foreign banks finance domestic investment? New evidence on
the importance of legal and financial systems”, Harvard Business School, working paper.
Eun, Cheol S and S Janakiramanan (1986): “A model of international asset pricing with a
constraint on the foreign equity ownership”, Journal of Finance, 41, 897-914.
Henry, Peter Blair (2000): “Stock market liberalization, economic reform, and emerging
market equity rices”, Journal of Finance, 55, 529-64.
Kaminsky, Graciela L and Sergio Schmukler (2002): “Short run pain, long run gain: the
effects of financial liberalization”, unpublished George Washington University and World
Bank working paper, June.
La Porta, R, F Lopez-de-Silanes, A Shleifer and R W Vishny (1997): “Legal determinants of
external finance”, Journal of Finance, 52, 1131-50.
——— (1998): “Law and Finance”, Journal of Political Economy, 106, 1113-55.
——— (2000): “Investor protection and corporate governance”, Journal of Financial
Economics, 58, 3-27.
——— (2002): “Investor protection and corporate valuation”, Journal of Finance, 57,
1147-70.
La Porta, R, F Lopez-de-Silanes and A Shleifer (2002a): “What works in securities law?”,
Harvard University, Yale University and Harvard University, working paper.
La Porta, R, F Lopez-de-Silanes and A Shleifer (2002b): “Government ownership of banks”,
Journal of Finance, February.
McCauley, Robert and Y C Park (2006): “Developing the bond market(s) of East Asia: global,
regional or national?”, this volume.
Miller, Darius P and John J Puthenpurackal (2002): “The costs, wealth effects, and
determinants of international capital raising: evidence from public Yankee bonds”, Journal of
Financial Intermediation, 11, 455-85.
Morck, R, B Yeung and W Yu (2000): “The information content of stock markets: why do
emerging markets have synchronous stock price movements?”, Journal of Financial
Economics, 58, 215-60.
Obstfeld, Maurice (1998): “The global capital market: benefactor or menace?”, Journal of
Economic Perspectives, 12, 9-30.
Rajan, R and L Zingales (1998): “Financial dependence and growth”, American Economic
Review, 88, 559-86.
Siegel, J I (2005): “Can foreign firms bond themselves effectively by renting U.S. securities
laws?”, Journal of Financial Economics, 75, no 2, February, 319-59.
Stapleton, R and M Subrahmanyan (1977): “Market imperfections, capital market equilibrium
and corporate finance”, Journal of Finance, 32, 307-319.
Stulz, Rene M (1999a): “International portfolio flows and security markets”, in Martin
Feldstein (ed) International capital flows, University Chicago Press, 257-93.
——— (1999b): “Globalization, corporate finance and the cost of capital”, Journal of Applied
Corporate Finance, 12, 8-25.
Stulz, Rene and R Williamson (2003): “Culture, openness, and finance”, Journal of Financial
Economics, 70, 313-40.

128 BIS Papers No 30

 Choice of currency
by East Asia bond issuers

David G Fernandez and Simon Klassen

Introduction
In discussing bond markets in Asia, academics and policymakers typically begin by noting
that the Asian crisis of 1997-98 in part resulted from the underdevelopment of the region’s
domestic bond markets and the resultant currency and duration mismatches. When
assessing the progress made in developing these markets in the post-crisis years,
academics and policymakers usually observe that, while several domestic currency
government bond markets have moved ahead, corporate bond markets have lagged (Asian
Development Bank (2002), Reserve Bank of Australia (2003)). The policy conclusion is
therefore often drawn: to prevent another Asian crisis, Asian bond markets must be further
developed.
This paper has two objectives, one straightforward and factual, the other more speculative.
First, we let the data on Asian bond issuance speak for themselves, finding that since 2000,
primary issuance by Asian corporates in local currency has far eclipsed US dollar-
denominated paper. We conclude that the post-crisis growth of the domestic currency
corporate market is underappreciated (this section expands on points made in Fernandez
and Klassen (2003)). We comb through the data on issuance to point out cross-country,
cross-sector and duration differences, but the overall message is that corporate issuance,
per se, has grown significantly. Second, looking forward, we argue that the problem of Asia’s
corporate bond market development will not be one of the supply of domestic currency
obligations. Rather, it will have more to do with demand: demand side factors that lead to the
contrast between a liquid US dollar bond market in Asia and the relatively illiquid local
markets. The difference is a by-product of financial market globalisation generally and is one
that has drawn investors in Asia towards structured products and away from “plain vanilla”
local currency issues.

About the data

The focus of this paper is on Asian corporate, financial and quasi-sovereign issuance data
from BondWare, a database compiled by Dealogic, an independent data distributor. The data
start in 1980 and cover an extensive range: 10,157 individual bond issues by a total of 2,388
issuers, for a total issuance amount of USD 667 billion. We look at the issuance patterns in
10 Asian economies excluding Japan - China, India, Indonesia, Hong Kong SAR, Korea,
Malaysia, the Philippines, Singapore, Taiwan (China) 1 and Thailand - and focus on the data
after 1998, when local currency issuance first started to capture the market’s attention.
Restricting ourselves to this data source when looking at Asian corporate bond issuance
introduces a bias into our analysis. One of the purposes of BondWare data is to permit the
ranking of bookrunners (the lead underwriters controlling the distribution of paper) by amount

1
Hereinafter Taiwan.

BIS Papers No 30 129

or number of deals underwritten); indeed, the data are effectively provided by the
bookrunners. For this reason, BondWare only captures those issues that the bookrunner
wants to advertise. This means that small-sized bonds, reverse inquiries (investor-driven
deals) under private placement and the like may be underrepresented in our sample. In
addition, we suspect that BondWare’s historical closeness to the US dollar market and
developed markets generally skews its coverage away from local currency bond markets.
This sounds like a serious problem, but in fact this is a bias that serves our purpose well.
We are interested in documenting the current development of the region’s local currency
corporate debt market. Issuance that goes unadvertised, is placed directly in a limited
number of hands, and hence is hidden from the BondWare database is likely similarly to be
hidden from the wider investor base. The bonds captured in BondWare therefore represent
the state of internationalised corporate bond issuance - the ideal case.
For comparable data on government issuance, we use our own data compilation, based on
the respective national treasuries. Issuance that is the direct result of sterilisation of central
bank purchases of foreign exchange is excluded (though McCauley (2003) makes interesting
points about the potential synergies between the two markets).

Government issuance has led the way for local bond market
development
Issuance of local currency government bonds has grown substantially since the Asian crisis
(Klassen (2004)). Taking two snapshots of issuance, one in 1999 and one in 2003, shows that
government issuance has grown roughly 40% over that period (Table 1). The amount of
renminbi-denominated bonds issued by the Chinese government remains the largest in non-
Japan Asia, but considerable increases in issuance have come out of other economies that
have run substantial fiscal deficits, such as Malaysia, the Philippines, India and Taiwan.
Korean issuance to refinance financial sector restructuring and for other needs has also risen.

Table 1
Government gross bond issuance
In billions of US dollars

Country Local 2003 Local 1999 USD 2003 USD 1999

China 45.5 48.6 1.0 –

Hong Kong SAR 1.7 0.8 – –
India 31.1 19.6 – –
Indonesia 1.4 – – –
Korea 28.8 15.8 1.0 –
Malaysia 12.6 3.7 – 1.0
Philippines 17.0 1.5 2.7 3.7
Singapore 4.9 3.7 – –
Taiwan, China 13.3 8.8 – –
Thailand 1.4 10.7 – –
Total 157.7 113.2 4.7 4.7
Source: National data compiled by JPMorgan Chase.

130 BIS Papers No 30

These snapshots of government issuance also show that sovereign funding done in US
dollars has risen. With the exception of the Philippines (which in certain years, such as 1999,
actually issued more in dollars than it did in pesos), the Asian sovereign financing mix is very
skewed towards local currency bonds. Sovereign dollar issuance from China and Korea in
2003 was clearly for benchmarking and profile purposes.
So, not only have local currency government bond markets grown in size over the past
several years, but sovereigns in Asia have displayed a clear preference for local currency
over US dollar funding. The mix between local and foreign funding by sovereigns in Asia over
the period 1998 to 2003 can be seen in Graph 1. First, there is a group of governments that
issue exclusively in local currency (Hong Kong, India, Singapore and Taiwan) and another
group that does so almost exclusively (China, Indonesia and Thailand). In the middle are
Korea and Malaysia, whose government issuance is still skewed towards local funding, but
who tend to have a more balanced mix. And, finally, there is the Philippines exception, which
was referred to above. Finally, we take note of the maturity of government issuance, because
it will be echoed in the next section on corporate issuance. For governments and quasi-
government entities that issue in both local and foreign currency, their issuance in US dollars
tends to be further along the yield curve, especially in the Philippines and Malaysia. In the
case of Thailand, dollar funding, like baht funding, tends to be of short maturity.
Among the various other characteristics necessary to make government bond markets
effective is liquidity. McCauley and Remolona (2000) show that, with respect to government
bond markets, larger size tends to lead to more trading and greater liquidity. When it comes
to the Asian government bond markets, it seems fair to conclude that size itself is not the
limiting factor in market development.

Size matters and Asia has it: domestic corporate bond markets have
grown considerably
Before the Asian crisis, the words “liquid” and “internationalised” could not have been used to
describe any of the corporate, local currency debt markets in non-Japan Asia. The
BondWare data (recalling the biases cited earlier) show that in 1998 corporations in Asia
excluding Japan issued a paltry USD 222 million-equivalent of local currency bonds
(Graph 2, with amounts converted at prevailing exchange rates). What is effectively a zero
line for corporate, local currency issuance across the region continued through the 1997-98
regional crisis. On the other hand, the same graph shows that there was some foreign
currency issuance before the crisis, though it largely came from one entity: the Singapore-
headquartered Asia Pulp and Paper. By the end of 1998, total US dollar corporate issuance
out of the region stood at USD 8.8 billion.
Another way to think about this very skewed currency composition of debt issuance is that
local currency bonds made up less than 2% of total corporate issuance before the Asian
crisis and a mere 6% during the crisis period. So, the problem of Asia’s corporate, local
currency bond markets before the regional financial crisis could not have been more basic -
they almost did not exist but for a few examples in Hong Kong SAR, Singapore and possibly
Korea. The obvious first step towards creating markets that could ameliorate the double
mismatch of currency and duration was for the domestic, corporate bond markets to grow
from their paltry size.
Looking at the data, it is immediately apparent that the growth in the size and the shift in the
composition of Asia’s corporate bond markets has been even more striking than those of
Asia’s government bond markets. Since the Asian crisis, local currency bond issuance by
corporations has soared: Graph 2 shows that, by 1999, corporate issuance in local currency
surpassed that in US dollars, and since then the race has not been a contest. Total local
currency bond issuance in non-Japan Asia was over 10 times higher in 2000 than it was just

BIS Papers No 30 131

two years earlier. This is not to say that US dollar debt issuance from corporations has stood
still since the crisis. Indeed, this asset class has expanded to such an extent that dedicated
indices, such as the JPMorgan Asia Credit Index (JACI), have been developed as
benchmarks for the investor base. Nevertheless, the growth story since the crisis is clearly in
the local corporate bond markets, with over 70% of bonds in non-Japan Asia issued in
domestic currencies since 1998 - a vast improvement from the single digit percentages cited
earlier (Graph 3).

Corporate issuance tends to mirror sovereign’s currency choices

Just as we have seen in the case of Asian governments, Asian corporations vary in their
choice of currency mix across economies. Like governments, Asia corporations in most of
the 10 economies show a clear preference for issuing in their own currencies. Indeed, there
is a striking similarity between government and corporate behaviour: the currency choice by
corporations tends to closely mirror that of their respective sovereigns (Graph 4).
For Thailand, Taiwan (China), India, Korea, Malaysia and the Philippines, the choice of
currency for bond issuance by sovereigns and corporations is very similar. At one extreme,
Thai and Taiwanese firms effectively issue all of their debt in local currency (just like their
sovereigns), while at the other extreme, Philippine firms raise most of their funding from the
US dollar market (just like their sovereign).
The exceptions are worth noting. Clearly, in the cases of Hong Kong and Singapore, the
local multinational firms have a natural need for some US dollar funding. In addition, the fixed
exchange rate policies of Hong Kong and Malaysia, and to a lesser extent Singapore, create
a natural tendency for firms to rely on US dollar financing. We would also recognise the
biases of the BondWare data in the case of China, where corporate funding in renminbi may
be a higher proportion than that shown if many issues occur “under the radar” in the sense
that they are placed with local banks and are not accessible to investors generally. In
addition, until recently, tax incentives actually encouraged Chinese issuers to borrow in US
dollars. Finally, for Indonesia, note that the numbers in dollar terms are very small and that
government issuance in local currencies has only just resumed, while small dollar issuance
by corporations became possible only in 2002.
This post-crisis transition to a preference for local currency funding by Asian corporations
also varies across countries. To highlight the variation, Graphs 5 and 6 display the dollar
amounts of domestic and foreign currency issuance by firms from Thailand and Singapore,
respectively, from 1994 to 2003. In the case of Thailand, the switch was dramatic. Pre-crisis,
the exchange rate stability, the interest rate premium on baht and other factors led Thai
corporations to fund themselves in US dollars. That funding source evaporated during the
crisis and since then almost all issuance has been in local currency, including substantial
amounts issued to repay foreign currency-denominated debts. In the case of Singapore, the
transition by firms was much more gradual and the outcome more balanced. Nevertheless,
the pattern of increasing amounts of issuance in local currency is also apparent.

Other characteristics of Asian corporate local bond issuance: maturity,

size, sectors
Along with overall market size, there are a host of other characteristics that are associated
with a well functioning corporate bond market: issuance across the yield curve, large
individual issuance size, and issuance from a variety of sectors, to name a few. We take a
last look at the BondWare data on Asian corporations and highlight these patterns. While the

132 BIS Papers No 30

amount of local currency issuance has clearly grown in many Asian countries, there are other
dimensions in which this supply could still develop.
Maturity: Graph 7 shows that the maturity of local currency corporate issuance is similar to
that seen in government bonds. That is, corporate issuance in local currency is heavily
skewed to the very short end. For issuance above five years, foreign currency bonds are
favoured over local bonds.
Size: Graph 8 shows that, in general, issuance in local currency tends to be done in smaller
sized bonds. On average, issue size for local currency issuance is less than half that for
foreign currency issuance. It should be noted that, over time, especially in the past two years,
that size gap is narrowing.
Sector: Graphs 9 and 10 show that, in both local and foreign currency bond markets,
issuance from financials dominates.

Prospects for the issuance of Asian local currency corporate bonds are
bright
So, the overall message regarding the supply of local currency corporate bonds in Asia is
that the situation today is already very positive. Governments had taken the lead after the
Asian crisis and corporates have followed.
And the future of corporate bond supply looks even brighter. Simply looking at scheduled
redemptions (Graph 11), reflecting the short-dated characteristic of local bonds pointed out
above, there is more issuance in the pipeline. Once the stock of debt has been built up, as it
has over the past several years, issuance tends to be perpetuated: supply begets more
supply.

Thoughts about the demand side of the market

In closing, we turn our thoughts to the much-neglected demand side of the equation with
respect to Asian corporate bond market development. In our view, the benefits of this fast-
growing local currency bond market that we have documented have not been spread evenly.
Much attention has rightly been paid to the importance of building a primary bond market for
Asia’s corporate issuers. But let us spare a moment to think about investors’ needs. Indeed,
even though significant progress has been made on the supply front over the past several
years, the constraint on developing Asian secondary bond markets looks likely to come from
issues on the demand side.
In general, the relationship between issuer and investors is not necessarily a zero-sum
game. True, for each issue, a lower coupon benefits the issuer at the direct expense of the
investor, but lower debt service costs for a firm reduce the likelihood of default, which
benefits both. There is also a virtuous cycle in which the lower cost of capital reduces
inflation pressures; lower inflation reduces financial volatility; lower volatility is rewarded
through lower borrowing costs. Many of these factors have been at work in Asia.
Another area of obvious benefit for both lenders and borrowers is the development of
secondary market liquidity. Investors benefit from the flexibility of being able to transfer risk
before maturity - they are prepared to pay a premium for this which then lowers the interest
costs for issuers. However, a different kind of liquidity - the excess deposit liquidity in the
Asian banking sector - already benefits issuers and, as such, the lack of secondary market
liquidity is primarily disadvantaging investors. Indeed, the clearest benefits for issuers have
come from more traditional demand side pressures: viz the excess liquidity in the region’s

BIS Papers No 30 133

banks (Graph 12). Comparing this increase in excess liquidity with the growth in issuance, it
can be seen that issuers have benefited through the increased demand.
The result has been a shrinking of credit spreads on Asian dollar bonds (Graph 13). While
the exact determinants of this spread tightening are difficult to pin down, we would certainly
attach significant importance to the excess liquidity in the banking sector that has prompted
the so-called “Asian bid”. Whatever the causes, lower spreads have reduced the opportunity
for investors to earn yield by taking credit exposure.
Ideally, this imbalance should be corrected as new issuers enter the market and widen
spreads. But the low yields have not enticed enough new entrants. One reason is that that
the universe of issuers in Asia is very narrow and issuers lack diversity. Additionally, low
bond yields have also had to compete with another form of capital-raising - equity financing.
With a high demand for global diversity in equity portfolios and the low actual supply, Asian
equities are arguably even more mispriced than credit spreads. Accordingly, firms find equity
issuance more attractive than bond issuance.
Investors, recognising the supply imbalance, have instead shifted the focus of their demand
away from “plain vanilla” corporate issuance. As an alternative, investors have set their
sights on structured products. Effectively, investors do not believe that current yields
adequately compensate them for taking the credit risk entailed in buying corporate bonds
and accepting the lack of secondary market liquidity. Instead, structured products which offer
principal protection and leverage market risk (either yield curve shape or volatility), are
perceived to offer a higher yield for the same liquidity constraints. In essence, structured
products work through taking leveraged bets on observable market prices.
Examples of structured products offered to Asian investors are now numerous:
• Convertible bonds: these pay coupons like a regular bond but the principal is
redeemed in equity subject to the performance of the stock and at the discretion of
the issuer.
• Cover call selling: the investor buys a standard bond but sells potential upside to
lock in a higher up-front yield.
• “Quanto” structures: the principal is issued and redeemed in one currency but the
coupon is indexed off another, higher-yielding, currency yield curve.
• Range accrual notes: these pay twice the current market coupon if an observable
market interest rate stays been two agreed levels, but half the coupon if the interest
rate moves outside the agreed range.
• Target redemption notes: these guarantee principal protection and a higher than
market yield but the redemption date is not fixed at issuance.
• Collateralised debt obligations: these tier both the returns and the risks of standard
bonds into low-risk and low-return tranches, on the one hand, and high-return, high-
risk tranches on the other, and may use derivative products to tailor characteristics
to demand.
A common trait across all of these products is that, so long as the investor holds it to
maturity, the principal is protected. Of course, this is the same as with a standard bond,
barring default. In contrast to the credit risk of direct lending to corporations, however, with
market-linked structures the investor limits credit exposure to the default risk of the issuing or
arranging banks and even this risk is sometimes reduced by a special-purpose vehicle that
segregates the original capital. These products’ returns can also be guaranteed in the
currency of the investor, meaning the risk of the currency mismatch between investor and
issuer is borne by the arranger or passed onto the market. And, although structured products
are often thought of as complex and confusing, they are indexed off tradable market prices,

134 BIS Papers No 30

and so could be considered more transparent than the domestic credit market, with its lack of
market standards and, by global standards, poor disclosure.
At this point it is worth discussing whether Asia is any worse off in the current situation:
issuers fund cheaply in the credit market, investors in structured products have their principal
protected and earn their target yield. The benefits of structured products are clear, they are
tailored to meet both sides’ cash flow requirements subject to their return and risk profiles.
Structured products can be designed to embody yield curve risk, option risk or direct credit
risk, alone or in combination. Although the risks are more complex, they are not necessarily
higher than the naked credit exposure of owning corporate paper.
Tellingly, just as the demand for corporate bonds pushed down credit spreads, buying of
structured products has flattened yield curves and lowered the premiums on financial
options, so the attraction of earning income through structured products is waning compared
with owning simple credit. Another concern is that, although we argued that credit spreads
are too tight and do not compensate for the illiquidity (cost of transferring the risk to someone
before maturity), structured products are even less liquid. A holder of a publicly tradable
corporate bond can sell it to any other investor in the market. Due to the customised nature
of structured products, the buyer’s recourse is to the original arranger, so the cost of
unwinding is much more than even the wide bid-offer spread on Asia corporate paper.
The other side of structured product buying is structured issuance. Similar to the zero-sum
nature of setting a coupon for issuer and investor, a structured product issuer is taking the
risks of intermediating between wholesale markets and buyers trying to achieve higher
returns. However, taking advantage of this potentially much cheaper funding alternative
requires from the issuer an appetite for risk and the sophistication to manage the yield curve,
option or credit risks. 2
And here is the kicker: regardless of what the threshold requirements of sophistication for
issuers to participate in this market might be, at the moment fully understanding structured
products appears beyond the scope of most Asian financial institutions. To date, only a
handful, helped by foreign banks, have issued structured products. So, it is fair to say,
structured products currently sold to Asian investors almost entirely benefit issuers
headquartered outside Asia. This is obviously not the ideal case, and brings us back to the
original problem of Asian savings exiting the region and Asian issuers being disadvantaged.
An obvious solution to this might be to encourage the region’s institutions to issue their own
structured products. Indeed, foreign expertise has already transferred pricing skills to the
region’s banks, which have then used their networks to market structured products to their
customers. But for all the merits of structured products, they do not represent the ideal
solution. They are designed to be customised for counterparties, not commoditised for
liquidity. Straight bonds are still needed for corporate benchmarking - as a derivative, some
structured products are meant to derive their value from the very cash bonds that they are
currently replacing.
So, indeed, Asia corporate bond markets have come a long way. The primary market is now
liquid and deep enough to cover most of the needs of the current issuer base. But
imbalances, especially the excess liquidity in the banking sector, have benefited issuers, the
supply side of the equation. In the future, the market’s success for issuers may be
constrained by the costs to investors. Investors, seeking alternatives to raise yields, are
increasingly turning away from standard corporate fixed income instruments and towards
investments such as structured products. Clearly this is not ideal as, collectively, the myriad

2
Although arguably, given the zero-sum nature of derivatives, the same level of sophistication should be
required for the Asian buyers of structural products.

BIS Papers No 30 135

risks involved are more complex than simple credit risks, and at the margin the benefits
accrue to the structured product arrangers at the expense of both issuers and investors.
Efforts are needed to restore the balance between supply and demand in Asia’s corporate
bond markets, primarily now helping investors though increasing the diversity and scope of
issuer alternatives.

136 BIS Papers No 30

 Chart 1
Currency mix of government bonds
Per cent of total, 1998-2003

100
Foreign
90
Local
80

70

60
50

40

30

20

10
0

PH MY KR ID TH CH SG HK TW IN

Source: BondWare, December 2003.

Chart 2
Currency mix of corporate bonds
USD billions

90
Local Foreign
80

70

60

50

40

30

20

10

0
1998 1999 2000 2001 2002 2003

Source: BondWare, December 2003.

BIS Papers No 30 137

 Chart 3
Asian corporate issuance by currency
Per cent, 1998-2003

USD
26%

Yen
3%
Euro
2%
Intra-Asian
< 1%

Local Currency
69%

Source: BondWare, December 2003.

Chart 4
Foreign currency mix of government and corporate bonds
Per cent of total, 1998-2003
100

90 Govt Corp
80

70

60

50

40

30

20

10

0
TH TW IN KR ID MY CH SG HK PH

Source: BondWare, December 2003.

138 BIS Papers No 30

 Chart 5 Chart 6
Thailand's switch to local financing … while Singapore’s was
was dramatic ... more balanced
USD millions USD millions
7,087
Domestic
4,272
Domestic Foreign (Inverted)
3,245
Foreign (inverted) 1,966
2,993
1,623 1,849 1,949
2,445

759

-300 -279 -300 -386

-609
-972 -1,130-1,200 -935
-1,479-1,445
-2,356
-2,148
-3,071
-2,989
1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

-6,002

1994

1995

1996

1997

1999

2000

2001

2002

2003
Source: BondWare, December 2003.

Source: BondWare, December 2003.

Chart 7
Maturity structure of corporate bonds
Per cent of total

90

80 Domestic Foreign

70

60

50

40

30

20

10

0
0-5 6-10 11-15 15-20 20+

Source: BondWare, December 2003.

BIS Papers No 30 139

 Chart 8
Average issue size
USD millions
350

Local Foreign
300

250

200

150

100

50

0
1998 1999 2000 2001 2002 2003

Source: BondWare, December 2003.

Chart 9 Chart 10
Local currency issuance Foreign issuance
USD billions, 1998-2003 USD billions, 1998-2003

70
200
180 60
160
50
140
120 40
100
30
80
60 20
40
20 10

0
0
Corp Financial Quasi
Corp Financial Quasi

Source: BondWare, December 2003. Source: BondWare, December 2003.

140 BIS Papers No 30

 Chart 11
Redemption schedule
USD billions

60 Local Foreign

50

40

30

20

10

0
2004 2006 2008 2010 2012 2014 2016 2018 2020 2022

Source: BondWare, December 2003.

Chart 12
Measures of excess liquidity
USD billions/per cent

350 94

1
Sterilized liquidity (LHS) 92
300 Loan/deposit ratio (RHS)
90

88
250
86

84
200
82

80
150

78

100 76
Jan 00 Jul 00 Feb 01 Aug 01 Mar 02 Sep 02 Apr 03 Nov 03

1
Combined size of Asian central bank issuance to mop up excess liquidity.
Source: BondWare, December 2003.

BIS Papers No 30 141

 Chart 13
Credit spreads
Basis points

400

JACI spread over UST

350 US BBB

300

250

200

150

100
Jan 00 Jul 00 Feb 01 Aug 01 Mar 02 Sep 02 Apr 03 Nov 03

Source: BondWare, December 2003.

References
Asian Development Bank (2002): Bond market development in East Asia: issues and
challenges, May.
Fernandez, D G and S Klassen (2003): “Asia Bond Fund crawls, but should it ever walk?”,
Global Data Watch, JPMorgan, 20 June.
Klassen, S (2004): Asian Government Issuance 2004, JPMorgan, 16 January.
McCauley, R N (2003): “Unifying government bond markets in East Asia”, BIS Quarterly
Review, December.
McCauley, R N and E Remolona (2000): “Size and liquidity in government bond markets”,
BIS Quarterly Review, November.
Reserve Bank of Australia (2003): “Bond market development in East Asia,” Reserve Bank of
Australia Bulletin, December.

142 BIS Papers No 30

 Asian bond issues in Tokyo:
history, structure and prospects

Fumiaki Nishi and Alexander Vergus

Overview of Tokyo’s capital market

This paper outlines the contribution made by the Japanese capital market to international
and in particular Asian bond finance. Drawing on one of the major global pools of savings,
the Japanese capital market has provided funding opportunities, both in debt and equity, for
international entities including sovereigns, government agencies, corporations and other
financing vehicles.
During the last several decades, Japanese money has played an important role in the
international capital markets. For example, during the 1980s, the movement of Japanese
money exerted crucial influence over the performance of US Treasuries, and all market
participants carefully monitored Japanese investors’ policy. In the area of primary markets,
Japanese securities firms monopolised the top positions in international bond underwriting
league tables for a considerable time.
However, with the burst of the bubble economy in the early 1990s, the relative importance of
Japanese money and capital markets started to decline. Japanese money retreated home,
where domestic government debt issuance increased dramatically in the 1990s. As a result,
the share of the yen has also been declining in international bond issuance. At the peak,
yen-denominated bonds accounted for about 17% of total international bond issuance, but
decreased to less than 2% last year.
However, Japan still has a huge capital market with abundant investor liquidity, so that with
right products capturing market trends there should be ample opportunities to mobilise
Japanese money for international financing. An example is the recent shift of investor
preferences from credit/spread instruments to foreign currencies and equity risk, following
the credit events of late 1990s.

Brief historical background

During late 1960s, Japanese foreign currency reserves started to grow and pressure to open
up the market and revalue the yen began to intensify. The exchange rate of the yen at that
time was fixed at 360 yen per US dollar. It was revalued to 308 yen in 1971 and moved to a
floating rate system in February 1973.
In order to divert the foreign exchange pressure, the Japanese government set out around
that time to open the Japanese capital market to foreign entities, allowing them to issue
yen-denominated bonds.
Japanese foreign currency denominated bond investment started at the same time. In 1970
the first “Samurai” - a yen-denominated bond publicly offered by an international borrower in
the domestic market - was issued by the Asian Development Bank.
Development of the Samurai market was followed by liberalisation of euroyen issuance, with
the first euroyen bond issue made by the European Investment Bank in 1977. Both
developments were aimed at reducing pressure for appreciation of the yen as well as at
liberalising the Japanese capital market. However, both the Samurai and the euroyen market

BIS Papers No 30 143

started off under strict government control and there were detailed guidelines on issuer
eligibility, bond structure and issuance amounts.
Such controls were gradually eased, attracting a wider range of market participants, and
eventually leading to full liberalisation of both markets.
The next section is devoted to the current Japanese bond market and its development,
focusing on how Asian and other foreign entities have been using the Japanese market
through Samurai and Euroyen bond issuance.

Japan’s domestic bond market

In the present-day Japanese bond market, government debt accounts for the lion’s share
both in terms of primary issuance and secondary turnover. As can be seen from Chart 1,
government bond issuance started to surge in the mid-1980s as fiscal policy was deployed in
order to stimulate the economy, reaching nearly 80% of all bond issuance in 2003.

Chart 1
Public debt and deficit levels since 1984
170 -9
General govt debt as a % GDP (lhs)
160
-8
General govt deficit as % GDP (rhs)
150
-7
140
130 -6
120
-5
110
-4
100
90 -3
80
-2
70
-1
60
50 0
CY84
CY85
CY86
CY87
CY88
CY89
CY90
CY91
CY92
CY93
CY94
CY95
CY96
CY97
CY98
CY99
CY00
CY01
CY02
CY03
CY04

Note: 2004 data are estimated.

Source: Ministry of Finance, Japan.

In particular, after the bubble burst at the beginning of the 1990s, Japan government bond
issuance accelerated, reaching an estimated 36.4 trillion yen of new bond issuance in
FY2003 and bringing the outstanding volume to approximately 500 trillion yen, or an
estimated 140% of GDP, at the end of FY2003.
Development of the corporate bond market in Japan has been a slow process, due mainly to
the strong influence of commercial banks interested in preserving the importance of lending
as the main channel for corporate financing. However, during the last decade, with banks
weakened by non-performing loans, corporate bond issuance has been increasing.
For a number of reasons related to events in Japan and internationally, the bulk of the
international yen bond market is currently taken by euroyen issuance, with the Samurai
market carving out only a very small share, as can be seen in the breakdown of new
issuance presented in Chart 2 below.

144 BIS Papers No 30

 Chart 2
FY2002 bond issuance by instrument

Source: Bank of Japan.

The rather conservative Japanese individual investor portfolio constrains the development of
Japan’s bond market. Household financial assets numbering 1,400 trillion yen
(US$ 13 trillion) are for the most part held in cash and bank deposits, complemented by
significant insurance and pension policy claims, leaving only a relatively small portion of the
market for direct securities purchases by Japanese households (Chart 3).

Chart 3
Japanese household financial assets,
breakdown and investment channels
As of end-September 2003

Foreign Foreign
Securities Currency
0.6% Deposits
Equities 0.3% Cash
6.4% 2.4%
Domestic
Bonds
4.2%

Others
16.0% Yen
Deposits
45.4%

Pension
9.0%

Insurance
15.6%

Source: Bank of Japan.

At present, current deposits enjoy unlimited government guarantees in case of bank failure;
however, the introduction of a ceiling on such guarantees (10 million yen per depositor per
bank), ie the so-called pay-off system, is planned to take effect in April 2005, and is expected
to cause significant movement of retail money out of deposits into other financial instruments.

BIS Papers No 30 145

Therefore, current major investors in the Japanese capital market are the banking and
insurance sectors, against the backdrop of risk-averse household saving behaviour. Banks
and insurers had traditionally channelled funds to the ever-expanding corporate sector;
however, slowing economic growth in the mid-1980s created industrial overcapacity and the
consequent shrinking of domestic demand for loans. Excessive property investments
followed, creating Japan’s notorious bubble economy of the second half of the 1980s. The
burst of the bubble in 1989-1990 has left a legacy of bad loans and huge loss of savings
value, the consequences of which Japan is still dealing with.
As can be seen in Chart 4, as a result of these trends, Japanese banks had to cope with a
continuously widening gap between liabilities (deposits) and assets (lending), causing them
to dramatically increase their bond holdings. To a lesser extent, the same applied to the
insurance sector. Recent bond holding trends by major investors are shown in Chart 5 below.

Chart 4
Asset-liability mismatch of the banking system
(Yen trn)
540
Deposits
520

500

480
App. 100
trillion yen
460

440 Lending

420

400
94 95 96 97 98 99 00 01 02 03

Source: Bank of Japan.

Chart 5
Bond holdings
(Yen trn) (Yen trn)
5,000 500
446
4,500 450 Large Banks 418
Private Investors
4,000 400 Insurance Co.s
Public Sector 348
318 330 328
3,500 350 298
3,000 2,869 300 256 270
2,500 2,537 250
1,756 2,141 191
2,000 1,554 200
1,584
1,500 1,375 1,460 150 116 114
1,000 100
1,612
500 890 1,186 1,351 1,291 1,270 50
719 819
- 0
FY94 FY95 FY96 FY97 FY98 FY99 FY00 FY01 FY98 FY99 FY00 FY01 FY02 FY03

Source: Bank of Japan.

Increased inflow of funds into the bond markets has tended to push down interest rates in
Japan since the peak of 1989. As can be seen in Chart 6, while reacting to political and other

146 BIS Papers No 30

factors, interest rates had followed a long-term falling trend, until finally hitting bottom in June
2003, and jumping since then back to the levels of around 2001. In another unique
development, as shown in Chart 6, spreads between banking interest rate swap contracts
and government bond yields also virtually disappeared around 2001, as markets attached
the same risk as the government to the banking system enjoying the government’s umbrella
guarantee. Interest rates still remain at very low levels which are expected to persist as long
as the Bank of Japan maintains zero short-term interest rates in a recovering economy that
still suffers from deflation.

Chart 6
Government bond yields and yen IRS rates
for 5 and 10 years since 2000

(%)
3.00

2.50

2.00

1.50

1.00

10yr JGB
0.50
10yr \/\ Swap
5yr JGB
5yr \/\ Swap
0.00
00/07 00/10 01/01 01/04 01/07 01/10 02/01 02/04 02/07 02/10 03/01 03/04 03/07 03/10 04/01

Source: Nomura.

The Samurai bond market

As can be seen in Chart 6, the amount of Samurai bond issuance accounted for a very small
share of issuance in 2003 - only 0.3%.
As we have mentioned above, the Samurai market was set up under pressure to alleviate the
external imbalances of the late 1960s, and the inaugural issue was made by the Asian
Development Bank in November 1970. The issue amount was 6 billion yen with a 7-year
maturity, and it was accepted very well in the market.
Initially, high credit issuers such as the ADB or other supranationals were given priority in
access to the market. In order to control the flow of issuance, eligibility criteria were
established.
For issuers unable to satisfy such eligibility criteria, there was another way to finance yen
through yen-denominated private placement bonds. They were allowed to target a limited
number of institutional investors, and issuance terms and liquidity were tightly controlled.
Yen private placement issuance started in 1972; however, with the liberalisation of the
Samurai and euroyen markets, the raison d’être for private placements diminished, and the
market shrank and was replaced by other means of targeting a limited circle of investors.
In the Samurai market, in July 1972 Australia issued the first sovereign bond. The market
continued to be gradually opened for more issuers through the easing of eligibility criteria and
the liberalisation of new types of bonds. Complicated eligibility criteria were subsequently
replaced by requirements for minimum credit ratings and finally abolished.

BIS Papers No 30 147

Later on, bond structures were diversified to allow dual currency or reverse dual currency
bonds, with the principal and interest payment in different currencies, while the issuance of
floating rate notes and index-linked structures were also liberalised.
As can be seen in Chart 7 below, the Samurai market developed gradually before peaking in
1996, and in the course of its development, it helped to liberalise Japanese bond market in
many ways.

Chart 7
Historical Samurai issuance
(Yen bn)

4,500

4,000 3,873.7

3,500

3,000

2,500
2,232.9
2,123.3
2,000
1,700 1,656.5
1,552.4
1,500 1,173
1,115 1,126 1,194 1,162.5

1,000 915
811.8
722 663 720 681 660.5
590 635 638.6
495 420
500 296 333 261
205
6 33 85 40 0 20 65
0
1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002

Source: Nomura.

Looking at the ratings breakdown of Samurai issuance during its most active period, one can
perceive the activity of AA borrowers with structured Samurais in 1994-1998 (popular as the
yen was steadily depreciating after hitting a peak in 1995), while between 1999-2001 the
market was open mainly for BBB, and more recently mostly for single A, ratings.

Chart 8
Samurai issuance trends by rating

350
300
250
200
150
100
50
0 AA
1994

A
1995

1996

1997

BBB
1998

1999

2000

BB
2001

2002

O ct-03

Source: Nomura.

148 BIS Papers No 30

Many borrowers have tapped this market, and for several countries the Samurai market was
at times the most important foreign bond market. Many issues originated from Asian entities.
Singapore opened the market for Asian sovereigns in 1976, while the Korea Development
Bank made its debut issue in January 1978, followed by multiple issues by various Korean
entities. The Asian presence in the Samurai market is shown in Chart 9 and Table 1 below.

Chart 9
Samurai bond issuance in Asia and share of the total

600 35.00%

30.00%
500

25.00%
400

20.00%
300
15.00%

200
10.00%

100
5.00%

0 0.00%
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
Oct- 03
Source: Nomura.

Table 1
Samurai bonds by Asian issuers, 1970 to date

Amount (yen bn) No of issues

South Korea 1,461 69

Australia 1,249 64
China 941 49
Malaysia 292 14
ADB 252 9
Thailand 215 16
Hong Kong 148 12
India 141 6
Philippines 90 4
New Zealand 45 2
Source: Nomura

BIS Papers No 30 149

 Table 2
Deregulation of the Samurai market
Volume1
Deregulation measures Remarks Issues1
(Yen bil)
1970 Supranational issuance allowed ADB issues the first Samurai bond 1 6
1972 Sovereign issuance allowed Australia becomes first sovereign
Yen private placements by non- Samurai issuer, Mexico and Brazil
residents started debut in 1973 6 85
1979 Corporate issuance starts Sears Overseas Finance NV issues
the first corporate Samurai bond 16 333
1984 Public issuer rating criteria Bank of China issues debut Samurai
broadened to single A 37 915
1986 Financial parameters and single A Samurai issuance decreases as
rating as criteria for private issuers Euroyen issuance is widely
liberalised 21 590
1988 Introduction of shelf registration First issue by Greece, later to
become one of the largest sovereign
borrowers 22 635
1989 First reverse dual currency Samurai Denmark issued the first reverse
issue dual Samurai Bond - main forex
exposure structure to date (currently
as PRDC) 47 1,126
1991 Sovereigns broadened to BBB, FRN First BBB-rated sovereign issue takes
issuance place in 1994 (Central Bank of Tunisia) 34 681
1992 Public sector broadened to BBB Ankara Municipality issues first BBB
public sector Samurai 53 1,700
1994 Private sector issuers broadened to PEMEX issues first BBB rated
BBB corporate Samurai 55 1,163
1995 First dual currency Samurai issue Popularity of dual currency structure
among retail investors as yen
depreciates 35 1,115
1996 Lifting of issuance eligibility criteria, Brazil issues the first sub-
but financial institutions remain shut investment grade bond in the
from the market Samurai market; market on the way
to record in 1996 153 3,874
1996 Foreign non-banks allowed to issue FMCC issues Samurai bond as the
for purposes other than lending first non-bank issuer
1997 Samurai issuance by overseas Citicorp becomes first commercial
commercial banks liberated bank to issue a Samurai 85 2,123
1998 Issuance of Nikkei-linked Samurai No issuance so far 14 205
bonds liberalised
1999 Foreign non-bankers allowed to Associates debut in 1999,
issue for general finance purposes Household
Finance - 7 issues since 2000 10 661
2003 Single A and higher rated foreign Several corporates have already
non-sovereigns not listed at TSE taken advantage of the measure
eligible for shelf registration 27 641

1 Excluding private placements.

150 BIS Papers No 30

Again, as can be seen, the gradual expansion of the market reached a peak in 1996 and was
arrested by the crisis of 1997, before only partially recovering around 1999, for a less
diversified range of issuers. There are several reasons for the relatively inactive market.
There has been criticism as to the rather rigid issue eligibility, the opaque pricing procedure
brought about by illiquid trading and cumbersome issue procedures. Some of these problems
have been solved. As can be seen in Table 2, all issuance eligibility criteria have been
abolished, and almost all the restrictions on types of bonds have been lifted.
However, there is still some room for improvement. The most important issue is that under
the Japanese Securities and Exchange Law, in order to make a public offering, the issuer
has to file a Securities Registration Statement with the Ministry of Finance in Japanese.
It used to take 30 days for the Securities Registration Statement to become effective,
hindering flexibility to promptly bring an issue to the market. In order to simplify this
procedure, the shelf registration system was introduced in 1988 and contributed to significant
shortening of the issuance timetable. However, there are still requests to shorten the
procedure as well as - since annual and semi-annual statements have to be filed in
Japanese - to allow issuers to use English to save both time and costs necessary for
translation.
Other practices offering room for improvement include the obsolete registration system,
whereby holdings by institutions are “recorded” by the paying agent bank, which also serves
as a “recording agency” for each issue. Retail investors, for their part, still receive expensive
printed bond certificates. Finally, the commercial law requirement to appoint commissioned
banks as trustees also increases issuance expenses in comparison to issuance using just
fiscal agents.

The euroyen market

Chart 10 below shows historical volumes of euroyen issuance.

Chart 10
Historical Euroyen Issuance

E uroyen Issuance
7,000 1,200

Japanese borrow ers

6,000 N on-Japanese borrow ers
N o.of Issues 1,000

5,000
bi
(Yenbn)

800
Amount(Yen

No. of Issue

4,000
Issue Amount

600
3,000
Issue

400
2,000

200
1,000

- -
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003

Source: Nomura.

BIS Papers No 30 151

The euroyen market made its appearance in April 1977, with an issue by the European
Investment Bank. At the initial stage, the market was highly regulated by the government and
remained very small since at the time Japanese monetary authorities considered that full
liberalisation of the euroyen market might cause difficulties in conducting its monetary and
foreign exchange policy.
In the autumn of 1983, the closed nature of Japan’s financial markets and the yen-dollar
problem escalated into a political issue. This led to the setting up in 1983 of the Japan-US
Yen-Dollar Committee, which presented in May 1984 its famous report.
This report resulted in drastic changes of the government policy. In December 1984, the
rating requirement was relaxed from AAA to single A, and non-Japanese houses licensed
under Japanese securities law were allowed to lead manage euroyen bonds. Such
deregulations accelerated issuance of euroyen bonds, and in 1985 the amount of euroyen
bond issuance exceeded that of the Samurai market. Also, by June 1987, three Japanese
rating agencies, in addition to three international rating agencies, were approved as official
rating agencies.
In 1987, according to International Finance Review, euroyen issue amount accounted for
17% of the total euro market issue amount. In June 1989, all eligibility criteria were lifted,
while in 1994, as the last major liberalisation measure, restrictions on flowback, ie the
secondary market sale of bonds issued offshore back into the Japanese domestic market,
were eased, before being fully lifted in 1995.
Since then, Euroyen issuance, with its relatively easier procedures, became even more
popular so that nowadays bonds are sometimes issued offshore even when the targeted
investor base is domestic Japanese investors. Table 3 presents the history of the
liberalisation of this market.
The market is now completely free and has established its position as one of the major global
capital markets. A wide range of issuers utilise this market, including Japanese companies
seeking funding abroad.

Other structures: Shoguns, Daimyos, globals and the MTN market

Originally both the Samurai and euroyen markets started to develop due to “gaiatsu”
(pressure from foreign countries) but internal efforts have also been made to develop the
market.
The “Shogun” - foreign currency denominated bonds issued and sold in the Japanese market
to provide a domestic instrument for investors interested in purchasing foreign currency
bonds, was introduced in 1985, with the IBRD as the inaugural issuer. The liquidity of the
market is very important to make the market attractive for both investors and issuers, and
efforts to integrate the Japanese market with the international market were subsequently
undertaken.
To that end, it was necessary to set up a system to sell and trade bonds not only in the
Japanese market but also in European and US markets. “Daimyo” bonds, which later
transformed into “global bonds”, were created based on the above concept. The idea was to
make registration necessary, to make the bond saleable internationally. At the initial stage,
supranationals were major issuers as they are exempt from registration requirements in
Japan and the US.

152 BIS Papers No 30

 Table 3
Deregulation of the Euroyen Market

Issues Volume (yen bn) Deregulation Measures

1977 2 30.0 Euroyen issuance by EIB

1978-83 22 340.0
1984 13 227.0 Broad relaxation of guidelines:
1. Broadening of issuers from supras and
governments to municipalities,
government agencies and private entities
2. Relaxation of criteria: for public
borrowers, from AAA to A; for private, A
and compliance with Samurai criteria
3. Quantitative relaxation: from 6-7 annual to
no limits on number and amounts of
issuance
4. Limits on lead managers: from domestic
only to domestic and foreign securities
houses
1985 66 1,445.7 New types of instruments allowed: FRN,
zero-coupon, discount, deep discount and
currency conversion bonds
1986 141 2,551.5 Quantitative criteria for corporates replaced
with rating only and rating criteria broadened
to A rated
No flowback period shortened from 180 to 90
days
Japanese rating agencies recognised
Issuance allowed for banks on condition that
proceeds would not enter Japan
1987 151 2,993.9 Tenors shortened from 5 to 4 years
Fitch recognised as rating agency
1988 224 2,213.0
1989 395 3,557.9 Lifting of tenor criteria
1990 512 4,980.9
1991 314 3,290.4 IBRD issues first global yen; the issue is
exempt from flowback limits
1992 250 3,328.0
1993 640 5,084.8 Full lifting of issuance criteria
1994 2,030 10,085.1 Lifting of flowback limitations for sovereign
issues
1995 3,015 11,515.6 Lifting of all flowback limitations
1996 5,345 14,937.4
1997 5,492 16,952.2
1998 NA 12,300.0 Under the new Foreign Exchange Law,
advance approval requirement replaced by
filing of post-transaction report

BIS Papers No 30 153

Global dollar bonds and global yen bonds were issued mainly during the 1990s. However, to
integrate disclosure requirements, taxation issues and the other administrative requirements
of all the markets is not easy, and the majority of issuers continued to be supranationals that
enjoy exemption from registration. As a result, as can be seen in Charts 11-13 below, the
number of bonds issued has not been that impressive. (We have also presented respective
issuance lists and an overview chart of different structures in the annexes).

Chart 11
Historical Shogun issuance
Shogun Issuance
(USD mil) (No. of Issues)
1000 10
881.2 Amount (USD mil.)
755
800 8 No. of Issues
8 674

600 6
5
400
180 198
175
200 2
1
1 1 50
0 0
85 86 87 88 89 90 91 92 93 94

Source: Nomura.

Chart 12
Historical Daimyo issuance

Daimyo Issuance
(Yen bn) (No. of Issues)
400 10
9 Amount (Yen bn)
305 No. of Issues
300 280
7
190
200 5 5
130 140
3 3
100
40
1
- 0
87 88 89 90 91 92

Source: Nomura.

154 BIS Papers No 30

 Chart 13
Historical global yen Issuance

Global Yen Issues since 1992

(Yen bn)
(No. of Issues)
1,600 1,459 25
Amount (Yen bil.) 23
1,400
No. of Issues 20
1,200
1,000 15
800 675
556 10
600 475 425 440
7 330 7290
400
6 5
200 2 2 3 5
0
92 93 94 95 96 97 98 99 00 01 02 03

Source: Nomura.

Other efforts have been made to develop new products to satisfy very liquid but conservative
domestic investors. Asset-backed securities, index-linked securities and other types of new
products have been marketed in Japan. But products offering foreign exchange risk and
equity risk have traditionally been, and still are, the most popular ones.
Also, many relatively small, tailor-made transactions targeting the specific requirements of
institutional investors are executed through issuers’ MTN programmes, with numerous small
MTN transactions adding up to an ample market volume, as presented in Chart 14. Again,
most issues tend to offer foreign currency exposure through some structure, with the “power
reverse dual currency” being the most popular recently.

Chart 14
MTN issuance
Total Amount and Number of Issues
3000

2000

1000

0
1998 1999 2000 2001 2002 2003
Total Size (Bil Yen) 663.81556 776.13905 988.079069 1329.221434 1120.95 1026.25
Total Number of Issues 426 635 1145 1360 1951 2552

Structure Type

R-FRN CMS Other

4.0% 1.4% 1.2%

PRD
93.4%

Source: Nomura.

BIS Papers No 30 155

The retail market: the Uridashi
In order to make the domestic public offering of foreign securities possible, in 1994 “Uridashi”
registration was introduced. Under this registration, bonds denominated in yen or another
currency and issued overseas can be sold in Japan through a secondary market registration
and selling contract between a domestic Uridashi agent and an overseas underwriter.
This registration has enabled securities firms to efficiently satisfy domestic retail demand for
both foreign currency fixed income and equity-linked products. Some yen straight bond
issuers have also sometimes opted for a “euroyen with an Uridashi” structure, notably the
Korea Electric Power Corporation.
The Uridashi registration most efficiently serves an issuance structure in which a top-rated
foreign, usually corporate, entity issues a foreign currency straight or structured (including
equity-linked) instrument. These products normally offer some risk other than the credit risk
of the issuer, so that issuers are limited to low-risk entities, while the risk offered is that of a
currency popular among retail investors, or of an equity exposure at times of attractive equity
market.
Naturally, in recent years, foreign exchange products have enjoyed massive retail demand
as Japanese interest rates reached virtually zero. The range of currencies tends to be
confined to those of the G7 countries. The US dollar recently conceded popularity to the
Australian dollar, with its high interest rates and the robust performance of the Australian
economy, the euro on its recent strengthening trend, and, to a lesser degree, the sterling and
the Canadian dollar (Chart 15). Some other currencies are also carving out a niche in this
market, most notably the South African rand.

Chart 15
Uridashi issuance 2000-2003

Forex Uridashi Bond Issues by Currency

Eqv USD bn
30.0

25.0 Others
EUR
20.0 AUD
USD
15.0

10.0

5.0

0.0
2000 2001 2002 2003

Source: Nomura.

Expansion of this market to Asian currencies is a natural issue at hand, and this will require
the resolution of a number of market-related, regulatory and technical issues, ie:
– First, acceptance by Japanese investors of AAA or AA rated (such as those issued
by the ADB) bonds denominated in Asian currencies;
– Next, opening of the market for Asian local entities issuing in local currencies; this
will require further improvement of credit stories across the region (the market
traditionally accepts AAA to AA issuance);

156 BIS Papers No 30

– In the secondary market, Japanese investors’ portfolios of foreign currency bonds
go through much more trading than those for yen bonds (other than Japan
government bonds) as investors tend to move to take profits or cut losses at times of
wide currency movements. Liquidity is therefore important, making the infrastructure
(ie market capacity, as well as such technical points as clearing and settlement
systems, hedging and money market tools, etc.) of the bond market of the
corresponding currency an important issue for the functioning of the market;
– The issue of correlation with the dollar may be the last point worth mentioning. As is
the case with Chinese yuan (renminbi), any prospects of Asian currencies
decoupling and appreciating versus the G3 group will justify the emergence of a new
market for these currencies.

Current issues
We have outlined above procedural issues still weighing mainly on the Samurai market.
However, they can and have been easily overcome, should this market be sufficiently
attractive for borrowers. Much more crucially, there are a number of important market-related
issues which have been affecting the yen markets in all their facets:
1. As Japan’s rating was downgraded to the AA/A category, it became difficult for
higher rated borrowers to tap the market. Among Japanese investors, it is invariably
difficult to sell any instrument whose terms are better than those for Japanese
government bonds.
2. Some emerging market sovereigns, as well as some companies hitherto perceived
as top-grade defaulted or suffered downgrades. Until 2001, sub-investment grade
issues had been very popular as investors were attracted by their yield pick-up amid
ever falling domestic interest rates. As a result, the Japanese market is extensively
exposed to the Argentine problem, with no prospect of resolution in sight, souring
the current market sentiment toward credit risk.
3. However, for some regular, especially Asian, visitors to the Samurai market, their
domestic liquidity became very high and the requirements for international fund-
raising diminished. In particular, some Asian countries, such as Thailand, Malaysia
and China, which used to be regular issuers in the Samurai market, have seen their
international funding requirements drastically reduced.
4. As absolute levels of interest rates elsewhere have been “catching up” with low yen
interest rates, the exchange risk associated with borrowing in yen started to
outweigh the attractiveness of yen for international borrowers.
5. The emergence of the euro as a single currency and of the associated single capital
market has attracted countries in other regions, in particular EU and EU accession
countries, to finance in the euro market, taking them further away from alternatives,
including the yen market.
6. As regards Asia, the position of the dollar remains very strong, with many currencies
pegged to the dollar and foreign reserves also still mainly in dollars. As is widely
known, recently Asian central banks, together with Japan, have been supporting the
dollar, adding record amounts of US Treasuries to their foreign reserves in the
process. It can be observed that not only the yen- but also the euro-denominated
issuance is also rather limited in the region.
7. Another obvious reason for the decline of the yen’s relative importance in the
international markets is that some Japanese demand has shifted to non-yen
currencies.

BIS Papers No 30 157

This has left international banks and companies, which require yen funding for their
Japanese domestic operations, as the only consistently present issuer category on the
Samurai market.
Among sovereign and government agency issuers, the market is currently confined to those
with BBB to A ratings, with borrowers from Korea, primarily the KDB, as the most consistent
visitors to the market, while the People’s Republic of China, the Malaysian Federation or
PETRONAS, the Kingdom of Thailand and the Republic of the Philippines are among other
potential visitors.
There are also higher rated issuers on the market, but they are essentially accessing the
market through structured types of bonds, similar to the MTN market.
For a wider community of issuers, the euroyen format, targeting investors throughout
international markets through issuance of large internationally liquid issues, still remains
attractive (while such issuance would not work through the Samurai format in terms of
flexibility, costs, etc.). Such corporates as GECC and IBM, and the Republic of Italy among
sovereigns, are most notable issuers on this market.
Any market has its cycles and, as the Japanese economy is at last showing signs of
sustainable recovery, Japanese investors have been regaining some of their confidence and
the international yen market has been gradually improving. Subject to global economic and
consequent investor liquidity, currency and credit events and other trends, the yen capital
market may regain some of its prominence in the years to come.
However, another possible reason for the retreat of the yen market may have been the
failure to make the yen a truly international currency, and the lack of medium- and long-term
policy to develop the Japanese domestic bond and euroyen markets into truly competitive
international markets. Without political and legal commitment from the government, such a
goal cannot be achieved, and in our view the authorities still have some role to play in the
further development of Asian bond markets.

158 BIS Papers No 30

 Annex 1:
The Japanese market for Asian equities

While the main focus of this paper is on bonds, the picture of capital flows from Japan to Asia
will not be complete without briefly mentioning the equity market. If the underlying theme of
this discussion is the flow of Japanese capital into Asian currencies or Asian credits, then
while such flow via bonds may have been limited, Japanese investors have made substantial
investments in Asian equities.
To briefly recall the history, the opening of the Japanese equity market started around the
same time as that of the bond market, and in 1972, the first foreign public equity offering in
Japan was made by a US company, General Telephone and Electronics, while in 1973 the
first listings on the Tokyo Stock Exchange were made by four US companies joined by their
French confrère.
Since then, although there have been ups and downs in the amount and number of
companies conducting equity financing in Japan or listed on the Tokyo Stock Exchange, the
Japanese market has over the years made an ample contribution to international and Asian
equity financing.
This contribution has been particularly visible in recent cases of major global offerings of
Chinese companies, on some of which it was reported that they “would not have succeeded
without the Japanese demand”. What in our view makes them worth mentioning is that
through these equity purchases Japanese investors have been aggressively investing in
Asian currencies and Asian corporate credits, implicitly taking Asian sovereign exposure as
well.

BIS Papers No 30 159

 Annex 2:
Issuance in various Japanese yen market segments

List of Shogun issues

Launch Amount Issue

Issuer Currency Coupon Maturity
date (mil) price

Jun-85 EIB (Shibo) ECU 50 9.2 99.75 10

Aug-85 World Bank USD 300 10.5 100.000 10
Aug-85 Sallie Mae USD 100 9.75 100.875 5
Victorian Public Authorities Finance
Oct-85 A$ 60 13.625 99.875 7
Agency
Oct-85 NSW Treasury Corp USD 100 10.375 100.875 7
Oct-85 Bank of China USD 100 10 100 10
Nov-85 Southern California Edison Company USD 100 10.5 100.65 8
Dec-85 CITIC USD 100 9.625 100.75 10
Mar-86 Farm Credit Corporation C$ 90 9.125 100.75 5
May-86 State Bank of NSW USD 100 7.125 100.9 7
Jul-86 Hydro Quebec C$ 150 9.27 100.875 10
Aug-86 Federal Business Development Bank C$ 75 9 100.95 5
Sep-86 World Bank C$ 100 9 100.85 10
Nov-86 National Bank of Hungary (Shibo) USD 55 8.875 100 7
Nov-86 NSW Treasury Corp USD 100 7.75 101.375 7
Dec-86 World Bank USD 200 7.375 101.125 7
Jan-87 Standard Oil USD 50 7.875 101.375 10
Feb-87 Ciba Geigy USD 50 7.5 101.355 7
Feb-87 FHLB USD 200 7.75 101.125 10
Mar-87 GTE Finance USD 100 8 101.3 7
May-87 Bell Canada C$ 100 9.875 101.4 9
Oct-87 Fannie Mae USD 200 10.1 100.19 7
Oct-88 EIB ECU 100 8 101.875 10
Oct-88 World Bank USD 100 9.04 100 10
Aug-89 P&O (Shogun CB) USD 175 6 100 15
May-91 EIB ECU 240 9 101.755 8
Jan-94 HNG USD 50 5 (A$) 100.6 4

160 BIS Papers No 30

 List of Daimyo issues

Launch Amount Issue Book-

Issuer Coupon Maturity
date (yen mil) price runner

04/22/87 World Bank 40,000 4.625 101.000 05/12/97 Daiwa

05/15/87 World Bank 40,000 4.250 100.375 06/03/97 Yamaichi
11/12/87 World Bank 50,000 5.625 101.125 11/27/97 Nomura
02/25/88 World Bank 100,000 5.125 100.250 03/17/98 Nikko
04/08/88 Inter-American Development Bk 20,000 5.000 100.000 04/14/98 Nikko
05/27/88 World Bank 60,000 5.125 100.000 03/17/98 Daiwa
08/06/88 World Bank 70,000 5.250 100.500 08/19/98 Yamaichi
09/08/88 ADB 30,000 5.500 100.250 09/26/98 Nomura
03/10/89 ADB 40,000 5.500 100.650 03/16/99 Nikko
04/01/89 ADB 40,000 5.125 100.000 04/14/99 Nomura
04/10/89 Inter-American Development Bk 25,000 5.125 100.375 04/19/99 Yamaichi
07/05/89 IFC 20,000 5.375 100.80 07/18/96 Nomura
09/05/89 World Bank 70,000 5.250 100.750 09/20/99 IBJ
10/17/89 Inter-American Development Bk 30,000 5.625 100.80 10/25/99 Nomura
11/02/89 African Development Bank 20,000 5.750 100.30 11/17/99 Nikko
12/14/89 World Bank 50,000 5.875 100.750 12/22/94 Yamaichi
12/21/89 IFC 10,000 7.100 100.80 12/28/94 Nomura
04/24/90 African Development Bank 40,000 7.250 100.500 05/09/97 Daiwa
06/01/90 IFC 15,000 8.000 100.500 06/14/96 Nomura
07/11/90 African Development Bank 30,000 7.250 100.375 07/16/97 Nomura
07/22/90 World Bank 30,000 7.000 95.000 07/30/97 Nomura
09/18/90 IFC 10,000 8.000 95.000 09/28/00 Yamaichi
10/19/90 ADB 30,000 7.500 100.500 11/09/00 Daiwa
12/18/90 Inter-American Development Bk 35,000 7.000 100.750 12/27/00 Daiwa
04/17/91 African Development Bank 40,000 7.250 100.875 04/24/98 Daiwa
01/23/92 ADB 50,000 5.625 99.60 02/12/02 Yamaichi
01/31/92 Inter-American Development Bk 50,000 5.375 100.450 02/25/97 Nikko
06/09/92 African Development Bank 40,000 6.200 99.80 06/18/02 Yamaichi

BIS Papers No 30 161

 List of global yen issues 1992-2003

Launch Amount Issue

Issuer Coupon Maturity Bookrunner
date (yen mil) price

03/10/92 International 250,000 5.250 99.230 03/20/02 IBJ-I/J-P-MORGAN(JB)/

Bank for NOMURA-SEC(JB)
Reconst
10/14/92 International 225,000 4.500 99.80 12/22/97 DAIWA-SEC
Bank for
Reconst
02/18/93 International 200,000 4.500 99.72 03/20/03 IBJ-I/MS-I(JB)/NIKKO-EUR(JB)
Bank for
Reconst
06/30/93 International 225,000 4.500 99.983 06/20/00 DAIWA-SEC/GS(JB)/IBJ-I(JB)
Bank for
Reconst
01/11/94 Italy 300,000 3.500 99.980 06/20/01 DAIWA-EUR/JPM(JB)
07/13/94 Italy 175,000 Floats 99.870 07/26/99 MERRILL/NOMURA(JB)
11/08/94 International 200,000 4.750 99.404 12/20/04 ML-I/NIKKO-EUR(JB)/
Bank for NOMURA-INT(JB)
Reconst
06/29/99 Inter- 100,000 1.900 99.449 07/08/09 MSDW/TMI(JB)
American
Develop-
ment Bk
07/19/99 General 50,000 Floats 100.000 07/26/02 BOTMI
Motors
Acceptance
Corp
10/05/99 Inter- 50,000 1.900 101.834 07/08/09 MSDW
American
Develop-
ment Bk
11/01/99 KFW 100,000 1.000 99.980 12/20/04 DEUTSCHE-ALEX-B
International
Finance Inc
11/09/99 Toyota 50,000 1.000 99.429 12/20/04 MERRILL/NOMURA-SEC(JB)
Motor Credit
Corp
11/12/99 General 50,000 1.250 99.72 12/20/04 BEAR/MERRILL(JB)
Motors Corp
11/30/99 Procter & 40,000 1.500 99.578 12/07/05 SALOMON-SMITH
Gamble Co
01/11/00 Westpac 50,000 .875 99.604 09/22/03 BOTMI
Banking
Corp
01/13/00 OKB 100,000 1.800 99.805 03/22/10 NOMURA

162 BIS Papers No 30

 List of global yen issues 1992-2003 (cont)

Launch Amount Issue

Issuer Coupon Maturity Bookrunner
date (yen mil) price

01/25/00 Bank of 100,000 Floats 99.968 02/01/02 DEUTSCHE-ALEX-B

Scotland
Treasury
Svcs
01/28/00 Ford Motor 100,000 1.200 99.855 02/07/05 GS/MSDW(JB)/BOTMI(JB)
Credit Co
02/03/00 Mellon Bank 10,000 1.400 Market 02/24/05 JPM
NA,
Pittsburgh, PA
02/18/00 Italy 100,000 1.800 99.882 02/23/10 MSDW/NOMURA-SEC(JB)
03/02/00 McDonald’s 15,000 2.000 99.927 03/09/10 MERRILL/MSDW(JB)
Corp
03/02/00 McDonald’s 15,000 1.250 99.859 03/09/05 MERRILL/MSDW(JB)
Corp
03/08/00 KFW 100,000 1.750 99.735 03/23/10 JPM/BOT-MITSUBISHI(JB)
International
Finance Inc
04/06/00 IBM Corp 100,000 .900 99.97 04/14/03 MSDW/TOKYO-MIT-SEC(JB)
04/14/00 Westpac 25,000 .875 99.605 09/22/03 NOMURA-SEC-INTL
Banking
Corp
04/27/00 Italy 50,000 .375 100.010 05/12/02 DEUTSCHE-ALEX-B
05/11/00 Toyota 50,000 .625 99.840 05/23/03 SALOMON-SMITH
Motor Credit
Corp
05/16/00 Procter & 50,000 1.500 100.162 12/07/05 SALOMON-SMITH/BOTMI(JB)
Gamble Co
05/16/00 Procter & 15,000 2.000 99.717 06/21/10 SALOMON-SMITH/BOTMI(JB)
Gamble Co
05/24/00 Province of 50,000 1.875 100.241 01/25/10 NOMURA-SEC
Ontario
06/28/00 Deutsche 90,000 1.500 99.805 06/15/05 DEUTSCHE-ALEX-B/
Telekom GS(JB)/MSDW(JB)
International
07/03/00 Japan 100,000 1.750 99.634 06/21/10 NOMURA
Develop-
ment Bank
07/10/00 Citigroup Inc 45,000 1.400 99.954 07/18/05 NIKKO-SSB
07/19/00 Morgan 9,400 .900 100.000 07/28/03 MSDW
Stanley
Dean Witter
07/31/00 Italy 50,000 .375 100.095 05/12/02 UBS-WARBURG
08/21/00 OKB 50,000 1.800 100.702 03/22/10 NOMURA

BIS Papers No 30 163

 List of global yen issues 1992-2003 (cont)

Launch Amount Issue

Issuer Coupon Maturity Bookrunner
date (yen mil) price

08/23/00 Deutsche 100,000 1.850 100.009 09/20/10 DEUTSCHE-ALEX-B/

Ausgleichs- NOMURA-SEC(JB)/
bank SALOMON-SMITH(JB)
03/09/01 Banque 20,000 4.200 100.000 03/17/31 ML-I
Centrale de
Tunisie
03/13/01 Canada 50,000 .700 101.027 03/20/06 NOMURA
04/26/01 Province of 50,000 1.600 99.138 05/09/13 JPM/NOMURA(JB)
Quebec
06/01/01 Toyota Motor 50,000 .750 99.592 06/09/08 DAIWA-SEC/NOMURA(JB)
Credit Corp
06/20/01 GE Financial 60,000 1.600 99.793 06/20/11 NIKKO-SSB/MORGAN-
Assurance STANLEY(JB)
10/04/01 Italy 100,000 .375 99.936 10/10/06 NOMURA/JPM(JB)
02/15/02 Japan 130,000 1.550 99.785 02/21/12 NIKSSE/NOMURA(JB)/UBS-
Finance Corp WARBURG(JB)
Mncpl Entpr
02/26/02 Textron 6,000 Zero 99.188 03/04/03 BOTMI
Financial
Corp
03/21/02 Italy 100,000 .375 99.844 10/10/06 NOMURA
04/26/02 IBM Corp 100,000 .400 99.968 05/10/04 NIKKO-SSB/BOTMI(JB)
06/20/02 Develop- 75,000 1.400 99.71 06/20/12 NOMURA/UBS-WARBURG(JB)
ment Bank
of Japan
11/27/02 Japan 70,000 1.550 105.813 02/21/12 NIKSSE/UBS-WARBURG(JB)
Finance
Corp Mncpl
Entpr
12/05/02 Develop- 75,000 1.700 99.767 09/20/22 NIKKO-SSB/UBS-
ment Bank WARBURG(JB)
of Japan
02/25/03 Develop- 30,000 1.700 105.803 09/20/22 NIKSSE/UBS-WARBURG(JB)
ment Bank
of Japan
06/24/03 Develop- 75,000 1.050 99.354 06/20/23 NOMURA-SEC/UBS-INV-
ment Bank BANK(JB)
of Japan
10/23/03 Citigroup 50,000 .800 99.853 10/30/08 NIKKO-CITIGROUP
11/19/03 Japan 130,000 1.350 99.86 11/26/13 MILTD/NIKKO-
Finance CITIGROUP(JB)/
Corp Mncpl NOMURA-INT(JB)
Entpr
12/10/03 Morgan 5,000 Floats 100.000 12/15/06 MORGAN-STANLEY
Stanley

164 BIS Papers No 30

 Annex 3:
Samurai and euroyen defaults

List of defaulted Samurai bond issues

Date of Launch Rating Amount Coupon

Issuer Maturity Bookrunner
default date (initial) (yen bn) (%)

14-Oct-00 15-Dec-94 HITIC A+(JCR) 14.5 5.000 26-Dec-01 Nikko

Feb-02 26-Jun-96 FIEC A+(JBRI) 14.0 4.100 24-Jul-06 Yamaichi
25-Mar-98 10-Jun-97 Peregrine BBB+(JBRI) 10.0 2.600 30-Jun-00 Nikko
Investment
25-Mar-98 10-Jun-97 Peregrine BBB+(JBRI) 10.0 6mL+90 20-Jun-00 Nikko
Investment
29-Aug-98 19-Jun-97 Dharmala BBB-(NIS) 15.0 5.450 9-Jul-02 Yamaichi
Intiutama Int’l
14-Oct-00 9-Sep-97 HITIC A+(JCR) 14.0 3.400 24-Sep-04 Nikko
20-Dec-03 5-Dec-96 Argentina B1/BB(M/SP) 50.0 5.000 20-Dec-02 Nikko
20-Dec-03 2-Dec-99 Argentina B1/BB(M/SP) 20.0 5.400 17-Dec-03 Nikko SSB
20-Dec-03 24-May-00 Argentina B1/BB(M/SP) 60.0 5.125 14-Jun-04 Nomura /
Merrill Lynch
20-Dec-03 22-Aug-00 Argentina B1/BB(M/SP) 61.5 4.850 26-Sep-05 Nomura /
Nikko SSB

BIS Papers No 30 165

 List of defaulted euroyen bonds

Date of Launch Amount Coupon Book-

Issuer Rating Maturity
Default Date (yen bn) (%) runner

20-Dec-03 14-Mar-96 Argentina B1/BB-(M/SP) 8.0 7.400 4-Apr-06 Yamaichi

20-Dec-03 25-Apr-96 Argentina B1/BB-(M/SP) 8.0 7.400 25-Apr-06 Nomura
20-Dec-03 15-May-96 Argentina B1/BB-(M/SP) 7.0 7.400 15-May-06 Nomura
20-Dec-03 24-Oct-96 Argentina B1/BB-(M/SP) 50.0 6.000 24-Mar-05 Nikko
– 4-Feb-97 Bakrie Int’l NR 14.0 0.000 12-Feb-00 Peregrine
Finance
20-Feb-02 20-Feb-97 APP Int’l Ba3(M) 17.0 4.800 20-Feb-02 NA
Finance
20-Dec-03 30-Apr-97 Argentina B1/BB(M/SP) 50.0 4.400 27-May-04 Yamaichi
1
5-Apr-00 10-Jun-97 Robinson NR 6.8 0.000 12-Jun-00 Societe
Dept Generale
Asia
– 11-Aug-97 Sierad NR 2.0 0.000 15-Aug-00 IBJ
Produce
20-Dec-03 23-Jul-99 Argentina Ba3/BB(M/SP) 18.0 3.500 11-Aug-09 Daiwa
24-May-02 24-May-00 Buenos B1/BB(M/SP) 3.0 4.250 27-May-03 NSSB
Aires
Province
24-Dec-03 31-May-00 Parmalat NR 30.0 3.650 1-Jun-20 NSSB
Fin BV
2-Dec-01 8-May-01 Enron Corp Baa1/BBB+(M/SP) 25.0 0.520 15-May-02 Merrill
Lynch
2-Dec-01 4-Jun-01 Enron Corp Baa1/BBB+(M/SP) 40.0 0.000 18-Jun-03 Merrill
Lynch
2-Dec-01 4-Jun-01 Enron Corp Baa1/BBB+(M/SP) 10.0 0.770 18-Jun-03 Merrill
Lynch
2-Dec-01 5-Jun-01 Enron Corp Baa1/BBB+(M/SP) 10.0 0.970 18-Jun-04 Merrill
Lynch
2-Dec-01 15-Jun-01 Enron Corp Baa1/BBB+(M/SP) 20.0 0.493 13-Jun-02 Merrill
Lynch
1
Date of bankruptcy

Bibliography
Japan Securities Research Institute (2003): Japan Securities Market in 2002
(ISBN 4-89032-814-9).

166 BIS Papers No 30

BIS Papers No 30

Annex 4:
Bond structures in the liberalisation of Japan’s capital market:
from Samurai via Shogun and Daimyo to euro with Uridashi and global

:
1970 First Samurai bond
ADB yen 6bn
1972 First sovereign
Samurai: 1977 First Euroyenbond (EIB)
Australia yen 10bn 1978 Foreign currency Euroyen: Euromarket yen
. denominated domestic denominated bonds
1979 First corporate bonds allowed
Heavily controlled issuance until 1983
Samurai:
1985 First Shogun bond: 1984 Broad deregulation
Sears yen 20bn
Forex bond fully placed .
. Further gradual liberation of issuer
domestically (IBRD) .
Various deregulation 1987-92 Daimyo bonds criteria, domestic placement,
.
. secondary flowback, etc.
. Domestic yen bonds
1989 First structured
. with euro settlement
Samurai (RDC) .
. 1990 First global bond ($)
Denmark yen 120bn ..
1993 Last Shogun with Japanese offering
. .
1994 Public offering through Uridashi
1996 Lifting of eligibility .
fixed price re-offer of eurobonds star 1992 First global yen bond:
criteria .
. Global syndication and
Record Samurai . . simultaneous offering in
issuance . . Japan, US and Euro
. market (IBRD)
1997 Asian crisis, first . . .
Samurai default 2000 Large exchangeable bond issuance .
. 2001- Surge in forex (AUD, USD, EUR) . .
2002 Issuance: yen 638bn Uridashi issuance . .
2002 Issuance: USD 24.5bn . 2002 Issuance: yen 556bn
2002 Issuance: yen 2,251bn
167
 Comments on Nishi and Vergus’s paper
“Asian bond issues in Tokyo: history,
structure and prospects”

Toshiharu Kitamura

As a discussant of the Nomura paper, “Asian bond issues in Tokyo: history, structure and
prospects”, I would like to focus on a few implications of Japanese experiences for emerging
Asian bond markets.

1. Division of labour in financial intermediation

Many studies indicate that, in East Asia, the banking sector has generally dominated the
function of financial intermediation (with the possible exception of the Philippines). Several
factors are responsible for the present situation.
(1) The presence of a government bond market is considered a precondition for the
development of a corporate bond market, but government bond markets in East
Asia have not been well developed.
(2) In contrast, the banking mechanism from deposit-taking to loan-making has been
strongly supported by the legal framework typified by the central bank and a variety
of policy measures in many East Asian countries, including Japan.
(3) As economic arguments suggest, information asymmetry existed tenaciously
between fund providers (investors) and fund raisers (corporations). In other words,
at earlier stages, fund raisers in general failed to provide relevant information to fund
providers, and only banks were in a position to be well informed about borrowers’
activities and thus able to monitor them. This was the case even in Japan until
recently.
Only top-level fund-raising institutions such as blue-chip sovereigns and well established
corporations have enjoyed direct access to bond markets. However, the so-called double
mismatch in 1997 and 1998 in East Asia revealed the shortcomings of financial
intermediation that is overly dependent on the banking sector. As a result, more attention has
been focused on fostering corporate bond markets.
Even so, bank domination will remain strong in the so-called middle market in many East
Asian countries, including Japan. Therefore, it follows that the basic framework for financial
intermediation in East Asia will be based on a double-channel approach, paying due
attention to both direct finance (capital markets, including bond markets) and indirect finance
(banking business). Emphasising only bond markets in East Asia does not adequately
address the region’s financial intermediation issues.

2. Business conflicts (business territorial disputes) and market

competition
In the background of the Tokyo Samurai bond market, which appears to have peaked in
around 1996, is the fact that, inside Japan, the traditional financial system has continued to
help the banking sector dominate the domestic corporate financing business, securing very

168 BIS Papers No 30

close relationships between individual banks and corporations. In general, the Japanese
bond market has been predominantly one of government-related bonds (more than 80% of
total yen-denominated bond issuance in 2002), and even the issuance of bank debentures
has considerably surpassed that of corporate bonds. In this setting, the position of securities
firms (investment banks) has been highly biased towards equity business. It is no
exaggeration to say that non-resident bond issuers took advantage of Japan’s abundant
financial resources through the Tokyo Samurai bond market, which has only a small share of
Japan’s entire financial market.
In contrast, the banking sector has successfully defended its vested interests in corporate
financing. This was especially true before the 1990s: the banking sector, notably the
Industrial Bank of Japan, had implicitly enforced rigid criteria for issuance eligibility and
emphasised the need for a fiscal agent for the public offering of corporate bonds.
Conversely, the securities firms had obstinately interfered in the bond offering business
engaged in by bank-affiliated securities subsidiaries both at home and abroad, particularly
limits on the lead-manager role. Thus, there had been a battle over the business territory
between the banking and securities business sectors in Japan during the 1980s. In these
circumstances, the slow process of deregulation of restrictive practices in bond issuance in
the 1980s and 1990s led many fund-raising institutions to use the Eurobond market outside
Japan. Accordingly, Japan’s domestic bond market was seriously hollowed out despite
abundant domestic financial resources.
Two lessons for Asian bond markets could be derived from these Japanese experiences.
The first concerns vested interests among related parties in financial markets. In many East
Asian countries, the banking sector has dominated and established deeply rooted vested
interests in the corporate financing business. The banking sector will naturally try to defend
its privileged position over the newly emerging financial channel. (Conflicts of business of this
type also existed in the United States, involving regulatory controversies, until 2000.) The
lesson for the development of Asian bond markets is that the supervisory authorities should
have a fair and clear policy perspective and strong commitment to achieving an efficient and
balanced financial intermediary process - one that balances the direct (capital market) and
indirect (banking business) finance channels.
The second lesson is related to the competition among bond markets: if a local bond market
fails to catch up with international standards, potential corporate issuers will quickly turn to
other bond markets. (International standards include disclosure requirements, accounting
standards, price transparency, trading and clearing/settlement procedures, and taxation
harmonisation, as well as related documentation requirements and effective enforcement.)
Financial markets are intensifying competition among themselves. (European stock markets
have already indicated some movement in this direction, and perhaps bond markets have
similar reasons to do so.) The Eurobond market and the Yankee bond market are already
widely used and have been predominant around the world. Similarly, in Asia, Tokyo, Hong
Kong, and Singapore have gained good bond market reputations. A bond issuer successful
in any local Asian bond market will compare emerging local markets with the leading
international bond markets. If there is any inefficiency in the local bond market, potential
bond issuers will be easily attracted to those sophisticated leading markets.
The dispersion of both the Shogun (foreign currency-denominated bond offerings in the
Tokyo market) and the Daimyo bond market (yen-denominated bonds offered in the Tokyo
market but settled in the Euro market) could be interpreted in line with this market
competition. The public offerings through Uridashi described in the Nomura paper could be
interpreted as a new approach rather than an heir to the Shogun and Daimyo bond markets.

BIS Papers No 30 169

3. Currency constraints
Basically, bonds are offered because the issuers are eager to raise funds and the investors
are seeking safe yet lucrative securities. However, for cross-border transactions, there arises
another factor, related to currency sovereignty. For yen-denominated bonds, the Samurai
and the Euroyen bond markets have coexisted since the 1970s. The Samurai market aims at
mobilising financial resources from domestic investors of the issuing market, whereas the
Euroyen bond market aims at mobilising financial resources from the broader base of
international investors. Behind them is a crucial currency issue for the Japanese monetary
authorities: whether the yen can be freely used outside Japan for financial transactions
among non-residents. The process of using the yen outside Japan was called the Euroisation
of the yen. (Nowadays, the common terminology for the process of using or not using an
individual national currency outside its original currency territory is the “internationalisation”
or “non-internationalisation” of the currency.)
Even the US monetary authorities were cautious about the free use of the US dollar outside
the United States in the 1960s, and the German monetary authorities were even more
cautious about the free use of the Deutsche mark outside Germany. Japan first faced this
issue towards the end of the 1970s. The non-internationalisation of a national currency is
generally intended to segregate a country’s domestic economic and financial transactions
from international turbulence arising from the free use of the national currency in transactions
outside its territory. However, Japan decided to internationalise the yen in the 1980s.
The currency internationalisation issue is not an easy one. For instance, the Swiss monetary
authorities have long been cautious about it, and in Asia, Singapore has had a long tradition
of non-internationalisation of the Singaporean dollar. The Singaporean approach has
impressed monetary authorities in many Asian countries, including Malaysia, Indonesia, and
Thailand. Taiwan, China has also strictly maintained this approach for many years. Most
recently, many Asian economists have focused on the offshore restrictions on the
deliverability of currencies (which give rise to non-deliverable forward (NDF) transactions in
several Asian currencies) - which are intended to ensure the exchange rate stability of those
currencies. (Offshore NDF transactions provide an offshore margin-based mechanism to
hedge those currencies that are considered “unhedgeable” due either to the absence of a
local forward market or to limited access to a local currency market.)
In these circumstances, monetary authorities in East Asia will likely remain conservative
about the free use of their currencies outside their territories. If so, Asian bonds denominated
in local currency will be transacted only inside the territory of the issuing market; thus, it
follows that there will be no cross-border transactions of local currency-denominated Asian
bonds.
If Asian bonds are denominated in a unit reflecting a regional currency basket such as an
Asian Currency Unit (ACU), the question remains whether or not those currencies that
compose the currency basket will become freely available for smooth bond trading across
borders. Another serious problem is the determination of parities among the component
currencies.
If local bond markets have to face severe competition with the sophisticated international
bond markets, and if the above-mentioned currency issue remains unsolved, Asian bond
markets may have to confine themselves to locally segregated bond markets where local
issuers raise funds only from local investors. This may sound most pessimistic, but it is a
reality learned from past experience.
A long-term policy perspective will have to be stressed for the successful development of
Asian bond markets.

170 BIS Papers No 30

 Recent development of the Korean bond market

Seongtae Lee

Ladies and gentlemen, I am very pleased to speak here today at this Asian Bond Market
Research Conference, co-hosted by Korea University and the BIS.
As you are well aware, over the past half century the Asian region has achieved
unprecedentedly high economic growth, and it has now emerged as among the key players
in the world economy. With this continuing remarkable economic growth, trade within the
Asian region is growing rapidly, and there is also a greatly increasing demand for financing.
However, regional financial markets have so far not developed sufficiently to meet this
demand.
Asia’s bond markets, in particular, are far less developed than those of developed countries.
According to IMF statistics, the amount outstanding of the primary bond markets of Asia’s
nine emerging market economies combined totalled USD 1.473 trillion as of the end of 2002.
This is only 8% the size of the US primary bond market, 12% that of the 15 EU countries
combined and 21% that of Japan. The ratio of Asia’s outstanding stock of bonds to GDP is a
mere 43%, far lower than the figures of 182% in the United States, 146% in the European
Union and 175% in Japan. In their secondary bond markets as well, almost all regional
economies have much smaller ratios of turnover to average outstanding stocks compared
with advanced nations.
There is an argument that the 1997 financial crisis was aggravated by these poorly
developed regional bond markets. The argument goes that because the financial structures
of Asian economies were overly reliant on banks and Asian bond markets were poorly
developed, the Asian countries had to rely for investment on short-term foreign borrowings,
which were inevitably withdrawn in a hurry once the regional crisis hit.
Recently, the Asian countries have accumulated significant amounts of foreign assets,
thanks to their continued current account surpluses. However, due to the underdeveloped
bond markets here, some say that there have been some inefficiencies in both the raising
and the managing of foreign funds in the region.
In the wake of the financial crisis, therefore, countries in the region came to realise the
necessity of developing their bond markets and related infrastructure. This enlightenment
has become the momentum driving the countries in the region to make the necessary efforts.
Let me now share with you the efforts Korea has made since the financial crisis in order to
develop its bond markets, and tell you what we have achieved so far.
Before the financial crisis, Korea did not have to issue large quantities of government bonds,
due to observing its balanced budget principle. Once the crisis had broken out, however, the
long-cherished principle of maintaining balanced budgets was abandoned, and the
government began to actively issue bonds, as a means of financing the soaring fiscal deficit
and supporting economic recovery. With technical support from the IMF and the World Bank,
Korea also embarked, in a step-by-step manner, on a series of institutional improvements.
These included the introduction of regular government bond auctions, the establishment of a
primary dealer system, the integration of different types of government bonds, the opening of
the Korean government bond futures market and the introduction of a fungible issue system.
For its part, the Bank of Korea (BOK), the nation’s central bank, has also made considerable
efforts to help develop the country’s bond markets, including the government bond market.
Before the crisis, Korean government bond issuance was to some extent not carried out in
accordance with the market mechanism. A portion of any bond issue was discretionarily

BIS Papers No 30 171

allocated for underwriting to certain financial institutions. The remaining part of the issue was
taken over by the BOK. The government also sometimes financed its fiscal needs by relying
on borrowings from the BOK rather than by issuing bonds. Since June 1998, however, the
BOK has not accepted government requests to underwrite government bonds and has
advised the government to issue all government bonds in the market, based on market
interest rates. In addition, the BOK has improved the infrastructure related to bond
transactions. It has introduced an electronic competitive bidding process for government
bonds using BOK-Wire and adopted the delivery-versus-payment system for government
bonds.
As a result of these efforts, the outstanding volume of Korean government bonds has grown
almost seven times since 1997, while the turnover ratio for these bonds has skyrocketed
23 times in the same period. In 2003 the issuance of five-year government bonds increased
substantially thanks to the government’s efforts to meet growing demand for longer-term
bonds on the part of institutional investors such as pension funds and insurance companies.
The five-year Korean government bond futures market was also set up in order to allow
hedging against the associated risks. In line with these developments, the proportion of five-
year government bonds in total bond issuance increased by 7 percentage points in 2003, to
36%.
Korea’s corporate bond markets are also in the process of becoming more advanced,
through the introduction of new financial instruments and the establishment of relevant
institutions that will support bond market development. Since 1999, new products such as
asset-backed securities (ABSs), mortgage-backed securities (MBSs), asset-backed
commercial paper (ABCP), collateralised loan obligations (CLOs) and collateralised bond
obligations (CBOs) have been introduced. After the financial crisis, the market structure for
corporate bonds shifted to focus on secured bonds rather than unsecured bonds, and this
change required more scientific analysis of issuers’ credit risk. In response to this need, the
setting-up of new credit rating agencies was allowed, and the fiercer competition among
credit rating agencies has led to a significant development of credit risk analysis techniques.
There have also been many improvements in terms of market behaviour. With the
implementation of mark-to-market bond valuation, rates of return on fund management are
now evaluated and disclosed in accordance with daily bond yields. The introduction of a
system for interest rate risk management has also made institutional investors respond more
sensitively to changes in bond yields. Accordingly, it is now becoming common practice for
market participants to closely watch the BOK’s monetary policy, along with economic
developments at home and abroad.
As I have just described, Korea’s bond markets have gone through remarkable changes, in
terms of both quality and quantity, since the financial crisis. The government bond market, in
particular, has been totally reshaped across the primary and the secondary markets, as well
as its infrastructure. We believe that Korea’s bond markets will develop further in the future
through continuous efforts to heighten bond market efficiency and to push on with corporate
restructuring.
There has been considerable discussion going on recently, in various international
cooperative forums such as ASEAN+3 and the EMEAP (Executives’ Meeting of East Asia
and Pacific Central Banks), of the development of regional bond markets. In ASEAN+3, in
particular, views on securitisation and credit guarantees have been discussed. In relation to
this issue, Korea has put together a two-tier securitisation draft proposal making the most of
the securitisation and credit guarantee mechanism. Other nations have also developed
plausible draft proposals in these fields. Meanwhile, the EMEAP, comprising 11 central
banks and monetary authorities in the East Asia and Pacific region, has raised a USD 1 billion
fund and invested it in dollar-denominated bonds issued by countries in the region. Taking
this arrangement one step further, the EMEAP is now actively discussing ways to expand the
fund and invest it in local currency denominated bonds in the region.

172 BIS Papers No 30

Under these circumstances, I think that this conference is very significant, in that it brings
together distinguished scholars and experts from both at home and abroad, to diagnose the
reasons why Asia’s bond markets have not been developed sufficiently and to propose new
visions and ideas on the ways to facilitate their development.
I hope there will be as many active discussions tomorrow as there were today, through which
a variety of epoch-making views will be advanced concerning the development of our
regional bond markets.
Thank you for your attention.

BIS Papers No 30 173

 Minding the gap in Asia:
foreign and local currency ratings

Kate Kisselev 1 and Frank Packer 2

Introduction
As governments embrace the goal of developing local currency bond markets as an
alternative to inflows of foreign capital, 3 rating agencies now commonly assign a domestic
currency rating to sovereigns in addition to a foreign currency one. In Asia, 18 major
sovereigns with foreign debt ratings now have a domestic currency rating from a major rating
agency. Usually the domestic rating is higher, reflecting the presumed greater ability and
willingness of sovereigns to service debt denominated in their own currency. However, the
gap between the two ratings is uniform neither across borrowers nor across agencies.
The distinctions between local and foreign currency ratings are likely to have increasingly
important implications for the development of capital markets globally and in Asia in
particular. The degree to which rating policies favour a particular currency of denomination
might provide significant incentives in terms of investor acceptance and market pricing.
Rating policies might reinforce government policy initiatives and regulations as well. 4
In this paper, we first provide a comparative overview of domestic and foreign currency
ratings globally and in Asia in particular. Asian credits are similar to the global sample in
terms of both the newcomer status of local currency ratings and the tendency for the
local/foreign currency rating gap to be largest in the lower investment grade/upper
non-investment grade region. However, differences of opinion among rating agencies
regarding the relative creditworthiness of local and foreign currency obligations are quite
pronounced in Asia. Within a linear regression framework, we then examine the determinants
of the difference between local and foreign currency ratings, and find evidence that
differences among agencies are driven by distinctions in their overall rating policy rather than
a distinct Asian factor per se. Other than the paper of Trevino and Thomas (2001), ours is

1
Former Vice President, Citigroup Global Country Risk Management.
2
This paper was completed in early 2005. Márcia Elyseau provided helpful research assistance. Thanks are
also due to Robert McCauley, as well as the discussant at the BIS-Korea University Asian bond markets
conference, Thomas Byrne, for useful comments at an earlier stage of the paper. Participants in the BIS
workshop on bond markets in Hong Kong, and the 17th annual Australasian Banking and Finance Conference
in Sydney in 2004, also provided useful comments. The views expressed do not necessarily represent those
of the Bank for International Settlements or Citigroup.
3
While the first Asian Bond Fund invested in dollar-denominated debt, East Asian central banks announced in
late 2004 the launch of a second fund with a mandate to invest in domestic currency denominated bonds. See
the press statement of the Executives’ Meeting of East Asia-Pacific Central Banks (EMEAP), 16 December
2004.
4
For the most part, regulations that key off agency ratings make little distinction between foreign and domestic
currency rated claims. Those exceptions that do exist favour domestic currency ratings and/or domestic
currency claims. For instance, under the standardised approach of Basel II, a new capital adequacy
framework for banks, in the case of foreign currency exposures to multilateral development banks whose
convertibility and transfer risk are “considered by national supervisory authorities to be effectively mitigated”,
the domestic currency rating may be used for risk weighting purposes instead of the foreign currency rating
(see Basel Committee on Banking Supervision (2004)). In addition, the framework gives national authorities
the general discretion to apply even lower risk weights to their banks’ exposures to sovereign (or central bank)
domestic currency obligations, which is not the case with foreign currency obligations.

174 BIS Papers No 30

the first to empirically estimate the determinants of the local and foreign currency gap for
sovereign credits.

Foreign and local currency ratings

Over the past few decades, the business of providing sovereign ratings has grown
considerably. As of 1985, only 15 countries obtained credit agency bond ratings to borrow in
international capital markets. Most of these countries were rated AAA; less financially strong
countries relied on bank finance or privately placed bonds. However, over the past
15-20 years, countries at the lower end of the credit quality spectrum have relied increasingly
on bond markets, and obtained a credit rating for that purpose.
Initially, most of the new sovereign ratings applied to foreign currency debt, as sovereigns
apparently felt little need to obtain a rating for domestic currency obligations. But sovereigns
gradually moved to having domestic currency ratings, a likely reflection of efforts to increase
the investor base for domestic currency bonds (Tables 1 and 2).
Pretty much the same story holds with Asian ratings. Among Asian sovereigns, only Japan
and Australia had foreign currency ratings as of 1975, but more than half of the 18 Asian
sovereigns had a foreign currency rating by 1990. And although none had a local currency
rating before 1990, the catch-up is now complete, which parallels the global rating
phenomenon (Tables 1 and 3).

What might drive the rating gap?

Rating agencies often give higher ratings to the domestic currency obligations of sovereign
states than to their foreign currency ones. This is a global phenomenon: for instance, the
average gap between Standard & Poor’s local and foreign currency ratings was 0.7 notches
globally in late 2004, while the Moody’s gap was 0.4 notches. Differences are often justified
in terms of the sovereign’s ability to tax and appropriate domestic currency assets, which is
often judged to be greater than in the case of foreign currency assets. In addition, while the
sovereign must generate foreign exchange to repay foreign currency debts, it can print
money to meet domestic currency obligations (see, for example, Fitch Investors Service
(2003)). 5
Following this logic, constraints on the sovereign’s ability to print domestic currency would
tend to reduce the justification for a rating gap. Prime examples would be sovereigns that
use the currencies of foreign countries, such as Panama and El Salvador. The countries of
the euro area are also special cases; here the delegation of monetary policy to the ECB has
greatly diminished the distinctions drawn between local and foreign currency debt. 6 Countries

5
Another frequently cited justification for notching is that the incidence of default on local currency debt has
been lower than that on foreign currency debt (S&P (2003)). However, this is usually based on the default
statistics, which include defaults on bank debt. As for the cases of default on rated bonds, the limited default
experience to date suggests that it is not obvious that default on foreign currency bonds tends to precede or
be more likely than that on domestic currency bonds (see Packer (2003)).
6
Though there was a difference of approach over whether foreign currency ratings should be upgraded or
domestic currency ratings downgraded, the major rating agencies eliminated or narrowed outstanding
domestic/foreign currency rating gaps for euro area countries ahead of and during the transition to the euro
(for further discussion, see McCauley and White (1997)).

BIS Papers No 30 175

whose local currency obligations are held by foreigners may also have smaller rating gaps. In
these countries, it is the local banking system, rather than the sovereign, that must hedge
foreign investments in local currency denominated sovereign debt, and the government may
be unwilling to print money if that would impose substantial costs on the banking system.
More generally, the frequent existence of significant political costs to high levels of inflation
should limit the applicability of the “printing press” argument for high domestic currency
ratings.
Another possible exception would be if foreign currency issuance is small relative to the total
debt outstanding of a sovereign. After all, one of the underlying principles of sovereign debt
analysis is that sovereign risk always depends on the willingness as well as the ability to pay.
Given a small enough burden, the sovereign might conceivably make an extra effort to avoid
default on foreign currency obligations. It is likely that the relatively small size of the
international bonds of emerging market countries in the early 1980s explains why the default
experience on bonds at that time was rather limited, despite a range of bank loan
restructuring programmes.
Another factor influencing the size of the gap is a purely technical one: there is no rating
higher than AAA (Aaa) in the rating agencies’ symbology. The additional credit standing that
a foreign currency AAA credit might gain by being denominated in domestic currency is
unobservable. In addition, countries that are AA+ can only be raised by one notch, and so
forth. Notching should thus become more pronounced and frequent as the foreign currency
rating drifts downwards from AAA and AA, which is in fact what we generally observe both
globally and in Asia. Little surprise, then, that countries such as Malaysia and the Philippines
have marked notching relative to highly rated Singapore, Australia and New Zealand, which
have little to no room for a notching-up on their domestic currency obligations.
On the other hand, it appears that the gap peaks in the mid-grade rating category BB. For
instance, according to Standard & Poor’s, 83% of all rated sovereigns in the BB category in
late 2004 had domestic currency obligations that were rated at least one notch higher than
foreign currency obligations (Table 4). By contrast, the relative advantage of domestic
currency obligations was much smaller for countries that are further below investment grade;
only 27% of countries in the B category enjoy a rating gap. For its part, Standard & Poor’s
posits that low-rated countries face risks, such as high degrees of social and political stress,
that would also impair their ability to keep servicing domestic obligations in circumstances
where foreign currency debts were allowed to default (S&P (2002)).
Asia does follow the same global hump-shaped pattern in the distribution of rating gaps, as is
evident in Table 4. The propensity for rating gaps to exceed one notch is noticeable in the A
range, where two out of three sovereigns - Korea and Malaysia - have large gaps.
Meanwhile, Indonesia has one of the lowest ratings among Asian sovereigns and gets only a
single one-notch improvement in the local currency rating from only one of the rating
agencies.

Differences among the rating agencies

There are surprisingly sharp differences among the rating agencies with respect to the
frequency at and degree to which domestic obligations are given favourable ratings. In
particular, Moody’s tends to notch up its domestic currency rating much less frequently than
the other agencies; for instance, in November 2004 it gave a higher domestic currency rating
on only 30% of its rated universe of sovereigns, compared with 44% for S&P (Table 5).

176 BIS Papers No 30

Moody’s also assigned a higher foreign currency rating than domestic currency rating in four
cases, 7 with a relatively small proportion of outstanding foreign currency debt relative to
foreign exchange reserves always cited as a reason (see Moody’s (2003c,d)). By contrast,
S&P did not assign a higher foreign currency rating to any sovereign, while Fitch assigned a
higher foreign currency rating only in the case of Japan.
Consistent with this global finding, Asian countries see far less notching from Moody’s than
from Standard & Poor’s (Table 6). In fact, the differences are starker in the case of Asia: the
average gap between S&P’s foreign and local currency ratings was 0.9 notches in Asia, even
wider than the 0.6 notch gap for non-Asian countries (Table 6). At the same time, Moody’s
actually notched Asian countries in the other direction, on average, with a mean Asian gap of
–0.3 notches compared to a gap of 0.5 notches for non-Asians. This pattern, that the Asia
subsample shows an accentuation of the differences in notching policies among the
agencies, has held since 1995, when Asian sovereigns began to receive foreign and local
currency ratings widely. In the regression analysis to follow, we explore possible reasons
why this may be the case.

Regression analysis
The previous sections present the stylised facts that, in the case of S&P ratings, there is
more likely to be a gap between the foreign currency and domestic currency ratings if a
country is in Asia, and that such a gap is likely to be larger if a country is in Asia rather than
elsewhere. In the case of Moody’s, rating gaps in Asia are smaller than elsewhere. Are these
facts simply the by-product of different observable endowments among the Asian economies
versus elsewhere, which might tend to magnify both the gap and the agency differences,
given the agencies’ respective rating technologies? Or rather, might there be an unobserved
factor common to Asia that is driving the results, reflecting rating agency biases and/or
omitted variables?

Previous literature
According to general descriptions of the rating process by the rating agencies themselves
(see Moody’s (2003a, 2004) and S&P (2002, 2004), sovereign local and foreign currency
ratings are based on a wide array of quantitative and qualitative factors that are intended to
capture political risk, income and economic structure, growth, monetary policy, budgetary
and public debt management, and external liquidity and debt. However, quantitative studies
of ratings - such as Cantor and Packer (1996), Moody’s (2003b) and Borio and Packer
(2004) - find that most of the variance in Moody’s and S&P ratings can be explained by a
relatively small number of variables. Typically, the debt burden itself, default history, per
capita income and economic growth are important as indicators of a country’s wealth and
ability to pay, and indices of political risk are also important, presumably because they proxy
for willingness to pay.
Partly because they have been around longer, the literature is more developed with regard to
the determinants of foreign currency ratings. Specifically, Cantor and Packer (1996) found
that per capita income, inflation, external debt, economic development, and default history
were particularly strong predictors of foreign currency ratings. A weaker relationship existed
between sovereign ratings and GDP growth and the fiscal balance, and there was no
statistical relationship between ratings and the external balance. Moody’s (2004) found that

7
The four countries are India, Lebanon (one notch), Turkey (two notches) and Japan (five notches).

BIS Papers No 30 177

per capita income, debt/exports, growth, external transfer risk, and government effectiveness
explain 91% of the variation in its own foreign currency ratings. In addition to many of the
above-listed variables, Borio and Packer (2004) also found corruption perceptions to have
significant explanatory power in predicting variation in a panel of foreign currency ratings.
The examination of the determinants of local currency ratings includes another study of
Moody’s (2003b), which found that government debt/government revenue, per capita
income, growth, and government effectiveness explained 87% of the variation in Moody’s
local currency ratings. In a study of the gap between local and foreign currency ratings,
which used a probit methodology, Trevino and Thomas (2001) looked at many of the
variables discussed above and also added measures of the term structure of bank debt, a
country’s share of bank lending, bank commitments, bank borrowing/ deposits, reserves and
IMF credit usage, rating agency dummy variables, and regional dummy variables. The
authors found regional and rating agency biases present even after the inclusion of these
variables.

Methodology
To address our questions about what is driving the prevalence and magnitude of rating gaps
in Asia and globally, we estimate regression models for the foreign currency ratings of
Moody’s and S&P, the local currency ratings of each agency, and the gap between the
foreign currency and local currency ratings. We also estimate a set of regressions where the
left-hand side variables are the differences between Moody’s and S&P’s local currency
ratings, foreign currency ratings, and notching gaps. In each of our regressions, we use a
fixed effects specification and examine the Asian countries’ fixed effects for evidence of an
unobserved common factor.
We proceed by identifying 61 variables that reflect political risk, default history, external debt
burden, macroeconomic performance, and government financial management, and we
collect annual data on these variables for the 101 countries that have both foreign currency
and local currency ratings at either S&P or Moody’s from 1995 to 2003. 8 Ratings are recoded
numerically with AAA and Aaa equal to 1, AA+ and Aa1 equal to 2, and so on. Each end-year
rating is assumed to be the function of explanatory variables from that same year, and the
candidate explanatory variables are listed in Table 7. In many cases, these variables may
capture overlapping aspects of ability and willingness to repay foreign or local currency debt,
so we pare the list of variables in each regression by identifying subcategories of variables
that may capture the same concept. These subcategories are also listed in Table 7.
We start by fitting a regression model to S&P foreign currency ratings. Within each
subcategory of variables, we test the fit of each variable separately. For example, we start by
testing the fit of each of the CPI-related variables. If no CPI variable is significant at the
.10 level, we proceed to the GDP growth subcategory, leaving out a CPI variable. In cases
where only one CPI variable is significant, we retain it while testing GDP growth variables. In
cases where several CPI variables are separately significant, we include them together in the
regression to see whether they are robust to one another’s inclusion. We then eliminate
variables that are not robust according to t-statistic, and retain robust CPI variables while
testing GDP growth variables. After moving through all subcategories in this way, we then
eliminate variables that are no longer statistically significant at the .10 level to arrive at the
final S&P foreign currency specification. Hausman tests in nearly every specification suggest

8
The sample criteria are that a country must have a foreign currency rating and a local currency rating from
either S&P or Moody’s at any time between 1995 and 2003, and all economic, political, and financial indicators
must be available from the sources listed in Table 7.

178 BIS Papers No 30

that fixed effects rather than random effects are present, so we retain a fixed effects
specification for consistency throughout. F-tests in every regression specification confirm the
importance of the fixed effects. All regressions are estimated in SAS with two-way (cross-
sectional and time-series) fixed effects, and we require countries to have more than one
time-series observation in each regression.
We follow the same procedure for Moody’s foreign currency ratings, except that the starting
point is the final S&P specification, for convenience. We first eliminate variables that are not
significant for Moody’s, and then move through the subcategories again, testing candidate
variables in the same way as we did in fitting the S&P model.
Our starting point for the S&P (Moody’s) local currency rating specification is the S&P
(Moody’s) foreign currency specification. The starting point for the S&P (Moody’s) rating gap
specification is the union of variables in the S&P (Moody’s) foreign currency and local
currency specifications. Similarly, the starting point for the S&P-Moody’s foreign currency
(local currency) rating difference specification is the union of variables in the S&P and
Moody’s foreign currency (local currency) rating specifications.

Rating regressions
The final regressions of S&P and Moody’s foreign currency ratings on full sets of explanatory
variables are shown in the first two columns of Table 8. In both regressions, all of the
explanatory variables are significant at the .05 level and take the expected signs. Per capita
GDP is significant at both agencies, with higher levels of income leading to better ratings.
While higher per capita GDP may well imply higher costs associated with default, this
variable is also likely to proxy for other indicators of development and creditworthiness.
M2/reserves is also significant at both agencies, and this variable captures monetary
volatility, excess monetary liquidity, and reserve volatility, so that greater variation in this ratio
leads to a worse foreign currency rating. Investment is significant at both agencies and has
the expected interpretation: higher rates of investment should generate the ability to repay
debts. Overall political risk is an important determinant of both agencies’ ratings; higher
levels of political risk are associated with worse ratings. The importance of political risk is
underscored by the fact that an additional source of political risk is significant in each
agency’s ratings, with control of corruption associated with better S&P ratings, and regulatory
quality associated with better Moody’s ratings. Finally, the time elapsed since the last default
on foreign currency obligations is also important at both agencies, with longer periods without
default associated with better ratings.
The most obvious difference between the agencies’ foreign currency rating methodologies is
in the treatment of debt and external vulnerability, two critical components of foreign currency
ratings. Standard & Poor’s appears to focus on total public sector indebtedness, with higher
public debt/GDP ratios resulting in worse ratings. It also considers exchange rate regime,
with pegged and managed floating regimes penalised by half a notch. The significance of the
exchange rate regime variable highlights the view that rigid exchange rates may be a direct
constraint on debt servicing capacity; if governments must use reserves to defend a
currency, less foreign exchange remains available for debt servicing. By contrast, Moody’s
appears to weight more directly the net external debt burden, as a fraction of exports.
Adding country-specific fixed effects to the regression improves the fit, as all of the cross-
sectional fixed effects are significant. It is noteworthy that although the average of the fixed
effects in both the S&P and Moody’s regressions implies ratings for Asian sovereigns that
are 1.4-1.5 notches better on average than for other countries, there is enough variation
within the country fixed effect coefficients that an F-test cannot reject the hypothesis of no
difference between Asian and non-Asian countries. Thus, higher credit ratings in Asia can be
explained more on the basis of better fundamentals in Asian countries and country-specific
factors than the result of an “Asia factor”.

BIS Papers No 30 179

Regressions of the local currency ratings of each agency on the explanatory variables are
presented in the final two columns of Table 8. All variables are significant at the .05 level and
all coefficients take the correct signs. As in the foreign currency rating regressions, per capita
GDP is an important explanatory variable for both agencies. Public debt/GDP is now
significant for both agencies as well. However, while Moody’s appears to weight more heavily
in its assessments the variables of economic growth and M2/reserves, S&P appears to
weight many more additional variables, including inflation, investment, political risk, control of
corruption, exchange rate rigidity, the nominal exchange rate change, and the time elapsed
since the last local currency default. The relative parsimony of the Moody’s regression bears
out the agency’s own finding that most of the variation in its local currency ratings can be
explained with just a few variables (Moody’s (2003b)).
The local currency regression results are consistent with our earlier findings regarding Asian
foreign currency ratings. While the average Asian fixed effect is 1.1-1.2 notches better than
that of non-Asian countries, this difference is not statistically significant, so we conclude that
better average Asian local currency ratings tend to be driven by fundamentals and country-
specific effects rather than an “Asia factor”.
One of the drawbacks of the above approach to discerning differences in rating agency
methodology is the fact that the agencies have rated different sets of countries over time,
and it may be these differences in samples, rather than differences in methodology, that
drive the regression results. To control for this effect, we impose the requirement that the
S&P and Moody’s samples be the same and rerun the regressions (Table 9). The
explanatory variables from the foreign currency rating regressions are highly robust, with all
variables still significant at the .10 level and all coefficients remaining at roughly the same
order of magnitude. The Moody’s local currency regression is also robust to the sample
change. The S&P local currency regression is slightly less robust, with three variables
becoming insignificant when we change the sample: the investment, political risk score, and
exchange rate rigidity variables are no longer significant at the .10 level.

Rating gap regressions

Next, we report the results of regressions of the gap between local and foreign currency
ratings of S&P and Moody’s on the explanatory variables (Table 10, first two columns). Our
convention is that a positive gap means that the local currency rating is better than the
foreign currency rating. Among the variables, inflation, M2, and M2/reserves are important
determinants of the gap between foreign and local currency ratings at both agencies. High
inflation tends to lead to a smaller gap between ratings, consistent with high levels of inflation
eroding the creditworthiness of the sovereign across the board and thus diminishing the
relative safety of local currency obligations. The positive coefficients on the change in M2 or
M2/reserves suggest that monetary expansion is initially associated with relatively safer local
currency obligations, though the negative coefficients on the volatility of the same variables
suggest that this effect has diminishing marginal returns, perhaps for the same reason that
high inflation is associated with a diminished rating gap.
Some variables explain rating gaps for Moody’s rating only. Per capita GDP is significant in
the Moody’s regression, with higher levels of income associated with smaller rating gaps.
This variable may well be picking up the fact that sovereigns with the best foreign currency
ratings cannot enjoy a rating gap because they are already at the top of the rating scale,
though it is not clear why such an effect would only hold for Moody’s. Saving is also
significant in the Moody’s regression, with higher saving rates permitting larger rating gaps.
When a sovereign can tap a large pool of local savings, Moody’s may view it as easier for the
sovereign to roll over local currency debt and avoid default. Public debt/GDP leads to smaller
rating gaps in the Moody’s regression, which suggests that Moody’s views very heavy debt
burdens as making default more likely on all debt, regardless of currency. Real effective
exchange rate overvaluation also leads to larger gaps at Moody’s, and this may illustrate the

180 BIS Papers No 30

fact that it is expensive to maintain an overvalued exchange rate, and using foreign
exchange to defend a currency reduces the availability of foreign exchange for repaying
debt.
A few variables enter uniquely into the S&P gap regressions as well. For instance, the
measures of regulatory quality and import cover lead to larger rating gaps in the S&P
regression, while the time elapsed since local currency default narrows the gap. The latter
result perhaps indicates that S&P is more likely to view longer default-free periods as an
important sign of creditworthiness on all obligations, which should diminish any foreign-local
currency differential.
In the gap regressions, the fixed effect for Asian countries is even more muted than it was in
the simple foreign and local currency rating regressions. Both the Moody’s and S&P
regressions suggest that the average expected Asian rating gap is nearly equal to the
expected gap in other countries. This finding supports the view that the different degree of
“notching” in Asia between local and foreign currency ratings can be better explained by
broadly applicable fundamentals than by Asia-specific factors.

Rating difference regressions

The regressions discussed above suggest that the rating agencies may weight variables
differently when they assess foreign and local currency creditworthiness. In order to further
investigate how the rating agency methodologies differ, we regress the difference between
S&P and Moody’s foreign currency ratings, and then the difference between their local
currency ratings, on the explanatory variables, with the results reported in the final two
columns of Table 10. Here, our convention is that positive differences imply a better rating by
Moody’s.
As in the gap regressions, inflation, M2, and M2/reserves are important determinants of the
differences between the agencies’ foreign and local currency ratings. S&P tends to weight
inflation more heavily as a negative factor in local currency ratings, but relatively less heavily
as a negative factor for foreign currency ratings. S&P also tends to weight monetary
expansion less heavily as a negative factor for both foreign and local currency ratings.
Additionally, control of corruption, domestic debt/GDP, budget revenue/GDP, short-term
debt/GDP, exchange rate changes, and time elapsed since default all appear to be weighted
significantly differently across the agencies.
Two final findings are that while average Asian fixed effects do not provide assistance in
predicting differences between Moody’s and S&P foreign currency credit ratings, there is a
significant Asian effect in the difference between S&P and Moody’s local currency ratings.
That is, S&P’s Asian local currency rating are significantly better than those of Moody’s, even
after controlling for differences in rating methodology.

Conclusion
This paper has analysed the patterns of the foreign and local currency ratings of S&P and
Moody’s both in Asia and globally, with a particular emphasis on whether ratings and the
gaps between foreign and local currency ratings are driven by the same factors in Asia as
elsewhere in the world, and whether the different rating agencies take the same approach to
ratings and gaps. We find that rating gaps in Asia can be explained by many of the same
factors that drive gaps globally, and that the evidence for an Asia-wide effect on ratings is
slim.
The local-foreign currency rating gaps of both agencies can be partly explained by inflation,
M2, and M2/reserves, but we also find evidence of a divergence across the rating agencies

BIS Papers No 30 181

in methodology. Namely, the gaps of S&P appear to depend more heavily on regulatory
quality, import cover, and time elapsed since default, while the gaps in Moody’s ratings are
better explained by per capita income, saving, public debt/GDP, and the real effective
exchange rate overvaluation. At the same time, an Asia factor may help to explain the
difference of local currency ratings between agencies, since S&P’s local currency Asian
ratings are significantly better than those of Moody’s even after controlling for differences in
methodology.
As domestic bond markets grow in importance, understanding local currency sovereign
ratings and what makes them different from foreign currency ratings will become increasingly
important. This paper suggests that there are important differences in the way rating
agencies view the relationship between foreign and local currency ratings, a fact that could
have implications for investors and regulators alike.

Table 1
Domestic and foreign currency sovereign ratings
Number of Asian sovereigns in parentheses

New domestic currency

New foreign currency ratings
ratings

Number of sovereigns
Pre-1985 15 (3) 0 (0)
1986-90 22 (8) 2 (0)
1991-95 20 (3) 32 (7)
1996-2000 55 (4) 65 (9)
2001-04 17 (0) 20 (2)
Total 129 (18) 119 (18)

Note: Sovereigns are deemed to have a rating if one of the three major agencies has a rating outstanding. The
United States did not receive a foreign currency rating until 1992.
Sources: Fitch Investors Service; Moody’s Investors Service; Standard & Poor’s.

Table 2
The credit quality of newly assigned sovereign ratings

New domestic currency

New foreign currency ratings
ratings

Median rating
Pre-1985 AAA ...
1986-90 A AA+
1991-95 BB+ AAA
1996-2000 BB BBB
2001-04 B+ B+

Note: Sovereigns are deemed to have a rating if one of the three major agencies has a rating outstanding.
Sources: Fitch Investors Service; Moody’s Investors Service; Standard & Poor’s.

182 BIS Papers No 30

 Table 3
Local and foreign currency credit rating of
selected Asian economies, November 2004

S&P Moody’s Fitch R&I

LC FC LC FC LC FC LC FC

Australia AAA AAA Aaa Aaa AAA AA+ AAA AA+

China BBB+ BBB+ – A2 A A– – A
Hong Kong SAR AA– A+ Aa3 A1 AA+ AA– AA AA–
India BB+ BB Ba2 Baa3 BB+ BB+ – BBB
Indonesia B+ B B2 B2 B+ B+ – B
Japan AA– AA– A2 Aaa AA– AA AAA AAA
Korea A+ A– A3 A3 AA– A – A–
Macau SAR – – A1 A1 – – – –
Malaysia A+ A– A3 Baa1 A+ A– – A–
Mongolia B B – – – – – –
New Zealand AAA AA+ Aaa Aaa AAA AA+ AAA AA+
Pakistan BB B+ B2 B2 – – – –
Papua New Guinea B+ B B1 B1 B+ B – –
Philippines BBB– BB Ba2 Ba2 BB+ BB – BBB–
Singapore AAA AAA Aaa Aaa AAA AAA AAA AAA
Taiwan, China AA– AA– Aa3 Aa3 AA A+ – AA
Thailand A BBB+ Baa1 Baa1 A– BBB – BBB+
Vietnam BB BB– – B1 BB BB– – –

Note: LC refers to local currency rating, and FC to foreign currency rating.

Sources: Fitch Investors Service; Japan Rating and Investment Information, Inc (R&I); Moody’s Investors
Service; Standard & Poor’s.

BIS Papers No 30 183

 Table 4
Domestic vs foreign currency rating
gaps by rating, November 2004
Asian countries in parentheses

Domestic currency debt Domestic currency debt

S&P foreign No difference rated higher by exactly rated higher by more
currency one notch than one notch
rating
Number of sovereigns

AAA 18 (2) 0 0
AA 8 (2) 2 (1) 0
A 7 (0) 8 (1) 5 (2)
BBB 2 (1) 6 (0) 6 (1)
BB 3 (0) 11 (2) 4 (1)
B 18 (1) 4 (2) 1 (1)

Note: Ratings indicate the broad letter grade category, eg AA stands for credits rated AA+, AA and AA–.
Source: Standard & Poor’s.

Table 5
Domestic vs foreign currency rating gaps, November 2004
Asian countries in parentheses

Moody’s S&P

Number of sovereigns

4-notch differential – –
3-notch 6 (0) 8 (0)
2-notch 7 (0) 8 (4)
1-notch 15 (2) 31 (6)
No difference 62 (10) 59 (6)
–1-notch 2 (0) –
–2-notch 1 (1) –
–3-notch – –
–4-notch – –
–5-notch 1 (1) –
Total 94 (14) 106 (16)

Sources: Moody’s Investors Service; Standard & Poor’s.

184 BIS Papers No 30

 Table 6
Notching of local currency credit rating
of Asian economies, November 2004

S&P Moody’s Fitch R&I

LC/FC gap LC/FC gap LC/FC gap LC/FC gap

Australia 0 0 1 1
China 0 – 1 –
Hong Kong SAR 1 1 2 1
India 1 –2 0 –
Indonesia 1 0 0 –
Japan 0 –5 –1 0
Korea 2 0 2 –
Macau SAR – 0 – –
Malaysia 2 1 2 –
Mongolia 0 – – –
New Zealand 1 0 1 1
Pakistan 2 0 – –
Papua New Guinea 1 0 1 –
Philippines 2 0 1 –
Singapore 0 0 0 0
Taiwan, China 0 0 2 –
Thailand 2 0 2 –
Vietnam 1 – 1 –

Average 0.94 –0.33 1.00 0.60

Average (ex Japan) 1.00 0.00 1.14 0.75

Note: LC refers to local currency rating, and FC to foreign currency rating.

Sources: Fitch Investors Service; Japan Rating and Investment Information, Inc (R&I); Moody’s Investors
Service; Standard & Poor’s.

BIS Papers No 30 185

 Table 7
Explanatory regression variables

Category Subcategory Variables

Macroeconomic Inflation Inflation over 1 year, 5 years and 10 years

Log inflation over 1 year, 5 years and 10 years
GDP Log per capita GDP
Per capita GDP
GDP growth 1-year rate
GDP growth 3-year rate
GDP growth 3-year average annual rate
Monetary M2: 1-year, 5-year, 10-year, % change
M2: log 1-year, 5-year and 10-year change
M2: 1-year, 5-year and 10-year volatility
M2: log 1-year, 5-year and 10-year volatility
Monetary/liquidity M2/reserves: 1-year, 5-year, 10-year, % change
M2/reserves: log 1-year, 5-year and 10-year change
M2/reserves: 1-year, 5-year and 10-year volatility
M2/reserves: log 1-year, 5-year and 10-year volatility
Saving/investment Saving/GDP
Investment/GDP
Political Political Transparency International Corruption Perceptions Index
Political risk score (Political Risk Services)
Government control of corruption (latest available,
Kaufmann
et al (2003))
Government effectiveness (latest available, Kaufmann
et al (2003))
Government regulatory quality (latest available, Kaufmann
et al (2003))
Government Government Public debt/GDP
finance finance Fiscal balance/GDP
Budget revenue/GDP
Domestic debt/GDP
Government debt/revenue
External position Debt Net debt/GDP
Net debt/exports
Gross debt/exports
Gross debt/GDP
Short-term debt/reserves
Short-term debt/GDP
Import cover
Exchange rate Nominal exchange rate 1-year change
Real effective exchange rate, % deviation from LT
average (JPMorgan Chase)
IMF exchange rate regime dummy variable (pegs and
managed floats are coded as 1)

Default Years since foreign currency default

Log years since foreign currency default
Years since local currency default
Log years since local currency default

Sources: Transparency International; Political Risk Services’ International Country Risk Guide; Kaufmann et al
(2003); EIU; Datastream; Standard & Poor’s; JPMorgan Chase.

186 BIS Papers No 30

 Table 8
The determinants of Moody’s and S&P’s
foreign and local currency ratings

Agency S&P Moody’s S&P Moody’s

Foreign Foreign Local Local

Dependent variable
currency rating currency rating currency rating currency rating

R-squared .9949 .9961 .9879 .9951

F-test of significance of
fixed effects, p-value <.0001 <.0001 <.0001 <.0001
Hausman test, fixed/
random effects, p-value .0068 .1642 .0026 .0004
Degrees of freedom 329 235 358 340

Independent variables Coeff t-stat Coeff t-stat Coeff t-stat Coeff t-stat

Macroeconomic
Inflation, log 1-yr 0.366 3.930
Inflation, 1-yr
Inflation, log 10-yr
Per capita GDP, log –3.016 –7.870 –2.231 –5.360 –1.752 –3.170 –0.985 –2.950
GDP growth, 3-yr avg –0.093 –3.460
M2, 10-yr % chg
M2, log 10-yr % chg
M2, 1-yr log volatility 0.334 2.600
M2, 5-yr volatility
M2/reserves, 5-yr log
volatility 0.463 2.680 0.908 4.950 0.734 4.840
M2/reserves, log 10-yr
% chg
M2/reserves, 1-yr volatility
M2/reserves, log 5-yr %
chg
Investment –0.053 –2.970 –0.083 –4.210 –0.064 –3.000
Saving
Political
Political risk score –0.034 2.560 –0.030 1.970 –0.045 –2.630
Regulatory quality –1.418 –6.300
Control of corruption –1.264 –4.120 –1.466 –3.700
Government finance
Public debt/GDP 0.041 7.040 0.072 9.690 0.037 7.300
Domestic debt/GDP
Budget revenue/GDP
External
Net debt/exports 0.004 2.420
Short-term debt/GDP
Import cover

BIS Papers No 30 187

 Table 8 (cont)
The determinants of Moody’s and S&P’s
foreign and local currency ratings

Agency S&P Moody’s S&P Moody’s

Foreign Foreign Local Local

Dependent variable
currency rating currency rating currency rating currency rating

Independent variables Coeff t-stat Coeff t-stat Coeff t-stat Coeff t-stat
External (cont)
Exchange rate rigidity 0.493 2.580 0.584 2.510
Exchange rate, 1-yr chg –0.012 –3.320
Real effective exchange
rate
Years since foreign
currency default, log –0.346 –2.910 –0.236 –2.790
Years since local currency
default, log –0.661 –2.110
Years since local currency
default
Time-series fixed effects
Year 1995 –0.143 –0.580
Year 1996 –0.197 –0.650 0.013 0.040 –0.854 –2.330 –0.151 –0.610
Year 1997 –0.031 –0.110 0.482 1.650 –0.613 –1.790 –0.049 –0.230
Year 1998 –0.255 –1.400 0.376 1.960 –0.461 –2.070 0.195 1.220
Year 1999 –0.298 –1.810 0.224 1.240 –0.154 –0.730 0.023 0.160
Year 2000 –0.196 –1.230 0.121 0.710 –0.237 –1.150 0.035 0.250
Year 2001 –0.244 –1.580 0.158 0.950 –0.123 –0.590 –0.003 –0.020
Year 2002 –0.341 –2.380 –0.259 –1.590 –0.327 –1.760 –0.029 –0.220
Cross-sectional fixed
effects
Argentina 42.066 11.990 36.030 9.340 30.755 6.210 20.740 6.650
Australia 38.690 10.100 29.766 7.150 27.594 5.110 8.888 2.590
Austria 24.271 4.310 7.661 2.100
Bahrain 37.817 10.290 26.200 5.090
Barbados
Belgium 21.543 3.750
Bolivia 31.343 10.230
Botswana 35.650 11.300 25.630 5.710 12.167 4.310
Brazil 39.252 11.950 34.015 9.460 27.670 5.950 19.606 6.750
Bulgaria 35.604 11.350 33.375 10.180 26.324 6.060 17.419 6.340
Canada 36.597 9.060 29.884 7.130 24.147 4.270 7.431 2.060
Chile 37.966 11.700 32.059 9.160 26.727 5.840 12.227 4.190
Colombia 33.313 10.380 26.989 7.660 23.193 5.430 11.423 4.040
Costa Rica 39.463 11.870 33.270 9.380 30.149 6.480 16.238 5.480
Croatia 36.586 10.760 32.284 8.980 23.328 4.780 13.537 4.480
Cyprus 37.123 9.760 31.503 7.860 23.764 4.450 11.859 3.480
Czech Republic 36.820 10.910 31.375 8.690 23.978 4.950 11.826 3.910

188 BIS Papers No 30

 Table 8 (cont)
The determinants of Moody’s and S&P’s
foreign and local currency ratings

Agency S&P Moody’s S&P Moody’s

Foreign Foreign Local Local

Dependent variable
currency rating currency rating currency rating currency rating

Independent variables Coeff t-stat Coeff t-stat Coeff t-stat Coeff t-stat
Cross-sectional fixed
effects (cont)
Denmark 38.360 9.380 29.897 6.880 26.363 4.600 7.895 2.150
Dominican Republic 38.269 12.400 32.239 9.560 30.655 6.800 18.846 6.970
Ecuador 38.642 12.060 34.761 10.510 25.128 5.550 19.327 6.970
Egypt 30.896 10.180 29.295 9.540 19.215 4.590 11.276 4.250
El Salvador 35.873 11.580 29.702 8.800 24.268 5.580 14.114 5.200
Estonia 36.972 11.500 29.650 8.300 27.612 6.030 12.079 4.190
Finland 36.491 8.890 27.110 6.100 26.513 4.700 6.196 1.690
France 34.486 8.610 27.636 6.530 22.736 4.050 7.248 2.010
Germany 35.456 8.850 24.001 4.270 7.269 2.010
Greece 23.261 4.310 9.722 2.800
Guatemala 16.280 6.170
Hong Kong SAR 42.158 11.200 32.084 6.070
Hungary 24.702 5.080
Iceland 41.818 10.380 31.264 7.140 27.720 4.870 8.048 2.230
India 31.870 12.330 21.837 6.190 15.682 7.020
Indonesia 32.841 11.540 29.574 9.980 23.548 5.990 17.839 7.220
Ireland 37.489 9.390 26.707 4.770 8.711 2.450
Israel 37.226 9.370 32.177 7.990 21.479 3.870 10.981 3.090
Italy 33.010 7.990 29.440 6.960 19.251 3.340 7.742 2.100
Jamaica 35.997 10.120 34.626 10.150 22.921 4.660 11.676 3.760
Japan 34.617 7.990 31.204 7.300 20.567 3.390 7.603 1.980
Jordan 34.562 10.920 32.587 10.230 22.675 5.220 15.437 5.550
Kazakhstan 35.176 11.550 28.188 8.300 26.046 6.260 14.228 5.450
Kuwait 37.615 10.110 25.484 4.810 14.352 4.270
Latvia 36.325 11.560 30.374 8.870 25.933 5.850 12.829 4.590
Lebanon 35.836 9.370 36.395 10.130 22.581 4.370 15.976 4.760
Lithuania 36.458 11.320 31.195 8.930 26.082 5.790 13.767 4.790
Malaysia 35.989 11.030 31.147 9.020 24.230 5.350 12.750 4.400
Malta 36.123 9.870 23.025 4.490 13.197 4.050
Mauritius 12.026 4.140
Mexico 39.169 11.620 32.694 9.030 27.604 5.880 15.226 5.130
Moldova 18.243 7.750
Mongolia 37.041 11.780 24.691 5.480
Morocco 33.859 11.220 30.190 9.710 22.095 5.310 14.429 5.420
Netherlands 37.108 9.170 26.405 4.710 7.186 1.990
New Zealand 37.625 10.050 28.299 7.000 26.987 5.130 8.080 2.400

BIS Papers No 30 189

 Table 8 (cont)
The determinants of Moody’s and S&P’s
foreign and local currency ratings

Agency S&P Moody’s S&P Moody’s

Foreign Foreign Local Local

Dependent variable
currency rating currency rating currency rating currency rating

Independent variables Coeff t-stat Coeff t-stat Coeff t-stat Coeff t-stat

Cross-sectional fixed
effects (cont)
Nicaragua 30.400 10.390 14.347 5.670
Norway 39.149 9.610 28.197 4.910 9.101 2.490
Oman 40.673 11.660 32.591 8.510 31.162 6.520 16.078 5.180
Pakistan 30.697 11.400 28.556 9.950 21.952 6.060 17.578 7.520
Panama
Papua New Guinea 32.810 11.660 22.335 5.810 15.894 6.590
Paraguay 37.156 12.600 27.770 6.800 18.626 7.290
Peru 36.977 12.040 32.898 10.050 28.206 6.640 15.282 5.620
Philippines 32.411 11.240 29.233 9.850 21.378 5.350 13.338 5.330
Poland 36.635 11.040 30.687 8.630 25.441 5.450 11.975 4.030
Portugal 34.721 9.280 23.997 4.550 8.481 2.540
Qatar 40.000 9.770 26.635 4.650 14.710 4.040
Romania 38.991 12.850 34.770 10.550 29.474 6.930 20.244 7.560
Russia 35.430 11.070 32.009 9.180 24.258 5.310 17.049 6.140
Senegal 24.887 6.810
Singapore 35.674 8.950 23.292 4.100 7.136 1.990
Slovakia 37.991 11.490 32.044 8.990 25.251 5.370 12.211 4.160
Slovenia 37.127 10.310 30.644 7.890 23.436 4.510
South Africa 35.953 11.180 29.310 8.460 24.724 5.500 11.810 4.120
Spain 35.348 9.100 28.584 7.000 24.312 4.530 7.837 2.270
Sweden 37.393 9.120 24.457 4.180 7.650 2.080
Switzerland 39.128 9.490 28.306 4.950 8.411 2.280
Thailand 33.952 10.990 30.294 9.300 24.058 5.640 12.422 4.540
Trinidad 37.049 10.550 32.732 8.840 24.834 5.090 13.992 4.500
Tunisia 34.659 11.100 22.762 5.280 13.575 4.920
Turkey 38.494 11.610 33.704 9.780 28.213 6.090 19.083 6.490
Ukraine 34.907 12.320 30.390 9.840 26.194 6.530 20.564 8.480
United Kingdom 37.322 9.360 29.444 7.000 26.323 4.720 8.137 2.300
United States 37.749 9.170 26.569 4.640 8.305 2.260
Uruguay 38.842 11.300 33.406 9.030 28.855 6.010 16.581 5.370
Venezuela 39.969 11.910 34.759 9.520 21.854 7.350
Vietnam 22.940 6.270
1 1
Asia average 35.473 3.31 29.931 3.57 24.272 4.551 11.963 2.931
Non-Asia average 36.897 3.471 31.457 3.581 25.381 4.851 13.124 3.071
1
Standard deviations are given for the average Asian and non-Asian fixed effect. Asian countries are shaded.

190 BIS Papers No 30

 Table 9
The determinants of Moody’s and S&P’s foreign
and local currency ratings, common sample

Agency S&P Moody’s S&P Moody’s

Foreign Foreign Local Local

Dependent variable
currency rating currency rating currency rating currency rating

R-squared .9954 .9959 .9880 .9947

F-test of significance of
fixed effects, p-value <.0001 <.0001 <.0001 <.0001
Hausman test, fixed/
random effects, p-value .0209 .1801 .0026 .0010
Degrees of freedom 219 220 269 274

Independent variables Coeff t-stat Coeff T-stat Coeff t-stat Coeff t-stat
Macroeconomic
Inflation, log 1-yr 0.308 2.830
Inflation, 1-yr
Inflation, log 10-yr
Per capita GDP, log –2.711 –5.950 –2.121 –4.980 –2.375 – 3.510 –0.873 – 2.280
GDP growth, 3-yr avg –0.084 – 2.700
M2, 10-yr % chg
M2, log 10-yr % chg
M2, 1-yr log volatility 0.311 1.920
M2, 5-yr volatility
M2/reserves, 5-yr log
volatility 0.713 3.230 0.995 4.970 0.909 4.910
M2/reserves, log 10-yr
% chg
M2/reserves, 1-yr
volatility
M2/reserves, log 5-yr %
chg
Investment –0.069 –3.000 –0.086 –3.920 –0.029 –0.940
Saving
Political
Political risk score –0.046 –2.760 –0.029 –1.820 –0.024 –1.090
Regulatory quality –1.497 –6.400
Control of corruption –1.140 –2.750 –0.972 – 1.960
Government finance
Public debt/GDP 0.045 6.770 0.073 7.730 0.040 6.430
Domestic debt/GDP
Budget revenue/GDP
External
Net debt/exports 0.004 2.480
Short-term debt/GDP
Import cover
Exchange rate rigidity 0.787 3.230 –0.005 –0.020

BIS Papers No 30 191

 Table 9 (cont)
The determinants of Moody’s and S&P’s foreign
and local currency ratings, common sample

Agency S&P Moody’s S&P Moody’s

Foreign Foreign Local Local

Dependent variable
currency rating currency rating currency rating currency rating

Independent variables Coeff t-stat Coeff T-stat Coeff t-stat Coeff t-stat
Exchange rate, 1-yr chg –0.012 – 2.800
Real effective exchange
rate
Years since foreign
currency default, log –0.325 –2.430 –0.221 –1.790
Years since local
currency default, log –0.660 – 1.720
Years since local
currency default
Time-series fixed
effects
Year 1995
Year 1996 –0.215 –0.600 0.059 0.190 –1.545 – 1.270 –0.214 –0.260
Year 1997 –0.058 –0.180 0.516 1.730 –0.973 –1.370 –0.248 –0.500
Year 1998 –0.213 –0.930 0.456 2.240 –0.646 – 2.370 0.220 1.220
Year 1999 –0.443 –2.170 0.255 1.380 –0.262 – 1.100 0.044 0.280
Year 2000 –0.266 –1.350 0.158 0.900 –0.272 – 1.180 0.071 0.460
Year 2001 –0.148 –0.780 0.196 1.150 –0.200 – 0.860 0.042 0.280
Year 2002 –0.263 –1.470 –0.177 –1.060 –0.376 – 1.830 –0.025 –0.170
Cross-sectional fixed
effects
Argentina 39.570 9.620 34.692 8.830 35.235 5.710 19.210 5.350
Australia 36.199 8.100 28.422 6.690 30.294 4.550 7.289 1.850
Austria
Bahrain
Barbados
Belgium
Bolivia
Botswana 28.629 5.140 10.885 3.360
Brazil 37.071 9.640 32.753 8.940 30.491 5.300 18.052 5.390
Bulgaria 33.426 9.130 32.276 9.680 29.534 5.510 15.930 5.020
Canada 33.786 7.190 28.552 6.680 26.811 3.860 5.802 1.390
Chile 36.166 9.590 31.014 8.710 29.023 5.160 10.972 3.280
Colombia 30.443 7.980 25.606 7.110 27.059 5.160 9.520 2.910
Costa Rica 37.060 9.570 32.133 8.910 33.356 5.830 14.695 4.300
Croatia 34.400 8.690 31.170 8.530 26.913 4.430 12.095 3.470
Cyprus 34.230 7.720 30.211 7.390 27.498 4.170 10.130 2.580
Czech Republic 34.822 8.830 30.308 8.260 27.211 4.490 10.487 3.020

192 BIS Papers No 30

 Table 9 (cont)
The determinants of Moody’s and S&P’s foreign
and local currency ratings, common sample

Agency S&P Moody’s S&P Moody’s

Foreign Foreign Local Local

Dependent variable
currency rating currency rating currency rating currency rating

Independent variables Coeff t-stat Coeff T-stat Coeff t-stat Coeff t-stat
Cross-sectional fixed
effects (cont)
Denmark 35.246 7.400 28.507 6.430 29.739 4.230 6.134 1.450
Dominican Republic 35.755 9.780 31.099 9.060 34.246 6.050 17.399 5.580
Ecuador 36.262 9.650 33.006 9.680 28.911 5.110 17.471 5.420
Egypt 28.747 8.190 28.293 9.050 22.728 4.400 9.921 3.220
El Salvador 32.775 8.880 28.547 8.290 27.874 5.150 12.751 4.080
Estonia 35.197 9.110 28.555 7.850 30.841 5.380 10.684 3.220
Finland 32.920 6.830 25.521 5.610 29.454 4.250 4.067 0.960
France 31.425 6.750 26.284 6.100 26.562 3.840 5.439 1.310
Germany 27.516 3.970 5.412 1.300
Greece
Guatemala
Hong Kong SAR
Hungary
Iceland 38.791 8.250 29.821 6.680 30.811 4.400 6.394 1.540
India 24.524 5.590 14.530 5.620
Indonesia 30.931 9.260 28.535 9.460 27.729 5.720 16.397 5.730
Ireland 30.222 4.380 6.914 1.680
Israel 34.039 7.380 30.900 7.530 25.650 3.780 9.224 2.240
Italy 29.818 6.220 28.038 6.500 23.316 3.290 5.695 1.340
Jamaica 33.626 8.170 33.567 9.650 26.204 4.280 9.918 2.740
Japan 31.765 6.340 29.971 6.900 23.786 3.180 5.841 1.300
Jordan 32.440 8.840 31.549 9.740 25.674 4.770 13.920 4.310
Kazakhstan 33.312 9.310 27.003 7.820 28.948 5.470 12.862 4.280
Kuwait 29.824 4.550 12.824 3.320
Latvia 34.623 9.380 29.303 8.410 29.186 5.220 11.496 3.580
Lebanon 33.029 7.470 35.215 9.610 26.295 4.100 14.101 3.600
Lithuania 34.212 9.080 30.055 8.460 29.583 5.260 12.306 3.720
Malaysia 33.847 8.910 30.055 8.550 27.586 4.950 11.355 3.400
Malta 26.762 4.200 11.673 3.110
Mauritius
Mexico 37.193 9.390 31.509 8.550 31.178 5.360 13.861 4.040
Moldova
Mongolia
Morocco 31.872 9.110 29.185 9.230 25.134 4.870 13.006 4.230
Netherlands 29.605 4.310 5.326 1.280
New Zealand 35.111 8.040 26.992 6.540 29.325 4.520 6.439 1.670

BIS Papers No 30 193

 Table 9 (cont)
The determinants of Moody’s and S&P’s foreign
and local currency ratings, common sample

Agency S&P Moody’s S&P Moody’s

Foreign Foreign Local Local

Dependent variable
currency rating currency rating currency rating currency rating

Independent variables Coeff t-stat Coeff T-stat Coeff t-stat Coeff t-stat

Cross-sectional fixed
effects (cont)
Nicaragua
Norway 31.563 4.460 7.284 1.730
Oman 38.029 9.320 31.326 8.000 34.929 5.960 14.607 4.100
Pakistan 28.435 9.000 27.448 9.390 25.254 5.660 16.048 5.920
Panama
Papua New Guinea 25.262 5.310 14.443 5.180
Paraguay 31.890 6.350 17.309 5.880
Peru 35.454 9.880 31.861 9.560 31.107 5.970 14.024 4.480
Philippines 30.846 9.170 28.249 9.360 23.465 4.760 11.988 4.140
Poland 34.677 8.970 29.568 8.180 28.262 4.880 10.558 3.090
Portugal 27.063 4.140 6.578 1.700
Qatar 30.920 4.380 12.818 3.040
Romania 36.963 10.400 33.667 10.040 33.541 6.370 18.908 6.140
Russia 33.202 8.820 30.807 8.690 28.237 4.940 15.516 4.840
Senegal
Singapore 25.594 3.650 5.476 1.320
Slovakia 35.933 9.310 30.909 8.520 28.566 4.830 10.749 3.180
Slovenia 34.746 8.280 29.469 7.470
South Africa 33.759 8.970 28.144 7.990 27.465 4.970 10.311 3.130
Spain 32.397 7.180 27.208 6.490 27.726 4.220 5.962 1.490
Sweden 27.250 3.790 5.706 1.350
Switzerland 31.229 4.460 6.731 1.580
Thailand 32.163 8.920 29.257 8.840 26.611 5.060 11.044 3.510
Trinidad 34.435 8.380 31.531 8.370 28.781 4.790 12.344 3.440
Tunisia 25.662 4.790 12.084 3.800
Turkey 36.244 9.370 32.518 9.250 31.240 5.420 17.487 5.150
Ukraine 32.890 9.850 29.594 9.320 29.730 5.920 18.295 6.480
United Kingdom 34.690 7.480 28.127 6.550 29.196 4.270 6.513 1.590
United States 29.777 4.260 6.786 1.600
Uruguay 36.342 9.060 32.170 8.540 32.392 5.510 15.040 4.230
Venezuela 37.592 9.500 33.532 9.020
Vietnam
Asia average 32.980 3.891 28.783 3.561 26.418 5.671 10.480 3.391
Non-Asia average 34.401 3.951 30.279 3.681 28.944 5.991 11.373 3.571
1
Standard deviations are given for the average Asian and non-Asian fixed effect. Asian countries are shaded.

194 BIS Papers No 30

 Table 10
The determinants of gaps between foreign
and local currency ratings, Moody’s and S&P
S&P-Moody’s S&P-Moody’s
Agency S&P Moody’s
difference difference
Foreign-local Foreign-local
Foreign Local
Dependent variable currency rating currency rating
currency rating currency rating
gap gap
R-squared .8981 .8659 .6900 .8804
F-test of significance of
fixed effects, p-value <.0001 <.0001 <.0001 <.0001
Hausman test, fixed/
random effects, p-value .0006 .0782 .0116 .0002
Degrees of freedom 319 199 170 128
Independent variables Coeff t-stat Coeff t-stat Coeff t-stat Coeff t-stat
Macroeconomic
Inflation, log 1-yr –0.174 –2.730 0.159 2.180
Inflation, 1-yr –0.028 –3.440
Inflation, log 10-yr –0.107 –1.690
Per capita GDP, log –2.608 –5.030
GDP growth, 3-yr avg
M2, 10-yr % chg –0.001 –3.350
M2, log 10-yr % chg 0.100 1.970
M2, 1-yr log volatility
M2, 5-yr volatility –0.011 –1.890 0.110 1.790
M2/reserves, 5-yr log
volatility
M2/reserves, log 10-yr %
chg 0.101 3.150 –0.072 –2.040
M2/reserves, 1-yr
volatility –0.006 –2.970
M2/reserves, log 5-yr %
chg 0.089 2.680
Investment
Saving 0.073 3.210
Political
Political risk score
Regulatory quality 0.509 2.950
Control of corruption –0.791 –2.350
Government finance
Public debt/GDP –0.010 –1.740
Domestic debt/GDP 0.033 3.000
Budget revenue/GDP 0.050 1.680 –0.041 –2.160
External
Net debt/exports
Short-term debt/GDP 0.000 1.850
Import cover 0.079 2.510
Exchange rate rigidity
Exchange rate, 1-yr
change –0.006 –2.770

BIS Papers No 30 195

 Table 10 (cont)
The determinants of gaps between foreign
and local currency ratings, Moody’s and S&P
S&P-Moody’s S&P-Moody’s
Agency S&P Moody’s
difference difference
Foreign-local Foreign-local
Foreign Local
Dependent variable currency rating currency rating
currency rating currency rating
gap gap
Independent variables Coeff t-stat Coeff t-stat Coeff t-stat Coeff t-stat
External (cont)
Real effective exchange
rate 0.018 2.460
Years since foreign
currency default, log
Years since local
currency default, log –0.567 –2.220
Years since local
currency default –0.248 –4.500
Time-series fixed
effects
Year 1995 0.501 1.990 –1.328 –4.840
Year 1996 1.228 6.650 0.338 1.390 –1.797 –3.810 –1.212 –4.470
Year 1997 1.078 6.090 0.325 1.560 –1.589 –4.060 –1.193 –4.800
Year 1998 0.852 5.170 0.330 1.900 –1.366 –3.920 –0.844 –4.230
Year 1999 0.468 3.040 0.109 0.690 –0.977 –3.310 –0.734 –3.840
Year 2000 0.485 3.170 –0.087 –0.580 –0.672 –2.990 –0.706 –3.710
Year 2001 0.358 2.450 –0.175 –1.160 –0.381 –2.110 –0.570 –2.990
Year 2002 0.232 1.660 –0.313 –2.210 –0.088 –0.570 –0.404 –2.140
Cross-sectional fixed
effects
Argentina 0.759 1.310 22.009 4.600 3.046 3.640
Australia 2.238 1.880 25.821 4.840 25.155 4.470
Austria 0.584 0.470 25.302 4.520 1.402 1.280
Bahrain 1.590 1.560
Barbados 3.833 3.940 –2.826 –3.420
Belgium
Bolivia 3.358 2.450 42.320 4.290
Botswana –0.087 –0.060
Brazil 2.769 3.530 19.932 4.550 1.767 1.420
Bulgaria 3.047 2.420 18.875 4.590 20.894 3.940
Canada 2.265 1.790 25.789 4.610 –0.696 –0.670
Chile 5.189 3.750 23.424 5.150 47.390 4.510 –1.081 –1.870
Colombia 7.109 4.430 23.118 5.440 46.876 4.410 –1.425 –2.150
Costa Rica 3.064 2.240 46.742 4.730 –0.117 –0.220
Croatia
Cyprus 2.330 2.350 23.800 4.630 9.962 3.890 –1.451 –1.590
Czech Republic

196 BIS Papers No 30

 Table 10 (cont)
The determinants of gaps between foreign
and local currency ratings, Moody’s and S&P
S&P-Moody’s S&P-Moody’s
Agency S&P Moody’s
difference difference
Foreign-local Foreign-local
Foreign Local
Dependent variable currency rating currency rating
currency rating currency rating
gap gap
Independent variables Coeff t-stat Coeff t-stat Coeff t-stat Coeff t-stat
Cross-sectional fixed
effects (cont)
Denmark 2.371 1.720 26.125 4.610 0.613 0.510
Dominican Republic –3.019 –8.150 1.003 1.470
Ecuador 2.741 4.550 19.607 4.600 –0.402 –0.470
Egypt 3.853 3.200 20.960 5.120 18.724 4.160
El Salvador 1.481 2.880 19.790 4.880 2.073 2.650 –0.156 –0.350
Estonia 0.155 0.190 20.124 4.540 4.253 2.940
Finland
France 0.626 0.640 25.154 4.520 0.992 0.880
Germany 0.318 0.310 1.911 1.850
Greece 3.866 2.680 26.046 5.020 41.298 4.310
Guatemala 3.722 2.620 44.976 4.490 0.138 0.310
Hong Kong SAR 2.185 1.750
Hungary 22.387 4.700 37.821 4.430
Iceland 4.079 3.530 2.739 3.390
India 3.778 3.370 –3.384 –4.790
Indonesia 3.437 3.080 17.240 4.420 14.085 4.290 –0.394 –0.510
Ireland
Israel 3.645 3.430 24.928 4.590 13.686 4.260 –2.407 –2.120
Italy 1.443 1.370 25.481 4.500 11.850 3.470 –1.920 –1.330
Jamaica 2.726 2.680 22.972 4.960 10.595 4.390
Japan 0.341 0.310 24.686 4.140
Jordan 4.157 3.850 18.214 4.240 13.281 4.110
Kazakhstan
Kuwait 0.923 1.360
Latvia
Lebanon 0.349 0.280 21.571 4.460
Lithuania
Malaysia 3.851 3.830 21.163 4.530 10.913 4.260 –1.115 –1.790
Malta 2.794 2.840
Mauritius 22.501 4.880
Mexico 5.713 4.190 44.597 4.540 –0.948 –1.630
Moldova 16.644 4.650
Mongolia 0.288 0 .460
Morocco 4.212 3.970 17.415 4.260 11.801 4.370
Netherlands 1.679 1.090 –0.247 –0.200

BIS Papers No 30 197

 Table 10 (cont)
The determinants of gaps between foreign
and local currency ratings, Moody’s and S&P
S&P-Moody’s S&P-Moody’s
Agency S&P Moody’s
difference difference
Foreign-local Foreign-local
Foreign Local
Dependent variable currency rating currency rating
currency rating currency rating
gap gap
Independent variables Coeff t-stat Coeff t-stat Coeff t-stat Coeff t-stat
Cross-sectional fixed
effects (cont)
New Zealand 1.681 1.430 24.492 4.770 24.177 4.550
Nicaragua 16.783 4.540
Norway 1.547 1.300 25.300 4.420 2.152 1.990
Oman 2.625 1.930 1.399 2.090
Pakistan 4.206 3.730 15.497 4.370 12.046 3.820 –3.639 –5.560
Panama 23.914 4.290
Papua New Guinea 3.835 4.010 –1.978 –2.350
Paraguay 5.162 3.620
Peru 3.440 2.460 21.425 5.220 44.160 4.430
Philippines 4.878 4.530 17.575 4.470 13.998 4.420 –2.444 –4.350
Poland 4.145 3.510 22.285 4.890 22.124 4.810 0.019 0.040
Portugal 0.947 1.080 23.108 4.570 5.684 2.970 –0.139 –0.140
Qatar 1.983 2.110 23.922 4.120 6.494 3.130 –0.242 –0.310
Romania 3.788 2.830 19.746 4.770 29.041 4.210 –0.782 –1.080
Russia
Senegal
Singapore –0.171 –0.180
Slovakia
Slovenia 3.348 4.250 3.548 2.550
South Africa 4.522 4.510 22.469 5.060 7.579 2.670 –0.385 –0.560
Spain 2.369 1.490 23.931 4.480 0.613 0.780
Sweden 1.151 1.310 27.070 4.780 1.363 1.140
Switzerland 1.498 0.940
Thailand 4.821 3.540 20.095 4.620 43.601 4.430 –1.826 –3.010
Trinidad 3.407 3.410 9.245 3.940
Tunisia 4.940 4.750
Turkey 2.416 1.960 19.034 4.130 20.297 4.570 –1.379 –1.430
Ukraine 0.766 1.250 16.151 4.210 0.485 0.440
United Kingdom 1.086 0.820 25.085 4.610 1.207 1.360
United States 1.551 1.120 0.662 1.310
Uruguay 3.325 2.420 21.675 4.670 44.227 4.540 –0.595 –1.020
Venezuela 19.667 4.290 0.016 0.020
Vietnam
Asia average 2.597 1.011 21.581 4.741 21.988 4.931 –1.857 0.601
Non-Asia average 2.579 1.041 21.883 4.701 20.109 4.611 –0.176 0.751
1
Standard deviations are given for the average Asian and non-Asian fixed effect. Asian countries are shaded.

198 BIS Papers No 30

References
Bank for International Settlements (2001): “The development of bond markets in emerging
economies”, BIS Papers, no 11, Basel, June.
Basel Committee on Banking Supervision (2004): International convergence of capital
measurement and capital standards: a revised framework, Bank for International
Settlements, June.
Borio, C and F Packer (2004): “Assessing new perspectives on country risk”, BIS Quarterly
Review, December.
Cantor, R and F Packer (1996): “Determinants and impact of sovereign credit ratings”,
FRBNY Economic Policy Review, October.
Claessens, S, D Klingebiel and S Schmukler (2003): “Government bonds in domestic and
foreign currency: the role of macroeconomic and institutional factors”, Discussion Paper no
3789, Centre for Economic Policy Research, February.
Eichengreen, B, R Hausmann and U Panizza (2003): “Currency mismatches, debt
intolerance and original sin: why they are not the same and why it matters”, NBER Working
Paper Series, October.
Executives’ Meeting of East Asia-Pacific Central Banks (2004): “EMEAP central banks
announce the launch of the Asian Bond Fund 2”, Press Release, 16 December.
Fitch Investors Service (2003): Presentation at Daiwa Securities Japan Symposium,
February.
International Monetary Fund (2003): “Local securities and derivatives markets in emerging
markets: selected policy issues”, IMF Financial Markets Quarterly.
Kaufmann, D, A Kraay and M Mastruzzi (2003): Governance matters III: governance
indicators for 1996-2002, World Bank.
McCauley, R N and G Jiang (2004): “Diversifying with Asian local currency bonds”, BIS
Quarterly Review, September, pp 51-66.
McCauley, R and W White (1997): “The euro and European financial markets”, BIS Working
Papers, no 41, May; also in Paul Masson, Thomas H Krueger and Bart G Turtelboom (eds)
(1997), EMU and the international monetary system, IMF, Washington DC, pp 324-88.
Merrill Lynch (2000): Bond index rules & definitions, 12 October.
Moody’s Investors Service (2003a): Sovereign bond defaults, rating transitions, and
recoveries (1985-2002), February.
——— (2003b): A quantitative model for local currency bond ratings, September.
——— (2003c): India: analysis, February.
——— (2003d): “Japan: a rating agency perspective”, Special Comment, February.
——— (2004): A quantitative model for foreign currency government bond ratings, February.
Packer, F (2003): “Mind the gap: domestic versus foreign currency sovereign ratings”, BIS
Quarterly Review, September.
Standard & Poor’s (2002): Sovereign credit ratings: a primer, April.
——— (2003): Foreign/local currency and sovereign/nonsovereign ratings differentials,
22 September.
——— (2004): Sovereign credit ratings: a primer, April.
Trevino, L and S Thomas (2001): “Local and foreign currency ratings: what determines
sovereign transfer risk?”, Journal of Fixed Income, June.

BIS Papers No 30 199

 Comments on Kate Kisselev and Frank Packer’s
paper “Minding the gap in Asia: foreign
and local currency ratings”

Tom Byrne

To understand the ratings gap phenomenon, or lack thereof, the historical dynamics of local
currency ratings need to be explored, in addition to a static look at key indicators as rating
drivers. I will also comment on a number of related points made by the authors.
In assigning local currency ratings, the practice in general is to notch up from the foreign
currency rating on the grounds that there is no foreign exchange constraint or risk of capital
controls being imposed on local currency payments. Governments have the power to tax,
and so can take on a higher local currency than foreign currency debt burden. Domestic
purchasers of government securities, especially banks, are a much more stable investor
base than foreign creditors. However, Moody’s is mindful that the recent historical record of
sovereign default, beginning with the 1998 Russian case, shows that governments have
defaulted on obligations in both foreign and local currency, or in one but not the other, as
well. This means that stylised approaches to rating local currency obligations are not reliable
or accurate. A government’s fiscal and debt position, economic growth prospects,
institutional strengths (in particular, Moody’s takes into account the World Bank’s indicator of
government effectiveness) and policy capabilities need to be taken into account on a case by
case basis.
Moody’s big push for expanding local currency ratings in Asia took place in the aftermath of
the 1997 Asian financial crisis. This was a time when a number of sovereign foreign currency
ratings were many notches (up to six, in the most extreme case, Korea) lower than their
pre-crisis peak. Of course, the lower foreign currency ratings reflected a diminished capacity
to service external debt. (It should be pointed out that, although ratings fell too rapidly, the
Asian financial crisis did not result in any government bond defaults.) Against this
background of “abnormally” low foreign currency ratings, local currency ratings for most of
the Asian emerging market economies were assigned.
Likewise, the assignment of local currency ratings in Asia was also constrained to some
degree by the 1997 crisis. Moody’s recognised that the fiscal cost of the crisis was not fully
reflected in government budgets or debt in early 1998, so we tried to anticipate future
fiscalisation from financial sector restructuring as well as the rise in government debt from
the shift to fiscal stimulus and large budget deficits as governments offset the contractionary
effect of the crisis. Nevertheless, Asian countries’ local currency ratings were generally
assigned two or three notches higher than their foreign currency government ratings in 1998.
The relatively quick adjustment in the external payments position of most countries affected
by the crisis, compared with their slower fiscal consolidation, explains the reduction in the
original notching differential. In general, multi-notch upward movement took place in the
foreign currency ratings, while upward movement in local currency ratings was not as
pronounced, or was absent. Korea and Thailand, for example, sharply reduced external debt
and accumulated large holdings of official foreign exchange reserves. Accordingly, Korea’s
foreign currency government rating was raised four notches, and Thailand’s three notches,
post-crisis. In contrast, Korea’s local currency government rating was raised only one notch
post-crisis, and Thailand’s has remained unchanged, reflecting the large accumulation of
domestic debt and less robust prospects for debt reduction over the long term.
The authors’ rejection of the argument that governments can easily inflate their way out of a
local currency debt problem is well taken. The historical record does not show that this

200 BIS Papers No 30

remedy is painless, either economically or politically, or readily taken. This view has carried
little weight in rating committee deliberations. In contrast, default may be better fiscal policy,
in extremis, than inflation.
The key indicators guiding the movement of local currency government ratings have been,
first, general government debt to GDP and, second, general government debt to general
government revenue. Both indicate the capacity of the government, immediately, and the
economy, ultimately, to service the government’s debt obligations. The former indicator is
especially relevant because it is unique to Moody’s rating consistency model that attempts to
explain the ordinal rankings of local currency government ratings.
There was an “Asian bias” embedded in Moody’s foreign currency ratings, although it is
difficult to quantify in a statistical sense. Moody’s Sovereign Risk Unit had for a long period
emphasised savings/investment balances as key rating support factors. Asian countries have
high saving rates compared with other regions. Large savings available to absorb
government debt make recourse to foreign borrowing less necessary, supporting local
currency ratings. Large savings can also be channelled into domestic investment. Economic
growth and the level of national income are also key rating determinants for both local and
foreign currency government ratings.
Lastly, regarding political risk, Moody’s has explicitly incorporated this into the ratings of
Taiwan (China) and Korea. In both cases, geopolitical risks constrain both the foreign and
local currency government ratings equally, with no ratings gap.

BIS Papers No 30 201

 Building infrastructure for Asian bond
markets: settlement and credit rating

Daekeun Park and Changyong Rhee 1

1. Introduction
The development of bond markets in Asia has recently emerged as an important policy
issue. During the Ninth APEC Finance Ministers Meeting in September 2002, it was agreed
that a regional bond market would be developed through securitisation and credit
guarantees. Since then, a number of meetings have been held on this and related issues.
The consensus for developing regional bond markets for Asian countries is a result of the
1997 Asian financial crisis, which the underdevelopment of the region’s bond markets is
thought to have greatly exacerbated. Firms that had long been dependent upon banks for
funds could not find alternative sources of financing when the crisis erupted. The idea of
creating regional bond markets is also promoted as a means of overcoming the double
mismatch problem that most Asian borrowers face when depending on short-term foreign
currency debt to fund long-term projects generating domestic currency revenues. This is also
considered one of the root causes of the 1997 crisis.
Development of a regional bond market is also seen as a way to facilitate the mobilisation of
East Asian savings within the region. The foreign exchange reserves of most Asian countries
have increased significantly since the financial crisis, boosted by the huge current account
surpluses triggered by the economic recession and sharp currency depreciations brought
about by the financial crisis. By the end of 2002, the Asian economies together held more
than half of the world’s foreign exchange reserves, and the bulk of these were invested in
safe and liquid assets such as US Treasury securities and supranational bonds. At the same
time, until regional bond markets are fully established, East Asian borrowers have to rely on
international financial markets for funding. East Asia as a whole can thus be considered an
importer of safe assets and an exporter of risky assets. As has been pointed out by Oh et al
(2003a), such a pattern of capital flows is not desirable in the sense that it deprives the
region’s financial markets and institutions of valuable opportunities to develop and could
render the countries in the region more vulnerable to financial crises.
Under the Asian Bond Market Initiative, ASEAN+3 has launched six working groups to study
various aspects of regional bond markets including securitisation, regional credit rating
agencies, regional clearing and settlement systems, regional credit guarantee agencies and
so on. 2 The paper focuses on the two topics that primarily address building institutional
infrastructure for the Asian bond market: securities clearing and settlement systems and
credit rating agencies.
Every financial transaction ultimately entails settlement of securities. In order for financial
markets to function properly, reliable and efficient financial substructures, including clearing
and settlement systems, must be established and supported by the legal and regulatory

1
We thank Haeil Jang at the Korea Securities Depository (KSD) for helpful comments and discussions.
2
EMEAP (Executives’ Meeting of East Asia and Pacific Central Banks) has also set up the Asian Bond Fund
(ABF) with contributions from the foreign reserves of each member bank. This fund is managed by the Bank
for International Settlements under the mandate to invest in dollar-denominated bonds issued by qualified
Asian issuers.

202 BIS Papers No 30

systems. An internationally compatible clearing and settlement system is required to promote
active cross-border transactions of bonds among Asian countries. Section 2 of this paper
discusses how a cross-border clearing and settlement system can be constructed for the
development of Asian bond markets. The paper attempts to determine if it is necessary to
establish a new Asian settlement system even though there are cross-border settlement
systems operated by international central securities depositories (ICSDs) such as Euroclear
and Clearstream.
The second financial infrastructure component that this paper addresses is the regional
credit rating system. The Asian financial crisis in 1997 made East Asian countries recognise
the importance of international credit rating agencies. These countries witnessed the
influence of the judgments of credit rating agencies in shaking domestic economies and
foreign markets to their foundations and in determining the direction and intensity of foreign
capital flows. Credit rating agencies are expected to become more influential because capital
adequacy for banks under the revised Basel Accord will take credit ratings into account (BIS
(1999)).
Currently, local currency denominated bonds in Asian countries are independently rated by
local credit rating agencies. Government bonds are given the highest credit ratings within
each country regardless of the sovereign credit ratings, making credit rating comparisons
among countries useless. To increase transaction volumes in Asian bond markets and cross-
border bond issuance and investment, a common regional credit rating system is needed.
Let us consider, for example, the two-tier securitisation process, which is already under
discussion as a strategy for creating the Asian bond market. In this process, each country
aggregates and securitises local currency denominated bonds; the junior bonds are
absorbed domestically, and the senior bonds are sold to the offshore special purpose vehicle
(SPV). The SPV, in turn, repackages these senior bonds and issues asset-backed securities
(ABSs). 3 For this two-tier securitisation process to work, the credit ratings of the bonds
issued in the first and second stages must be based on a common standard, regardless of
the country of issuance. 4
Section 3 of this paper takes up the question of whether there is a need to establish a new
regional credit rating institution in Asia. The alternative would be for the local or global credit
rating agencies to continue to do the job.
This paper makes different recommendations regarding the settlement and clearing system
and credit rating. We recommend that a new regional institution (a regional ICSD dubbed
AsiaSettle) be created to serve as the regional clearing and settlement system. For the credit
rating system, in contrast, we recommend that a common standard and methodology be
established through coordination among the existing local credit rating agencies.
Establishment of a regional ICSD is recommended even though ICSDs such as Euroclear
and Clearstream already handle cross-border settlements. Establishment of AsiaSettle would
solve the third time zone problem, which the existing ICSDs cannot solve, and the process of
establishing AsiaSettle with the support of the Asian governments itself would provide an
opportunity for Asian countries to relax and harmonise financial regulations and open up their
domestic markets.

3
For a discussion on the pros and cons of two-tier securitisation, please refer to Oh and Park (2003).
4
The need for a common rating standard will become a more concrete problem when the New Basel Accord is
adopted. This is because when capital adequacy regulations consider credit ratings, the financial supervisory
institutions must clarify how the ratings of foreign bonds by foreign agencies will be rated by domestic
standards. For example, when a Japanese financial institution holds a bond that a Korean agency has rated A,
there is a question as to whether a Korean A-rated bond equals a Japanese A-rated bond when assessing the
additional risk factor.

BIS Papers No 30 203

Conversely, it is unrealistic and not recommendable to establish a regional credit rating
agency through government support. In the credit rating business, advanced scientific and
analytic methods are important, but so are value judgments based on local knowledge. Thus,
if the regional rating agency is to rate local bonds in each country, it must build up local
databases, human resources and other infrastructure. The costs of building this infrastructure
would not be negligible. In addition, for a credit rating business to be successful, the most
important factor is its reputation for being impartial and accurate. It is questionable whether
an agency established with government support and not exposed to market competition
would be competitive and acknowledged by the market as impartial. Also, if a regional
agency were established with government support, there would be the risk that the agency
would crowd out private local credit rating agencies. Therefore, it is preferable for the
governments to promote collaboration among local and global rating agencies in establishing
a common rating system and meeting new business needs. The Association of Credit Rating
Agencies in Asia (ACRAA) has already started to standardise the rating systems with a view
to developing the Asian bond market.
This paper is organised as follows. Section 2 describes the current cross-border clearing and
settlement system in Asia and the role of the existing ICSDs. It evaluates the pros and cons
of various possibilities for setting up a cross-border clearing and settlement system in Asia
and proposes the establishment of AsiaSettle as a regional ICSD. A detailed explanation of
the benefits of creating a new regional ICSD is offered.
Section 3 discusses the credit rating system. Like Section 2, it deals with the current status of
credit rating agencies in Asia and the role of the global credit rating agencies. Section 3 goes
on to take up the question of whether establishment of a new agency or collaboration among
the existing local agencies would be more appropriate. This section also discusses how the
collaboration should be arranged if this second option is considered more appropriate.

2. Building a settlement infrastructure for Asian bond markets

2.1 Cross-border securities settlement in Asia

Cross-border trades of securities can be settled through four different channels: (i) through a
local agent, (ii) through a global custodian (GC), (iii) through an ICSD or (iv) through a
national central securities depository (NCSD) that has a link with the NCSD in the country of
issue. 5 All of these channels are utilised in varying degrees for cross-border settlement in
Asia. 6
Historically, local agents have been used most frequently in cross-border settlement,
especially when securities settlement must be made in a country that has no linkage
between its NCSD and other NCSDs or between its NCSD and ICSD. However, using a local
custodian has one major disadvantage in that investors must designate a separate local
custodian for each country where investments are made, and sometimes the fees charged
by local custodians can be significant.
Due to this cost disadvantage, institutional investors have increasingly used global
custodians (GCs) that provide settlement and custody services in multiple markets through a

5
In addition to these channels, cross-border trades can be settled through direct membership in the NCSD of
the country of issue. According to BIS (1995), however, this channel is seldom utilised since CSDs typically
prohibit foreign residents from becoming participants.
6
BIS (1995) and the Giovannini Group (2001) compare various methods of cross-border securities settlement.

204 BIS Papers No 30

single gateway by integrating services performed by a network of subcustodians, including its
own local branches. GCs can offer cost advantages through economies of scale and scope.
Another important advantage of using GCs is the availability of integrated multicurrency
banking and cash management services since most GCs are large international commercial
banks. Most settlements of Asian securities are made through GCs, not ICSDs, using
international currencies such as the US dollar.
In fact, the business base of GCs lies in the inefficiency of the international financial market
due to differences in the trading, clearing and settlement systems of each country. GCs
provide investors with the convenience of a single interface for their international security
transactions. The convenience of a single gateway, however, must come at a price since
GCs also have to hire local agents themselves. In addition, the quality of their services differs
widely by region depending upon the quality of service provided by the local agents.

Table 1
Countries with settlement linkages to Euroclear1

Region Countries Number

Asia Pacific Australia, New Zealand, Hong Kong SAR, Indonesia, Japan, 9
Malaysia, the Philippines, Singapore, Thailand
Europe Belgium, Finland, France, Germany, Greece, Ireland, 16
Luxembourg, the Netherlands, Norway, Portugal, Spain,
Sweden, Switzerland, United Kingdom, Austria, Italy
America United States, Argentina, Canada, Mexico 4
Others Russia, South Africa 2

Total 31
1
Includes all countries linked to Euroclear through specialised depositories, common depositories or clearing
depositories.
Source: Korea Securities Depository.

Another way to conduct cross-border settlement is to use an ICSD. As a matter of fact,

ICSDs are the natural channels of cross-border settlement for securities like eurobonds that
use ICSDs as the central depository. However, even when Asian bonds are deposited in the
NCSD of the country of issue and not in an ICSD, their settlement can still be executed
through an ICSD. In this case, the ICSD should be linked to the NCSDs of individual
countries or to the custodian banks that are members of the NCSDs. Countries that have
these linkages with ICSDs are called clearing members. Table 1 shows the 31 clearing
members of Euroclear. The table shows that the coverage of Asian countries by Euroclear is
quite limited. As of 2002, only seven Asian countries were clearing members of Euroclear.
Cross-border settlement can also be conducted by using an NCSD that has a bilateral
linkage with the NCSD of the country of issue. Currently, however, there are only a few
bilateral linkages between Asian NCSDs. Some of these include the linkages between
Australia and New Zealand, Hong Kong SAR and New Zealand, Korea and Hong Kong and
Japan and Hong Kong. Except for the linkage between Australia and New Zealand, trading
volumes are quite minimal.

BIS Papers No 30 205

2.2 The case for a regional clearing and settlement system in Asia
An informal working group has been organised under the Asian Bond Market Initiative of
ASEAN+3 to review and propose the clearing and settlement system for Asian bond markets.
Though this group has been in operation for some time, it has yet to make a concrete
proposal for the clearing and settlement system, although there have been a few
suggestions. These include utilising the existing ICSDs, establishing a regional settlement
system by linking NCSDs and creating a new regional ICSD. In this subsection, we will
review the pros and cons of using the existing ICSDs. In particular, we will discuss the third
time zone problem that arises when investors use the existing ICSDs located in the
European time zone to settle securities trades denominated in Asian currencies.

2.2.1 Limits of using an international central securities depository (ICSD)

Asian bonds will be issued in the offshore market or in the domestic market of the regional
financial centres in Asia and will be denominated in the currencies other than the currency of
the country of issue (Park and Park, 2003). As a result, Asian bonds are likely to be
international bonds like eurobonds. Since securities traded in the eurobond markets are
mostly deposited in and settled through ICSDs such as Euroclear or Clearstream, the same
ICSDs may also serve as the depository for Asian bonds. Then, Asian bonds denominated in
Asian currencies can be settled through the existing ICSDs in the same way eurobonds
denominated in Asian currencies are currently settled through the ICSDs.
Currently, Euroclear offers investors a choice of the currency of settlement, but as shown in
Table 2, the range of choices is very limited. 7 Only 32 currencies for 42 countries are
available for settlement, and of these, only nine are Asian currencies. The currencies of
Korea, China, India and Taiwan, China (hereinafter referred to as Taiwan) are not included,
but this is not because they are not internationalised. The Malaysian ringgit and the
Singapore dollar are settlement currencies in Euroclear even though they are not
internationalised.
Most Asian currencies are not Euroclear settlement currencies because there are some
limitations on their convertibility or substantial legal uncertainties regarding the application of
regulations on foreign currency transactions. In Korea, for example, omnibus accounts
(accounts for large groups of investors) are not permitted. This is a major reason that the
Korean won is excluded from the list. 8 Non-resident investors in Korea are required to report
their individual identities when opening Korean won-denominated accounts. This regulation
prohibits ICSDs from opening omnibus accounts with the NCSD in Korea. If an ICSD has an
omnibus account in its own name and manages all the internal transactions among its
members, the government fears that it will be unable to monitor individual transactions. This
regulation, however, subjects foreign investors to onerous procedural requirements and does
not permit protection of investors’ anonymity. For these reasons, Euroclear does not
designate the Korean won as a currency of settlement. 9

7
Euroclear (2002) lists settlement currencies and cash correspondents.
8
Oh et al (2003b) present the reasons why ICSDs do not include the won among settlement currencies.
9
The Indonesian rupiah is a currency of settlement in Euroclear, but its use became somewhat restricted after
the financial crisis in 1997. The restriction is not due to exchange rate or credit risk to Euroclear: as the
settlement of Euroclear is done via the RTGS and DVP systems, Euroclear is not subject to any exchange
rate or credit risk. The restriction was introduced due to increasing uncertainty with regard to regulation on
capital transactions in Indonesia.

206 BIS Papers No 30

 Table 2
Settlement currencies of Euroclear

Region Country

Asia Australia (AUD), New Zealand (NZD), Hong Kong (HKD), Nine currencies
Indonesia (IDR), Japan (JPY), Malaysia (MYR), Philippines of nine countries
(PHP), Singapore (SGD), Thailand (THB)
Europe EUR (Austria, Belgium, Finland, France, Germany, Greece, 15 currencies of
Ireland, Italy, Portugal, Spain, Luxembourg, the Netherlands), 26 countries
Norway (NOK), Sweden (SEK), Denmark (DKK), Switzerland
(CHF), the United Kingdom (GBP),
[Republic of Croatia (HRK), Czech (CZK), Republic of Iceland
(ISK), Slovakia (SKK), Estonia (EEK), Hungary (HUF),
Lithuania (LTL), Latvia (LVL), Poland (PLN)]1
North/South USA (USD), Argentina (ARS), Canada (CAD), Mexico (MXN) Four currencies
America of four countries
Middle East and South Africa (ZAR), [Kuwait (KWD), Israel (ILS)]1 Three
Africa currencies of
three countries
Others Gold (XAU)2 One currency
Total 32 currencies of 42 countries
1
Countries in [ ] are not clearing members of Euroclear, but their currencies are designated as currencies of
settlement. Russia is a clearing member of Euroclear, but the Russian ruble is not a currency of settlement
(payments are settled in US dollars). Gold is converted into one of the currencies of settlement and then settled
according to its value in the currency in question.
Source: Korea Securities Depository.

In addition to the limitation on the currencies of settlement, there is also limitation on the
countries of settlement. If we exclude Australia and New Zealand, only seven countries in
Asia are directly or indirectly connected to Euroclear: Hong Kong, Japan, Singapore,
Thailand, the Philippines, Malaysia and Indonesia, with some restriction on Malaysia and
Indonesia. Other Asian countries such as Korea, China, Taiwan, India and Pakistan are not
clearing members of Euroclear. The low coverage of Euroclear in the Asian region indicates
that there is potential demand for a regional ICSD, and we will take this issue up in the next
section.
The selection criteria for clearing members are not identical to those for currencies of
settlement. All four cases are possible if we compare Table 1 and Table 2. First, countries
such as Japan and Thailand are clearing members of Euroclear, and their currencies are
designated currencies of settlement. Second, Russia is a clearing member of Euroclear, but
its currency is not a settlement currency. Third, there are countries such as Korea that are
not clearing members and whose currencies are not designated as currencies of settlement.
Fourth, countries such as Croatia, the Czech Republic, Israel and Iceland are not clearing
members, but their currencies are used for settlement.

2.2.2 The need to create a regional clearing and settlement system

The low coverage of the Asian countries and currencies by existing ICSDs provides a case
for creating a regional ICSD for Asian bond markets. Besides the limited coverage, there is
another reason for creating a regional ICSD; that is, the time zone problem.

BIS Papers No 30 207

Since most Asian currencies are not internationalised, the payment settlement of Asian
bonds denominated in local currencies must be finalised in the local market, even though
securities settlement can be done through ICSDs located in Europe. However, due to the
difference in time zone between Europe and Asia, real-time settlement of Asian bonds is not
possible, and there have been calls to establish a regional clearing and settlement system
within Asia, the third time zone, in order to cover the non-business hours of the two other
time zones, Europe and the Americas.
To illustrate the third time zone problem, consider the settlement process of an Asian bond
denominated in Hong Kong dollars. 10 Hong Kong is seven hours ahead of Brussels, where
Euroclear is located. Assume that the settlement date of the bond transaction is 2 October in
Brussels. In order to finalise the settlement by that date, Euroclear currently mandates that
the buyer and the seller deposit money and securities in the common depository of Euroclear
in Hong Kong, HSBC Bank, by 1 October, which is a day before the settlement date. After
receiving a notification from HSBC overnight, Euroclear Bank in Brussels completes the
security settlement by 9 am on 2 October (4 pm in Hong Kong). Then, the seller in Hong
Kong can withdraw Hong Kong dollars and the settlement can be finished by 2 October.
Instead of depositing money and securities a day before the settlement date, if the buyer and
seller want to settle securities using the real-time gross settlement (RTGS) system on
2 October in Belgian time, the seller may not be able to withdraw money by 2 October. For
example, by the time the RTGS settlement is completed by 3 pm on 2 October, it is already
10 pm in Hong Kong, and the bond seller has to wait until the next day to withdraw his
money. This is one reason why Euroclear mandates that traders deposit money and
securities a day in advance of the settlement date for bonds that are denominated in Asian
currencies. If bonds are denominated in European currencies or the US dollar, securities and
payment settlement can be completed on the same day through the RTGS system as there
is no time zone difference, and the time difference between Europe and the Americas works
favourably between security settlement and payment settlements. The time zone problem
implies that investors have to bear the extra cost of losing liquidity for a day when trading
Asian currency denominated bonds. If there were a regional ICSD within Asia, investors
would not face this extra cost. The benefit of solving the third time zone problem can be
significant considering that the major investors for Asian currency denominated bonds are
institutional investors located in Asia.
The time zone problem may also hinder efforts to reduce the settlement cycle. The recent
movement to do so has been motivated by a report by the Group of Thirty (1988) that
recommended that the settlement cycle be reduced to T+3 in order to reduce settlement risk.
The settlement cycle and settlement amount are the key determinants of settlement risk.
Accommodating the report’s recommendation, major countries have reduced the settlement
cycle to T+3. Some, including the United States, plan to reduce the settlement cycle further
down to T+1.
If trades of Asian bonds denominated and settled in Asian currencies are settled through a
regional settlement system, such as a regional ICSD located in Asia and operating in the
Asian time zone, then investors would not have to deposit securities or money one day in
advance of the settlement date. This would not only facilitate liquidity management by
investors but would also make it possible to reduce the settlement cycle down to T+1.

10
For the detailed settlement procedure, refer to Euroclear (2003).

208 BIS Papers No 30

2.3 Proposals for establishing a regional clearing and settlement system in Asia
The previous subsection presented the reasons for creating a regional clearing and
settlement system for Asian bond markets. There are a couple of proposals to create a
regional clearing and settlement system: the bilateral linkage model and the regional ICSD
model. In this subsection, we will review these two proposals.

2.3.1 Model I: Bilateral linkages between NCSDs for cross-border settlements

If bilateral linkages can be established among Asian NCSDs, cross-border settlement is
possible without ICSDs. In fact, the European Central Securities Depositories Association
(ECSDA) once proposed a bilateral linkage model for cross-border settlement in Europe.
Figure 1 shows two pan-European bilateral linkage models that ECSDA has studied. One is
the Eurolinks Real-time Network model (Spaghetti Model), and the other is the European
Financial Superhighway (Cannelloni Model). The former connects all NCSDs with each
other, while the latter uses major NCSDs as pivots to connect other small NCSDs. Both
models emphasise the need to strengthen mutual linkages among NCSDs.

Figure1
Bilateral linkage models of ECSDA

( 1 ) E urolinks R eal- T im e N etw ork ( 2) E uropean F inancial S uperhighw ay

( S paghetti M odel) ( Canneloni M odel)

Minor Minor
CSD CSD

CSD CSD
m ajor
CSD
CSD CSD
Minor
Minor
CSD
CSD m ajor m ajor
CSD CSD
Minor Minor
CSD CSD
CSD CSD
m ajor
CSD
CSD CSD
Minor Minor
CSD CSD

Within Asia, Hong Kong has shown the greatest interest in bilateral linkage models. The
Hong Kong Monetary Authority (HKMA) has proposed the establishment of AsiaClear, a
regional settlement institution, by linking the clearing and settlement systems of member
countries in Asia in the manner of the internet. That is, the HKMA defines AsiaClear not as a
single hub institution, but as a common network among individual NCSDs in Asia. Thanks to
advances in information technology, HKMA believes that linking NCSDs is now feasible in
virtual space. For this reason, disagreement over where to locate AsiaClear can be
finessed. 11 In fact, the HKMA has been actively pursuing linkages with other Asian countries;
it now has links with Australia, New Zealand and Korea, and soon will have one with China. 12

11
HKMA (1997a) researched the state of financial market and IT development in Hong Kong that might enable
Hong Kong to function as a financial hub in Asia.
12
See HKMA (1997b, 1997c, 1998, 1999, 2002) on the linkages between HKMA and other NCSDs in Asia.

BIS Papers No 30 209

However, there are a number of problems with applying bilateral linkage models to Asia.
First, it is an inefficient method compared with settlement through ICSDs. Transaction costs
in bilateral linkage models would likely be high as each NCSD has to open accounts in the
NCSDs of all counterparties. Second, these models could only handle securities registered in
both NCSDs being used for a transaction. Third, the initial setup costs of establishing
bilateral linkages can be high if countries do not share standardised settlement platforms. 13
However, the most serious bottleneck in applying bilateral linkage models is that bond
markets in Asian countries are at such greatly varying stages of development that they
cannot be readily linked to each other. Among Asian NCSDs, only seven countries (Australia,
Hong Kong, Japan, Korea, Malaysia, New Zealand and Singapore) are using RTGS and
delivery versus payment (DVP) systems. Table 3 shows the wide difference among Asian
NCSDs with regard to compliance with the recommendations of G30/ISSA, which renders
the building of bilateral linkages among them difficult. 14
Different legal systems are another factor. Unlike Europe, where the legal systems of each
country are relatively similar, Asian countries have much more varied historical backgrounds,
cultures and legal systems, which make it difficult to standardise linkages among Asian
NCSDs.

2.3.2 Model II: Building a regional ICSD - AsiaSettle

Oh et al (2003b) propose that AsiaSettle, a regional ICSD, be established as a regional
clearing and settlement system in Asia. According to this proposal, AsiaSettle will serve as
the hub for the cross-border settlement of Asian bonds with NCSDs serving as the
subdepositories. Unlike Euroclear, which is for the most part indirectly linked to NCSDs
through custodian banks, AsiaSettle will be directly linked to NCSDs. The direct linkage
model has the cost advantage over the indirect linkage model. For the cash settlement,
AsiaSettle will be directly linked to the central banks of Asian countries rather than to
custodian banks. The direct linkage to the central banks has the advantage of reducing the
settlement cycle as well as the settlement cost. Oh et al (2003b) discuss in detail the pros
and cons of alternative models of operation and governance structure for AsiaSettle.
The AsiaSettle model belongs to the hub and spoke model as far as settlement for Asian
bonds is concerned. The advantage of this model lies with the low setup cost since it fully
utilises the existing settlement infrastructure. The public characteristics of AsiaSettle may
also serve as a positive factor in building the direct linkages to NCSDs, considering Asian
governments’ interest in promoting Asian bond markets. 15 AsiaSettle can be established as a
private agency funded by the NCSDs and the central banks of Asian countries, or as a
multilateral agency.
AsiaSettle’s raison d’être has been set forth above in subsection 2.2: the limited coverage of
settlement services in Asia by existing ICSDs such as Euroclear and the time zone problem.
There are more reasons to create a regional ICSD for Asian bond markets: harmonisation of
regulations and introduction of a central counterparty (CCP).

13
Park and Hong (2001) discuss the advantages and disadvantages of bilateral linkage models.
14
For more details, see ISSA (2002).
15
Direct linkage means that an ICSD has its own omnibus account in a local NCSD. Indirect linkage means that
an ICSD is linked to a local NCSD through a third party such as a specialised or common depository. It is
more common for Euroclear to have indirect linkages with NCSDs.

210 BIS Papers No 30

 Table 3
NCSDs of Asian countries:
compliance with G30/ISSA recommendations

BA CN HK IN ID JP KR MY PA PH SG TH TW

Trade comparisons between

direct market participants by T+0
Matched trade details should be
linked in the settlement system
Indirect market participants to
achieve affirmation by T+1
Central depository, broadest
possible participation
Widest possible range of
depository eligible instruments
Immobilisation/ dematerialisation
to the utmost extent possible
Compatible rules and practices
in case of municipal CSDs
Real-time gross settlement
system
Trade netting system as per
Lamfalussy recommendations
Delivery vs payment (DVP) as
defined by ISSA
Same day funds for securities
settlement
Same day funds for the servicing
of securities portfolios
A rolling settlement system
should be adopted by all
markets
Final settlement for all trades by
T+3
Securities lending and borrowing
should by encouraged
Existing regulatory and taxation
barriers should be removed
ISO Standard 7775
(Securities Messages)
ISO Standard 6166
(ISIN Numbering System)

BA: Bangladesh, CN: China, HK: Hong Kong SAR, IN: India, ID: Indonesia, JP: Japan, KR: Korea,
MY: Malaysia, PA: Pakistan, PH: Philippines, SG: Singapore, TH: Thailand, TW: Taiwan, China.
Source: Korea Securities Depository.

BIS Papers No 30 211

As previously discussed, the low coverage of ICSDs in Asia is partly due to the existence of
complex regulations and legal uncertainties involving cross-border trading of securities.
Setting up AsiaSettle through the cooperation of Asian governments offers a great
opportunity to open domestic markets and harmonise bond market regulations across Asia.
The existing ICSDs are private entities, and Asian governments have had no incentive to
ease regulations to increase business flows for the ICSDs unless it is very much in their
national interest. However, if a regional ICSD were designed under the consent of Asian
governments, the governments would face incentives to ease and harmonise regulations
regarding cross-border trading and settlement of securities in order to promote Asian bond
markets. Thus, AsiaSettle could be an effective catalyst for easing regulations and opening
up local markets in Asia, and consequently in developing the Asian bond market.
Harmonisation of regulations, in turn, would enhance the cost efficiency of AsiaSettle. One of
the criticisms of AsiaSettle is that it would entail the additional cost of creating a new
institution. Contrary to the common belief, the cost of building the IT infrastructure for
AsiaSettle would not be very high. Instead, a large portion of the cost would arise from the
need to study the legal and regulatory environment for cross-border securities trading and
settlement in each country. It goes without saying that harmonising the regulations would
greatly reduce the need for research and consultation on legal and regulatory matters.
Admittedly, harmonisation of regulations enhances the cost efficiency of the bilateral linkage
model as well.
Creating a regional ICSD has the advantage of introducing central counterparty services in
the clearing and settlement of Asian bonds. A CCP is a special financial institution that
stands between the seller and buyer in each trade. It replaces the original contractual
obligations to deliver and pay with equivalent obligations by the CCP. 16 As a result, a CCP
replaces several counterparty exposures with a single one and reduces settlement risks. 17 A
CCP can benefit the capital markets by offering standardised processing that translates into
lower operating costs and anonymity among participants. Moreover, a CCP minimises the
value and volume of settlements through multilateral netting. For example, the gross amount
of securities settlement at the Depository Trust & Clearing Corporation (DTCC) in the United
States in 2000 was about 722 billion US dollars, but after multilateral netting, the net amount
of settlement shrank to only 22 billion US dollars.
Taking advantage of the benefits of a latecomer, the clearing and settlement system for
Asian bond markets could be designed so that AsiaSettle provides CCP services itself or by
setting up a subsidiary. By providing CCP services for bond settlement, AsiaSettle could
enhance the efficiency of Asian bond markets and differentiate itself from existing ICSDs. In
particular, the multilateral netting function of AsiaSettle could be expected to reduce foreign
exchange (FX) transaction costs in settling Asian bonds denominated in Asian currencies.
Since most Asian currencies are not internationalised and their exchange rates vary, it is
more likely that the payment settlements for Asian bonds will be effected by international
currencies such as the US dollar. However, if CCP services are provided, the volume of FX
transactions could be reduced significantly through multilateral netting, and settlement costs
could be significantly lowered.

16
This is known as “novation”. DTCC (2000) has an overview of the current development of the CCP industry.
17
The DVP system can also reduce settlement risks, but it cannot effectively cover replacement risk. A CCP can
cover principal as well as replacement risk.

212 BIS Papers No 30

 Table 4
Local credit agencies in Asian countries

Affiliation
Country Rating agencies (operation, Major stockholders
capital)
Korea KR Fitch Hanil Cement, Korea
Development Bank, Fitch
KIS Moody’s Moody’s
NICE R&I Domestic bank
SCI JCR SB Partners
Japan R&I Nikkei 56.5%
JCR
Moody’s Japan KK Moody’s Moody’s
S&P Japan S&P S&P
Fitch Japan Branch Fitch Fitch
China China Chengxin International Credit Rating Fitch 30%
Co Ltd with-
drawal
Fitch Ratings Hong Kong Limited Fitch Fitch subsidiary
Dagong Global Credit Rating Co Ltd Moody’s Moody’s
S&P office S&P S&P
China Lianhe Credit Rating Co Ltd
India The Credit Rating Information Services of S&P 9.68% acquired by S&P in
India Ltd (CRISIL) 1997
Investment Information & Credit Rating Moody’s Moody’s, Central Bank,
Agency Ltd (ICRA) public financial institutions
Credit Analysis and Research Limited (CARE) IDBI, Canara Bank, UTI
Fitch Ratings India Pvt Ltd Fitch Fitch subsidiary
Indonesia PEFINDO S&P S&P
PT Kasnic Credit Rating Indonesia
Singapore Moody’s Singapore Pte Ltd Moody’s Moody’s
S&P office S&P S&P
Fitch Ratings Singapore Pte Ltd Fitch Fitch
Bangladesh Credit Rating Information & Services Ltd JCR-VIS, Join venture between
(CRISL) RAM JCR-VIS and RAM
Malaysia Rating Agency Malaysia Berhad (RAM) Fitch Fitch 4.9%, minor shares
held by other banks
Malaysian Rating Corp Berhad Fitch Affiliate, but Fitch does not
seem to hold any shares
Pakistan JCR-VIS Credit Rating Co Ltd IIRA, JCR 15%, VIS 67.5%,
CRISL KSE 12.5% ISE 5.0%
The Pakistan Credit Rating Agency (Private) No longer
Ltd, (PACRA) affiliated
with Fitch

BIS Papers No 30 213

 Table 4 (cont)
Local credit agencies in Asian countries

Affiliation
Country Rating agencies (operation, Major stockholders
capital)
Philippines Philippine Rating Services Corp (PhilRatings) S&P
Fitch Ratings Manila Representative Office Fitch Fitch affiliate
Taiwan, Taiwan Ratings Corp (TRC) S&P Domestic banks and
China financial institutions
Fitch Ratings Taipei Representative Office Fitch Fitch subsidiary
Moody’s Chinese Taipei branch Moody’s Moody’s
Thailand Thai Rating & Information Services Co Ltd
(TRIS)
Fitch Ratings (Thailand) Ltd Fitch Fitch affiliate
Sri Lanka Fitch Ratings Lanka Ltd Fitch Fitch affiliate

3. Building a common credit rating system for the Asian bond

market

3.1 Credit rating agencies in East Asia

3.1.1 Local credit rating agencies

As seen in Table 4, most East Asian countries have their own local credit rating agencies,
which are responsible for rating bonds or examining the credit ratings of bank loans. In
Korea, for example, there are three credit rating agencies that are allowed to rate all types of
bonds, and one agency that rates only commercial paper (CP) and asset backed-securities
(ABSs). In Japan, two local agencies, R&I and JCR, are currently in business and global
credit rating agencies such as Moody’s, S&P and Fitch have also established themselves in
the Japanese market. Both R&I and JCR represent themselves as global credit rating
agencies and provide ratings for samurai and sovereign bonds. With the exception of Japan,
the credit rating agencies of East Asian countries are primarily focused on rating domestic
bonds, and most of them are linked to global credit rating agencies through capital
investment or operational collaboration. The last two columns of Table 4 indicate the
collaborative ties between local and global rating agencies and their major shareholders.

3.1.2 Global credit rating agencies in Asia

In response to the rapid expansion of the financial markets in Asia, global credit rating
agencies such as Moody’s, S&P and Fitch have branched out into Asia. Table 5 shows the
number of employees at the global credit rating agencies in Asia as of 2001. S&P has
234 employees, Moody’s 99 and Fitch 66.

214 BIS Papers No 30

 Table 5
Number of employees at global
credit rating agencies in Asia, 2001
S&P Moody’s Fitch

Japan 80 70 33
Hong Kong SAR 30 13 17
Singapore 24 10 4
Australia 100 6 12
Total 234 99 66
Source: Korea Investors Service.

Table 6 shows the number of issuers that have been directly rated by global credit rating
agencies in Asia. In 2001, Moody’s rated the highest number of issuing companies, followed
by S&P. However, ratings have been highly concentrated only in those countries where the
bond markets are relatively more developed, such as Japan, Australia, Korea and Hong
Kong.

Table 6
Number of issuers rated in 2001
S&P Moody’s Fitch Total

Japan 297 504 71 872

Hong Kong SAR 227 226 178 631
Singapore 53 89 6 148
Australia 47 38 3 88
Total 624 857 258 1,739
Source: Korea Investors Service.

3.1.3 Association of Credit Rating Agencies in Asia (ACRAA)

The Association of Credit Rating Agencies in Asia (ACRAA) is an organisation whose
members currently consist of 20 rating agencies in 11 Asian countries for mutual cooperation
and joint research. Currently, Japan’s JCR chairs the ACRAA, and PhilRatings of the
Philippines is the executive director. Every credit rating agency in Asia is eligible for ACRAA
membership. To join, an agency must receive the Executive Committee’s recommendation
and then the Board of Directors’ approval. The ACRAA holds a meeting every year at which
member credit rating agencies discuss various issues concerning mutual collaboration. The
ACRAA also offers educational programs to member agencies biannually.

BIS Papers No 30 215

 Table 7
ACRAA members
As of February 2006

Country Credit rating agency Notes

Bangladesh Credit Rating Information & Services

Credit Rating Agency of Bangladesh (CRAB)
China Dagong Global Credit Rating
Shanghai Far East Credit Rating
India Credit Rating Information Services of India Limited Board of directors
(CRISIL) Board of directors
Investment Information & Credit Rating Agency (ICRA) (Vice Chairman)

Credit Analysis and Research

Indonesia PEFINDO Credit Rating Indonesia
PT Kasnic Credit Rating Indonesia
Japan Japan Credit Rating Agency (JCR) Board of directors
(Chairman)
Korea Seoul Credit Rating & Information
Korea Investors Service
Korea Ratings
Malaysia Rating Agency Malaysia Berhad Board of directors
Malaysian Rating Corporation Berhard (MARC)

Pakistan JCR-VIS Credit Rating

The Pakistan Credit Rating Agency (PACRA)
Philippines Philippine Rating Services Corporation (PhilRatings) Board of directors
Taiwan Taiwan Ratings
Thailand TRIS Rating

Source: ACRAA.

3.2 The need for a regional credit rating agency

As the number of international bond investors increases in Asia, the need for common
regional credit rating standards becomes greater. The simplest means to meet this need
would be to utilise the global credit rating agencies, rather than establish a new regional
institution. Considering the fact that cross-border bond investment in Asia is still quite small,
utilising global agencies would be more cost effective. However, there are some limitations to
using the global credit rating agencies if Asian countries want to promote cross-border bond
investment in local currency denominated bonds.
The number of local currency denominated bonds in Asia being rated by global credit rating
agencies is very limited. This number may increase in the future as the Asian bond market
becomes larger, but the rating standards of global agencies, which are designed mainly for
developed countries, may not be adequate to differentiate the credit ratings of the Asian
bonds that are highly concentrated in low credit ratings.
Table 8 shows the current sovereign credit ratings for foreign currency debt assigned by a
global credit rating agency. Major developed countries in North America and Europe have

216 BIS Papers No 30

ratings of AAA, other European countries and Japan have AA ratings, and emerging
economies have ratings varying from A to B. Because there is little possibility of corporate
bonds being rated above the sovereign, most Asian bonds will likely be rated lower than
BBB. Therefore, unless the global rating system’s lower credit ratings are broken down into
more specific levels, the ratings of Asian bonds will have a very concentrated distribution,
providing limited if any differentiation.

Table 8
Sovereign foreign currency credit ratings
of Standard & Poor’s, June 2004

Credit rating Country

AAA Australia, Austria, Canada, Denmark, Finland, France, Germany, Ireland, Isle of
Man, Liechtenstein, Luxembourg, Netherlands, Norway, Singapore, Sweden,
Switzerland, United Kingdom, United States
AA+ Belgium, New Zealand, Spain
AA Bermuda, Italy, Portugal
AA– Andorra, Japan, Slovenia, Taiwan
A+ Hellenic Republic, Hong Kong SAR, Iceland, Kuwait, Qatar,
A Botswana, Chile, Cyprus, Malta, Saudi Arabia
A– The Bahamas, Bahrain, Barbados, Czech Republic, Estonia, Hungary, Korea,
Lithuania, Malaysia
BBB+ China, Latvia, Poland, Slovak Republic, Trinidad & Tobago
BBB Oman, South Africa, Thailand, Tunisia

BBB– Bulgaria, Croatia, Kazakhstan, Mexico

BB+ Egypt, El Salvador, Russia
Colombia, Costa Rica, India, Jordan, Morocco, Panama, Peru, Philippines,
BB
Romania
BB– Grenada, Guatemala, Vietnam
B+ Belize, Benin, Brazil, Cook Islands, Ghana, Senegal, Turkey
B Burkina Faso, Cameroon, Indonesia, Jamaica, Mali, Mongolia, Pakistan, Papua
New Guinea, Ukraine
B– Bolivia, Lebanon, Suriname, Uruguay, Venezuela
CC Dominican Republic
CCC+ Ecuador
D Argentina, Paraguay

Source: S&P homepage, www.standardandpoors.com.

In order to solve this concentration problem, the global agencies could adopt a regional
rating system separate from the global rating system. However, it is doubtful whether they
will develop such a new system for a market that is relatively small. It might create the
unwanted impression that the agencies are adopting a double standard. In addition, the
credibility and accuracy of credit ratings depend on a detailed awareness of local knowledge
along with scientific methodology. If global rating agencies plan to expand into rating local

BIS Papers No 30 217

bonds, they must establish a database and build up local human resources, which would
entail significant infrastructure building costs. Because of these limitations, direct involvement
of the global agencies in the local rating business has not been extended beyond the
collaboration stage with local agents except in Japan.
Another problem with utilising the global rating agencies is that they are often criticised for
their lack of impartiality. In 2003, S&P lowered the ratings of major German companies.
Germany protested the decision and thereafter started to discuss the possibility of
establishing its own domestic credit rating agency. 18 Ferri and Liu (1999) also argue that
companies in developing countries are more conservatively rated than companies in
developed countries. Together with the criticism that the global credit rating agencies lack
awareness of each country’s specific economic situation, this argument is reinforced by the
fact that there are many black boxes in their credit rating methodology.
An alternative to utilising global rating agencies in adopting a common credit rating system is
for local agencies to jointly establish a regional credit rating agency. If the New Basel Accord,
which mandates that the credit risk adjusted capital adequacy is to be adopted in the future,
and the financial supervisory institutions of each country were to encourage domestic
financial institutions to use the credit ratings of this new institution, the regional agency would
have a good business outlook, as well.
However, there are serious drawbacks in establishing a new regional credit rating agency
through the sponsorship of each country’s government. As is the case with the global rating
agencies, the regional rating agency must build up the infrastructure, including a database
and human resources, if it is to rate local bonds in each country, and the costs of building this
infrastructure would not be negligible. In addition, the agency must build a reputation for
impartiality and accuracy to compete with global agencies. But it is doubtful whether a
government-supported agency, which would not be exposed to competition, would be
acknowledged by the market as impartial and efficiently run. Also, if a regional agency were
established through government support, it would definitely crowd out the private business of
local credit rating agencies.
As seen in Section 2, the establishment of a regional clearing and settlement system with
government support is justifiable in that it addresses problems associated with market failure;
it would increase efficiency by concentrating transaction volumes and alleviate the third time
zone problem. However, establishing a regional credit rating agency is a different problem,
even if the need for a new common credit rating system in Asia, which can effectively
differentiate Asian bonds concentrated in low credit ratings, is accepted. Considering the
drawback of establishing a new regional agency, it would be better for the governments to
promote cooperation among local and global rating agencies in establishing a common rating
system and meeting the new business needs. The ACRAA has already started standardising
the rating systems in anticipation of the development of the Asian bond market.

3.3 Standardisation for introducing a regional credit rating system

The bond markets of Asian countries differ in size and degree of development, as do their
credit rating systems and agencies. Clearly, adopting common credit rating standards

18
In February 2003, when S&P lowered the rating of three major companies in Germany (Thyssen Krupp, Linde
and Deutsche Post), Europe questioned the fairness of the decision. In August 2003, even though the capital
of Munich Re, an international reinsurance company based in Germany, had been raised, its credit ratings
were lowered from AA– to A+. Germany questioned the credit rating capabilities of the global credit rating
agencies, which led to discussion about establishing a new credit rating agency. Gerke and Pellens (2003)
argue that the global agencies failed to reflect the difference in pension reserve methods in Germany.

218 BIS Papers No 30

through mutual cooperation among these agencies will not be an easy task. Despite the
difficulties, the ACRAA recently announced a plan for mutual cooperation. There is an
ongoing effort to improve the expertise of analysts of member agencies through mutual
education programs as well as exchanges on rating standards. Furthermore, a best practices
committee was set up to determine the common standards for credit ratings, and members
are actively working on standardising credit rating concepts and sharing credit rating
methodologies.
The process of standardising the credit rating systems in Asia that the ACRAA is promoting
will occur in several stages. The most basic stage involves sharing basic rating concepts and
offering mutual training through which the member agencies share rating definitions, default
concepts and other basic credit rating concepts.
In a more advanced stage of standardisation, agencies could share rating methodologies
and a joint committee for a common rating system could be considered. At that stage, if
needed, a regional credit rating agency could be established to rate offshore bonds issued in
Asia. Instead of governments taking the initiative and financing it, such an agency would
draw on the voluntary participation of local credit rating agencies from each country. This
would prevent disagreements between the regional and local agencies because the roles of
each could be determined beforehand. Moreover, a great deal of overlap in infrastructure
investment could be avoided by building an information hub that contained each country’s
database.
It is expected that actually establishing a common credit rating system for Asia or a regional
credit agency will require a considerable amount of time. It is imperative in the meantime that
the local credit rating agencies cooperate with each other in order to develop the Asian bond
market. For example, if Indonesian bonds were to be pooled together and issued as Korean
won-denominated CDOs in Korea, a structure should be set up such that the Korean and
Indonesian credit rating agencies could trust the quality of the each other’s ratings, even if
there is no regional credit rating system. In other words, there must be a certain degree of
standardisation between the two countries’ credit ratings. The current effort of the Asian
governments to develop the Asian bond markets is a valuable opportunity to promote joint
cooperation among the local credit rating agencies.

4. Conclusion
This paper discussed the issues involved in a building infrastructure for Asian bond markets,
namely establishing a regional security settlement system and credit rating agencies. As for
a clearing and settlement institution for the Asian bond market, we propose establishing a
regional ICSD dubbed AsiaSettle by linking the central banks and NCSDs of each country. At
the initial stage, AsiaSettle would perform as the clearing and settlement system for local
currency-denominated government bonds of Asian countries. The focus in the early stages
on government bonds is extremely important; because the supply of high-quality bonds in the
private sector is low, high-quality government bonds would be an indispensable catalyst for
the development of the Asian bond market. We also discussed the necessity that AsiaSettle
also function as the central counterparty for the exchange of government bonds and possess
Electronic Communication Networks (ECN) platform capabilities. Furthermore, we discussed
the desirable governance structure of AsiaSettle and proposed that AsiaSettle be established
as an institution owned by each country’s NCSD and central bank, or as a new multilateral
agency for Asia.
As for a regional credit rating system, there is a great need for a common credit rating
system amongst the Asian countries to develop the Asian bond market. However, unlike the
clearing and settlement system, it is not recommended that the regional credit rating agency
be established through government support. This recommendation recognises the high costs

BIS Papers No 30 219

of building a centralised agency equipped with an extensive database and specialised local
human resources to handle the credit rating of local bonds. Moreover, it is questionable
whether an agency established through government support could compete with private
agencies in acquiring and retaining a reputation for impartiality. Therefore, this paper opts for
harmonisation, in which local credit rating agencies and global credit rating agencies
coordinate in building a common credit rating system. The ACRAA, an organisation of Asia’s
credit rating agencies, is currently undertaking such a harmonisation process.
Whether a credit rating or a settlement agency, the argument for building a regional
institution is mistakenly seen as one of market protection. However, building regional
infrastructure must be seen not as an attempt at protectionism, but rather as a catalyst for
opening the underdeveloped Asian bond markets and removing local restrictions on
developing an international bond market in Asia.
There is no denying that the best way to begin developing the Asian bond market is to
develop each country’s domestic bond market and open it up to foreign investors. In other
words, the optimal method of developing cross-border trading in Asia is for Asian countries to
open up their domestic bond markets to enable Asian issuers to issue bonds in any country
of their choice and to enable investors to invest in bonds in the domestic market of any
country. However, the bond markets of East Asian countries are at greatly varying stages of
development. Some are much more liberalised than others, and different kinds of capital
controls are imposed. Some Asian countries do not even have the economies of scale to
support all the components of the bond market infrastructure, such as a settlement and
depository system, a primary dealer system, credit rating agencies, bond pricing agencies
and credit guarantee agencies, which are needed to develop domestic bond markets. It is,
therefore, very unrealistic to expect every Asian country to develop and open up its domestic
bond markets in the near future, unless there is political pressure to develop the Asian bond
market.
The failure to develop the Asian bond market in the 1990s is a good example. To many, the
recent discussion on Asian bond markets seems to be a repetition of the old bond market
idea from the early 1990s. The launching of the dragon bond initiatives in the early 1990s
sparked discussion in Asia on the development of the Asian bond market which continued
through the end of the decade.
In retrospect, however, the Asian bond market initiatives of the 1990s were merely talk
without action. Their proponents failed to establish a consensus on their benefits. There was
skepticism about the growth potential of the Asian bond market due to the reluctance of
Asian countries to liberalise and open up their domestic capital markets for fear of creating
market distortions and making themselves vulnerable to speculative attacks. The skeptics
also did not believe that the Asian bond market would attract much attention because there
were already well established, efficient international bond markets such as the eurobond
market.
The situation changed greatly during the Asian financial crisis, and we are finally seeing
some meaningful action towards establishing Asian bond markets at least on the government
level. A consensus among Asian economies has emerged that regional bond markets should
be promoted in order to facilitate the recycling of regional savings and to prevent the
recurrence of financial crises as explained in the introduction. For these reasons, Asian
countries finally agree on the importance of developing the Asian bond market, and this is a
great opportunity for each country to ease government regulations and open domestic
markets to international investors. The attempt to build regional infrastructure for the Asian
bond markets should not be interpreted as an effort to close off and protect the Asian market,
but rather as an opportunity to open and develop it through harmonisation of regulations,
policy coordination and improvements in legal structures.

220 BIS Papers No 30

References
Bank for International Settlements (1995): “Cross-border securities settlements”, March.
——— (1999): “A new capital adequacy framework”, June.
Depository Trust & Clearing Corporation (2000): “Central counterparties: development,
cooperation and consolidation”, October.
Euroclear (1999): “The hub and spokes clearance and settlement model”, May.
——— (2002): “Quick cash card”, June.
——— (2003): “Quick guide to the Euroclear system”, February.
Ferri, Giovanni and Li-Gang Liu (1999): “How do global credit rating agencies rate firms from
developing countries?”, mimeograph, the ADB Institute and University of Bari.
Gerke, Wolfgang and Bernhard Pellens (2003): “Pension provisions, pension funds and the
ratings of companies: a critical analysis”, Thyssen Krupp AG.
Giovannini Group (2001):”Cross-border clearing and settlement arrangement in the
European Union”, November.
Group of Thirty (1988): “Clearance and settlement systems in the world’s financial markets”,
special report.
Hong Kong Monetary Authority (1997a): “Financial technology infrastructure for Hong Kong”,
December.
——— (1997b): Hong Kong Monetary Authority annual report 1997.
——— (1997c): “HKMA introduces securities lending programme and linkage with the
securities depositories in Australia”, press release, October.
——— (1998): “Bilateral linkage between the central securities depositories in Hong Kong
and New Zealand”, press release, April.
——— (1999): “Hong Kong and South Korea signed agreements for a reciprocal bilateral
linkage between HKMA and KSD”, press release, September.
——— (2002): “HKMA and China Government Securities Depositary Trust & Clearing agree
to establish central securities depositary link”, press release, January.
International Securities Services Association (2002): “Compliance to the ISSA
recommendations 2000”, April.
Oh, Gyutaeg, Daekeun Park, Jae-Ha Park and Doo Yong Yang (2003a): “How to mobilize
the Asian savings within the region: securitisation and credit enhancement for the
development of East Asia’s bond maket,” KIEP Working Paper 03-02.
Oh, Gyutaeg, Daekeun Park, Jae-Ha Park and Changyong Rhee (2003b): “Building a
settlement infrastructure for the Asian bond markets: AsiaSettle”, paper presented at
“Workshop on East Asia’s strategy for regional and global financial cooperation” (hosted by
Asia Pacific School of Economics and Government, Australian National University and
Northeast Asian Institute of Business and Economic Research, Korea University), on
11 October 2003.
Oh, Gyutaeg and Jae-Ha Park (2003): “Fostering an Asian bond market using securitisation
and credit guarantee”, paper presented at the ASEAN+3 Informal Session of Finance
Ministers and Central Bank Deputies on “Fostering Asian bond markets”.
Park, Yeong-Suk and Jeong-Hoon Hong (2001): “An establishment proposition for cross-
border settlement system which is suitable to the globalization or unification of securities
markets”, Korean Securities Association.
Park, Yung-Chul and Daekeun Park (2003): “Creating regional bond markets in East Asia:
rationale and strategy,” paper presented at Second Finance Forum of Pacific Economic
Cooperation Council, Hua Hin, Thailand on 7-10 July.

BIS Papers No 30 221

 Comments on Daekeun Park and Changyong Rhee’s
paper “Building infrastructure for the Asian bond
markets: settlement and credit rating”

Tom Byrne

I shall limit my remarks to Section III of Park and Rhee’s paper.

The authors conclude by arguing that “harmonisation” of local and global credit rating
agencies is necessary to develop an Asian bond market. If what the authors mean is
constraining competition between and the independence of credit rating agencies, I disagree.
While the authors reject, rightfully, going down the route of government-supported rating
agencies, they do not adequately appreciate how global agencies function or their role in the
market. From an Asian-centric point of view, the authors’ basic gripe is that global credit
rating agencies are ignorant of local conditions, implying that they currently lack and will
never have the capability to deliver accurate and objective rating opinions on local currency
bond issuers. Their argument is convoluted, however, and does not appreciate that factors
that affect credit fundamentals tend to be universal. The authors should have stuck more
closely to the solution they cite in the opening paragraph of the subsection “The need for a
regional credit rating agency”. That is, “[t]he simplest means to meet this need would be to
utilise global credit rating agencies, rather than establish a new regional institution”. To which
I would add, use global agencies along with competitive local agencies.
The fact is that global rating agencies are already increasingly expanding ratings in local
markets, in Asia and elsewhere, employing local staff along with experienced staff from their
home offices. Expansion includes assigning ratings on an agency’s global scale, and also
rating local corporations on country-specific rating scales. Moody’s has ownership stakes in
local rating agency affiliates that use indigenous rating scales. Moody’s also uses national
scale ratings where it does not have affiliates so as to fit into local credit rating systems. A
national scale does not necessarily convey the same information as Moody’s global scale
rating symbols - the probability and expected severity of default. Rather, national rating
scales are essentially ordinal, and issuers are notched down from the best issuer (Aaa by
definition), providing a relative ranking of creditworthiness. Local rating scales are not as
powerful as a rating agency’s global scale. One important reason for such local scales is that
capital market regulators in many emerging market countries require that a corporation
receive an investment grade rating in order to issue. Therefore, regulators themselves have
historically limited the universe of ratings. Thus, the national authorities themselves restrict
the downward range of ratings on the part of either a local or a global rating agency, and
foster a concentrated distribution of relatively high ratings.
Another mistaken notion of the authors is that global rating scales are not suitable for
emerging markets. The authors claim that because the sovereign ratings for emerging
market sovereigns are lower than in advanced economies (as laid out in Table 8), emerging
market corporations are destined to have ratings concentrated in lower rating levels. This is
not necessarily so, but to the extent that it is so, it reflects an assessment of credit
fundamentals, not rating methodology biases. The fact is that the sovereign defaults in recent
history (starting with Russia in 1998) occurred with non-investment grade emerging market
government bonds (at the time), not investment grade rated government bonds. So the
global rating scale functions as intended. The table lists foreign currency government ratings,
but Moody’s rating practice allows in select cases a corporation to pierce the foreign currency
rating of the government. This happens mostly in countries that are assigned low ratings.
Moreover, Moody’s does not constrain the local currency ratings of fundamentally strong

222 BIS Papers No 30

corporations by the local currency rating of the government (because the possibility of the
imposition of foreign exchange controls is moot for local currency obligations).
Credit rating agencies in the United States developed in an environment of competition. I do
not think that the authors’ stress on “harmonisation” and coordination “in building a common
credit rating system” captures the essence of how rating agencies have contributed to the
efficiency of markets. The credit rating industry has its roots in the Anglo-Saxon tradition of
free speech and constitutional guarantees of freedom of the press. Credit rating agencies
exist in a publishing culture that values objectivity, accuracy and attentiveness to investor
needs. A rating agency sells its judgments in a competitive marketplace that rewards
credibility, and an agency accepts the reality that market forces will punish bad performance.
Recently, in the United States, the number of rating agencies recognised by the market
regulator increased to four from three, in an effort to heighten competition and accuracy, but
not to foster harmonisation. A danger in the authors’ prescription is that rating agencies might
be appropriated either implicitly or explicitly by governmental authorities. In this event,
companies would be buying a licence to issue bonds, rather than buying the credibility of a
credit rating agency’s opinion. Investor confidence would be sacrificed, along with efficiency
in allocating capital and pricing risk into credit decisions. Rather, the authors should urge an
improvement in accounting and disclosure practices of issuers in Asian bond markets - just
like efforts under way in the US capital market at present.

BIS Papers No 30 223

 Creation of a regional credit
guarantee mechanism in Asia

Gyutaeg Oh and Jae-Ha Park

1. Introduction

The development of the bond market has become one of the most significant policy goals in
Asia. It is widely considered an important step towards preventing another financial crisis.
There is a region-wide realisation that heavy dependence on bank-intermediated financing,
especially on foreign currency short-term financing, was one of the main causes of the Asian
financial crisis. Asia’s dependence on bank-dominated financial systems supported rapid
economic growth, but left the corporate sector over-reliant on short-term bank loans, which
made the financial system - and entire economies - vulnerable to external shocks. The
development of local currency bond markets has, therefore, become one of the important
policy initiatives in Asia to prevent another financial crisis in the region.
The importance of developing the region’s bond markets has also increased recently as the
need to recycle the vast amounts of accumulated official foreign exchange reserves directly
into the region has risen. Since the financial crisis, Asian countries have accumulated
substantial foreign exchange reserves, partly as a result of huge current account surpluses
reflecting high personal savings and subdued investment demand. This increase in foreign
exchange reserves was, initially at least, an intentional policy in response to Asian countries’
realisation that a lack of foreign currency liquidity caused the crisis. Unfortunately, however,
Asia could not benefit much from these reserves since most of them have been invested in
developed markets such as the United States and Europe. Capital flows from developing
economies where investment returns are higher than in mature economies. The reserves are
recycled back into the region in the form of risky assets such as equities and foreign direct
investment. There are thus huge missed opportunities for capital market development in
Asia. Until the Asian bond markets are fully established, East Asian borrowers will have to
turn to the international financial markets. In order to facilitate the recycling of regional
savings and to prevent the recurrence of a financial crisis, both Asian policymakers and
economists concur that sound and liquid bond markets must be developed.
With this recognition Asian countries have stepped up their collaborative efforts to develop
and strengthen the bond markets in the region through diverse forums such as ASEAN+3,
APEC, EMEAP and ACD. Since late 2002, experts and policymakers in Asian countries have
exchanged their views and ideas on various issues relating to the development of bond
markets. One of the most important policy issues that have been intensively discussed by
policymakers and markets experts is how to develop and strengthen the credit guarantee
mechanism in the region. Development of a credit guarantee market is thought to be critical
for the development of the regional bond market, as one of the most critical factors hindering
the development of the regional bond markets is the credit quality gap between the low credit
ratings of issuers and the minimum credit requirements of investors.
This paper is organised as follows. The next section reviews the recent progress of
discussions on the development of the Asian bond markets. In particular, we discuss the
efforts by ASEAN+3, APEC and EMEAP. Section 3 provides an overview and discusses the
characteristics of the Asian bond markets. In addition, we explain why progress in the
development of the bond markets in Asia has been limited. Section 4 explains the rationale
and background for creating a new regional credit guarantee mechanism, focusing on the

224 BIS Papers No 30

credit quality gap problem and the limits of the existing guarantee institutions in meeting the
need for guarantees in Asia. Section 5 presents possible options for a new guarantee
institution including attributes, business strategy and institutional forms. Section 6 concludes
with a discussion of policy implications.

2. Progress of discussions on the development of Asian bond

markets

A. ASEAN+3: Asian Bond Market Initiative (ABMI)

An informal meeting of the ASEAN+3 Finance Ministry Deputies and Central Bank Deputies
(AFDM+3) was held in Tokyo in November 2002 to discuss specific ways to develop the
Asian bond markets under the ASEAN+3 framework. A month later, in Chiang Mai, Thailand,
a comprehensive plan for the development of the regional bond market, the Asian Bond
Market Initiative (ABMI), was endorsed.
On 28 February 2003, an informal session on “Fostering Bond Markets in Asia” was held by
AFDM+3 in Tokyo, Japan. Various proposals were presented by member countries to
contribute to the development of the bond markets in the region. The delegates agreed to
further study those proposals in depth to achieve tangible results as soon as possible.
Reflecting the proposals and opinions of the delegates, six working groups of volunteers
were established to conduct detailed studies on various aspects of bond market development
(see Table 1).

Table 1
The six working groups of the ABMI

Working group Chair country

1. Creating new securitised debt instruments Thailand

2. Credit guarantee and investment Korea and China
mechanisms
3. Foreign exchange transaction and
settlement issues Malaysia
4. Issuance of bonds denominated in local
currencies by multilateral development banks
(MDBs), government agencies and Asian
multinational corporations China
5. Local and regional rating agencies Singapore and Japan
6. Technical assistance coordination Indonesia, the Philippines and Malaysia

The working groups analyse two areas: (i) facilitating market access through a wide variety of
issuance and (ii) creating an environment conducive to developing bond markets. The issues
related to market access include: (i) bond issuance by Asian governments to establish
benchmarks, (ii) bond issuance by Asian governments’ financial institutions (governments) to
finance domestic private enterprises, (iii) creation of asset-backed securities markets,
including collateralised debt obligations (CDOs), (iv) bond issuance by multilateral financial
institutions and government agencies, (v) bond issuance for funding foreign direct investment
in Asian countries and (vi) issuance of bonds in a wider range of currencies and introduction
of currency basket bonds. The issues concerning the creation of an environment conducive

BIS Papers No 30 225

to active participation by both issuers and investors are: (i) provision of credit guarantees,
(ii) improvement of the credit rating system, (iii) establishment of a mechanism for
disseminating information, (iv) improvement of the settlement system and (v) strengthening
of the legal and institutional infrastructure for bond market development.
The Working Group on Credit Guarantee and Investment Mechanisms chaired by Korea and
China focuses on ways to promote the use of credit guarantee mechanisms in Asia. Many
government delegates and market experts agreed on the need to study this issue in a more
comprehensive and systematic way to produce a practically workable proposal. Also, the
Asian Development Bank (ADB) decided to contribute to the creation of an appropriate
regional credit guarantee mechanism to support the development of the domestic and
regional bond market in ASEAN+3 countries.

B. APEC: securitisation and credit guarantee

The development of the regional bond markets has long been seen by APEC economies as
an important objective in the broader effort to promote greater openness, diversity and
competitiveness in regional financial markets. This position was reaffirmed most recently in
the APEC Finance Ministers’ Joint Ministerial Statement of September 2002 and the APEC
Leaders’ Declaration of October 2002 in Los Cabos, Mexico. The objectives of this initiative
are to identify impediments to the development of securitisation and credit guarantee
markets within the APEC economies and to propose appropriate solutions to remove them.
Securitisation coupled with credit enhancement offers significant benefits for developing
markets to the extent that it helps reconcile credit and liquidity mismatches between issuers
and investors and can also facilitate balance sheet restructuring.
This APEC initiative on the development of securitisation and credit guarantee markets is
highly action-oriented, and it is co-chaired by Hong Kong SAR, Korea and Thailand. This
initiative involves (i) holding policy dialogues for the APEC economies to exchange views on
the use of securitisation and credit guarantees at the national and regional levels and
(ii) sharing experience among APEC economies in identifying impediments and developing
detailed action plans. The first policy dialogue was held in April 2003 in Seoul. So far, nine
APEC member economies have participated in the initiative, either by sponsoring experts or
by seeking expert advice on how to remove impediments in their markets. They are
Australia, China, Hong Kong, Japan, Korea, Mexico, the Philippines, Thailand and the United
States. Through collaboration among expert panels, domestic interdepartmental taskforces
and private sector advisory groups, action plans at the economy level have been drafted to
remove impediments to the development of securitisation and credit guarantee markets in
individual APEC economies.

C. EMEAP: Asian Bond Fund (ABF)

The ABF is the first fund of its kind in the region. The first phase of the initiative, ABF1, was
launched in 2003 and is fully invested in US dollar-denominated bonds in the EMEAP
economies. Since then, EMEAP has been working on the second phase of the project:
broadening the ABF to cover bonds denominated in local currency, or ABF2. Both phases of
the initiative are aimed at promoting the development of the bond market by improving the
domestic and regional bond market infrastructures.

1. ABF 1: the critical first step

The establishment of ABF1 was announced in June 2003. All 11 EMEAP central banks
invested in ABF1 at its launch, which had a capitalisation of about US$ 1 billion. The fund is
now fully invested in US dollar-denominated bonds issued by sovereign and quasi-sovereign
issuers in eight EMEAP economies (China, Hong Kong, Indonesia, Korea, the Philippines,

226 BIS Papers No 30

Malaysia, Singapore and Thailand). The developmental benefit of ABF1 is more than the
first-round demand effect of US$ 1 billion invested by the central banks. Indeed, the seed
money invested by EMEAP central banks serves to attract additional money from the private
sector, thereby deepening and broadening the demand in the markets. The promotional
effect of ABF1 would generate second-round investor and issuer interest in the Asian bond
markets, broadening the investor base and increasing market liquidity over time.
Furthermore, the ABF1 initiative is a milestone in regional central bank cooperation. As
noted, ABF1 is the first of its kind in Asia, and its success is as symbolic as it is material. The
successful launching of ABF1 not only sent a strong message to the financial markets that
the regional authorities are committed to stepping up their cooperative efforts in promoting
bond market development, it also paved the way for the development of ABF2. The
remarkable one-year time frame from the initial discussions to the actual commitment of
funds and the subsequent launching of ABF1 testifies to the rapport and sense of ownership
among EMEAP members, which will prove valuable towards the development of ABF2.
ABF2 will involve many more complex and technical issues than ABF1, and the precedent of
ABF1 should be very helpful in efforts to garner political support and commitment in the
challenges to the development of ABF2.

2. ABF2: the bold new second phase

Building on the momentum of developing ABF1, EMEAP has proceeded to study the
feasibility and design of ABF2. Owing to the complexity of the project and the likelihood of
opening up the funds for private sector investment in the future, the EMEAP Group has
appointed financial advisers from the private sector to advise on the design and structure as
well as the construction of benchmark indices for ABF2.
In April 2004, the EMEAP Group issued a press release setting out the basic design and
latest thinking behind ABF2. It was proposed that ABF2 would consist of two components: a
Pan-Asian Bond Index Fund (PAIF) and a Fund of Bond Funds (FoBF) (Figure 1). While
many issues regarding ABF2, such as fund size and detailed fund structure, have yet to be
determined by EMEAP after having taken into account such factors as market conditions, the
latest thinking on ABF2 is described below.
The preliminary framework of ABF can be described as follows. PAIF is a single bond index
fund investing in local currency-denominated bonds in EMEAP economies. It will act as a
convenient and cost-effective investment fund and new asset class for regional and
international investors who wish to have a well diversified exposure to bond markets in Asia.
The FoBF is a two-tier structure with a parent fund investing in a number of country sub-
funds comprising local currency-denominated bonds issued in the respective EMEAP
economies. While the parent fund is confined to EMEAP investment, the country sub-funds
are intended to provide local investors with low-cost and index-driven investment vehicles
and at the same time give regional and international investors the flexibility to invest in the
Asian bond markets of their choice.
The ABF2 funds are intended to be passively managed against a set of transparent and
predetermined benchmark indices, covering local currency bonds issued by sovereign and
quasi-sovereign issuers in EMEAP economies. ABF2 is being designed in such a way that it
will facilitate investment by other public and private sector investors. In addition to attracting
additional money into the bond market, as in the case of ABF1, ABF2 seeks to achieve a
larger and longer-lasting positive impact on regional bond market development. Several
features of the design of ABF2 are conducive thereto.

BIS Papers No 30 227

 Figure 1
ABF2 framework

EMEAP
central banks

Fund of Bond Funds

(FoBF)
Pan-Asian Bond Index Fund
(PAIF)
Parent fund

Country Country Country

sub-fund sub-fund sub-fund

Underlying bonds

ABF2 is likely to have a long-lasting impact on market development in addition to its effect on
demand from the seed money invested by EMEAP. In the process of the development of
ABF2, individual EMEAP economies can leverage the interest and momentum generated
from the collective investment in ABF2 to further develop their domestic bond markets as
appropriate. For instance, the appropriate EMEAP members can work with the relevant
authorities to improve market infrastructure by identifying and minimising the legal, regulatory
and tax hurdles in their markets, thereby facilitating the creation of fixed income products in
the region. The momentum and political impetus generated from the development of the fund
could perhaps help harmonise regulatory procedures, eg cross-registration of bond funds,
and the cross-issuance and trading of bonds, and help address the issue of fragmentation of
regional bond markets.
ABF2 will also improve bond market infrastructure by encouraging the development of
transparent, replicable and credible bond market indices for use in the PAIF and the FoBF
country sub-funds. There is currently a lack of low-cost, passively managed index bond funds
in most EMEAP economies, and most of the existing funds available in the markets are
actively managed. Furthermore, the performance yardsticks for these funds vary
substantially. It is hoped that the ABF2 bond indices would be widely adopted by private
sector fund managers as benchmark indices for their fixed income products. For instance,
private sector participants could clone or customise (by means of adding in corporate issues,
etc) these transparent benchmarks and facilitate the setup of low-cost index bond funds.

228 BIS Papers No 30

Derivative products might also be created based on such indices, which should greatly
increase market liquidity.
At the same time, ABF2 would also seek to encourage the development of exchange-traded
bond funds (ETFs) in the region. ETFs are new to Asia, but have become increasingly
popular and are a fast-growing fund category in the United States and Canada. The
development of ETFs will help promote product diversification for both institutional and retail
investors, increase market liquidity, provide low transaction cost bond fund products and
make the price setting process more transparent. Thus, the seed money invested by EMEAP
would also give impetus to product innovation that might otherwise not occur.

3. Asian bond markets: overview and factors hindering their

development

A. Overview of the Asian bond markets

The financial structure of most of the Asian countries is dominated by banks. Usually, the
size of outstanding bank loans is much larger than that of outstanding bonds. Prior to the
Asian financial crisis, Asian countries pursued prudent fiscal policies with balanced budgets.
As a result, the government bonds outstanding were very limited. In addition, corporate bond
markets were not developed in most of the countries because of the limited supply of bonds
and underdeveloped infrastructure. However, since the 1997-98 financial crisis, government
bond issues have increased sharply to finance budget deficits resulting from expansionary
fiscal policy, bank recapitalisation and rising social safety net expenditures. Corporate bond
issuance has also increased in most of the countries. In some countries like Korea, corporate
bond issues increased dramatically, mostly led by rapid growth in the asset-backed securities
(ABS) market.
Asian bond markets accounted for 24% of the world bond market in 2001 with US$ 6.8 trillion
in capitalisation. In recent years, Asian countries have sought to improve their bond markets,
yet much remains to be done. In fact, the Asian bond market is in its infancy when compared
to the US bond market with US$ 15 trillion capitalisation. Table 2 shows that the ratios of the
bond markets as percentages of GDP in most Asian countries are less than 70%, except for
Japan. The Japanese bond market accounts for about 85% of the Asian bond market,
illustrating the fact that the Asian bond market is still in its initial development stage. Table 2
also shows a high percentage of government bond issuance in Japan, China and Singapore,
of corporate bond issuance in Korea, Taiwan (China) and Malaysia, and of financial
institutions’ bond issuance in Hong Kong.
East Asian countries’ bond issuance from 1997 to 2001 is shown in Table 3. The issuance of
bonds generally increased during these years, but except for Korea and China, the size of
the bond issue is still very small. In addition, the proportions of Asian stock markets to GDP
are relatively higher than those of the bond markets. For instance, the capitalisation of the
Hong Kong stock market is three times the size of GDP, while its bond market is only 38.1%
of GDP. Indonesia’s bond market remains at only 5.1% of GDP, and that of China is only
18.0%.

BIS Papers No 30 229

 Table 2
Asia’s bond market capitalisation, 2001
In billions of US dollars and percentages

Financial
Total Governmen Corporate
institution
bonds t bonds bonds
s’ bonds (B)/(A) (A)/GDP
outstandin outstandin outstandin
outstandin
g (A) g (B) g (D)
g (C)

Philippines 21.6 20.5 – 1.1 94.9 30.3

Thailand 36.2 18.4 12.1 5.7 50.8 31.5
Hong Kong 43.6 14.6 24.0 5.0 33.5 26.6
SAR
Singapore 52.2 29.0 17.5 5.7 55.5 61.0
Malaysia 82.8 32.0 8.4 42.4 38.6 94.0
Taiwan, China 124.3 54.5 15.1 54.6 43.8 44.0
Korea 292.7 77.3 97.9 117.5 26.4 69.3
China 403.0 201.3 191.3 10.4 50.0 34.8
Japan 5,816.9 3,904.7 1,211.7 700.5 67.1 139.4
Total 6,873.3 4,352.1 1,578.0 942.9 63.3 104.8
United States 15,366.6 4,271.9 8,658.2 2,436.4 27.8 152.4

Sources: World Federation Exchange (2001); FIS; BIS.

Table 3
Bond issuance of East Asian countries
In billions of US dollars

1997 1998 1999 2000 2001

China 10.27 14.39 19.53 25.34 20.83

Hong Kong 3.52 5.03 5.71 6.09 6.33
SAR
Indonesia 0.78 0.57 1.01 0.98 0.75
Korea 23.55 23.83 30.64 37.54 38.14
Malaysia 4.04 3.60 4.15 6.39 7.25
The 2.04 1.58 2.50 2.42 2.45
Philippines
Singapore 2.69 2.90 3.68 4.48 5.68
Thailand 1.74 2.28 3.68 4.08 4.27
East Asia 48.62 54.17 70.90 87.31 85.69

Source: Ismail Dalla, “Asset-backed securities market in selected East Asian countries”, World Bank, 2002.

230 BIS Papers No 30

B. Factors hindering development of an Asian bond market
As mentioned above, capital movement within the East Asian region is limited. Even with
substantial foreign exchange reserves, current account surpluses, a surge in exports and
high personal savings, not enough of the surplus capital has been circulated within the
region. Important insights into this situation can be gleaned by examining the supply of, and
demand for, capital in East Asia. On the supply side, low-rated assets in the region and
capital controls and regulations have restricted foreign investment in domestic assets. On the
demand side, there is a lack of expertise among institutional investors and risk-averse
behaviour among regional investors.
A wide credit quality gap exists between the issuers and the minimum requirement of the
investors as the credit ratings of many East Asian countries are below investment grade,
which on the whole discourages international investors. In addition, there are only a limited
number of large, reputable firms which can issue high-rated bonds. The low credit ratings of
bonds issued by East Asian governments or corporations have been one of the major factors
hindering the development of the bond market in the region. The low credit ratings are mainly
due to widely perceived political and commercial risks in East Asian countries. The low credit
ratings of major East Asian countries make it difficult for international investors to invest
without constraints. The bulk of Japan’s overseas capital investment, for example, is in non-
Asian bonds because the low credit rating of major East Asian countries means that Japan
has few options for portfolio investment in East Asia.
East Asia’s capital controls and regulations have been relaxed gradually but they remain a
major obstacle impeding capital movement within the region. Capital controls and regulations
create distortions in international capital flows. They often take the form of restrictions on
foreign financial institutions entering the domestic financial market or a cap on foreign equity
ownership in domestic financial institutions. Such restrictions have long since been removed
in advanced economies such as Japan, Korea, Hong Kong and Singapore, but many East
Asian countries still control capital flows in many ways. Some notable features of capital
controls are: (i) more restrictions on capital outflows than on capital inflows and (ii) relatively
stronger control on capital inflows for bond investment than on capital inflows for equity
investment.
On the demand side, institutional investors in East Asia are largely underdeveloped. Most of
the pension funds, mutual funds and insurance companies are small and incapable of
expanding their cross-border portfolios. Institutional investors are in their early stage of
development in most East Asian countries. Four factors can account for the weak institutional
investor base. First, bank-dominated financial intermediation in most of East Asia hinders the
development of institutional investors, with the extensive branch networks of banks tapping
high domestic savings. Second, corporate governance of family-controlled companies and
the emphasis on the expansion of capital through business profits or bank loans tend to
discourage the growth of institutional investors. Third, the absence of a long-term capital
market and lack of long-term investment products complicate portfolio management by
institutional investors. Fourth, the development of the institutional investor base is impeded
by government legislation or decrees, such as those restricting pension funds or imposing
rigid investment criteria on insurance companies.
Risk-averse behaviour by regional investors is another factor limiting capital movement within
East Asia, and was reinforced by the financial crisis. Both public and private investors from
Japan, China, Korea, Hong Kong and Singapore who are able to invest abroad have been
reluctant to take risks as they consider East Asian investment to be riskier since the financial
crisis. Also, East Asia shows a tendency to export risky assets and import safe assets. This
is because the increase in accumulated foreign exchange reserves is creating a situation that
structurally forces East Asian countries to manage their assets safely. East Asia’s relatively
weaker capability to evaluate and manage risks also discourages East Asian investors from
taking risks.

BIS Papers No 30 231

4. Rationale for creating a new regional credit guarantee mechanism

A. Credit guarantees and securitisation as a solution to the credit quality gap

Securitisation coupled with credit guarantees has been suggested by many experts and
economists as one means of narrowing the credit quality gap between the low credit ratings
of issuers and the minimum credit requirements of investors. A credit guarantee mechanism
is an agency or institution that guarantees bonds by holding a certain sum of money at all
times to immediately pay investors in the event that an issuer defaults. Two types of
guarantees provided by existing guarantee institutions are commercial and political risk
guarantees. Commercial guarantees protect against defaults caused by a company’s
bankruptcy. Political guarantees protect against defaults due to currency inconvertibility,
expropriation, war and social disorder or breach of contract by the government.
A credit guarantor is able to close the gap between the low credit ratings of issuers and the
minimum requirement of investors by enhancing issuers’ credit. The guarantor is in a way
“renting” its high credit rating to the issuer to match the minimum requirement of the
investors. Hence, securitisation supported by credit guarantees enables borrowers to issue
asset-backed securities at much higher credit ratings than they could on their own.
Combined with credit guarantees, these securities could attain a credit quality acceptable to
asset managers by qualifying for investment grade ratings from the international rating
agencies.
Most bond-issuing Asian firms have poor credit ratings. There are a limited number of large,
reputable firms, but they have been migrating to the global bond markets. Some small and
medium-sized enterprises (SMEs) have good credit records but are not capable of raising
capital from either the local or regional bond markets. Moreover, many corporations in
developing countries are rated low because their sovereign ratings are low. These
corporations also have low credit ratings due to the poor quality of institutions and
information disclosure. The supply of top-grade Asian bonds is very limited, pointing to the
need to devise a practical and viable mechanism to increase the supply of investment grade
Asian bonds denominated in local currencies.

B. Example using credit guarantees and securitisation

The figure below illustrates one example of how credit guarantees and securitisation can be
used to facilitate the financing of developing countries’ SMEs by mitigating the credit quality
gap problem. This example entails implementing a two-tier securitisation process; one in
borrowing countries and the other in capital-abundant countries. A two-tier process is
necessary because of differences in the financial and legal systems among the participating
countries in the region.
As a first step, government financial institutions or agencies in capital-importing countries
would securitise loans or bonds issued by SMEs in local currencies. Then, in a capital-
abundant country, a special purpose company (SPC) could be established to securitise the
underlying assets, which are composed of the senior tranches from the capital-importing
countries. The junior tranches, on the other hand, are assumed by local institutions in the
capital-importing countries, which select the firms eligible for securitisation. Some senior
tranches may be sold to local investors, but the remainder would be transferred to an SPC in
the capital-abundant countries. Additionally, senior tranches could be backed by credit
guarantees either from local credit guarantee agencies or government institutions in the
capital-abundant countries. In this process, the newly created credit guarantee institution
would provide guarantees for senior tranches.
It should be noted that cooperation among participating institutions is critical for the smooth
functioning of this system. The coupon rates of both underlying assets and asset-backed
securities, fees for underwriting and credit guarantee and the portion of senior tranches

232 BIS Papers No 30

compared to respective junior tranches are notable examples of many areas in which
cooperation among participating institutions would be critical. Furthermore, because this
proposal allows for the adoption of securitisation with proper risk-sharing among financial
institutions, investors and guarantee companies, moral hazard is expected to be minimal.

Figure 2
SME financing using credit guarantees and securitisation

Issued in local currency

Issued in A currency

Country A Senior
SPC in Senior
Junior country D
SME loans (capital-
abundant
country) Junior

Country B Senior
Credit guarantee
Junior
SME loans
Regional credit

Guarantee institution
Country C Senior

Junior
SME loans

Figure 3
SME financing with currency swaps

Issued in local currency

Issued in A currency
Swap
Country A
SPC in Senior
Senior country D
SME loans (capital-
Junior abundant
country) Junior
Swap
Country B
Credit guarantee
Senior
SME loans
Junior Regional credit

Swap Guarantee institution

Country C

Senior
SME loans
Junior

BIS Papers No 30 233

A slight modification of the first proposal is illustrated by Figure 3. This modified proposal is
nearly the same as Figure 2, the only difference being that it calls for providing currency
swaps to investors who are not willing to assume currency risk. A government agency (GA)
or other proper institution in country A would provide the SPC with currency swaps and then
hedge the currency risk using back-to-back swaps with swap dealers. This would be done
through the currency swap market, if one exists in the developing country, or with the help of
the developing country’s central bank if there is no swap market.

C. Limits of existing guarantee institutions in meeting the guarantee needs in

Asia
In the near future, the greatest demand for guarantees in Asia is expected to be for local
currency-denominated asset-backed securities and infrastructure revenue bonds guarantees.
Guarantees for non-traditional ABSs such as SMEs, CDOs and non-performing loans
(NPLs), and for mid-market, near investment-grade or unrated issues are in especially high
demand. To meet these demands, currency risk or transfer and convertibility risk, regulatory
and institutional infrastructures and technical assistance need to be taken care of. In this
section, the existing guarantee institutions are examined to determine if they meet the
guarantee needs in Asia.

1. Multilateral financial institutions

There are eight multilateral institutions that offer guarantee services to private and/or public
projects: the Multilateral Investment Guarantee Agency (MIGA), the International Finance
Corporation (IFC), the Asian Development Bank (ADB), the International Bank for
Reconstruction and Development (IBRD), the International Development Association (IDA),
the Inter-American Development Bank (IADB), the European Bank for Reconstruction and
Development (EBRD) and the African Development Bank (AFDB).
The guarantee operations of the existing multilateral institutions are limited. Most of the
institutions’ capital has been used for other types of assets; offering guarantees is not their
main line of business. For instance, the IBRD’s guarantee business accounts for only 1.4%
(US$ 1.58 billion) of its roughly US$ 116 billion in development-related assets, and most of
its resources are used for loans. Among all the existing multilateral institutions, MIGA is the
only one that performs guarantee operations as its main line of business. MIGA, however,
only covers political, not commercial, risks. Hence, it can be concluded that the existing
multilateral institutions do not meet Asia’s demand for specialised guarantee services in an
efficient and focused manner.

2. Private guarantee companies

Like the multilateral agencies, the guarantee operations of private institutions are also limited
in size and business scope. The US monoline insurance companies tend to overprice Asian
risk: they do not split the savings in the issuer’s cost resulting from credit enhancement with
the Asian issuer. According to the Hong Kong Monetary Authority, studies of six structured
deals done by US monolines in Hong Kong in the period 1994-99 show that there were no
clear cost advantages, and most of the savings resulting from credit enhancement of the
issuer were taken by the monolines. According to the insurers, this is because the small and
underdeveloped Asian bond markets prevent US monolines from enjoying economies of
scale. The reason for this is that the credit rating agencies consider the Asian financial
guarantee business to entail greater political and commercial risk than the US financial
guarantee business. Also, the deal closures were delayed because all guarantee decisions
were made at the US headquarters. The issuers, especially first-time issuers, who used the
US monolines’ guarantee services did not use their services to save money but to enhance
marketability and reach a wider pool of investors. Due to the lack of funding efficiency, Asian

234 BIS Papers No 30

issuers tend to rely more on bank loans and senior/subordinate structure than guaranteed
bonds.
Additionally, the US monolines cannot satisfy the demand for guarantee services in Asia
because they do not cover Asia’s needs for local currency-denominated guarantees and for
near-investment bonds and non-traditional ABSs such as SMEs, CDOs and NPLs.
Monolines have conservative underwriting guidelines and do not issue guarantees for these
riskier assets because they need to maintain their AAA credit ratings.

3. Existing Asian guarantee institutions

In Asia, there is a credit quality gap in intraregional intermediation in long-term credit, and the
existing private and multilateral guarantors do not fill this gap. Many of the existing
guarantors lack focus on Asian countries, whose long-term borrowing and lending needs
remain unmet. The unmet needs of long-term lending became even greater due to failure to
invest during the crisis years. Many infrastructure projects were shelved for lack of long-term
financing, and thus infrastructure remains woefully inadequate. However, the potential
demand for well structured, long-term bonds of acceptable risk is rapidly growing, as
witnessed by the ever growing trade surpluses and international reserves of the region.
Most Asian countries, except for Japan, Korea and China, do not have local credit guarantee
institutions. Even the existing ones, however, cannot meet the demand for credit
enhancement and guarantees in the Asian bond markets, mostly because they have no or
only sub-investment ratings themselves.

4. The experience of ASIA Ltd

A regional multilateral guarantee agency called Asian Securitization and Infrastructure
Assurance Pte Ltd (ASIA Ltd) was established in 1995 in an attempt to overcome the Asian
bond markets’ unmet demands for credit enhancement and guarantees. It was the first
regional credit guarantee agency established in Asia (located in Singapore) with more than a
commercial objective - to facilitate the development of fixed income markets in Asia. Its
shareholders included CapMAC Asia Ltd, Apmac Investment Pte Ltd, the Asian Development
Bank (ADB), Employees Provident Fund of Malaysia, American International Assurance Co
Ltd, Kookmin Bank, Netherlands Development Finance Co and Deutsche Investitions- und
Entwicklungsgesellschaft mbH (DEG). ASIA Ltd had total paid-in capital of US$ 150 million at
the time of establishment. Initially, ASIA Ltd also intended to set up local agencies in the
second phase of its development to cater to local currency-denominated bonds, while the
Singapore main office would handle non-local currency- (notably US dollar-) denominated
bonds. However, the Asian financial crisis led to the downgrade of its claims-paying ability
rating in 1998, and currently it is in a run-off mode.
ASIA Ltd applied zero-loss underwriting standards to its operations with a concentration of
business in the credit rating range of BBB+ to BBB–. Rating requirements for ASIA Ltd were
such that a maximum of 25% of its guarantee portfolio in non-investment grade bonds would
still allow it to keep its A rating. After the onset of the financial crisis and particularly the
downgrades of Indonesian and Korean credit risks, the claims-paying ability of ASIA Ltd was
downgraded by Standard & Poor’s from A to BB in January 1998. At this time, the agency
had to stop writing new business. As of end-1999, the company’s US$ 934 million credit
insurance portfolio consisted mostly of sovereign, asset-backed, infrastructure and financial
institution debt obligations throughout Asia. In terms of geographical distribution, as of end-
1999, 24% of ASIA Ltd’s guarantees and assumed reinsurance were outstanding to South
Korea, 11.4% to Malaysia, 10.3% to Indonesia, 9.1% to Hong Kong and 20.9% to OECD
countries (except South Korea). As of the end-March 2000, ASIA Ltd still had outstanding
guarantees totalling US$ 924 million.

BIS Papers No 30 235

Following the downgrade of its claims-paying ability, ASIA Ltd sought to raise additional
capital to restore its rating to A. The shareholders were, however, unable to agree on the
terms of a recapitalisation plan due to dissent over broadening the geographical coverage of
ASIA Ltd. In April 1999, ASIA Ltd contracted out its responsibility for day-to-day operations,
including surveillance of its credit insurance portfolios, to an affiliate of MBIA, and it is now
effectively under the management of MBIA Singapore Pte Ltd. Inclusive of the Reinsurance
Treaty with ERC Frankona Ruckversicherungs Aktiengesellschaft of US$ 100 million, ASIA
Ltd now has total claims-paying resources of approximately US$ 250 million.
In retrospect, the company’s business model had many structural problems which made it
vulnerable to the shocks. First of all, at the time when ASIA Ltd was building its book of
business in 1996-97, its risk management practices appeared to be prudent based on the
assumption that country risk among Asian countries was not highly correlated. However, the
reality turned out to be completely different; ie the crisis swept through most of the East
Asian countries at varying intensities. Second, the institution was prohibited from providing
any direct guarantees to non-Asian economies and certain developed Asia-Pacific
economies (Japan and Australia). The geographical restriction not only significantly reduced
the potential business available to the company but also exacerbated the correlation risk and
concentration problem among Asian countries. Third, in relation to the second problem, the
business deal flow of the company was too small and limited since its business area was
limited to developing countries. It could not follow the detailed local market situations of each
developing country. Therefore, the company was forced to engage in overly risky business.
There should have been some close links or mechanisms for collaboration with local financial
institutions and guarantee agencies. Fourth, the initial capital size was too small to absorb
the big risks inherent in the provision of guarantees to developing Asian countries. What
made the situation worse was that additional callable capital was not injected from
participating shareholders during stress periods, contrary to the original agreement. Fifth, the
initial credit rating, single A, was too low to do business.
With high demand for credit enhancement and guarantees in the Asian bond markets, a new
and more fitting multilateral guarantee institution that can meet the Asian bond markets’
current needs should be built. In light of the experiences of ASIA Ltd, the required level of
capital should be much higher, and the return on capital should be lowered. The
development of a regional credit guarantee institution would involve a stronger commitment
to capital injection from participating countries during stress periods. Thus, the credit
guarantee institution should initially be a “public” but “commercially viable” entity. Moral
hazard problems should be mitigated by instituting an appropriate ownership structure.
Exposure to each economy should be limited through structuring and reinsurance. Cross-
border risks should also be limited through an emphasis on local currency business.

5. Considerations for the creation of a new guarantee institution

A. Attributes of a regional guarantee institution

A new regional guarantee institution clearly needs to be established to satisfy the unmet
demands for Asian guarantee services. The objectives of the institution would include:
(i) development of regional capital markets for stable access to long-term funds to facilitate
private sector and infrastructure development, (ii) promotion of transparent and cash flow-
based lending by facilitating securitisation and bonds issues, (iii) catalysing investment by
providing comfort to investors through guarantees for new products and issuers,
(iv) providing cost savings to issuers in the region by sharing the reduction in spreads
resulting from credit enhancement, (v) improving the liquidity of bonds through diversification
of investment products and (vi) creating an environment in the capital markets conducive to
accelerated development of economies and establishment of bond markets.

236 BIS Papers No 30

To ensure the success of the new regional guarantee institution, contributions and
cooperation among the ASEAN+3 member countries, proper pricing and business scope,
exemption from regulation, risk diversity and sound underwriting policies are critical. The
institution should also have a sufficiently high credit rating to foster the development of Asian
bond markets. Acceptance by bond investors and sufficient initial paid-in capital and callable
capital are other factors to be considered for successful operation of the institution.

B. Possible structure of a regional guarantee institution

1. Business strategy
The primary guiding principles of the business strategy of the new regional guarantee
institution would be independence and prudence. The institution should not be influenced by
the main issuers or investors, or the country where it is located. The institution should be an
entity staffed by first-rate professionals who are able to analyse complex deal structures.
Finally, the institution should be prudent in risk-taking in order to remain sound and to attain
success.
The institution would provide credit risk guarantees, political risk guarantees and advisory
and structuring services. Under the credit risk guarantees, the institution would provide local
and convertible currency (US dollar, euro and yen) guarantees, partial credit guarantees,
political risk guarantees and guarantees of loans to bond issuers. The products would be
offered in three different phases. During the first phase, the institution would provide credit
enhancement or guarantees primarily for ABSs and infrastructure revenue bonds, and
advisory and structuring services. During the second phase, the institution would add
products for SME CDOs in close cooperation with local credit guarantee agencies. Finally,
during the third phase, the institution would provide a full-fledged range of products including
guarantees for municipal bonds. The institution would not provide guarantees to sovereign
bonds without conditionality. Because of these strict conditions, the institution’s guarantee
business for sovereign bonds might prove to be very limited.
Underwriting would be done according to the reinsurance model based on cooperation with
local guarantee institutions. Instead of taking on the zero-loss policy, the institution would
take on the risks with sufficient provisions. Because the institution would enter business lines
perceived as risky, such as local currency-denominated bond guarantees, the reinsurance
model would mitigate the risks. The role of the local institutions would be strengthened if the
institution adopted the reinsurance business model. In other words, the first loss is assumed
by the local guarantee institutions to the extent that they are able to assume it. The institution
itself might take the burden of the second loss, thus alleviating both the credit risk and the
currency risk.
The target market is ASEAN+3 countries. The institution would operate in all of these
member countries within two years of its inception. However, there would be country, obligor,
institutional and regional concentration limits based on the Fitch’s Emerging Market CDO
Model. The country concentration limits would be based on the country’s ratings: higher-
rated countries have higher concentration limits as their macroeconomic risks are expected
to be lower than the others. The industry concentration limits would vary depending on
whether the industry is global or local. Each obligor should represent no greater than a
certain portion of the total asset pool.
To maximise the benefits of the institution to the Asian region, it is hoped that the new
institution will be rated AAA (both local and foreign currency rating) by one or more
international rating agencies such as Moody’s Investors Service, Standard & Poor’s or Fitch
Ratings. A AAA credit rating extends the scope of benefit even to countries like Japan that
have credit ratings above AA. A AAA rating also increases the spread between rates with
and without guarantee, benefiting both the issuers and the institution. The challenge in

BIS Papers No 30 237

obtaining a AAA rating will be attracting enough capital. Nonetheless, the institution should
strive to get a AAA rating.
Adequate pricing would ensure a stable and sufficient return on capital for the institution.
Moreover, the spread may be split between the institution and the issuers so that there will
be adequate demand for the guarantee business.
The institution’s marketing strategy would be to produce continuous deal flows that meet its
strict underwriting standards through cooperation with intermediaries such as investment
banks and commercial banks, and other institutions including the rating agencies,
reinsurance companies and governmental and other regulatory authorities. In addition, to
build firm investor acceptance for securities guaranteed by the institution, local, regional and
international cooperation between regional and local credit guarantee facilities would be
essential and actively pursued.
The main objective of the investment portfolio would be to serve as a source of internal
liquidity for the institution. The preservation of its capital would be imperative, and the rate of
return should be reasonable. The annualised total rate of return of the portfolio would be
compared with inflation as measured by the consumer price index with respect to the
appropriate currencies, with the expectation of earning a positive real rate of return.
Investments would include primarily fixed income securities issued by governments of major
industrial countries such as government bonds of the United States and European countries.
This will diversify the portfolio risk of the institution.
With the aim of exercising due financial prudence and maintaining a sound capital structure,
the institution would adopt standard financial strategies and policies. At all times, the
institution would maintain adequate liquid resources. The guarantee leverage shall be up to
20 times its capital, surplus and reserves. This limit is considerably lower than those of
monoline bond insurers (with a limit of about 100 times), or ASIA Ltd (with a limit of up to
40 times), which shows that the sponsors/shareholders of the institution would operate in a
conservative manner.

2. Type of institutional form

The institution may be either public or private. There are five possible institutional options
depending on the legal status, organisational form and shareholders. The first option is a
multilateral institution, which would be established through a treaty prepared and signed by
participating shareholder countries. The second option is a private corporation with sovereign
shareholders including governments or government-related institutions. The third option is
also a private corporation, but the shareholders could include private as well as public
shareholders. One example of this model is ASIA Ltd. The fourth option is a purely private
corporation whose shareholders are private investors. The final option is an association of
existing local guarantee institutions and companies in the region such as the existing Asian
Credit Supplementation Institution Confederation (ACSIC).
A multilateral institution would have many advantages over a private corporation. First, it
would be relatively easy to enlist sovereign shareholders through the ASEAN+3 grouping,
even with a low expected return on capital in the initial period. The political and commercial
risks in Asia are considered very high compared to advanced countries and even developing
countries in other regions, and it would also be very difficult to maintain zero-loss
underwriting guidance in Asia. It would, therefore, be very difficult to find interested private
investors since the return on capital would naturally be lower than that of private companies.
Second, the injection of callable capital could be done quickly and with stronger commitment
from shareholders during periods of stress. Third, the tax-exempt status of a multilateral
institution could be a factor in the new institution’s price competitiveness. In addition, the
provision of guarantees would be more efficient and competitive if the new multilateral
institution utilises a trust fund and closely collaborates with local guarantee institutions.

238 BIS Papers No 30

However, it should also be noted that a multilateral institution would have shortcomings. The
most serious is the long period of time required to establish this type of institution considering
the long and complex process of coordinating and reconciling the wide-ranging economic,
administrative and political interests of participating countries. A multilateral institution might
also by its very nature be more bureaucratic than a private company.
A private corporation with sovereign shareholders including governments or government-
related institutions might share many advantages with multilateral institutions since the
shareholders would be public. However, it might have a shortcoming in that callable capital
might not be injected during periods of stress. Using an existing local guarantee institution
would be the least feasible option since most are undercapitalised, might lack the capability
and expertise needed to guarantee bonds and have no structure, rating and value as a bond
insurer to investors.
As discussed, each option has its own advantages and problems. However, considering not
only the nature and prospect of the new guarantee institution, but the economic and financial
environments in the region, a multilateral institution is believed to be the best option. To
ensure intergovernmental support for the new institution by, and regional financial
cooperation among, the ASEAN+3 countries, a multilateral institution is obviously a better
model than the others.
A multilateral institution would also be easier to establish than a private company in light of
the attractiveness of the new institution to prospective investors. The decision on whether to
establish a private or multilateral agency would depend on the attractiveness of its return on
capital to private investors. The return on capital measures how effectively the company uses
capital in generating profit, and is calculated by subtracting total costs from total revenues.
The credit rating agencies consider the Asian financial guarantee business as involving
greater political and commercial risks than the US financial guarantee business. As a result,
the capital charge for the financial guarantee business is two to three times higher, which
reduces the operating leverage by half or two thirds (from 100 to around 33). The much lower
rate of return on capital in the Asian financial guarantee business means that it does not
attract private equity. In conclusion, the tremendous social benefits and demands in the
Asian market suggest that the financial guarantee business should be multilateral rather than
private.
A multilateral institution would also be optimal considering that the mandate of the new
institution would be partly developmental in that it would facilitate development of bond
markets and grant small and medium-sized enterprises wider access to financial markets.
Finally, and most importantly, from a practical perspective, the new regional guarantee
institution should take the multilateral institution form since a long time might pass before the
new institution can generate steady cash flows, and its return on investment might initially be
lower than private investors would require.

6. Conclusion
This paper reviews recent efforts by Asian countries to develop the regional bond markets,
focusing on the creation of a credit guarantee institution. Establishment of a regional credit
guarantee institution is proposed as a viable solution to deal with the credit quality gap
between the low credit ratings of issuers and the minimum credit requirements of investors.
Combined with securitisation, credit guarantees are urged as an effective means to solve the
problem, among the most serious difficulties in the development of the bond markets in Asia.
The existing guarantee institutions, including multilateral institutions and private guarantee
companies, are not in a position to satisfy the already high demand for guarantees in Asia,
which is expected to grow over time. It is, therefore, believed that a new and more fitting
guarantee institution that can meet the Asian bond markets’ current needs should be

BIS Papers No 30 239

established. Towards that end, important lessons can be gleaned from the experience of
ASIA Ltd. Foremost among these, the new institution should be much more heavily
capitalised, and the expected return on capital should be lower. The development of a
regional credit guarantee institution would also entail a stronger commitment to capital
injection from participating countries during stress periods. Thus, the institution should
initially be a public but commercially viable entity. Moral hazard should be mitigated by
instituting an appropriate ownership structure, and its exposure to any particular country
should be limited through structuring and reinsurance. Cross-border risks should be
minimised through an emphasis on local currency business. Although we presented some
preliminary ideas on the institutional structure and business operations of the new regional
credit guarantee institution, we believe further, in-depth study is necessary. Among other
issues, we should seriously consider how to prevent moral hazard and make the new
institution commercially viable in the long term. For that purpose, close discussion and
cooperation among Asian countries is indispensable.
The new guarantee institution is expected to greatly contribute to the development of the
Asian bond markets. Other necessary components of the infrastructure and systems should
be introduced and developed to support bond market development, including but not limited
to credit rating systems, clearing and settlement systems, a centralised depository system
and liberalising exchange and capital controls.

References
Executives’ Meeting of East Asia-Pacific Central Banks (2002): “Payment systems in EMEAP
Economies”, July.
Hong Kong Monetary Authority (2001): “Study on the use and provision of credit guarantee
facilities in Asia”, unpublished report, March.
Ito, T (2003): “Promoting Asian Basket Currency (ABC) bonds”, December, pp 5-6.
Kaminsky, G L and C M Reinhart (1999): “The twin crises: the causes of banking and
balance-of-payments problems”, American Economic Review, vol 89, no 3, pp 473-500.
Moody’s Investor Service (2004): “2003 reviews and 2004 outlook Asian structured finance:
market faces mixed prospects in 2004 after issuance plunged in 2003”, Hong Kong.
Oh, G, D Park, J-H Park and D Y Yang (2003): “How to mobilize the Asian savings within the
region: securitization and credit enhancement for the development of East Asia’s bond
maket”, KIEP Working Paper, 03-02.
Oh, G and J-H Park (2003): “Developing the Asian bond markets using securitization and
credit guarantee”, KIF Financial Economic Series, 2003-04.
Oh, G, J-H Park, D Park and C Rhee (2004): “Building a settlement infrastructure for the
Asian bond markets: AsiaSettle”, KIF Financial Economic Series, 2004-02.
Park, Y-C (2004): “Asian bond market development: creation of a regional credit guarantee
institution”, February.
Park, Y-C and D Park (2003): “Creating regional bond markets in East Asia: rationale and
strategy”, paper presented at the Second Finance Forum of the Pacific Economic
Cooperation Council, Hua Hin, Thailand, 7-10 July.
Yoshitomi, M and K Ohno (1999): “Capital account crisis and credit contraction”, ADB
Institute Working Paper, no 2.

240 BIS Papers No 30

 Comments on Oh and Park’s paper
“ASEAN+3 regional guarantee mechanism”

Guorong Jiang 1

The joint paper by Professor Oh and Mr Park on the regional guarantee mechanism for Asia
provides a comprehensive discussion of credit guarantee mechanisms and securitisation in
the context of the ASEAN+3 Asian Bond Markets Initiative. The purpose of credit guarantees
is to bridge the perceived credit quality gap between the generally low credit rating of many
issuers in the region and investor demand for high-grade bonds. The authors propose that a
new regional credit guarantee agency be established, preferably in the form of a multilateral,
public sector organisation with an AAA rating. My main question concerns its financial
viability, even if profit maximisation were not its goal. In this regard, my comments focus on
the possible costs associated with extending the guarantee coverage to non-investment
grade credits, the use of securitisation, the dilemma of risk concentration and the difficulties
in risk mitigation.
The first question is whether the proposed regional credit guarantee agency would be
financially viable, being possibly the only financial guarantor in the world that would
insure a substantial sum of non-investment grade credits. Existing financial guarantors
are considered by the authors as inadequate in the Asian context - they generally are
available only for credits rated BBB or above before insurance. In fact, about three quarters
of the credit enhancements are on credits rated A or above before the guarantees. This
chosen risk profile excludes many Asian corporate credits, which are mostly non-investment
grade - about two thirds of the credits in Asia are non-investment grade and thus are not
potential customers of the existing financial guarantors. It is proposed that the new credit
guarantee agency would distinguish itself from existing guarantors by extending its coverage
to non-investment grade credits, so that a substantial number of Asian corporate credits
could benefit from its credit enhancement services. This could involve substantial costs for
the agency.
• First, to maintain the agency’s AAA rating, the leverage ratio needs to be
substantially reduced in order to cover lower-quality credits.
• Second, the premium charged by the agency needs to reflect the risk the agency
undertakes. However, this could be constrained by the need to provide incentives
for issuers to actively use the credit enhancement services to achieve a lower
funding cost, and the need to be competitive with other competing financing
channels such as bank loans.
• Third, the loss rates could be high, because the default rate is substantially higher
for non-investment grade credits compared to investment grade credits.
• Finally, a guarantee programme run by the public sector frequently suffers from
ultimately costly moral hazard and adverse selection problems. Hong Kong SAR’s
use of co-insurance in small and medium-sized enterprise (SME) financing is a
classic response to these problems. The Special Finance Scheme for SMEs was
launched in August 1998 to help address the liquidity crunch in the aftermath of the
Asian financial crisis. The government provided a 50-70% guarantee on loans
extended to SMEs by commercial banks. Thus, the credit assessments were
performed by banks without government interference. Recently disclosed data show

1
The views expressed here are those of the author and do not necessarily reflect those of the BIS.

BIS Papers No 30 241

 that out of about 12,000 approved guarantees to SMEs, over 1,700 loans (or 14% of
the total) went into default. Out of total guarantees of HKD 5.8 billion
(USD 0.7 billion), HKD 435 million worth of claims have been filed by banks with the
government, a default rate of 7.5%. Whether co-insurance would work as well in
economies with more corruption or weaker financial systems is an open question.
The second question is whether securitisation can serve to narrow the credit quality
gap between issuers and investors, thus mitigating the need for the credit guarantee
agency to move down the credit spectrum. Using senior-subordinate tranches,
securitisation could offer bonds with a credit rating higher than the underlying assets.
However, it should be noted that securitisation can repackage credit risks but cannot reduce
such risks, which show up in the subordinate (or equity) tranches. The lower the underlying
asset quality, the larger the equity tranche. In general, the subordinated tranches are illiquid
and hard to market to investors. In Korea’s corporate bond securitisation, they were bought
by government-supported agencies. Even if equity tranches are sold, possibly with deep
discounts, they shrink the size of senior tranches available to a broader range of investors
and raise the cost of funds created through securitisation.
The third question is how realistic it is for the credit guarantee agency to mitigate its
risk by diversifying its market coverage. Instead of focusing on insuring only Asian
credits, the proposed credit guarantee agency could cover credits from the developed
countries. This is indeed one of the lessons from ASIA Ltd, the first and failed Asian credit
guarantee company with heavy public sector involvement. During the Asian financial crisis,
credit risks between Asian economies became highly correlated and ASIA Ltd was
downgraded and went out of business. The key lesson is the need to diversify outside Asia to
reduce risk concentration. However, this conclusion raises two questions. First, it is hard to
justify a publicly funded regional institution writing insurance that is not closely related to the
aim of developing local bond markets. Second, a more practical issue is how this newly
established agency could compete outside Asia with existing guarantors in their home
markets.
The fourth question is how realistic some of the proposed risk mitigation options for
the credit guarantee agency are. One proposed option is cooperation with local guarantee
agencies. To the extent that they compete for business, it is hard to imagine why local
guarantee institutions would assume first loss and source possible deals for the regional
agency, unless it is a public or publicly owned entity not operating strictly in line with
commercial principles. The second option is to establish trust funds as shock absorbers with
contributions from regional governments and donors. This would certainly help shift some of
the costs out of the agency, but the overall support or subsidy needed for writing credit
guarantees is not reduced. To the extent that the regional credit guarantee agency would
have an expected low return that would not be attractive for private sector investors, or even
incur large losses to be absorbed by trust funds under adverse market conditions, it implies
that the initiative could potentially require significant public subsidy to support bond financing.
Whether such subsidies are the best use for public money is beyond the scope of the paper;
in practice, burden-sharing among Asian governments for financing the agency could be an
important issue to consider.
The final question is how serious an impediment to bond market development the
perceived credit quality gap between issuers and investors is. Strong demand from
investors for high-grade bonds should exert heavy market pressure on corporate issuers to
improve their creditworthiness through greater reliance on equity financing. Looking forward,
the credit quality gap is likely to narrow, with or without the credit guarantee agency. With
rapid growth and prudent macroeconomic and financial policies, sovereign and corporate
credit fundamentals have indeed improved substantially over the past year, evinced by
record upgrades in the region by international rating agencies. Improving corporate credits in
the region would certainly benefit the proposed credit enhancement agency.

242 BIS Papers No 30

 Promoting the Asian bond market

Kap-Soo Oh

Introduction
In dealing with the aftermath of the 1997 Asian financial crisis, many Asian countries came to
appreciate just how important it is for economic development and growth to keep their
financial system safe and sound. Many of the discussions and debates on the Asian bond
market point to the pivotal role that Asia’s bond market can play in further enhancing the
safety and soundness of the region’s financial system. And it is clear that, for Asia’s bond
market to grow, close cooperation and collaboration among the region’s financial regulators
will be critical.
There follow some thoughts on recent developments in the Asian bond market, on some of
the steps that we can take to further promote its growth, and on the role financial regulators
can play in this endeavour.

Recent developments in the Asian bond market

ASEAN+3, APEC, EMEAP and other regional groups have led most of the region-wide
efforts to promote the Asian bond market. The member countries of ASEAN+3 adopted the
“Chiang Mai Initiative”, the first major regional financing agreement, in May 2000. The
agreement strongly supports bilateral swap agreements among the member countries to
pre-empt liquidity shortages and other significant financial distress in the region. The
agreement provided a much needed boost to countries seeking region-wide swap
agreements, and as of October 2003 there were 13 swap agreements totalling approximately
USD 32.5 billion, or around double the number and amount before the agreement.
Along with these efforts, a number of specific steps designed to expand both demand and
supply in the Asian bond market are in the works as a way to further strengthen financial
cooperation among the Asian countries. To stimulate demand, EMEAP agreed to launch the
Asian Bond Fund (ABF) last year and approximately USD 1 billion is currently managed by
the fund. ASEAN+3 adopted the “Asia Bond Market Initiative” last August and has been
working on specific proposals to expand supply in the Asian bond market as well. Moreover,
efforts are under way to build a framework that can be used to support and stimulate the
local bond markets throughout the region and promote cross-border trading of local currency
bonds. Various studies on the feasibility of introducing a regional currency are also in the
works. Studies on improving the market and institutional infrastructure for Asia’s bond
markets - such as proposals on pooling the region’s corporate bonds and securitising them,
harmonising the region’s accounting standards, foreign exchange trading, and new
settlement systems - are also gaining attention.
For its part, Korea has been taking a number of concrete steps to contribute to the growth of
the Asian bond market. First, we are focusing on building an efficient infrastructure for the
bond market and improving international harmonisation, while at the same time enhancing
the ability of domestic financial institutions to bring greater depth to the bond market. At the
same time, we are aggressively moving forward with our efforts to enhance the transparency
and efficiency of our financial market and attract leading international financial institutions
and organisations to Korea. I would also like to add that we are making a concerted effort to
provide support and actively participate in all collective cooperative endeavours throughout
Asia to contribute to the success of Asia’s bond market.

BIS Papers No 30 243

Where the Asian bond market is headed
Although the depth and maturity of the bond market vary from country to country, and cross-
border bond investments remain limited, many expect the Asian bond market to show a
much more robust growth in the near future as the region’s financial system advances and
the local bond markets gain strength. And there is no doubt that working together to promote
economic development and prosperity will contribute to the economic and financial stability of
the region.
In promoting a vibrant regional bond market, we face a number of challenges that we must
overcome. First, we need to take concerted steps to facilitate the development of such key
infrastructure and mechanisms as a centralised clearing and settlement system, credit
guarantee institutions and credit rating systems in the bond market. As a part of this
endeavour, we may need to carefully weigh the merits of coming up with a blueprint for a
new architecture aimed at integrating the local markets closer together and improving the
overall efficiency of the region’s bond market.
Second, the regulatory regime and information systems of each country must continue to
improve so as to ensure efficient and transparent cross-border financial transactions. This
effort will most likely entail harmonising the region’s market systems and regulations through
close collaboration and cooperation among the Asian regulators.
Third, securitisation and other structured financing need to be more actively encouraged.
Although many Asian countries do utilise asset securitisation and structured financing in their
capital markets, only a few of them are at a mature stage. So it will be a worthwhile
endeavour to streamline the existing market structures and come up with some form of
uniform standards that can be used to further promote the region’s securitisation market.
Lastly, the traditional reliance of many of the regional economies on bank-centred financial
systems for capital will need to be shifted to the bond market.

The role of the regulators

In order to ensure a sound development of Asia’s financial market, the role that the region’s
regulators play and the efforts they make will be crucial. One of the key tasks they face is to
move forward with building an advanced regulatory system so as to eliminate frictions
between international standards and the local financial systems. This will mean, among other
things, strengthening the existing prudential regulation and supervision to preserve the safety
and soundness of local financial institutions and financial markets, and raising accounting,
disclosure and governance standards to enhance market transparency and ensure orderly
and disciplined market conduct.
Apart from close cooperation and collaboration on harmonising the region’s regulatory
structures for the bond market, the regulators will need to work on further market opening. An
environment must be created that allows foreign financial service providers to enter into the
market and to compete freely with others.

Concluding remarks
In the light of the overall situation in Asia, the financial infrastructure and the bond market, it
is clear that the challenges lying ahead will not be easy to overcome. One possible approach
may be to come up with and implement a step-by-step road map for the pan-Asia bond
market based on the development of the local bond markets in the region.

244 BIS Papers No 30

It is also important to bear in mind that, in order for Asia to firmly take its place in the global
economy as the most dynamic region in the 21st century, creating and maintaining efficient
and cooperative relations will be more important than ever. From this perspective, the recent
efforts led by ASEAN+3, APEC and other regional groups to promote an Asian bond market
will surely provide a significant impetus for the region’s financial market and help it move up
to a new level.
Korea will remain an active partner and participant in the drive for a robust, vibrant Asian
bond market, and to do its part, the country will forge ahead with the integration and
harmonisation of its bond market with the rest of Asia, further develop its market
infrastructure, and contribute to the growth and success of the Asian bond market.

BIS Papers No 30 245

 Identifying impediments to cross-border bond
investment and issuance in Asian countries 1

Atsushi Takeuchi 2

1. Introduction
The Asian currency crisis clearly demonstrated the need to develop well functioning local
bond markets in Asia. Before the crisis, companies (including banks) in Asia, regardless of
whether they had earnings in foreign currencies, often funded their business activities in
foreign currencies, taking advantage of the low interest rates at the time. Such funding
typically took the form of short-term bank lending, rolled over repeatedly. The funds raised in
this way were often spent to finance domestic fixed (long-term) investment producing local
currency cash flow, thereby creating the so-called “double mismatch” of maturity and
currency on the companies’ balance sheets. This system collapsed when foreign lenders
ceased to roll over their loans to the borrowers in Asia when market perceptions suddenly
changed, aggravating the crisis.
Since then, Asian governments have made great efforts to foster bond markets in their
respective countries by conducting a series of market reforms. Many of the reform efforts are
concentrated on government bond markets, which is quite understandable given that a well
functioning government bond market is considered a precondition for the development of the
corporate bond market. Such efforts also reflect an increased need for Asian governments to
finance their expanded fiscal spending after the crisis. 3 As a result, the size of local bond
markets in Asia in terms of outstanding amounts has more than doubled since 1998 and is
now estimated to exceed $1 trillion (excluding Japan).
Nonetheless, local bond markets in Asia have room for further improvement in many
aspects. Most notably, secondary markets are not liquid in many countries, which is
evidenced by the low turnover ratio of government bonds. While bid/ask spreads are already
narrow for some countries, they may be biased given the low trading volume (Table 1). A
1998 APEC study pointed out that inactive secondary markets were attributable to a number
of factors such as a lack of reliable benchmark yield curves, a lack of local institutional
investors, underdeveloped trading, clearing and settlement systems, a lack of liquidity, a lack
of committed market-makers, long settlement periods and the absence of bond lending
programmes. Although many of these factors have since seen dramatic improvements, they
remain impediments to active bond trading in many countries.

1
The markets covered in this paper are those of China, Hong Kong SAR, Indonesia, Japan, Korea, Malaysia,
the Philippines, Singapore and Thailand, unless explicitly stated otherwise.
2
The views expressed here are entirely the personal opinions of the author and do not reflect the official views
of the Bank of Japan. This paper is prepared for information only. Although the author has endeavoured to
provide accurate and timely information, readers are reminded that there is no guarantee that all information
provided is accurate and up-to-date.
3
There are a number of other reasons to develop local bond markets in Asia. Perhaps the most important is
that development of the bond market introduces “credit risk culture” into the region, which is necessary for the
efficient allocation of resources.

246 BIS Papers No 30

 Chart 1
Outstanding amount of debt
securities in Asian countries

($ billions)
1,200
China Hong Kong SAR
Malaysia Philippines
1,000
Singapore South Korea
Thailand
800

600

400

200

0
1996-03 1997-03 1998-03 1999-03 2000-03 2001-03 2002-03 2003-03

Source: BIS.

Table 1
Liquidity indicators in government bond markets

Turnover ratio1 Bid/ask spread

(times) (bps)

China 0.4 –
Hong Kong SAR 15.6 5-10
Indonesia 0.5 –
Korea 9.6 1
Malaysia 3.7 3-5
Philippines – 25-50
Singapore 5.0 5
Thailand 2.5 2-3
Japan 6.9 7
United States 22 3
1
Ratio of turnover to average outstanding stock in 2002.
Sources: RBA (2003); Mohanty (2002).

BIS Papers No 30 247

A related issue is extremely limited foreign (non-resident) participation. In fact, foreign
investors are virtually non-existent in local bond markets in Asia. This constitutes a clear
contrast with foreign investor participation in local stock markets (Table 2). Similarly, there
are only four bond markets in the region where non-resident issuers are meaningfully
present, namely Japan, Korea, Hong Kong SAR and Singapore. Among the four markets,
Hong Kong seems to be most able to attract foreign issuers. About one third of outstanding
Hong Kong dollar issues belong to foreign entities. Nonetheless, the amount of such
non-resident issuance is still limited compared with the eurobond market. For example, the
amount of issuance of samurai bonds (1.3 trillion yen) stood at only one tenth of that of
euroyen bonds in 2002.

Table 2
Foreign investor participation
in local markets

Bonds1 Stocks2

Hong Kong SAR <1 41

Indonesia <1 na
Korea 2 37
Malaysia 1 22
Thailand 1 29
Japan 5.8 32
United States ? ?
1
Amount of holdings as a percentage of total domestic debt outstanding (end-2000) except Japan (end-2003)
2
and Thailand (as a percentage share of total turnover value during 1999). Percentage share of turnover
value for Hong Kong (2001), Japan (2003) and Thailand (1999). Amount of holdings as a percentage of total
stocks outstanding for Korea (2001) and Malaysia (1998).
Sources: Mihaljek et al (2002); Daiwa Research Institute (2002); Kadir (2001); Prasarn (2001); Bank of Japan.

After discussing the rationale for promoting cross-border bond investment and issuance in
Asian local bond markets, this paper tries to identify the specific factors that are hampering
such investment and issuance, taking stock of previous studies and research done by
investment banks, academics and public authorities. The paper also offers some suggestions
on how to tackle this issue with a view to creating integrated regional bond markets in Asia.

2. Rationale for promoting foreign participation in local bond

markets

2.1 Benefits to local bond market

In Asia, local banks continue to play an important role in bond markets. They are the largest
bondholders in many countries, though local institutional investors are gradually gaining
ground. This concentration of investors is likely to accentuate volatility in the market during
times of stress. In this respect, foreign investors can help broaden the investor base and add
liquidity to the secondary market. There is a negative image associated with foreign investors
in Asian countries. Some believe that foreign investors tend to show herding behaviour
because of a lack of information and poor understanding of the market, thereby inducing

248 BIS Papers No 30

volatility in the market. While such a hypothesis is very difficult to verify, past studies and
research do not provide any evidence to support it (IMF (2003a)).
Non-resident issuers also contribute to the development of local bond markets by providing
an additional supply of bonds. Bonds issued by blue-chip multinational enterprises or
international organisations can create a new asset class for such markets, thereby offering
diversified investment opportunities to local investors. In addition, foreign market players
(both investors and issuers) are often eager to increase the efficiency of local markets and
therefore are vocal advocates for market reforms. They are the ones who offer global
perspectives when conducting such market reforms and bring internationally acknowledged
best practices into local markets.

2.2 Benefits to foreign investors and issuers

Opportunities provided by local bond markets in Asia to foreign investors should not be
underestimated. As a general characterisation of emerging local bond markets, JPMorgan
(2002) states that “emerging local markets are one of the final frontiers of untapped
diversification opportunity”. In recent years, Asian local bonds have shown strong
performance. A simulation by Consing (2003) shows that the HSBC’s Asian Local Currency
Bond Index (ALBI) 4 accumulated a return of 35.7% for the period between 29 December
2002 and 4 September 2003, thereby significantly outperforming US Treasuries (with similar
duration) with a return of 20.4% (Table 3). This was attributable to an excess return of 11.2%
on capital gains and carry income and a 4.1% excess return from currency gains. Reviewing
the performance data for the first half of 2003, Barclays Capital (2003a) finds that a basket of
Asian local bonds is a more efficient investment than individual bond holdings due to the less
than perfect correlation between local bond markets. In other words, the total volatility of a
basket of Asian local bonds in terms of standard deviation of daily percentage changes is
lower than the sum of individual market volatility.

Table 3
Performance of HSBC ALBI

Return

ALBI 35.7%
US Treasuries (similar duration) 20.4%
ALBI excess returns over US Treasuries 15.3% point
Of which:
Attributable to capital gains and carry income 11.2% point
Currency gains 4.1% point

Source: Consing (2003).

There are already encouraging signs that Asian local bond markets are beginning to gain
attention from foreign investors. Investment banks now produce reports on Asian local bond

4
The HSBC ALBI tracks the total US dollar return performance of liquid bonds denominated in local currencies
in mainland China (though the current weighting for China is zero), Hong Kong SAR, India, Malaysia, the
Philippines, Singapore, Taiwan (China) and Thailand.

BIS Papers No 30 249

markets more frequently than ever. Seminars and workshops dealing with this topic are
attracting quite a few participants from the private sector. Furthermore, Deutsche Bank
(2003) observes that most of the world’s largest bond funds have now set up offices in
Singapore and Hong Kong, even though many are satellite offices doing research on hard
currency credit only.
Foreign issuers also benefit from bond issuance in Asian local bond markets. The local bond
issuance can give multinational enterprises long-term financing for their business activities in
the country. In fact, such financing needs are increasing as direct investment in the region
continues to grow. In addition, foreign issuers enjoy diversification of their funding base.

3. Overview of impediments

3.1 Impediments to foreign investor participation in local bond markets

There are many empirical studies aimed at explaining the extent and determinants of
investors’ position in international equities. On the contrary, perhaps due to the lack of
available data, only a few studies have been conducted to explore the factors that determine
investors’ position in international bonds. Burger and Warnock (2003) found that country
weights in US investors’ foreign bond portfolios were related to the openness of capital
accounts and potential diversification benefits as indicated by historical correlations. They
also found that emerging market countries with more stable policies as evidenced by price
stability and stable exchange rates have greater bond market development and higher US
participation.
The following sections focus on structural impediments to foreign investor participation.
Capital control is probably the most critical issue, as shown in the above-mentioned study,
since it directly limits foreign investors’ market access. Availability of hedging instruments
(currency derivatives and interest rate derivatives) is another significant factor, given that
hedging instruments make it possible to reduce performance volatility. There are also factors
that affect foreign investor participation indirectly through an increase in costs and risks for
investment. Among others, taxation, market infrastructure (such as the disclosure framework
and price transparency), creditor protection and clearing and settlement systems are often
cited as determining factors. On top of that, the common problem across the region is the
language barrier, which makes it difficult for foreign investors to find accurate and timely
regulatory information and costly for them to fulfil documentation requirements. All of the
above are interrelated and naturally relevant to the development of local bond markets in
general. Hereafter, this paper describes the factors that are of particular relevance to foreign
investor participation in Asian local bond markets.

Capital controls
Controls on capital transactions are a broad concept which includes controls on capital and
money market instruments, derivatives and other instruments and credit operations. After the
1997 currency crisis, Asian countries adopted various controls on capital transactions and
still maintain many of them. The IMF’s Annual Report on Exchange Arrangements and
Exchange Restrictions offers comprehensive information on this subject.
As for control over foreign ownership of local bonds, China is the only country that imposes a
restriction (Table 4). The launch of the QFII (Qualified Foreign Institutional Investors) system
effective December 2002 was undoubtedly a significant first step toward financial
liberalisation. The total quota authorised for 10 QFIIs (as of 19 November 2003) amounts to
$1.7 billion. However, there are many restrictions that make this system difficult to use. QFIIs
are allowed to invest in government and corporate bonds listed on China’s securities
exchanges only (ie they are not allowed to participate in interbank markets where secondary

250 BIS Papers No 30

market liquidity is more ample). Furthermore, even though the regulation allows QFIIs to
conduct repurchase transactions in government bonds and to trade corporate bonds, the
Shanghai and Shenzhen exchange markets announced that such transactions were not to
be conducted temporarily due to technical reasons (PricewaterhouseCoopers (2003)).

Table 4
Selected capital control
measures in Asian countries (1)

Foreign ownership
Note
of local bonds

China QFIIs only. QFIIs must satisfy various requirements. For

instance, QFIIs must set up special renminbi
accounts with domestic banks and use the services
of domestic securities companies. Closed-end
QFIIs may only remit capital after three years, in
instalments of no more than 20% of the total each
time, at intervals of one month or more. Other QFIIs
may only remit capital after one year, in instalments
of no more than 20% of the total, and at intervals of
three months or longer. SAFE must approve all
repatriations.
Hong Kong SAR No restriction.
Indonesia No restriction. No non-resident person may purchase more than
1% of any mutual fund.
Japan No restriction.
Korea No restriction. In order to invest in listed bonds, foreign institutions
must apply for an Investment Registration
Certificate (IRC) from the Financial Supervisory
Service before investing in the market.
Malaysia No restriction.
Philippines No restriction. Registration with the BSP is necessary if the foreign
exchange needed for capital repatriation and
remittance of dividends, profits and earnings that
accrue thereon is purchased from the banking
system.
Singapore No restriction.
Thailand No restriction.

Sources: IMF (2003b); PwC (2003); BONY (2002, 2003); JPMorgan (2002).

Whether foreign investors can obtain local currency credit from local financial institutions is
another important factor. Foreign investors may find it easier to invest in local bonds if they
have free and timely access to local currency credit. As shown in Table 5, most Asian
countries set a limit on the extension of local currency credit to non-residents.
Restrictions on foreign exchange transactions are also relevant. Even when foreign
exchange transactions for foreign investors to purchase local bonds are permitted,
documentation requirements for approval or reporting may be quite onerous.

BIS Papers No 30 251

 Table 5
Selected capital control
measures in Asian countries (2)

Local currency credit facilities by

Other limitations for non-residents
residents to non-residents

China Financial institutions authorised by the

PBC may lend to overseas institutions
or contract overseas credits.
Hong Kong No restriction.
SAR
Indonesia Not allowed, unless it is related to Foreign exchange transactions must be
economic activities in Indonesia, such executed through banks incorporated in
as transactions involving rupiah- Indonesia and licensed by Bank
denominated securities. Indonesia.
Japan No restriction.
Korea Credit and loans of more than Won purchases by foreign investors
W100 million per borrower must be associated with a special
denominated in local currency and securities purchase transaction.
granted by institutional investor require
MOFE approval.
Malaysia Banking institutions as a group may Ringgit purchases by foreign investors
extend ringgit overdraft facilities not must be associated with securities
exceeding RM 500,000 in aggregate to transactions. There is no restriction on
non-residents, provided the credit the conversion of ringgit funds in
facilities are fully covered at all times by external accounts (funds belonging to a
fixed deposits placed by the non-resident individual or corporation,
non-resident with the banking maintained with a financial institution in
institutions extending the credit Malaysia) into foreign currency for
facilities. repatriation. Investors may bring funds
into and repatriate funds out of the
country (in foreign currency only).
Philippines No restriction so long as the facilities
do not involve foreign exchange
purchased from the banking system.
Singapore No restriction except that financial
institutions in Singapore may not
extend SGD credit facilities exceeding
SGD 5 million to non-resident financial
entities where they have reason to
believe that the proceeds may be used
for speculation. Overdrafts are
prohibited in all cases.
Thailand The BOT does not allow domestic
financial institutions to extend direct
baht-denominated loans to non-
residents. Credit facilities provided by
domestic financial institutions to foreign
investors are capped at THB 30 million
per each baht cash account and THB
50 million per counterparty.

Sources: IMF (2003b); PwC (2003); BONY (2002, 2003); JPMorgan (2002).

252 BIS Papers No 30

Availability of hedging instruments
In the absence of hedging instruments, foreign investors have no option but to take both
interest rate exposure and foreign exchange rate exposure when they purchase local bonds.
Therefore, the availability of hedging instruments, for currency risk in particular, is crucial for
foreign investors. After the Asian crisis, the offshore foreign exchange derivatives market for
Asian local currencies disappeared due to the introduction of capital controls (whereas
offshore non-deliverable forwards markets emerged for some currencies). Nonetheless, most
countries have onshore foreign exchange forwards and swap markets, although the hedging
costs are high (Barclays Capital (2003b)) and tenors are relatively short (Table 6).

Table 6
Overview of currency risk hedging instruments
Non-resident access to
Onshore FX forward Offshore market
onshore FX forwards
China Up to four months Not allowed NDF liquid
Hong Kong Liquid No restriction None
SAR
Indonesia Liquid Allowed to hedge principal NDF liquid
and coupon
Korea Liquid Allowed to hedge principal NDF liquid
and coupon
Malaysia Illiquid Prior approval required None
Philippines Liquid Prior approval required NDF illiquid
Singapore Liquid Allowed to hedge principal Deliverable forward
and coupon illiquid
Thailand Liquid Allowed to hedge principal Deliverable forward
and coupon illiquid

Source: Barclays Capital (2003b).

Taxation
Taxation has a significant impact on the development of bond markets in general. Capital
gains taxes create a disincentive to trade bonds frequently, thereby reducing arbitrage
opportunities. Moreover, withholding taxes on interest income to foreign investors reduce the
returns from holding bonds. They may also create market fragmentation through distortion if
the treatment of withholding tax is different depending on the types of bonds or investors.
Therefore, it is important to adopt tax policies that are compatible with market development
while not seriously compromising the principles of good taxation.
None of the G7 countries charge withholding tax on interest income obtained from
government bonds held by foreign investors. On the other hand, withholding tax on interest
income is charged in a number of countries in Asia, though the tax rates are not very high
(Table 7). Although there are tax treaties among Asian countries that reduce tax burdens, tax
reclaim procedures are complicated in many countries.

BIS Papers No 30 253

 Table 7
Tax treatment on returns from foreign
investors’ holdings of local bonds
Withholding tax on
Capital gains tax Any other tax
interest income
China No withholding tax on 33%, if bonds are not
interest rate income. held until maturity (plus
5% profit tax).
Hong Kong No withholding tax on No capital gains tax.
SAR interest income.
Withholding tax on Capital gains tax.
interest income.
Indonesia 20% of income earned. No capital gains tax. 0.1% of gross sale
proceeds is withheld by
the broker as income tax
for securities transactions
executed on the
exchange.
Japan No withholding tax for No capital gains tax.
JGB, if a number of
requirements are
satisfied.
Korea 27.5% of income earned. The lower of 11% of
gross sales proceeds or
27.5% of net capital
gains.
Malaysia 15% of income earned. No capital gains tax.
Philippines 20-32% of income No capital gains tax.
earned
Singapore 15% of income earned. No capital gains tax.
Resident investors are
exempt from withholding
tax.
Thailand 15% of income earned. 15%.
Sources: BONY (2002, 2003); JPMorgan (2002).

Clearing and settlement of bonds

Scripless settlement, delivery versus payment (DVP) and short settlement cycles are
desirable from the viewpoint of efficiency and risk reduction. In this respect, clearing and
settlement systems in Asia are fairly well developed (Table 8). In fact, securities settlement
systems in the region satisfy most of the G30 recommendations (IIMA (2003)). A real
problem arises from the requirements to appoint a local custodian in order to settle local
currency bonds (Table 9). For example, foreign investors complain that documentation
requirements for participation in BOJ-NET through a third-party intermediary are onerous (in
large part relating to the withholding tax exemption procedure).

254 BIS Papers No 30

 Table 8
Selected features of clearing, settlement
and custody in Asian countries (1)1
Real time gross
Scripless settlement Settlement cycles
settlement/DVP
China Yes No T+0 or T+1
Hong Kong SAR Yes Yes T+0 or T+1
Indonesia Yes (paper Unknown T+3
remains)
Japan Yes (paper Yes T+3
remains)
Korea Yes Yes T+1
Malaysia Yes Yes (for direct T+1
participants
only)
Philippines Yes Unknown T+0 or T+1
Singapore Yes Yes T+1
Thailand Yes Yes T+2
1
Government bonds.
Sources: BONY (2002, 2003); IIMA (2003).

Table 9
Selected features of clearing, settlement
and custody in Asian countries (2)1
International linkage of CSD Custodian
China CMU (Hong Kong, planned) Local custodian
Hong Kong SAR Clearstream (ICSD), Euroclear (ICSD), KSD (Korea), CMU
CDC (China, planned), AustraClear (Australia) and
AustraClear (New Zealand)
Indonesia None Local custodian
Japan None Local custodian
Korea CMU (Hong Kong) Local custodian
Malaysia None ADI
Philippines None Local custodian
Singapore Clearstream (ICSD) and Euroclear (ICSD) MAS, local
custodian
Thailand None Local custodian
1
Government bonds.
Sources: BONY (2002, 2003); IIMA (2003); Citigroup (2003a).

BIS Papers No 30 255

3.2 Impediments to foreign issuer participation in local bond markets
Issuance of local bonds by non-residents is allowed in most Asian countries except China
and Thailand. Nonetheless, as noted above, actual issuance by non-residents is negligible in
several countries. This is attributable to various factors. For instance, local rating or local
listing requirements, use of local law as governing law and preparation of documentation in
the local language all increase costs of issuance. Furthermore, if the approval process and/or
issuance procedure takes a long time, there is a risk that market conditions might change.
The availability of investment tools for the funds raised by non-resident issuance of bonds is
another factor.

Table 10
Factors affecting issuance of local bonds by non-residents (1)1

Issuance of local bonds by

Local rating/local listing Governing law
non-residents

China Not allowed Not required/required Chinese law

Hong Kong Allowed Not required/not required English law
SAR
Indonesia Allowed Required/required Indonesian law
Japan Allowed Not required/not required Japanese law
Korea Allowed (prior report to the Required/required Korean law
MOFE and FSC necessary)
Malaysia Allowed (approval required) Required/required English/New York/
Malaysian law
Philippines Allowed (only after the proper Required/not required Philippine law
license to do business in the
country is secured from the
appropriate government agency,
provided payment for the sale or
issue does not involve the
purchase of foreign exchange
from the banking system)
Singapore Allowed (when the Singapore Not required/not required English/New York/
dollar proceeds of an initial Singaporean law
public offering are used
offshore, they must be swapped
or converted into foreign
currency upon drawdown by the
issuer)
Thailand Not allowed Required/not required Thai law
1
Corporate bonds.
Sources: IMF (2003b); Citigroup (2003b).

256 BIS Papers No 30

 Table 11
Factors affecting issuance of local bonds by non-residents (2)1

Documentation Typical duration of

Time required to obtain approval
language issuance process

China Chinese 2-6 months 2 weeks

Hong Kong English A few weeks needed only 2-4 weeks
SAR for retail issues
Indonesia Bahasa 45 days upon completion of 4 months
documentations
Japan Japanese 1-2 weeks 2-3 months
Korea Korean na 2 weeks
Malaysia English 14 days for non-equity 4-8 weeks
linked issue
Philippines English 2 months 8-10 weeks
Singapore English None 2-4 weeks
Thailand Thai 2 weeks 8-12 weeks
1
Corporate bonds.
Sources: IMF (2003b); Citigroup (2003b); Mitsubishi Securities (2003).

4. Way forward

4.1 Regional cooperation

There are already several regional initiatives dealing with Asian bond market development.
Ministries of finance and central banks of ASEAN+3 countries are pursuing the “Asian Bond
Markets Initiative (ABMI)”, a comprehensive approach to developing bond markets in Asia.
Working groups have been created on a voluntary basis for six areas: creating new
securitised debt instruments, credit guarantee mechanisms, foreign exchange transactions
and settlement issues, issuance of bonds denominated in local currency by multilateral
development banks (MDBs), foreign government agencies and Asian multinational
corporations, local and regional rating agencies, and technical assistance coordination. Asia-
Pacific Economic Cooperation (APEC) has also discussed the harmonisation of bond market
rules and regulations across the region. Meanwhile, the Asian Cooperation Dialogue (ACD)
aims to promote public awareness of the various initiatives as well as to provide political
support for them. The Executives’ Meeting of East Asia-Pacific Central Banks (EMEAP), a
group of 11 central banks from the region (Australia, China, Hong Kong SAR, Indonesia,
Japan, Korea, Malaysia, New Zealand, the Philippines, Singapore and Thailand), launched
the Asian Bond Fund (ABF), an index bond fund of US dollar-denominated Asian bonds
managed by the Bank for International Settlements, in July 2003, and it was welcomed by
the ACD. The EMEAP central banks are now studying the ABF2, a bond fund of local
currency-denominated Asian bonds. All these regional initiatives indicate that there is a
political will to enact market reforms for the development of bond markets.
Against this background, consideration could be given to the creation of a forum for
securities regulators, ministries of finance and central banks of ASEAN+3 countries with a
view to promoting regional cross-border transactions in local bonds. Discussion topics could

BIS Papers No 30 257

include harmonisation of rules and regulations, regional tax arrangement and any other
issues that are of relevance to cross-border local bond transactions.

4.2 Private sector involvement

It should be noted that private sector involvement is key to the development of regional bond
markets, since market participants are the users of such markets. There should be regular
dialogue between the public and private sectors on the national as well as regional level.
Impediments to cross-border transactions of bonds must be identified specifically through
such dialogue. The public sector must carefully listen to the private sector when exploring
new infrastructure projects such as a regional credit guarantee mechanism or regional
securities settlement system. In particular, the public sector should be mindful of the risk of
crowding out the private sector.
In this context, it would be meaningful to conduct a survey among market participants of the
impediments to cross-border bond transactions. At the same time, self assessment could be
done by public authorities on the same subject. This may serve as a first step toward
identifying impediments, thereby forging common understanding of the issues across the
public and private sector.

References
Asia-Pacific Economic Cooperation (1998): “Questionnaire survey on developing domestic
bond markets: summary of responses”, APEC Collaborative Initiative on Development of
Domestic Bond Markets, December.
Asian Development Bank (2003): “Harmonization of bond market rules and regulations”,
Background Study for APEC Finance Ministers’ Process, August.
Bank of New York (2001): “China”, Securities Market Report, December.
——— (2002a): “Indonesia”, Securities Market Report, September.
——— (2002b): “Philippines”, Securities Market Report, December.
——— (2003a): “Singapore”, Securities Market Report, January.
——— (2003b): “Hong Kong”, Securities Market Report, August.
——— (2003c): “Malaysia”, Securities Market Report, August.
——— (2003d): “South Korea”, Securities Market Report, August.
——— (2003e): “Thailand”, Securities Market Report, August.
Barclays Capital (2003a): “High returns with low volatility: building baskets of local currency
Asian bonds”, Asia Rates and Credit Research, June.
——— (2003b): “The Asian bond market: from fragmentation to aggregation”, Asia Rates
and Credit Research, September.
Bekaert, G and C Harvey (2002): “Emerging markets finance”, Journal of Empirical Finance,
December.
Burger, J and F Warnock (2003): “Diversification, original sin, and international bond
portfolios”, Board of Governors of the Federal Reserve System International Finance
Discussion Papers, January.
Citigroup (2003a): “Guide to Asian bond and currency markets 2003”.
——— (2003b): Presentation before members of EMEAP Working Group on Financial
Markets, September.

258 BIS Papers No 30

Consing R (2003): “Creating an efficient Asian bond market - the private sector
perspective”, presentation material for BOAO Forum for Asia Annual Conference 2003,
November.
Deutsche Bank (2003): “Asian local bond markets”, June.
Edison, H and F Warnock: “Cross-border listings, capital controls, and equity flows to
emerging markets”, IMF Working Paper (WP/03/236), December.
Fabella, R and S Madhur (2003): “Bond market development in Asia: issues and
challenges”, ERD Working Paper, no. 35, January.
Herring, R and N Chatusripitak (2000): “The case of the missing market: the bond market
and why it matters for financial development”, paper prepared for the Asian Development
Bank Institute/Wharton Seminar on Financial Structure for Sustainable Development in
Post-Crisis Asia, Tokyo, 26 May 2000.
Institute for International Monetary Affairs (2003): “Settlement systems of East Asian
economies”, study commissioned by Ministry of Finance of Japan, September.
International Monetary Fund (2002): “Emerging local bond markets”, chapter IV in Global
Financial Stability Report, September.
——— (2003a): “Local securities and derivatives markets in emerging markets: selected
policy issues”, chapter IV in Global Financial Stability Report, March.
——— (2003b): “Annual report on exchange arrangements and exchange restrictions”,
August.
Jiang, G, A Lee, N Tang and E Law (2003): “Hong Kong dollar debt market development in
2002”, Hong Kong Monetary Authority Quarterly Bulletin, March.
JPMorgan (2002): “Local markets guide”, Emerging Markets Research, June.
Kadir, A (2001): “Capital market development in Malaysia”, in Bond Market Development in
Asia (OECD), pp 388-403.
Mihaljek, D, M Scatigna and A Villar (2002): “Recent trends in bond markets”, BIS Papers,
no 11, June.
Mitsubishi Securities (2003): “Endate Kisai Shijo Hikaku: Samurai Sai vs. Euroyen Sai”
(comparison of issuance: samurai bonds vs euroyen bonds), presentation before the Study
Group on internationalization of Japanese Capital Markets, May.
Park, Y and D Park (2003): “Creating regional bond markets in east Asia: rationale and
strategy”, paper presented at the 2nd Annual Conference of PECC Finance Forum, July.
Prasarn, T (2001): “Capital market development in Malaysia”, in Bond Market Development
in Asia (OECD), pp 404-418.
PricewaterhouseCoopers (2003): “A PRC business and tax review of the qualified foreign
institutional investors (“QFII”) Scheme”, February.
Reserve Bank of Australia (2003): “Bond market development in east Asia”, Reserve Bank
of Australia Bulletin, December.
Taniguchi, T (2003): Keynote address at ASEAN+3 high-level seminar Fostering Bond
Markets in Asia, March.
Tran, H and T Roldos (2003): “Asian bond markets: the role of securitization and credit
guarantees”, background paper prepared for the 2nd Annual Conference of PECC Finance
Forum, July.

BIS Papers No 30 259

Turner, P (2003): “Bond market development: what are the policy issues?”, paper prepared
for a joint World Bank/IMF/Brookings Institution conference on the Future of Domestic
Capital Markets in Developing Countries, April.
Watanabe, K, H Akama and J Mifune (2002): “The effectiveness of capital controls and
monitoring: the case of non-internationalization of emerging market currencies”, EMEAP
Discussion Paper, January.

260 BIS Papers No 30

 A survey on hedging markets in
Asia: a description of Asian derivatives
markets from a practical perspective

Martin Hohensee and Kyungjik Lee

Introduction
Since the Asian crisis, there has been tremendous growth in Asian bond markets that, in
many cases, has not been matched with commensurate growth in derivative products. For
those markets that grew out of the financial sector recapitalisation prior to the middle of 2003,
there was precious little experience of bearish bond market movement, and therefore
perhaps inadequate appreciation of the importance of hedging instruments at that time.
However, while liquid hedging instruments may not be strictly necessary for the formation of
a bond market, they are recognised as critical to its long-run success. The movement of
Asian central banks to better develop fixed income markets should therefore also endeavour
to strengthen the development of derivatives markets. As a first step in this effort, we present
below a description of the current state of hedging markets in the region.

The current state and future potential of Asian derivatives markets

Many Asian currency and interest rate derivatives markets are still in the very early stages of
development, while others boast a relatively broad range of derivative products. We find that
there exists a strong inverse relationship between market sophistication and regulatory
restrictions. The two top financial centres in the region, Hong Kong SAR and Singapore,
undoubtedly have the most advanced derivatives markets with the least regulation, while at
the other end of the spectrum are China and Indonesia who lag behind most of their Asian
neighbours. Other countries fall somewhere in between, as depicted below.

Level of restrictions vs market sophistication

Hong Kong SAR

Singapore
Market sophistication

Korea
Taiwan, China
India
Thailand Malaysia
Philippines
Indonesia
China

Level of restrictions

Source: DB Global Markets Research.

BIS Papers No 30 261

The development of the underlying bond market is surely one of the key factors in the growth
of the derivatives market. Since the financial crisis, Asian local bond markets have grown
rapidly and the size of nine East Asian local bond markets was estimated to be USD 1.2trn at
the end of 2002. In comparison with the more developed economies, however, bond markets
remain small. The total East Asian bond market, in fact, amounts to only about 20% of the
Japanese market. The average size of the bond market in those countries was 46% of GDP,
compared to 169% in Japan and 156% in the United States, suggesting continued high
growth potential in the years to come. 1 In order to grow an institutional investor base,
particularly an international one, the depth and breadth of the bond market will need to
increase, inevitably entailing improvements in the liquidity and diversity of available hedging
tools. Regulators in many countries have been moving toward increasing support for
derivative products, but there are still apparent inconsistencies in policy approaches. In many
countries, the ability to maintain currency stability has been an important regulatory
consideration since the 1997 Asian crisis. Different policy reactions to the currency
speculation during the crisis and the subsequent high level of FX volatility have set the pace
for the capital market liberalisation. For example, under the strict IMF program, Korea has
been aggressive in liberalising capital flows while Malaysia reversed its policy to a tightly
controlled capital account regime. 2 Although several countries still maintain strict FX
regulations, strong regional recovery and rapid accumulation of the FX reserves bode well for
further liberalisation of the Asian currency markets.

Asian FX reserves
USDbn
300
China Taiwan, China
Hong Kong SAR South Korea
250 Singapore India
Thailand Indonesia
200 Malaysia Philippines

150

100

50

0
1990 1992 1994 1996 1998 2000 2002

Source: CEIC and IMF.

FX derivatives
The degree of regulatory restrictions varies widely across the region; however, the common
purpose of FX control is to protect the local currency from speculation. Therefore, the most
frequent form of FX regulation is to impose hedging requirements on the FX trades. The
authorities often apply stricter rules to non-resident trades. Any restrictions on free capital

1
ADB (2003).
2
Park and Bae (2002).

262 BIS Papers No 30

mobility can create inefficiencies by not allowing market participants to take full advantage of
pricing differences. This is unfortunate since foreign participation could be an important
catalyst for the growth of Asian bond and derivatives markets, especially in the early stages.
Since restrictions on FX derivatives transactions have a direct result on the formation of other
derivatives markets, they represent an appropriate place to start our survey of derivatives
markets.

Hong Kong
Despite the fixed exchange rate regime, Hong Kong has the most liquid and efficient
FX market in the region. There are no restrictions on the onshore FX derivatives market in
Hong Kong. Foreigners can freely trade in the onshore market, so there has not been a need
for a non-deliverable forwards (NDF) market in HKD. The FX forward market is used to
hedge currency exposure on liabilities and assets, as well as to speculate on the yield curve
movements. The market is very liquid up to one-year maturity. The one-year USD/HKD
forward, which generally reflects market sentiment on the currency peg, is watched closely
by the market and the Hong Kong Monetary Authority (HKMA). Despite the fixed exchange
rate, the one-year USD/HKD premium has actually been quite volatile in recent years. Local
economic news, comments from senior officials in Hong Kong or China on the subject of the
linked exchange rate, and comments from international rating agencies are some of the
factors that have influenced the forward premium in the past. In 2003 alone, it went from
+350 pips when SARS hit Hong Kong in early 2003, to below –600 pips because of a flood
on liquidity into the economy following the September 2003 G10 meeting which increased
expectations of a regional currency realignment.

One-year USD/HKD forward premium

pip
600 One-year USD/HKD premium

400

200

–200

–400

–600

–800
Apr 00 Oct 00 Apr 01 Oct 01 Apr 02 Oct 02 Apr 03 Oct 03

Source: DB Global Markets Research.

Singapore
The Monetary Authority of Singapore (MAS) has continued to drive the expansion in the FX
derivatives market through liberalisation. For example, in 2002, the MAS revised Notice 757
to allow non-residents to transact freely in SGD FX options without filing documentation on
the purpose of each transaction. However, a few basic restrictions still apply to protect the
currency from speculation. For example, banks’ lending to non-resident financial institutions
cannot exceed SGD 5m per entity.

BIS Papers No 30 263

Korea
In 1999, Korea lifted most of the existing restrictions on the FX market, but has recently
reversed some of this liberalisation. Prior to 1999, all forward FX transactions required
certification as hedges against expected current account transactions. The onshore forward
market is highly liquid out to one year, but supporting documentation is required if there is to
be a physical delivery. Unlike Singapore or Hong Kong, Korea maintains an NDF market
where onshore banks are also permitted to participate with other banks or corporates. No
prior reports or approvals are required for NDF transactions. The regulatory environment
suddenly tightened in January 2004 in order to discourage offshore investors’ speculative
demand for KRW. The government put a ceiling on domestic financial institutions’ NDF
positions with non-residents. Local financial institutions were originally banned from
increasing net long dollar positions via NDF trades by more than 10%, or decreasing net
short positions by more than 10%, based on their position on 14 January and 16 January
2004 respectively. Later, the restriction was mitigated for net short dollar positions. For FX
options, the up-front premium should be under 20% of option notional amount, otherwise
prior approval from the Bank of Korea (BoK) is required. If there is to be a physical delivery at
the options maturity, the underlying trade documentation must comply with regulations on
forward transactions. The currency options market is also liquid, volume of about USD 250m
traded in the non-deliverable options (NDO) market.

Taiwan, China
The primary goal of the Central Bank of China (CBC) is to prevent speculative activity in the
currency market to maintain the financial stability of Taiwan, China (hereinafter referred to as
Taiwan). Maintaining export competitiveness against countries such as Japan and Korea is
also a major concern for the authorities. The CBC manages the currency via capital controls
and direct intervention, which is supported by its substantial foreign exchange reserves.
Onshore FX spot, forwards and options trades must be explicitly approved by the relevant
regulatory authority. Non-residents are not allowed to access the onshore deliverable market,
preventing arbitrage flows in the onshore FX and interest rate markets. Liquidity in the NDF
market is relatively good, but onshore banks are not allowed to book an NDF trade with
onshore corporates.

Thailand
In January 1998, the Bank of Thailand (BoT) lifted a number of measures imposed previously
to stem FX speculation. In particular, domestic financial institutions were allowed to engage
in spot transactions involving THB with non-residents so that the onshore and offshore spot
THB FX markets became reunited. To safeguard against speculation in the FX market,
however, credit facilities provided by each financial institution to non-residents where there is
no underlying trade or investment activity in Thailand are subject to a maximum outstanding
limit of THB 50m per counterparty. There are no size restrictions on hedging short THB
forward positions, however. In an effort to curb speculative capital inflows, the BoT recently
introduced new measures. Effective as of 12 September 2003, onshore banks were banned
from borrowing more than THB 50m from each non-resident without underlying transactions
for less than three months, including transactions similar to borrowing such as buying FX
forwards and selling/buying FX swaps. On 14 October 2003, additional measures were
introduced to prevent THB speculation. The key changes were: (i) the outstanding
non-resident THB balance cannot exceed THB 300m per account and (ii) non-resident THB
accounts are locked in for at least six months without interest.

Malaysia
In Malaysia, strict capital controls were imposed in September 1998 to insulate Malaysia
from the financial crisis. The government still maintains some strict regulations on the

264 BIS Papers No 30

currency market. Hedging of foreign exchange exposure must be related to trade activities
and cannot exceed a tenor of one year. Non-residents can only access the onshore forward
market for the purpose of purchasing MYR securities listed on the KLSE. Financial hedging
(for example, hedging of profit repatriation, loan payment) is not allowed without Bank
Negara Malaysia (BNM)’s prior approval. An NDF market is not regularly available for MYR.

Indonesia
Under the IMF program, Indonesia’s FX policy has been directed at preventing extreme volatility
in the currency. The Bank Indonesia (BI) tightened capital control in January 2001 by issuing
regulations prohibiting IDR transfers to offshore entities, unless supported by underlying trade or
investment transactions. These restrictions quickly spawned the development of an NDF market
starting in February 2001. Main hedgers are corporates, but hedge trades are somewhat
uncommon, due to high negative carry.3 Liquidity in the onshore and offshore market is generally
low. With the exit from the IMF program at the end of 2003, and inflation trending lower, the Bank
Indonesia may move back to a more liberal exchange rate regime.

Philippines
In the Philippines, Banko Sentral Ng Pilipinas (BSP) operates a managed floating exchange
rate regime in which capital controls, close observation of market positions and intermittent
interventions are used to prevent speculative activity. Corporates are the main hedgers for their
USD loans. Liquidity in both the spot and forward markets is generally quite low. Poor liquidity
and instability in the Philippines have led to squeezes in the peso market. The basic principle
of the BSP in managing foreign exchange is that, outside of the banking system, foreign
currency may be freely bought and sold against the peso. Hence, there is no prohibition
against, for example, exporters selling their USD directly to importers or even to private
investors. FX regulations therefore focus primarily on bank transactions, and the specific rules
and restrictions depend on the nature of the transaction as well as the type of counterparty. In
general, onshore banks may buy foreign currency from both onshore and offshore
counterparties, without prior BSP approval or any documentation requirements. An onshore
interbank FX options market does not exist and even the offshore market is mainly inactive.

India
Historically, India experienced substantial ongoing fiscal deficits and persistent current
account deficits. Consequently, the Reserve Bank of India (RBI) maintained capital controls
in an attempt to prevent speculative activity in the rupee market. In recent years, however,
the underlying economic conditions have changed dramatically. India now runs a surplus
current account and substantial capital inflows have allowed foreign exchange reserves to
balloon. The RBI has therefore eased several restrictions and allowed FX derivative trades
for hedging purposes, and the trend towards further flexibility is expected to continue. Most
recently, RBI decided to permit foreign currency/rupee options as of 7 July 2003 in order to
further develop the derivatives market in India and expand the spectrum of hedging products
for currency exposure. Authorised dealers can offer plain vanilla European options and
customers can purchase call or put options. The writing of options by customers is not
permitted. As in the FX spot and forward markets, customers who have genuine foreign
currency exposures are eligible to enter into options contracts, and authorised dealers can
use the product for the purpose of hedging trading books and balance sheet exposure. FX
options trading is still thin, however.

3
The academic evidence strongly rejects the hypothesis that the forward exchange rate is an unbiased
predictor of the future spot rate. Currencies that trade at a forward discount, on average, weaken less than the
amount implied by the forward discount, which is known as the “forward rate bias”.

BIS Papers No 30 265

266

Table 1
A summary of Asian FX derivatives markets

HKD SGD KRW TWD THB MYR INR IDR PHP CNY

FX forward
Restriction None None Minimal Prior Hedging Onshore Hedging Offer Offer Restricted
approval only only only restricted restricted
Liquidity Good Good Good Good Good Good Good Average Average
Trade size USD 20-50m USD 25m USD 10-20m USD 1-10m USD 5-20m USD 10-20m USD 1-5m USD 0.5-30m USD 2-5m
B/as spread 0.0005-0.0010 0.0002-0.0010 0.3-1.0 0.002-0.020 0.01-0.100 0.0005 0.03-0.05 10-80 0.02-0.25
Daily volume USD 3-5bn USD 6bn USD 1bn USD 500-700m USD 300-600m USD 100-200m USD 150m USD 200m USD 75m

FX options
Restriction None None Minimal Prior Offer
approval restricted
Liquidity Good Good Good Good Hedging Hedging Average
only Market only Market Market
non-existent non-existent non-existent
Trade size USD 25-50m Up to USD 20-50m USD 20m No interbank Just allowed USD 0.5-3m
USD 30m
B/as spread 0.2-0.5 vol 0.5-1.0 vol 1 vol 0.3-1.0 vol 5-7 vol
Daily volume USD 50-100m USD 50 m USD 50m USD 50m USD 20m

NDF
Restriction None Interbank None None None Interbank
only only
BIS Papers No 30

Liquidity Market Market Good Good Market Market Good Average Average Good
non-existent non-existent non-existent non-existent
Trade size USD 5-10m USD 3-10m USD 3-5m USD 3m USD 2-5m USD 5-10m
B/as spread 0.5 won 0.03-0.05 0.05-0.20 20-100 0.15-0.50 0.001-0.005
Daily volume USD 1bn USD 300m USD 20-50m USD 40-60m USD 20-30m USD 50m
BIS Papers No 30

Table 1 (cont)
A summary of Asian FX derivatives markets

HKD SGD KRW TWD THB MYR INR IDR PHP CNY

NDO
Restriction None Interbank None None None None Interbank
only only
Liquidity Market Market Good Good Good Market Average Average Average Average
non-existent non-existent non-existent
Trade size USD 20-50m USD 30m USD 20m USD 5m USD 5m USD 5m USD 10-20m
B/as spread 1 vol 0.7 vol 1 vol 2-4 vol 7 vol 5 vol 1-2 vol
Daily volume USD 250m USD 150m USD 50m Irregular Irregular Irregular USD 250m
Source: DB Global Markets Research.
267
China
Foreign exchange policy in China is determined by the People’s Bank of China (PBOC) and
managed by the State Administration for Foreign Exchange (SAFE). Despite becoming a
major player in global commerce, the Chinese economy in aggregate remains relatively
closed. However, strong global pressure on its currency will likely lead to a more flexible
currency regime in China sooner or later. Although China prefers gradual reform, the officials
made it clear that they wish to liberalise the capital account with the goal of developing a
sophisticated financial sector and an independent monetary policy. Currently, designated
onshore banks and authorised foreign banks are allowed to participate in the FX spot market.
For current account items, relevant documentation is required for transactions of over
USD 1m cumulative amount, while capital account items require pre-approval from the local
SAFE. Only the Bank of China can offer CNY deliverable forwards with the tenor out to 6M.
An active market for CNY NDFs exists, however, onshore corporates are not allowed to take
part in the market.

Exchange-traded rate futures markets

Appreciating the importance of adequate hedging instruments for dealers and investors,
particularly in bear markets, many countries in the region have devoted effort to developing
exchange-traded derivatives markets. As of March 2004, six Asian countries have interest
rate and/or bond futures markets, namely Hong Kong, Singapore, Korea, Taiwan, Malaysia
and India. These efforts have met with varying degrees of success, as described below.

Hong Kong
The three-year HKD bond futures is one of the few that is settled by physical delivery (the
other being the 10-year TWD contract). Exchange fund notes (EFNs) issued by the HKMA
are reference instruments for the contract. On the Hong Kong Futures Exchange one-month
Hibor futures, three-month Hibor futures and three-year EFN futures are traded. The liquidity
in each of these contracts has declined in recent years, disappointing expectations, probably
due to limited foreign participation and already good liquidity in the OTC derivatives market.

Table 2
Open interest for HKD interest rate futures

One-month Three-month Three-year

Hibor futures Hibor futures EFN futures

End-2001 750 58,830 375

End-2002 20 13,806 0
11 March 2004 55 6,172 300

Source: Hong Kong Exchanges and Clearing Limited.

Singapore
Currently, there are two exchange-traded interest rate futures products in Singapore: three-
month swap offer rate (SOR) futures and the five-year bond futures. The SOR is an FX
forward-implied interest rate calculated from three-month USD/SGD forwards, and an official
fixing is provided by the Association of Banks in Singapore. Trading in three-month SOR

268 BIS Papers No 30

futures has been sporadic, and the market has been inactive recently with open interest only
at 5,000 contracts in 2002. Possible reasons could be: (i) local corporates are generally
cash-rich and do not need to borrow much short-term cash, and/or (ii) with rates trending
lower for the past couple of years, hedging borrowing needs has not been a real concern.
The five-year bond futures market was launched in June 2001, but the trading has slowed to
nearly nothing. One explanation may be that the investor base in government bonds is
skewed towards banks, who have strong credit and ample access to OTC derivatives. In
recent years, the basis risk between bonds and swaps has been small, and spreads have
been tight. Hedging in the OTC market therefore has provided an efficient, liquid and cost
efficient hedge.

Taiwan
The Taiwan Financial Futures Exchange introduced government bond contracts only in early
2004 but these have not taken off. Whether the reason is the sharp liquidity and accessibility
differences between on- and off-the-run bonds, reflecting institutional investors’ propensity to
hold bonds to maturity, high transaction costs or other factors, is not clear. Neither a
narrowing of the maturity range for deliverable securities nor the removal of the penalty for
cash settlement has improved liquidity. At the short end of the maturity spectrum, the futures
contract based on 30-day commercial paper has also fizzled. In this case, the choice of
30- rather than 90-day paper seems unfortunate in retrospect. The floating leg of interest rate
swaps, for instance, is based on 90-day yields.

Malaysia
In March 2002, the Malaysian Derivatives Exchange (MDEX) launched Malaysia’s first bond
futures contract, with the five-year government bond being the underlying security. The
contract has similar specifications to the five-year bond futures in Singapore, using a basket
of bonds in pricing and being cash settled. Activity in the contract is not great with average
daily volume or open interest rarely moving above 1,000 contracts. The use of the futures
contract as a hedge for cash bond positions is small. Other futures products in Malaysia are
three-year and 10-year bond futures, launched in September 2003, and three-month KLibor
futures, launched in March 1996. As in the case of the five-year bond futures, liquidity
remains poor.

Singapore five-year bond futures Malaysia five-year bond futures

Monthly volume 25,000 Five-year MGS futures: open interest

25,000 Five-year SGS futures monthly open interest Five-year MGS futures: volume

20,000
20,000

15,000
15,000

10,000 10,000

5,000 5,000

– –
Apr 01 Nov 01 May 02 Dec 02 Jun 03 Jan 04 Feb 02 May 02 Sep 02 Dec 02 Mar 03 Jun 03 Oct 03 Jan 04

Source: DB Global Markets Research. Sources: Bloomberg; DB Global Markets Research.

BIS Papers No 30 269

Korea
The shining success story among the exchange-traded futures markets is clearly the cash
settled three-year Korean Treasury bond (KTB) futures contract. To some degree, the
success of the three-year bond futures in Korea could actually be attributable to lack of
transparency in the bond market and limited availability of alternative hedging tools. Over
time, improved liquidity in the underlying bond market and strong foreign interest have
effectively accelerated the growth of the futures market. According to the World Bank, 4 the
three-year KTB futures is the sixth most active interest rate futures contract by trading
volume. Banks and ITCs are the two main participants in the market with a combined share
of 50-60%. Despite the fact that foreigners trade only around 10% of the total volume, they
seem to play an important role in catalysing the development of the market. Their activities
are closely monitored by domestic players, since they are often viewed as more
sophisticated. Foreigners may find it easier to trade futures than cash bonds due to taxation,
leverage and liquidity issues. The fact that foreigners’ net open interest is as much as 30% of
the market compared to less than 1% holding in cash bonds would strengthen the argument.
In value terms, foreign net open interest typically ranges between 0.5 to 2.0% of the
government bond market.

Table 3
Foreign participation in Korean three-year bond futures

Foreigners’ net Typical market Value of foreigners’ Cash bond % to cash

open interest open interest net open interest outstanding market

10-30% 60,000 contracts KRW 0.5-2.0trn KRW 90trn 0.6-2.2%

Source: Korea Futures Exchange.

Three-month CD futures, one-year MSB futures and five-year KTB futures are also listed on
the Korea Futures Exchange (KOFEX), but they are totally inactive at the moment.

India
India launched 10-year bond futures in June 2003. One unique feature is that the settlement
price is based on the value of the notional bond, derived from the zero-coupon curve which
the National Stock Exchange of India (NSE) publishes every day. To date, this product has
been a failure. In fact, in less than three months after the launch, trading in bond futures
literally stopped. There may be several ways to explain this. First, the unique settlement price
calculation adds complication, and is seen by many investors as an opaque “black box”.
Moreover, the basis risk between the NSE curve and the cash prices could also be
substantial and unpredictable. Finally, restrictions on short selling and requiring financial
institutions to use derivatives only for hedging purposes could account for the inactivity of the
product. In this case, the absence of speculators may be cheating the market out of badly
needed liquidity.

4
OECD-World Bank Annual Bond Market Forum, 3 June 2003, Emerging Derivative Markets presented by
Oliver Fratzscher.

270 BIS Papers No 30

 Table 4
Indian bond futures trading

Jun 03 Jul Aug Sep Dec Feb 04

Number of days 5 23 20 22 22 19
Number of contracts
traded 9,768 963 50 0 0 0
Source: National Stock Exchange of India.

Taiwan
The Taiwan Futures Exchange (TAIFEX) launched 10-year bond futures in January 2004.
The contract is settled with physical delivery of bonds maturing between seven and 11 years,
much like the very successful 10-year Japanese Government bond futures contract. In order
to prevent short squeezes on the cheapest and to deliver bonds around the final settlement
date, TAIFEX simultaneously opened a Bond Lending Center. TAIFEX hoped that by
providing effective hedging tools, the launching of bond futures would attract more foreign
funds into the Taiwan bond market. This may eventually take place; however, the initial
market reaction was not very enthusiastic. So far, open interest has stayed below 1,000
contracts.

OTC interest rate derivatives 5

Interest rate swaps

Hong Kong
Based on the most recent survey done by the Hong Kong Monetary Authority (HKMA),
interest rate swaps (IRSs) represent the largest segment (over 73%) of interest rate
derivative transactions in Hong Kong. With about HKD 9.1bn in daily turnover, the liquidity in
the IRS market is far better than the Exchange Fund Note (EFN) market whose daily
turnover is about HKD 2.5bn. HKD swap spreads are among the widest in Asia. While EFNs
are consistently well supported by strong demand from local banks, HKD swaps are much
more sensitive to the volatile risk premium of the currency peg. In addition, the 10-year HKD
swap spread tends to be persistently tight, especially right after the issuance of a new 10-
year EFN. In order to reduce the interest cost of the 10-year paper in an environment of
upward sloping yield curve, the HKMA has consistently swapped the 10-year EFN into a
floater after the issuance, creating downward pressure on the 10-year part of the HKD swap
curve.

Singapore
Interest rate swaps are also more liquid than government bonds in Singapore. The floating
leg of an IRS trade is the swap offer rate (SOR) posted by the Association of Banks in
Singapore. (It is an FX-implied rate, in contrast to a Libor-type fixing in typical G7 trades.
Many countries in Asia are following this model for developing their own interest rate swap

5
Cross-currency swap, a hedging instrument for FX risk, is included here due to its similar trading patterns.

BIS Papers No 30 271

market, for lack of a well developed money market.) Corporations are the main hedgers in
the swap market, while hedge funds often come in for speculation as well as relative value
investments. The MAS has recognised that hedge funds can offer diversification from more
traditional funds, and help Singapore develop into an international financial centre. This
recognition led to the MAS introducing legislation to allow retail hedge funds to be offered to
the local market. Swap spreads have been generally market-directional and more so for
long-dated tenors. In other words, swap spreads tend to widen when rates rise and tighten
when they fall. There also exists an overnight interest rate swap (OIS) market, with the first
trade done in April 2000. In 2002, the government removed the restrictions in the cross-
currency swap market. Previously, cross-currency swap could not be traded without
economic activity associated with the trade.

Korea
During the first half of 2003, interest rate swaps held 88% share of total OTC interest rate
derivatives trades in Korea. 6 However, compared to bond futures, turnover of interest rate
swaps is relatively small and volatile. Very recently, daily trading volume has decreased to
less than KRW 0.3trn.

Table 5
Daily turnover for KRW bonds, futures and swaps
KRW trillion

Government Three-year bond Interest rate

All bonds
bonds futures swaps

2002 5.9 2.0 2.3 0.8

1H 2003 6.6 2.3 2.5 0.9
Sources: Financial Supervisory Service; KoreaBondWeb.

The relative illiquidity of swaps partly stems from the underdeveloped short-term benchmark
rates in Korea. Interest rate swaps use the three-month CD rate as the floating rate, but CDs
are not liquid in the secondary market. The BoK plans to introduce Libor-type interbank rates
in 2H 2004, which will help step up the liquidity of swap trades. IRS trades up to 10 years
with a bid/ask spread of 5 bps up to five years and 10 bps for longer tenors.
Trading in OTC derivatives increased in 2001, fuelled by the increased interest in structured
products. Since then, in order to enhance returns, investors welcomed structured products
whose performance was linked to future market direction or the shape of the yield curve. The
most popular products were dual index floaters and inverse floaters, based on constant
maturity treasury (CMT) or constant maturity swap (CMS) rates as the reference rates.
The Korean IRS market is highly driven by technicals that may not be reflected in the cash
market. The spread between bond yields and swaps (the bond/swap basis) can be quite
volatile, as seen in the chart below. Swaps, therefore, can be a relatively risky hedging tool
for investors.
A USD/KRW cross-currency swap (CCS) can be structured out to 10 years, but one to five
years are more actively traded tenors. Exporters’ USD forward selling and the central bank’s

6
Financial Supervisory Service, internal data.

272 BIS Papers No 30

FX intervention are the main drivers in the short end of the cross-currency curve, while
external debt issuers and investors generally trade in the long end.
The spread between the CCS and IRS rates (the IRS/CCS basis) has also been volatile in
recent years. For the past couple of years, the CCS market has been dominated by domestic
investors’ KRW-receiving demand to swap their offshore investment into KRW. Many swap
dealers who could not hedge their position in the CCS market hedged their position in the
relatively more liquid IRS market, running this IRS/CCS basis risk in order to increase market
share with asset swapping investors. With continued receiving demand in the CCS market,
however, dealers were eventually squeezed out of this basis risk, in late February 2003. As a
result of massive unwinding demand, - paying IRS and receiving CCS - five-year IRS/CCS
basis had widened to almost 200 bps before it came down to more reasonable levels.
Foreigners can participate in the swap market using non-deliverable IRS or CCS and current
participation is estimated to be 15%.

Korea five-year swap spread Korea five-year IRS/CCS basis

bp bp
50 200
Five-year swap spread Five-year IRS/CCS basis
40
30 150
20
10 100
0
–10 50
–20
–30 0
–40 Dec 00 Jun 01 Dec 01 Jun 02 Dec 02 Jun 03 Dec 03
Jan 02 Jul 02 Jan 03 Jul 03 Jan 04

Source: DB Global Markets Research. Source: DB Global Markets Research.

Taiwan
The Taiwan swap market has been affected by some of same dynamics that have affected
Korea’s. Interest rate swaps trade up to 10 years and the floating rate index is the Taiwan CP
rate. Currency swaps are quoted against six-month US Libor and the approval from CBC is
required for onshore entities to transact in the onshore CCS market. In recent years, with
interest rates declining, investors increased buying offshore debt and swapped cash flows
into TWD to achieve their yield targets. These activities, combined with multilateral entities’
swapping TWD-denominated debt into major currencies, tended to widen out the IRS/CCS
basis. Foreigners can participate in non-deliverable IRS or CCS market where cash flows are
net settled in USD, although liquidity is generally low.

Thailand
In Thailand, the dynamics of the bond/swap basis is slightly unusual. First, the onshore IRS
market adopts the onshore FX forward-implied yield (THBFIX) as the floating rate index. The
basis between onshore CCS and IRS curves is thus minimal, and any cross-currency asset
or liability swapping activity, therefore, directly affects the government bond/IRS basis. Since
the cross-currency swap uses USD Libor as its floating rate index, the onshore bond/swap
basis should theoretically be exactly equal to the Kingdom of Thailand’s USD swap spreads.
Following this theoretical argument, the THB IRS curve should be below the government
bond curve to reflect a positive USD Libor margin for the Kingdom. However, this turns out
not to be the case. An asset swap of Thailand’s offshore 2007 bond has always offered
onshore investors with a yield enhancement opportunity over the onshore market.

BIS Papers No 30 273

This is because the linkage between onshore and offshore capital markets is tightly
regulated. Onshore investors have only a limited ability to asset swap USD paper into THB
on a case-by-case basis, while foreign entities are currently prohibited from issuing bonds in
the onshore market. Only a few state-owned enterprises have been allowed an opportunity to
liability swap THB borrowings into USD. With only limited ability to engage in asset/liability
arbitrage, the market cannot force a convergence of credit pricing in THB and USD. As long
as cross-border restrictions on issuers and investors exist, those who are granted permission
will systematically have an opportunity to create cheap THB assets from USD assets, or
cheap USD liabilities from THB liabilities. The easing of the regulatory environment has been
slow so far.

KoT 2007 swapped into THB minus

interpolated onshore government bond yield
bp
450
400
350
300
250
200
150
100
50
0
Aug 00 Feb 01 Aug 01 Feb 02 Aug 02 Feb 03 Aug 03 Feb 04

Source: DB Global Markets Research.

Malaysia
With strict government regulation of the currency market, a cross-currency swap market does
not exist in Malaysia. The only active OTC derivative product in the onshore market is the
interest rate swap. Liquidity has improved over the years, but the average trade size is still
small at MYR 10m. The most liquid part of the curve is up to three years with a bid/ask
spread of 5 bps. Commercial banks, finance companies and merchant banks are the
predominant users of the swap market. Given the illiquid nature of the long-term fixed rate
assets such as commercial loans, banks with an expectation of higher interest rates are likely
to hedge in the swap market rather than to sell such assets outright.

Indonesia
In Indonesia, there is no developed IRS market trading onshore at the moment. However,
banks are interested in hedging their liabilities, from fixed to floating or floating to fixed
depending on their situation. As a result, some swap trades were done using the IDRFIX 7
and three-month sale offering of the Bank Indonesia Certificate (SBI) rate as floating rate.

7
IDR interest rate fixing = {([1+(FWD/SPOT)] x [1+(SIBOR x DAYS/360)] – 1} x 360/DAYS
where FWD = average of the offered side of the FX forward points from banks of each day
SPOT = average of mid spot rate provided by each bank
SIBOR = US interest rate for the respective tenor derived from Sibor
DAYS = number of days for each tenor.

274 BIS Papers No 30

Philippines
No offshore IRS market exists for the Philippines yet, but CCS does trade offshore, albeit
with low liquidity. Onshore IRS was launched on 18 August 2003 among six foreign banks
(Deutsche Bank, HSBC, Standard Chartered, Citibank, JP Morgan and ING). The market is
still very illiquid with trading volume of around PHP 200m a week and a bid/ask spread
ranging between 50-70 bps. But by the end of the first half of 2004, at least four local banks
are expected to join the market, which should improve the depth of the market. Standard
tenors are one to five years and the typical trade size is PHP 50m. The floating leg is the
three-month PHIREF or Philippine Interbank Reference Rate. The PHIREF is the implied
peso rate derived from all dealt USD/PHP swap and forward interbank transactions.

India
In India, interest rate swaps were first allowed in July 1999. Since then, the OTC
derivatives market has grown rapidly. Major hedgers are corporates and financial
institutions (FIs). In the past, most hedging came from large borrowers converting fixed
debts to floating. Historically, one of the major hurdles in developing the IRS market has
been the lack of a liquid money market curve. Due to limited credit appetite and the capital
constraints that banks face, the overnight market remains the deepest and the most liquid
in the short end of the curve. The overnight rate has also been the most widely accepted
benchmark for floating rate bond issues. As a result, overnight index swap (OIS) with the
floating rate indexed to the overnight NSE Mumbai interbank offer rate (Mibor) was the
most natural market to develop, and today enjoys high liquidity. Another interest rate swap
indexed to the three-month or the six-month Mumbai interbank forward offer rate (Mifor) is
also highly liquid, in fact far more liquid than OIS with 85% of the total swap trading
volume. Although all of the cash flows are settled in INR, Mifor swap effectively becomes a
CCS trade since the Mifor is essentially FX implied yield derived from onshore USD/INR
forwards.
An increasing number of banks, primary dealers and corporations are actively participating
in the two swap markets. Both the OIS and Mifor curves are active up to five years with a
bid/ask spread of around 5 bps up to five years and 15-20 bps for longer tenors. Lack of
participation from large players, such as public sector banks (PSBs), mutual funds and
insurance companies, is often pointed to as the hurdle to further development. In addition,
corporations can use OTC derivatives only for hedging purposes while no such restrictions
apply in the case of exchange-traded derivatives. One specific case is that corporates
cannot cancel and rebook a currency swap.

China
Although offshore non-deliverable CCS trading is not entirely impossible for China, the
main issue in the onshore Chinese market is how to develop the underlying bond market.
At this point, neither an IRS nor a CCS market exists onshore.

BIS Papers No 30 275

276

Table 6
A summary of Asian IRS and CCS markets

HKD SGD KRW TWD THB MYR INR IDR PHP CNY

Interest
rate swap
Floating Hibor SOR Three- Three- THBFIX Three- O/n Mibor Three-
rate month CP month CP month month
Klibor PHIFEF
Floating Qtr-Act/365 Semi-Act/ Qtr-Act/365 Qtr-Act/365 Semi-Act/ Qtr-Act/365 Daily-Act/ Qtr-Act/360
legal basis 365 365 365
Fixed legal Qtr-Act/365 Semi-Act/ Qtr-Act/365 Qtr-Act/365 Semi-Act/ Qtr-Act/365 Daily-Act/ Qtr-Act/360 Market
basis 365 365 365 Market
non-
non-existent
existent
Active One- to One- to Two- to Two- to Two- to One- to One- to One- to
tenors 10-year 10-year five-year five-year five-year three-year five-year five-year
Trading HKD 200m SGD 10-30m KRW 10bn TWD 300m USD 10m MYR 10m INR 250m PHP 50m
size
Bid/ask 10bp 2-5bp 5-10bp 5-10bp 10bp 5-15bp 5bp 50-70bp
spread
Daily HKD 10-15bn SGD 500-700m KRW 100-300bn TWD 1-2bn USD 15-20m MYR 25m INR 4-6bn PHP 50m
volume

Cross-
currency
swap
Market Market Market
Floating Hibor vs Six-month Six-month Six-month Six-month Six-month Three-
BIS Papers No 30

non- non- non-

rate Libor USD Libor USD Libor USD Libor USD Libor existent USD Mifor existent month T-bill existent
Floating Qtr-Act/365 Semi-Act/ Semi-Act/ Semi-Act/ Semi-Act/ Semi-Act/ Qtr-Act/360
legal basis vs 360 360 360 360 365
Qtr-Act/360
BIS Papers No 30

Table 6 (cont)
A summary of Asian IRS and CCS markets

HKD SGD KRW TWD THB MYR INR IDR PHP CNY

Fixed legal na Semi-Act/ Semi-Act/ Semi-Act/ Semi-Act/ Semi-Act/ Qtr-Act/360

basis 365 365 365 365 365
Active Two- to Two- to Two- to Two- to Two- to Two- to One- to
tenors 10-year 10-year five-year five-year five-year five-year five-year
Market Market
Trading USD 30m SGD 20m USD 10m USD 10m USD 10m INR 250m USD 3-5m
non- non-
size
existent existent
Bid/ask 10bp 2-4bp 5-10bp 10bp 10bp 5bp 60-110bp
spread
Daily USD 30-50m Volatile USD 100m USD 20-30m USD 15-20m INR 10-12bn Volatile
volume

Source: DB Global Markets Research.

277
Interest rate options
According to a recent BIS report, interest rate contracts account for 71.8% of the global OTC
derivatives market. 8 Particularly, interest rate swaps remain by far the largest single group of
products in the OTC derivatives market, representing 56% across market risk categories. In
the G10 countries that the BIS report covers, interest rate options are also popular, with 16%
share in the total OTC derivatives market. However, except for a few countries such as Hong
Kong and Singapore, interest rate options rarely trade in Asia. Among interest rate options
products, swaptions are more actively traded in Asia than caps/floors, probably due to a lack
of reliable benchmark short-term rates. Hong Kong has the most liquid interest rate options
market in the region with average daily trading volume around HKD 500-600m. In Singapore,
since the interest rate volatility market has begun to pick up since 2002, rates are quoted
regularly in the broker market. The typical daily volume is about SGD 50-100m with foreign
participation at about 5-10%. Although the market quotes swaptions with expirations out to
five years, three-month and six-month swaptions are the most actively traded. Interest rate
options markets also exist for KRW, TWD and THB, but trades are infrequent and volatile.
KRW swaptions trade somewhere between two to three times a day to three times per week,
and the trade size ranges from KRW 50bn to over KRW 1trn. Average trade size for TWD
swaptions is TWD 1bn and trade happens about 10 times a month. For THB, trades are
highly infrequent, probably five times a month, and THB 1trn is the average size. 9

Credit derivatives 10
Since the mid-1990s, banks have been turning to credit derivatives to more actively manage
the concentration and correlation risk inherent in their loan portfolios. But before the Russian
default in August 1998, many investors viewed credit derivatives as a curious but highly
specialised and exotic corner of the bond market. When suddenly faced with the prospect of
deteriorating credits and bond market illiquidity worldwide following Russia’s default,
investors could see in a very tangible way the attraction of a market where one can buy
protection to reduce risk, and sell protection to diversify a dangerously concentrated portfolio.
According to the latest British Bankers Association (BBA) credit derivatives report
2001/2002, 11 the total notional outstanding for credit derivative products stood at
USD 1.19trn at the end of 2001. By the end of 2004, the total notional outstanding is
expected to increase to USD 4.8trn. Banks are the most important players in this market,
capturing 52% of the protection buying and 39% of the protection selling. Among a variety of
products, single name credit default swaps (CDSs) continued to be the most popular product
with 45% of the market share in 2001. Sovereign single-name CDSs represents the most
actively traded credit derivative instrument in emerging markets. There is an active broker
market for Asian sovereign CDSs. Two-way pricing is generally available for the most liquid
sovereign names in the one to 10 year range. The standard inter-dealer notional trade
amount tends to be USD 5m.
Credit derivatives provide many investment and hedging opportunities that are impossible in
the cash market. Some general uses of credit derivatives in emerging markets include:

8
BIS (2003).
9
An estimate from Prebon Yamane Hong Kong.
10
Xu and Wilder (2003).
11
BBA publishes its report every two years. The latest available numbers are for the year 2001.

278 BIS Papers No 30

• Given the difficulty in shorting bonds due to underdeveloped repo markets, hedging
through credit derivatives is especially attractive for emerging markets investors.
• Credit derivatives allow investors to exploit illiquid points in the yield curve and
provide access to points on the yield curve where bullet bonds do not exist in the
cash market.
• Investors can sell credit protection and earn a risk premium on an unfunded basis.
• Emerging markets investors can use credit derivatives to exploit relative value
opportunities, express a directional view on a particular market, take advantage of
changes in the shape and structure of the yield curve and efficiently monetise
certain risk characteristics or pricing inefficiencies prevailing in different markets.
• Banks can use credit derivatives to hedge loan portfolio exposure, resulting in a
more efficient use of credit lines and economic capital.
• Credit default swap premiums quoted in the dealer market provide a very accurate
indicative measure of sovereign credit risk. Debt originators can use default swap
pricing as a tool to assist in pricing new bond issues, while potential investors can
use default protection premiums as a benchmark for measuring the fair value of new
debt issues.
CDSs allow counterparties to buy and sell protection on the credit risk inherent in a bond,
loan or guarantee/swap counterparty exposure. A key benefit of CDSs is the ability to create
maturity and credit exposure that are unavailable in the cash market. Aside from CDSs,
credit-linked notes (CLNs), total return swaps (TRSs) and synthetic collateralised debt
obligations (CDOs) are the other major credit derivative products. CLNs can be viewed as a
securitised default swap, whose performance depends primarily on the performance of the
imbedded default swaps. CLNs appeal to investors who seek customised credit exposures
but cannot deal directly in the credit default swap market. TRSs are most widely used in
emerging markets where leveraged investors seek to synthetically own and fund high
yielding investments without the overhead of investing directly in the underlying assets. In a
synthetic CDO, securitisation and credit derivative technologies are combined to provide a
more efficient way to structure CDOs, where the reference pool is a portfolio of protection
selling positions through CDSs.

Table 7
Emerging market sovereign credit default swaps

Tier 1 Tier 2 Tier 3

Most liquid Less liquid Sporadic activity

Mexico, Brazil, Colombia, Chile, Peru, Panama, Poland, Ecuador, Qatar, Czech
Venezuela, Russia, Turkey, Hungary, Croatia, South Africa, Republic, Israel, Romania,
Bulgaria, Philippines, Korea, China Lithuania, Slovenia, Morocco,
Malaysia, Thailand Tunisia, Egypt, Slovakia

Source: DB Global Markets Research.

An active credit derivatives market can improve the stability and efficiency of the financial
system by its pricing and diversification of credit risk. It also provides tools to securitise credit
risk that will help develop the overall bond market. The flip side is the potential moral hazard
of the protection buyers and the difficulty of the stakeholders in monitoring the activities. At
present, there are no developed local currency credit derivatives markets in Asia.

BIS Papers No 30 279

Conclusion
Growing local bond markets and the rapid accumulation of FX reserves highlight the potential
for active intra-regional local currency bond investment in Asia. For this purpose, the
availability of standard hedging tools surfaced as a particularly important factor to access
each bond market. So far, we have observed a couple of basic patterns in the growth of
Asian FX and interest rate derivatives markets.
• FX market liberalisation, although it should be the first step to facilitate cross-border
trades, remains incomplete in many countries. Many Asian central banks continue to
prefer the accumulation of FX reserves. Although underlying economic conditions
are ripe, the signs of further FX deregulation are yet to be seen in some countries.
Free capital mobility would be helpful in improving market efficiency by attracting
foreigners to the local markets.
• Exchange-traded futures are open in six countries, but except for the three-year
bond futures in Korea, liquidity is extremely limited. Two main issues are the lack of
a broad investor base (including foreign participation) and the market-unfriendly
structure of some products.
• In the OTC market, swap markets are either underdeveloped or inactive in many
countries. Reliability in reference rates and regulatory restrictions could be the main
reasons.
The choice between diversity and focus should be based on the maturity of each market.
While advanced markets like Hong Kong and Singapore may continue to develop a variety of
hedging tools, most other Asian countries may want to focus on a small number of basic
products. As mentioned before, interest rate options trade frequently only in Hong Kong and
Singapore. Although many Asian countries have recently launched an array of derivative
products, virtually none of them has good liquidity. Generally, investors would welcome a
simple and straightforward product. In those countries, focusing on a few liquid products
might be more effective in accelerating market development.
In this regard, the importance of exchange traded markets should not be underestimated, at
least for the initial stage. An exchange product can provide players with transparency and an
efficient price discovery process, and allow derivatives dealers to hedge their OTC positions
more efficiently. In addition, financial instability may arise in the OTC market when shocks
such as credit events and sharp movement in underlying asset prices occur. When
underlying asset prices change rapidly, the size and configuration of counterparty exposures
can become unsustainable and provoke a rapid unwinding of positions. In general, liquidity in
vanilla derivatives markets is likely to drive product innovation. Major benefits of exchange-
traded derivatives compared to OTC products are:
• no credit exposure with the settlement guaranteed;
• transparency and anonymity;
• lower transaction cost;
• less severe information asymmetry as OTC derivatives activities tend to concentrate
in major institutions.
Development of bond and derivatives markets are mutually reinforcing processes. By
definition, “derivatives” cannot exist without “underlyings”. The transparency and liquidity of
the underlying markets are the fundamental success factors for derivatives markets. A
reliable benchmark yield curve should be the main focus for nearly all Asian fixed income
markets, but probably with higher priority to China, Indonesia and the Philippines. Pricing of a
derivative product is subject to manipulation without solid reference rates, which will
subsequently limit the liquidity of the derivative product. Liquid short-term benchmark rates
are especially important for pricing swaps.

280 BIS Papers No 30

Finally, the problem of broadening the investor base is a common theme in most Asian
derivatives markets. In the case of Singapore and Hong Kong, liquidity in futures markets
remains low as there are other liquid OTC derivatives. If somehow more participants become
interested in futures, increased liquidity will likely bring more participants into the market in a
virtuous circle. In the futures market, the general issue is the lack of ability for market makers
or dealers to hedge their position in an efficient manner. For example, in Malaysia, if
investors were to sell dealers a futures position, there is no real way for the dealers to hedge
their long position by shorting it in the underlying bond market or in the swaps market, which
is only liquid out to three-years. Currently, investors in Malaysia are mainly banks and the
Employee Provident Fund (EPF), with life insurers and asset managers being small
investors. The Malaysian government may need to encourage hedge funds and more asset
managers to enter the market just as neighbouring Singapore did. In India, foreign
institutional investors (FIIs) are not allowed to trade in interest rate derivatives while they are
allowed to trade in equity derivatives. Moreover, large corporates who have the requisite
skills to manage or take risk are still subject to strict rules on hedging requirements, which is
hampering the improvement in market liquidity. Policy consistency is another important issue.
Many Asian governments aggressively introduced exchange-traded derivatives, but they also
maintain a variety of policies that discourage the demand for derivatives, such as capital
controls, restrictions on short sales and a lack of accounting or legal standards. In order to
bring investors into the derivatives market and facilitate sustainable market growth, the
authorities should consistently improve market infrastructure, enhance monitoring and
supervision systems and provide a fair and reliable regulatory environment.

References
Akhtar, S (2003): Asian financial markets emerging trends, The European Banking and
Insurance Fair, ADB, 28 October.
Asian Development Bank (2003): Harmonization of bond market rules and regulations,
August.
Bank for International Settlements (2003): OTC derivatives market activity in the first half of
2003, 12 November.
Fabella, R and S Madhur (2003): Bond market development in East Asia: issues and
challenges, January, ADB.
Fratzscher, O (2003): Emerging derivative markets, OECD-World Bank Annual Bond Forum,
World Bank, 3 June.
Hohensee, M et al (2003): Asia local bond markets, Deutsche Bank, June.
Park, Y C and K-H Bae (2002): Financial liberalization and economic integration in East Asia,
PECC Finance Forum Conference, August.
Redward, P (2002): Asian currency handbook, Deutsche Bank, August.
Xu, D and C Wilder (2003): Emerging markets credit derivatives, Deutsche Bank, May.

BIS Papers No 30 281

 Comments on Martin Hohensee and
Kyungjik Lee’s paper “Survey of
hedging markets in Asia”

Aaron Low

Current growth in global derivatives markets makes this a timely and important topic. The
authors have chosen well. With a diverse region and an increasing variety of instruments,
such a study is not easily undertaken. The paper is structured along different instruments
and makes specific assessments which render it easy to read like a good textbook. Rather
than addressing the specifics, which were amply discussed, I would like to concentrate more
on general market and development issues in the following sections. In particular, the first
section concerns market factors while the second section is focused on the regulatory impact
and implications. I conclude with a wish list of questions that may allow us to understand the
future direction of derivatives in Asia.

1. Hedging volumes and market factors

The growth in global derivatives has surpassed that in most other instruments in interest rate
space and has increased “spanning” in the global fixed income world. Given the fledgling
nature of Asian bond markets and the plain vanilla nature of Asian foreign exchange
instruments, it is not surprising that the authors find a great disparity in regional derivatives
market development and a general lack of open interest in futures contracts. It was also
emphasised that regulatory restrictions, a lack of market sophistication and low turnover in
the underlying instruments contributed to the slow pickup in derivatives activity. While these
are important factors, I would like to point out two more structural market drivers - correlation
and volatility.
Asian interest rates are highly correlated with US Treasuries, which is natural since monetary
policy has generally targeted currency levels. Foreign exchange trading volumes have
recovered slowly but are still driven by large regional currencies like the yen and the
renminbi. Not surprisingly, it is more efficient for large corporations to hedge these major
currency drivers (the high correlation across regional currencies makes it cheaper to hedge
exposures on a net basis and through the most liquid instrument) rather than independently
on a gross basis using less liquid Asian bilateral indirect cross rates. With increasing
convergence of monetary and fiscal policies, this trend in correlation looks likely to remain
intact.
The second factor, market volatility of both interest rates and currencies, has also diminished
since the crisis. A good recent example would be the Chinese renminbi. Before the G7 Dubai
meetings, we found that deal flows in the renminbi NDFs and Hong Kong dollar forwards
were relatively thin. This changed immediately after the Dubai meetings with a significant
pickup in activity with strong flows from both hedgers and speculators. Since then, flows and
positions have generally kept pace with volatility.
Also observed, as noted in the paper, is the important role played by the underlying
instruments. The recent success of the Korean swap markets arose from an increase in
corporate dollar issuance and the subsequent swapbacks into local currencies, which led to
further swap trading when cross-currency swap curve spreads widened. But I would disagree
with the authors that the potential for derivatives depends primarily on the liquidity of the
“underlying”. This is evidenced by the success of “non-underlying”-based instruments like

282 BIS Papers No 30

weather derivatives and the recently introduced economic derivatives. What is more
important in our view is the acceptance of the underlying reference rate or instrument.

2. Regulatory controls and restrictions

Differences in regulatory restrictions also affect the attractiveness of markets. The authors
make some comments on the differences in regulatory restrictions but could have gone
further in providing some assessment of costs and benefits, not necessarily quantitative, of
those regulations. Derivatives markets are generally supported by (but not restricted to)
several legs - hedging, speculative flows and arbitrage. Asian capital controls, some of which
were put in place to stem speculative flows, also serve to limit arbitrage possibilities and
leave hedgers as the main generators of flows. Understandably this makes liquidity difficult to
sustain, especially when hedgers tend to be on the same side of the trade (as in the case of
Korean corporations issuing dollar bonds and swapping back into won), creating price
distortions and reducing hedging efficiency.
Capital account closure is arguably beneficial in times of market stress, but it would be
difficult to argue that such closure would be beneficial if extended indefinitely. Rules and
regulations serve some market conditions well but do not provide first-best solutions all the
time.
Regulatory restrictions give rise to offshore instruments, which tend to circumvent some
regimes and thereby to extend market participation. In many cases, the offshore-onshore
spreads are indicative of market frictions with asymmetric volumes. This split in liquidity pools
certainly does not help in overall trading efficiency and price discovery, creating entrenched
clientele segments in the process. With further capital market development, such
segmentation will probably be reduced, as witnessed by the elimination of many of the Asian
A and B class equity share structures. It would certainly be interesting if the authors could
document examples of trade volumes before and after imposition and removal of restrictions.
Such an extension would provide an interesting assessment of the impact of specific market
rules and regulations.

3. Conclusions
The paper is an excellent survey of regional market structures for futures, swaps and credit
default swaps. There are good examples, highlighting the areas of success as well as the
instruments or countries that have lagged in performance or activity. However, the
arguments are rather dispersed, and the paper lacks a cohesive proposal on how we can
improve liquidity and participation for derivatives trading in the region. Should we concentrate
on a smaller but more focused subset of instruments to pool liquidity? Are markets too
fragmented to benefit from any returns to scale? Should more players be involved and who
should be targeted? Is there sufficient awareness or should there be more education in the
marketplace on the availability and suitability of such instruments for various types of
investors and traders? These are questions the answers to which may point to avenues for
growth. Ultimately, such growth may fortify the entire asset class.

BIS Papers No 30 283

 Clearing, settlement and depository issues 1

Francis Braeckevelt 2

1. Executive summary
Once neglected as a boring but necessary element of dealing in the capital markets, the
settlement process has caught the attention of both the public and the private sector.
The rise of emerging markets, the growth of financial markets, the increased focus on cross-
border activity and financial market deregulation in different parts of the world made the
settlement process considerably more complex but also made investors fully aware that
operational support systems form a critical part of an effective and efficient capital market.
Today, the creation of a robust clearing and settlement environment has become the topic of
many discussions and the recent technological innovation allowed for important
rationalisation, integration and consolidation trends to emerge. The current fragmented
infrastructure is increasingly perceived as a source of cost inefficiencies and significant risk.
As a result, new models are being developed that aim at mitigating risks and containing or
reducing costs.
In Section I, we highlight some of the current thoughts in today’s clearing and settlement
debate, the generally accepted roles and responsibilities of financial intermediaries, as well
as recommendations made by various national, supranational or private organisations on
how to develop robust and efficient clearing and settlement systems.
In Section II, we assess some selected central securities depositories in Asia. The most
noteworthy observations we have made with regards to the settlement environment in Asia
can, in our opinion, best be summarised as follows:
• The main barriers to developing an efficient bond market in Asia are mainly related
to the trading environment (ie liquidity constraints and foreign investor restrictions)
rather than to infrastructure issues.
• Today’s clearing and settlement infrastructure in Asia is very fragmented. Even
though the current setup operates well and conforms with the criteria outlined in the
US Investor Act of 1940, the infrastructure is not cost efficient and does not mitigate
risks in the settlement process in a comprehensive manner.
• Because of the central role central securities depositories play, it is important that
the intermediaries be structurally, financially and operationally sound. This entails
proper supervision by the public sector, an adequate capital base, stringent risk
management tools (audits, insurance, etc) and business recovery plans.
• Central securities depositories should continue to develop and implement true
delivery versus payment (DVP) systems and provide intraday settlement finality.

1
The information provided is derived from data received from various sources and from the respective
depositories and is believed to be reliable and accurate.
2
Francis Braeckevelt is responsible for Investor Services Product Management in Asia Pacific.

284 BIS Papers No 30

• The future Asian infrastructure will probably evolve gradually, based on the current
setup. We anticipate further consolidation within the respective markets, with a
gradual expansion of cross-border links.
• Alternatively, a “central utility” concept could be reviewed. This utility could act as a
central access point to various markets, providing multicurrency, DVP settlement,
ideally complemented by an automated securities lending and borrowing facility.

2. Introduction
Asia’s domestic markets continue to recover from the fallout of the severe 1997 financial
crisis and are, today, characterised by more stability and an expectation of growth. GDP in
Asian countries is expected to expand considerably over the coming years and to outpace
the growth in other regions, creating important funding and financing requirements. With the
1997 events fresh in mind, governments and private sector participants are acutely aware of
the dangers of relying on short-term capital, bank financing and capital inflows denominated
in foreign currency and subject to foreign exchange fluctuations.
Asian economies are therefore expected to limit their exposure to these traditional funding
sources by supplementing them with domestic currency financing alternatives.
A key feature in creating an investor-friendly environment is the development of an efficient
capital market, and more specifically a strong and liquid domestic debt market, allowing
market participants (both borrowers and investors) to attract and invest in longer-term
financial products.
Therefore, critical elements required for organising a sound capital market are:
1. the creation of a liquid government bond market, providing basic investment and
funding possibilities and a credible benchmark yield curve to price corporate debt;
2. adequate credit rating coverage, by either global (S&P, Moody’s, Fitch) or local
credit rating agencies;
3. the implementation of a harmonised taxation and regulatory environment;
4. the establishment of operational infrastructure based on efficient and sound clearing
and settlement mechanisms, central depositories and derivatives markets, as well
as securities borrowing and lending and repurchase agreement facilities.
The main drivers of an effective capital market are pricing and demand and supply
considerations. Increasingly, however, market participants appreciate the importance of an
adequate support infrastructure. In other words, where capital markets provide the
fundamental infrastructure to bring investors together, the clearing and settlement
infrastructure ensures the effectiveness and efficiency of the system.
The clearing and settlement process is a series of complex tasks that start with trade
confirmation and continue through the clearing process up to the actual settlement of a trade.
The successful functioning of this system or series of systems is largely dependent on the
close interaction of a number of intermediaries, each responsible for a distinct part of the
process.

3. Definitions
The following list presents some of the definitions commonly used in discussion on clearing
and settlement topics.

BIS Papers No 30 285

Clearing. Generally, clearing refers to the process of comparing trades before settlement
date or the determination of the net obligations of the broker participants (for both securities
and cash). In certain publications, clearing may be used synonymously with settlement.
Settlement. The settlement process refers to the exchange of cash and securities on the
contractual settlement date. The settlement date can be agreed upon at trade execution or
can be prescribed by local trading conventions. Settlement may be processed on a
provisional or a final basis.
Settlement finality. The exchange of cash and securities is final when a settlement can no
longer be unwound. Finality eliminates the main legal risks of payment and settlement
systems, reduces systemic risk and ensures the smooth operation of a system.
Provisional settlement. Provisional settlement allows for onward delivery of securities
which were not received on a final and irrevocable basis. Systemic risk is introduced in the
system if the unwinding of a specific settlement has a cascading effect on other previously
settled transactions.
Gross settlement. Gross settlement systems settle transactions on an instruction by
instruction and real-time (RTGS) basis throughout the day. RTGS systems are costly due to
the need for collateral or available cash balances to cover payment obligations during the
day or for securities lending programmes to cover short securities positions. RTGS systems,
however, typically reduce systemic risk.
Net settlement. In net settlement systems, obligations are settled at the end of the business
day on a net basis. The net process is subject to potential systemic risk, due to the contagion
effect where incoming funds are relied upon to make onward payments when a participant
cannot meet his obligations. As there is no requirement to post collateral or keep cash
balances readily available during the day, net systems tend to be less costly.
Central securities depository (CSD). A CSD is either the physical entity or the system that
facilitates the settlement and safekeeping of securities and ensures the reconciliation of
participant accounts. Securities can be safekept in immobilised or dematerialised form.
Settlement generally occurs in book entry form.
International central securities depository (ICSD). An ICSD is a depository settling trades
in international and various domestic securities, usually through direct and indirect links with
agents in the domestic markets. The best known ICSDs are Euroclear Bank and Clearstream
International. The eurobond market developed in part in response to operational and
regulatory inefficiencies in domestic bond markets.
Central counterparty (CCP). A CCP acts as counterparty to every buy and sell trade, a
process known as “novation”. This process concentrates counterparty risk and provides
multilateral netting.
Vertical integration. Vertical integration refers to the merger of institutions providing
different services in the value chain (eg trading, clearing, CSD). Vertical integration offers
advantages of scope.
Horizontal integration. Horizontal integration refers to the merger of institutions providing
similar services (eg clearing services for equities, derivatives and fixed income instruments).
Horizontal integration offers advantages of scale.
Continuous Linked Settlement (CLS™). CLS is a unique process that enables cross-
border currency transactions to be settled intraday. CLS enables settlement to be final with
payout from central bank funds. As it is a real-time, global settlement system, it will
significantly reduce the settlement risk caused by delays arising from differences in time
zones, legal jurisdictions and operating procedures.

286 BIS Papers No 30

Delivery versus payment. Delivery versus payment (DVP) is the simultaneous, final,
irrevocable and immediately available exchange of securities and cash on a continuous basis
throughout the day.
Straight through processing. Straight through processing (STP) is defined in many
different ways by different segments in the financial industry. In general, it is considered to be
a process that improves the efficiency of the securities industry by eliminating trade and
settlement failures, reducing manual processing, decreasing settlement time, etc.

Section I: Current trends in clearing and settlement

1. Clearing and settlement infrastructure

1.1 Introduction
It is difficult to unambiguously define the scope or the roles and responsibilities of a clearing
and settlement intermediary by analysing the current operating models. Market practices,
CSDs, clearing houses, CCPs and other parts of the clearing and settlement infrastructure
have developed at different rates, resulting in distinct regulatory, tax and technical
environments.
In the infrastructure section, we will cover:
• basic clearing and settlement services;
• the central counterparty concept;
• the clearing and settlement service model;
• rationalisation/integration/consolidation: public or private initiatives?

1.2 Basic clearing and settlement services

The analysis of various domestic market systems does identify some common and basic
roles and responsibilities of clearing and settlement providers.
These services include but are not necessarily limited to:
• trade matching;
• trade confirmation;
• clearing activities (ie netting of obligations);
• cash flow distribution;
• trade settlement (either on a final irrevocable or provisional basis);
• registration (beneficial owner’s name or in a nominee name);
• safekeeping of assets and holdings;
• messaging/reporting;
• other services, such as account reconciliation, which are often provided in
conjunction with local or global custodians.

1.3 Central counterparty

The use of a central counterparty is most prevalent in clearing activities, where, through
novation, counterparty risk is effectively mitigated on a centralised and cost-efficient basis. It

BIS Papers No 30 287

is commonly recommended that clearing houses be organised as separate legal entities from
the securities depositories. Even though the CCP can be closely related to a CSD, each
fulfils different functions and has a distinct role and responsibilities in the clearing and
settlement process.
A central counterparty should maintain an adequate capital base and strong risk
management tools to mitigate its own risks and deal with adverse situations. These tools
may include, but are not limited to, margin call procedures, settlement of margin calls in
central bank money, exposure limits, guarantee funds, securities lending and buy-in
procedures or the use of collateral or daylight overdraft arrangements.
Introducing a central counterparty has the most tangible effect where trading anonymity is
required. In market segments where trading occurs over the counter or through market-
makers, traders can effectively mitigate the exposure and risk through a careful selection and
due diligence of their counterparties.
The concept has gained importance worldwide and central counterparties have been
introduced in a number of Asian markets. As illustrated in Annex 4, countries like China (all
bonds), Indonesia (corporate bonds), Malaysia (corporate bonds) and Thailand (all bonds)
have integrated a central counterparty into the clearing and settlement infrastructure.

1.4 The clearing and settlement service model

Traditionally, clearing and settlement systems have been geared towards the domestic
market, organised around instrument types (equities, derivatives, fixed income, etc) and
primarily focused on overcoming imperfections and providing tailor-made solutions to market-
specific issues.
However, this resulted in a highly fragmented infrastructure, typically exposed to significant
operating inefficiencies, technical inconsistencies, high cost structures or disparate
regulatory environments.
Today, market participants recognise the need for a rationalised or integrated infrastructure
allowing for an efficient exchange of securities and payments, while at the same time
ensuring that risks are mitigated and costs are reduced or contained.
In a fragmented model, several intermediaries each provide part of the clearing and
settlement services. The challenge, therefore, is to effectively integrate this infrastructure by
introducing expensive but required regulatory and operational changes.
In a cross-border trading and settlement context, integration also involves creating full
interoperability between different domestic systems and providing access to systems outside
the domestic market.
Recent technological innovation has made this integration process a more realistic exercise
as countries continue to adopt international standards and conventions (International
Securities Identification Number (ISIN) standards or Bank Identification Codes (BIC)), and
have considerably increased investments in straight through processing (STP) solutions
(SWIFT or FIX (Financial Information eXchange) protocols).

1.5 Rationalisation/integration/consolidation: public or private initiatives?

With all the challenges and expenses to overcome, it is important to identify and assign
responsibilities to the various participants driving the rationalisation process forward.
Looking at the various initiatives that have been or are being introduced, the success of the
clearing and settlement infrastructure rationalisation process is highly dependent on the
interaction between the private and public sectors. This close cooperation will help achieve
risk reductions, cost controls and infrastructure efficiencies, as well as system stability and
robustness.

288 BIS Papers No 30

In general, the private sector focuses primarily on:
• analysing and improving market practices and technical requirements;
• the consolidation process itself.
The public sector, in contrast, tends to be predominantly involved in issues pertaining to:
• the harmonisation of tax legislation;
• the reconciliation and harmonisation of national laws in a cross-border environment;
• the creation of an integrated regulatory and oversight framework;
• the implementation of the most cost-efficient, competitive and sound infrastructure;
• the oversight of the private sector (intervening if required) to ensure the process is
resulting in increased efficiencies;
• the creation of regulations that provide incentives for providers to innovate.
As a result, capital markets have evolved significantly over time. In a traditional
infrastructure, intermediaries are closely related to the government or the main capital market
participants and operate as a natural or a de facto monopoly. In the government bond
market, for example, central banks generally provide the central securities depository
functions. In the more recent models, intermediaries are operating in a market-based
environment characterised by rationalisation, integration and consolidation.

2. Clearing and settlement infrastructure

2.1 Introduction
Seeing the importance of the clearing and settlement process in maintaining the systemic
soundness of the global financial markets, the operating environment has been analysed by
many international committees and working groups from both the public and private sectors.
As a result, a range of projects have been initiated to develop and promote securities
clearing and settlement systems that can operate in a stable and robust environment while at
the same time reducing risks and costs for their participants.
The operating, monetary, regulatory and infrastructure observations of individual markets are
also relevant in a regional context. The barriers encountered in the rationalisation or
integration process have not halted the progress, but have in some instances slowed it down,
resulting in regional clearing and settlement systems that are currently at varying stages of
development, sophistication and integration.
We will review the clearing and settlement infrastructure and recent developments in:
• the United States;
• Europe;
• Asia.

2.2 United States

In the United States, a relatively homogeneous clearing and settlement infrastructure has
been created thanks to the single currency and harmonised regulatory and tax environment.
US Treasuries are generally settled on an RTGS basis through the Federal Reserve book
entry systems and are held in dematerialised form. Settlement occurs on a rolling TD +1
basis, unless otherwise negotiated at the time of trading. The transfer of securities and cash
occurs immediately upon settlement, on a final and irrevocable basis.

BIS Papers No 30 289

To maintain confidence in the US financial markets, the Federal Reserve Bank continues to
participate in private sector initiatives and to explore ways of streamlining the design and
operation of the clearing and settlement process.
The most recent initiatives are related to CLS procedures and the introduction of Fedwire-like
services in the settlement procedures of additional instrument types.

2.3 Europe
In Europe, the introduction of the euro provided important momentum to the rationalisation
process of stock exchanges, payment systems and securities clearing and settlement
structures. Apart from the single currency, the following developments also accelerated the
process:
• the creation of a single integrated and efficient European capital market;
• an increase in international capital movements;
• technological progress;
• financial deregulation.
The European public institutions involved in this rationalisation effort are the European
Commission, the European Parliament (as the European legislator) and the European
System of Central Banks/Committee of European Securities Regulators (ESCB/CESR),
which combines both central banks and securities regulators. The main requirements that
have been identified for successfully constructing an integrated clearing and settlement
model are:
• the removal of technical, legal and fiscal barriers, to lower the costs and reduce
inefficiencies of cross-border settlement;
• the removal of competitive distortions/unequal treatment of entities performing
similar clearing and settlement activities;
• the creation of clearing and settlement industry standards to ensure a sound system
in which risks can be mitigated, reduced or controlled;
• a market-led rationalisation process with oversight from the public sector, which
covers:
– following through on changes in local laws and regulations, if required;
– remaining vigilant to a particular intermediary emerging as a monopolistic
entity and ensuring that a system for balancing stakeholders’ interests is
provided for;
– paying special attention to the soundness of the clearing and settlement
system and to the low-probability catastrophic risks that can introduce
systemic risks.
The diagrams in Annex 1 represent the current and future state of the European clearing and
settlement system, illustrating the complex nature and the extent of the consolidation and
integration.

2.4 Asia
Unlike the United States or Europe, Asia does not have a single currency or a unified market.
This lack of a homogeneous regulatory and operational environment results in a wide range
of different market practices and allows “domestic” systems to exist side by side. Where the
clearing and settlement systems supporting equity trading in Asia have evolved, the fixed

290 BIS Papers No 30

income markets remain significantly less sophisticated, from both an execution and a support
perspective.
Despite the many differences, however, Asia is subject to the same general trends as the
United States and Europe. Technological advances, deregulation and globalisation increase
the need to introduce or develop sound capital market and operating systems. The current
highly fragmented infrastructure, however, is prone to adverse risk factors and high costs.
Considering the current constraints, it may not be possible or necessary for Asia to introduce
a fully integrated US- or European-type clearing and settlement model. It would be beneficial
for Asia, however, to move towards a gradual rationalisation of the domestic and regional
infrastructure, introducing interoperability between domestic and cross-border systems as
well as leveraging the various global and private sector initiatives like CLS.
In the interim, the predominantly domestic infrastructure should focus on increasing
transparency and implementing optimal execution and operational mechanisms. This will
allow the current system to increase cost efficiency, provide investors with fast and robust
execution and processing mechanisms and align the local market with international
standards.
Certain countries have started to integrate parts of the clearing and settlement infrastructure,
achieving economies of scale (horizontal integration) or scope (vertical integration).
Examples of horizontal integration can be found in Korea, where the Korea Securities
Depository (KSD) acts as the central depository for all instruments in the market, or in
Australia, where government bonds were transferred to Austraclear in 2002. The benefits of
vertical integration are clearly illustrated in Indonesia where, since the introduction of the
Bank Indonesia - Scripless Securities Settlement System (BI-SSSS), securities are cleared
and settled in scripless form at Bank Indonesia (BI) and payments occur through the central
bank’s RTGS system. This is an encouraging trend provided the market remains vigilant in
assessing the associated risks, related to the creation of a de facto monopolistic situation
(higher costs) or to the introduction of contagion risk through increased system
concentration.
We therefore conclude that due to the specific developments in the Asian markets, domestic
systems will probably prevail, but that some form of integration can optimise the services
offered to both local and foreign investors. Another option that could possibly be introduced
in an Asian context is the utility-type approach, where a central, preferably user-owned, entity
provides multicurrency central depository services.

3. Regulations concerning clearing and settlement structures

3.1 Introduction
The challenges that regulators and market participants face in building a robust and efficient
infrastructure are numerous and complex. Many working groups and regulators have
contributed extensively to the creation of an operational and regulatory framework in which
this process can be conducted.
In the next section, some recommendations and research papers published on the subject
have been brought together. The main reports that will be briefly discussed are:
• Global publications:
– G30 recommendations;
– CPSS/IOSCO report.
• European publications:
– Giovannini report(s);

BIS Papers No 30 291

 – ESCB/CESR report;
– the Lamfalussy process (“Committee of Wise Men”).
• US rules and regulations:
– Rule 17f-7 of the Investment Company Act of 1940.

3.2 Global recommendations

3.2.1 G30 recommendations (Annex 2)
The Group of Thirty (G30) is a private, not-for-profit, international institute composed of
senior representatives from the private and public sectors and the academic world. The G30
has been instrumental in creating a common understanding about the structure of a secure
and efficient clearing and settlement infrastructure.
The G30 issued a first report on clearing and settlement in 1989, with the principal objective
of improving local market practices. A second report (Plan for Action) issued in January 2003
contains 20 recommendations designed to create a robust interoperable global network,
mitigate risk and improve governance.

3.2.2 CPSS/IOSCO recommendations (Annex 3)

In December 1999, a task force launched by the Bank for International Settlements (BIS)
reviewed and consolidated the work performed by the International Organization of
Securities Commissions (IOSCO) and the Committee on Payment and Settlement Systems
(CPSS).
As a result, in 2002, the task force issued 18 recommendations to enhance the safety of, and
reduce the system risk inherent in, central securities depositories.

3.3 European initiatives

3.3.1 Giovannini report(s)
The first Giovannini report was issued in November 2001 and identified 15 technical, market
practice, legal and fiscal barriers to creating a low-cost cross-border settlement system. Out
of these 15 barriers, 10 were deemed to be of a legal and regulatory nature (and not due to
technical constraints or market practice), emphasising the importance of close interaction
between the public and private sectors.
A second Giovannini report, issued in April 2003, presented a road map for removing the
barriers identified in the first report, and assigned action and follow-up responsibilities.
Finally, the report contained a detailed description of possible consolidation models and
policy responses.

3.3.2 ESCB/CESR report (European System of Central Banks/Committee of European

Securities Regulators)
In early 2002, the ESCB/CESR consulted the financial industry on how the global
CPSS/IOSCO recommendations for securities settlement systems should be adapted and
strengthened to apply to the European marketplace.
An extra standard has been added to the existing 18 CPSS/IOSCO recommendations
covering the “custodians operating systematically important systems” notion.

3.3.3 The Lamfalussy process (“Committee of Wise Men”)

The Committee of Wise Men was appointed in July 2000 by the European Council of Finance
Ministers (ECOFIN) out of concern that the existing regulatory and legislative framework

292 BIS Papers No 30

hampered the growth and competitiveness of European securities markets. The committee
was mandated to identify all obstacles in the securities markets that impede cross-border
transactions and to suggest remedial actions.
The final report, issued in 2001, found that important gaps persist in European legislation and
that the conventional legislative process of the EU is too slow, complex and cumbersome.

3.4 US rules and regulations

3.4.1 Rule 17f-7 of the Investment Company Act of 1940
In 2000, the US Securities and Exchange Commission (SEC) adopted Rule 17f-7, which
governs the circumstances under which US investment companies may hold securities
through the facilities of non-US central securities depositories.
The key requirement of Rule 17f-7, by reference to Rule 17f-4, is that an investment
company must receive from its “primary custodian” an analysis of the custody risks
associated with maintaining assets with each depository it uses.
Rule 17f-7 does not provide specific guidance concerning the content of these analyses of
depository custody risk, but in the release announcing the adoption of the rule, the SEC
stated:
“As a general matter, we expect that an analysis will cover a depository’s
expertise and market reputation, the quality of its services, its financial strength,
any insurance or indemnification arrangements, the extent and quality of
regulation and independent examination of the depository, its standing in
published ratings, its internal controls and other procedures for safeguarding
investments, and any related legal protections.”
The Bank of New York, as a primary custodian, reviews all depositories in its global custody
network on their eligibility under Rule 17f-7 of the Investment Company Act of 1940. The
eligibility requirements as detailed in the rule are that depositories must:
• act or operate a system for the central handling of securities or equivalent book
entries in the country;
• be regulated by a foreign financial regulatory authority;
• hold assets for the custodian that participates in the system under safekeeping
conditions no less favourable than the conditions that apply to other participants;
• maintain records that identify the assets of each participant and segregate the
system’s own assets from the assets of participants;
• provide periodic reports to its participants with respect to its safekeeping of assets,
including notices of transfers to or from any participant’s accounts;
• be subject to periodic examination by regulatory authorities or independent
accountants.

Section II: Analysis of Asian settlement infrastructure

1. Asian infrastructure review: methodology

1.1 Introduction
In Section II, we shift our attention and assess the situation in Asia, identifying areas of
strength or potential improvement.

BIS Papers No 30 293

Similar to the other regions, the clearing and settlement infrastructure in Asia has been
developed and is being continuously refined to reduce the inherent clearing and settlement
risks and to contain or lower the cost structure.
When reviewing the specific Asian situation, we focus on risk management aspects, as this
constitutes a critical component of the investment decision process. In addition, investors’
awareness of market infrastructure risk continues to grow.

1.2 Framework and scope of the Asia review and analysis

• As a global custodian, the Bank of New York (BNY) is most intimately involved with
the settlement and safekeeping functions in the domestic custody markets. As such,
our analysis will focus on central securities depositories and not on the clearing
infrastructure.
• The review pertains to the fixed income market (government and corporate bonds)
only.
• The analysis is based on BNY’s research on the central depositories’ compliance
with Rule 17f-7 of the US Securities Investment Act of 1940. Generally, the
information provided is derived from data received from the respective depositories
and is believed to be reliable and accurate.
• The decision to use the 17f-7 rule criteria in our analysis does not imply any opinion
on the value and usefulness of the other recommendations as described in the first
section, but is made primarily based on BNY’s expertise in this area.

1.3 Risks associated with central securities depository infrastructure

In general, we believe investors should review and be aware of five broad risk categories
when assessing whether a central securities depository meets their infrastructure risk
tolerance. Even though each risk category is reviewed separately, it should be noted that
risks are typically cumulative and should therefore be looked at on an aggregate basis.
Risk is only one, albeit critical, part of every due diligence. In addition, it is important to
remain vigilant to other factors that may impact the depository activity, like scalability and
remaining excess capacity. This becomes especially important if a depository is considering
expanding its responsibilities and activities, either by accepting new instruments, through the
consolidation of depository functions within a given market, or if cross-border activity is
expected to add substantial volumes to the domestic activity.
The five broad risk criteria to be reviewed are:

1.3.1 Market risk

Market risk is probably the broadest category and covers generic elements related to the
market as a whole. In the context of the Asian depositories review, this section will focus on:
• the organisational structure of the depository;
• immobilisation/dematerialisation of securities.

1.3.2 Legal risk

The regulatory environment and the legal structures covering central securities depositories
are critical to ensure the safety of the assets and contract enforceability. Features of the legal
risk review include:
• regulations and governance rules covering the depository activity;
• enforcement history in case of non-compliance;

294 BIS Papers No 30

• compulsory use of the depository;
• the liabilities assumed by the central securities depositories;
• the asset segregation policies of the depository system;
• recourse options when accounts are blocked and assets frozen;
• central depository asset lien provisions.

1.3.3 Credit risk

Credit risk is a third important aspect of the risk framework and is considered as a
combination of:
• principal risk: counterparty default.
• replacement risk: in case of failed settlement, counterparties may be compelled to
acquire securities in the marketplace, where prices may have fluctuated.
• liquidity risk: required funding is not available to fulfil the payment obligations.
When looking at credit risk in an infrastructure context, we will focus on:
• the membership criteria established by the depository;
• the compliance monitoring process;
• the disciplinary action available to the depositories when rules are breached;
• the existence of guarantee funds and insurance policies;
• the credit facilities extended by the depositories to their participants.

1.3.4 Operational/custody risk

This section describes how depositories handle operations risks, and reviews:
• the depositories’ performance and functionality;
• guarantee funds and insurance policies;
• the availability of delivery versus payment (DVP) settlement.

Table 1
The BIS settlement models

Model 1 Systems that settle transfer instructions for both securities and funds on a trade by
trade (gross) basis, with final (unconditional) transfer of securities from the seller to the
buyer (delivery) occurring at the same time as final transfer of funds from the buyer to
the seller (payment).

Model 2 Systems that settle securities transfer instructions on a gross basis, with final transfer
of securities from the seller to the buyer (delivery) occurring throughout the processing
cycle, but settle funds transfer on a net basis, with final transfer of funds from the buyer
to the seller (payment) occurring at the end of the processing cycle.

Model 3 Systems that settle transfer instructions for both securities and funds on a net basis,
with final transfers of both securities and funds occurring at the end of the processing
cycle.

BIS Papers No 30 295

1.3.5 Systemic risk
Eliminating systemic risk allows depositories to ensure the soundness and robustness of
their systems, especially to cope with adversities or business interruptions. The question of
whether a contagion effect can occur through the unwinding of settled transactions is
addressed in the operational/custody risk section, where we review the settlement methods.
In this section, we will focus on the central securities depositories’ business recovery plans.

2. Asian infrastructure depository analysis

2.1 Introduction
We reviewed the risk management tools and processes adopted by selected central
securities depositories in a country by country analysis. We believe the group of selected
central securities depositories, although not exhaustive, is representative for the Asian region
and includes a mixture of entities serving both the government and corporate bond market
segments.

Table 2
Country by country assessment: markets and depositories

Country Central securities depository

Australia Austraclear

China CSDCC Shanghai and Shenzhen1

Hong Kong SAR Central Monetary Unit (CMU)

India Reserve Bank of India (RBI)

National Securities Depository Limited (NSDL)

Indonesia Bank Indonesia (BI)

Japan The Bank of Japan (BOJ)

Korea Korea Securities Depository (KSD)

Malaysia Bank Negara Malaysia (BNM)

Philippines The Bureau of Treasury (BTR)

The Philippine Central Depository Inc (PCD)2

Singapore The Monetary Authority of Singapore (MAS)

Taiwan, China Central Bank of China (CBC)

Taiwan Securities Central Depository (TSCD)

Thailand The Bank of Thailand (BOT)

The Thailand Securities Depository Co Limited (TSD)
1
The current QFII scheme in China may be subject to further changes by the regulators and may affect the
risk assessment of the China Securities Depository and Clearing Corporation Ltd (CSDCC). 2 The PCD
received a temporary operating licence for debt securities in March 2002 and is expected to become a
subsidiary of the Fixed Income Exchange, when established (due date 2004). Even though part of the initial
analysis, all observations concerning the PCD have been excluded from the results, as the PCD is currently not
used for the safekeeping or settlement of fixed income instruments.

296 BIS Papers No 30

2.2 Government bond trading mechanisms
Even though trading falls outside the scope of this analysis, we found that in certain cases
the trading and settlement environments are highly correlated and that trading restrictions
may largely affect the mechanics of the settlement process and the attractiveness of a given
market.
When appraising the trading process in the government and corporate bond market, we
found that fixed income markets are generally less developed compared to the more active
domestic equity markets.
Some of the major limitations to foreign investor participation in the markets reviewed are:

Liquidity constraints
Bond issues are, for a variety of reasons, taken up by local investors and held to maturity. In
addition, because of the countries’ budgetary surpluses, Asian governments have
traditionally not issued large quantities of government paper. Both of these trends have
affected the development of the market infrastructure considerably and have negatively
impacted market liquidity.
• Australia: Commonwealth Government Securities are popular amongst banks/local
institutions and form a large part of their statutory liquidity requirements.
• Korea: Government bonds are acquired by large investment trust companies, banks
and life insurance companies to satisfy reserve requirements. The same investors
also acquire corporate bonds.
• Malaysia: Both Malaysian Government Securities (MGS longer tenures) and private
debt securities (PDS) are typically held till maturity. The local Employees Provident
Fund (EPF) is a big investor in the fixed income market.
• Singapore: About 80-90% of all corporate bonds are acquired by banks and
insurance companies and are typically held as proprietary positions until maturity.
• Thailand: The Bank of Thailand (BOT) generally issues BOT bonds only to financial
institutions, which use them to satisfy reserve and liquidity requirements.

Foreign investor restrictions

In selected countries, foreign investor regulations, restricting foreign participation in the
market, still apply. Most of these require foreign investors to obtain prior approvals from the
central bank (mainly foreign exchange-related) or from the regulators (investor status).
• China: Foreign participation is only allowed through the QFII (qualified foreign
institutional investor) scheme and is restricted to exchange-listed Treasury and
convertible bonds. Stringent capital controls are still in place.
• India: Foreign institutional investors (FIIs) must be registered with the Securities and
Exchange Board of India (SEBI) and the Reserve Bank of India (RBI). Approved
investors are allowed to invest up to 30% of the total portfolio in debt instruments.
Separate approval is required to start a 100% debt fund.
• Indonesia: Foreign participation is subject to prior approval from Bank Indonesia (BI)
and/or Bapepam, the capital markets supervisory agency. Foreign exchange
transactions must be approved and supported by underlying trade evidence.
• Japan: No direct restrictions apply but foreign investors are subject to stringent
foreign indirect participant and qualified foreign intermediary rules.

BIS Papers No 30 297

• Korea: Investors must obtain an investment registration certificate from the Financial
Supervisory Service (FSS). Korean won can only be purchased to meet exact
payment obligations.
• Philippines: Inward/outward remittances of funds are subject to central bank
approval.
• Taiwan, China (hereinafter Taiwan): Until the new FINI rules replaced the existing
QFII/non-QFII regulations, foreign investors were only allowed to hold a single cash
account. As this account was generally used for equity-related transactions, access
to the fixed income market was effectively barred. Foreign participation in the repo
market is capped and foreign exchange transactions must be reported to the central
bank.
• Thailand: Same day and next day foreign exchange transactions are prohibited.

2.3 Current clearing and settlement infrastructure in Asia

Before taking the analysis to an individual country level, let’s look at the clearing and
settlement infrastructure in the region as a whole.
Annex 4 presents a summary of the clearing and settlement infrastructure in selected Asian
countries for government and corporate bonds. For each country, the central securities
depositories, the payment system operator, the central clearing counterparty (if applicable)
and the clearing houses (if applicable) for government and corporate bonds have been
tabulated.
Without focusing on specific roles or responsibilities, some interesting trends and
observations about the organisation of the Asian domestic market infrastructure appear.
• A positive trend that has developed in most countries is the creation of formal
central securities depositories to settle and safekeep all types of securities. In some
countries, physical scrip still exists, but these instances are generally exceptional,
and physical securities are being gradually phased out.
• Payments are processed either electronically or by cheque. Cheques are used in
China (partly), Taiwan (partly) and India (across instruments). A further move
towards the implementation of electronic payment systems, in our view, should be
encouraged.
• Central clearing counterparties are not common. In countries where the concept has
been introduced, CCPs remain closely linked to either the exchange or the clearing
house. The feasibility and implications of introducing a central counterparty need to
be explored in further detail, taking into account specific market mechanics and
conditions.
• Securities clearing generally occurs through separate entities (India, Hong Kong
SAR, Philippines, Malaysia), licensed clearing banks (Taiwan), depositories (Taiwan
- selected corporate bonds; Thailand, corporate bonds) or the exchange (Korea).
• Usually, different institutions assume the role of CSD, payment system operator,
CCP and clearing house, depending on the underlying instrument, resulting in a
highly fragmented environment, which may lead to significant market inefficiencies.
Exceptions are Australia, China and Korea, where infrastructure is consistent across
segments.
• Within a specific asset class, infrastructure may differ based on the trading methods
(on exchange or OTC), eg government bonds in Hong Kong or corporate bonds in
Malaysia, further contributing to the market fragmentation and potential
inefficiencies.

298 BIS Papers No 30

• Even though payment systems may also differ across segments, the cash
infrastructure is less fragmented due to central banks’ involvement in the payment
side. In Malaysia and Thailand, for example, although instruments are held in
different depositories, payments are processed through the central bank’s RTGS
system (RENTAS and BAHTNET respectively).
• Government bonds are commonly deposited at central banks or their affiliates (Hong
Kong, India, Indonesia, Malaysia, Philippines, Thailand), providing the market
participants with an additional level of comfort with the depository infrastructure.
• In the other markets, fixed income instruments are held at depository and clearing
corporations (China, Hong Kong (selected issues) and Korea), limited liability
companies (Australia) or licensed clearing banks (Taiwan).
• Due to the similar trading conventions, the corporate bond intermediaries are
generally more closely aligned with the equity than with the government bond
infrastructure.
In summary, the domestic clearing and settlement infrastructure in Asia is very fragmented
and differs significantly across markets and instruments. Central banks play an important role
as depositories for government bonds, while the corporate bond market infrastructure follows
the equity market practices and infrastructure setup more closely.

2.4 Risk reviews of the selected Asian central depositories

In our assessment, the organisation and structure of the central securities depositories have
been tested against the various risk categories discussed in Section II.1.3:
• market risk;
• legal risk;
• credit risk;
• operational/custody risk;
• systemic risk.
Before going into the specifics of each risk component, it is important to highlight a unique
aspect of the relationship between the CSD, the local custodian (domestic settlement and
safekeeping agent) and the global custodian.
Even though the assets of a particular beneficial owner are held at a CSD, this investor has
not necessarily entered into a direct contractual relationship with the market provider. If a
beneficial owner uses a global custodian for its back office operations, these services are
governed by a specific global custody agreement.
Generally, to complete the chain through to the CSD, two more contractual/service
relationships have been set up to service the assets. First, there is the link between the local
and global custodian, based on a legal contract, and secondly, the relationship between the
CSD and the local custodian, generally covered by a participant agreement.
Considering the breadth and width of these cascading relationships, it is critical to adopt a
strict and continued due diligence process. This review and due diligence practice is
important not only with regard to the entities with which a direct contractual relationship has
been established, but also, and probably even more so, with regard to the other
intermediaries (such as CSDs, payment systems, etc) with which no legal link exists. To do
this effectively, market specific features must be reviewed carefully and local legal opinions
may be required in certain cases to validate generally accepted practices in a local market
context.

BIS Papers No 30 299

2.4.1 Market risk
As part of the investment decision, investors participating in cross-border trading have to be
aware of the market risks that investing in foreign markets and jurisdictions may entail. As
elements related to the organisation and governance of the securities depository may
contribute to the overall market risk, this section will focus on some of those specific
elements.

The organisational structure of the depository

Understanding the ownership structure of a central securities depository is important for
assessing its financial strength and support structure. Central banks or government-owned
entities usually receive the implicit or explicit backing of the government, whereas it is
important to analyse the financial statements for private entities to fully assess the financial
soundness of these providers and their ability to deal with default or other adverse event risk.
From the selected central securities depositories:
• three are organised as for-profit private limited liability companies (Austraclear
(Australia), NSDL (India), TSD (Thailand));
• three are non-profit organisations partly or wholly owned by the exchange (CSDCC
(China), KSD (Korea), TSCD (Taiwan); the KSD has the widest ownership
structure);
• nine central banks, units of a central bank or entities fully owned by the government
provide government bond depository services and operate on a non-profit basis
(CMU (Hong Kong; cost recovery basis), RBI (India), BI (Indonesia), BOJ (Japan),
BNM (Malaysia), BTR (Philippines), MAS (Singapore), CBC (Taiwan), BOT
(Thailand)).

Immobilisation/dematerialisation of securities
Physical securities are almost non-existent in today’s markets. In general, fixed income
paper deposited at the CSDs is either held in dematerialised or immobilised form, or a
combination of the two (depending on the issuer or instrument):
• Securities are held in dematerialised form at the CSDCC (China), NSDL and RBI
(India), BI (Indonesia; for current issues), BOJ (Japan), BNM (Malaysia), BTR
(Philippines), MAS (Singapore; insignificant portion of physical shares still exists),
CBC (Taiwan), TSD and BOT (Thailand).
• Securities are held in immobilised form at the TSCD (Taiwan).
• Instruments deposited at the following depositories are held in either immobilised or
dematerialised form, depending on the issuer or the instrument: Austraclear
(Australia), CMU (Hong Kong; EFBs and EFNs are dematerialised, other
instruments are held in immobilised form at two external settlement banks), KSD
(93% is dematerialised, the remainder is held in immobilised form).

2.4.2 Legal risk

Regulations and governance rules covering central depository activity
Considering that the rules under which the depositories are governed may affect the
outcome of legal disputes or other adverse situations, it is important to gain a fair
understanding of the depositories’ legal foundation and oversight structure.
• Central banks and other government-owned institutions, which are widely
represented in our selection, are generally self-regulated entities, incorporated
under the central bank laws of their respective countries and subject to oversight

300 BIS Papers No 30

 and supervision of government-linked commissions or ministries. Examples are
CMU (Hong Kong), RBI (India), BI (Indonesia), BOJ (Japan), BNM (Malaysia), BTR
(Philippines), MAS (Singapore), CBC (Taiwan) and BOT (Thailand).
• Supervision of the non-central bank entities (Austraclear (Australia), CSDCC
(China), NSDL (India), KSD (Korea), TSCD (Taiwan), TSD (Thailand)) generally falls
under the purview of the financial market regulators or government entities
(ministries or government departments). A common regulator is the Securities and
Exchange Commission (SEC) or its local equivalent.
Even though the financial regulators in each market keep close oversight over the
depositories, it remains prudent to ensure that specific contracts entered into by investors or
their representatives are enforceable under the local laws and regulations.

Enforcement actions in case of non-compliance

If central securities depositories fail to comply with the statutory or regulatory requirements,
the supervisory authorities may resort to remedial or enforcement actions such as imposing
fines or restricting, suspending or terminating the activities of the depository. All depositories
in the review stated that no such action had been brought against them in the past three
years.

Compulsory use of the depository

The use of central securities depositories for safekeeping and settlement is required either by
law or by local market practice. This improves transparency and reduces the risks of
settlement fails due to timing issues related to the withdrawal or lodgment process.
The use of the central depository for safekeeping is:
• not compulsory: none;
• compulsory by law: CSDCC (China); CMU (Hong Kong; EFB/EFN); BI (Indonesia;
verbal confirmation); BNM (Malaysia); BTR (Philippines); KSD (Korea); CBC
(Taiwan; government bonds issued after September 1997);
• consistent with prevailing market practice: Austraclear (Australia); CMU (HK; other);
RBI and NSDL (India); MAS (Singapore); TSCD (Taiwan); TSD and BOT (Thailand).
The use of the central depository for settlement is:
• not compulsory: none;
• compulsory by law: CSDCC; CMU (EFB/EFN); RBI; NSDL; BI (verbal confirmation);
BOJ; BNM; BTR; PCD; KSD; TSCD; TSD; CBC (government bonds issued after
September 1997);
• consistent with prevailing market practice: Austraclear; CMU (other); MAS; BOT.

The liabilities assumed by central securities depositories

The liability question is important in view of the compulsory nature of safekeeping and
settlement through most depositories. Apart from the fact that none of the depositories
assume liability for force majeure, political risks or acts of God, liability language may differ
substantially across central depositories:
• The CSDCC (China), RBI and NSDL (India), BI (Indonesia), BOJ (Japan) and KSD
(Korea) assume liability for their own performance.
• The TSCD (Taiwan) assumes liability for its own performance to the extent that the
loss is covered by insurance, and MAS in Singapore assumes liability for its own

BIS Papers No 30 301

 performance, limited to an amount not exceeding the value of the non-recoverable
amount plus interest.
• Other depositories do not assume liability for their own performance: Austraclear
(Australia), CMU (Hong Kong: “unless the error is caused by wilful default or gross
negligence of the HKMA or its servants or agents”), BNM (except for loss of
participant’s securities) and BTR (Philippines).
• In addition, the CMU (Hong Kong), which safekeeps physical certificates for bonds
and money market instruments at two subcustodians, states that, in the event of a
loss, the CMU will in the first instance recover its own damages from the payment
received from the subcustodian, and subsequently distribute compensation to
affected CMU members.
• The Bank of Thailand (BOT) states that it is liable for direct losses related to
reconciliation errors with registrars and theft of securities, but that it does not
assume liability for its errors and omissions or failure of its systems, and that it is
immune from legal action in its own jurisdiction.

The asset segregation policies of the depository system

In countries where nominee registration is very common, it is important to understand the
account setup and asset segregation policies adopted by the depositories, as this may have
important consequences in cases of default or insolvency.
• In all countries reviewed, the depository’s proprietary assets (if any) are separated
from the participant’s assets. Austraclear, the BOJ and the KSD, explicitly state that
they do not hold any proprietary assets in the depository system.
• In most countries, participants are required to segregate their proprietary and client
assets. Exceptions are Australia (Austraclear), where segregation at the participant
level is optional, and Indonesia (BI), where segregation is not allowed.
• In general, setting up omnibus accounts is a common market practice. Assets in
omnibus accounts are usually held on a fully fungible basis and are not identified as
resident/non-resident holdings, or are not linked to a specific individual investor.
• In China (CSDCC), India (NSDL) and Korea (KSD), where foreign investors are
subject to prior investment approval from the authorities or regulators, assets are
unambiguously linked to the accounts of an approved underlying foreign investor.
• In Japan, the account structures at the BOJ allow for the identification of non-
resident holdings and the beneficial owner’s tax status and liabilities.
• In Malaysia (BNM), assets are identified as either resident or non-resident holdings.
• In Thailand (BOT), proprietary assets are segregated from client assets. Client
assets must be segregated per underlying client.

Recourse options when accounts are blocked or assets are frozen

Accounts may be blocked or assets frozen when disciplinary action (default/rule violations) is
brought against a depository participant. It is important for the underlying investors to
understand how this affects the protection of and access to assets:
• In Australia (Austraclear), Hong Kong (CMU) and the Philippines (BTR), access to
the account by the underlying client is subject to the approval of an
administrator/liquidator.
• In China (CSDCC), India (RBI) and Japan (BOJ), prior approval from the securities
depository is required.

302 BIS Papers No 30

• In Korea (KSD), approval is granted by the securities regulators (FSS).
• In Indonesia (BI), Malaysia (BNM) and Thailand (TSD), access is subject to the
relevant (bankruptcy) legislation.
• In India (NSDL), Taiwan (TSCD) and Thailand (BOT), assets are unambiguously
linked to the underlying investor through the segregated account setup. Client
accounts, therefore, remain fully accessible at all times.

Central depository asset lien provisions

Generally, a depository does not hold liens on assets, except in cases like:
• the payment of fees/expenses;
• securities encumbrance issues (Austraclear);
• settlement defaults (Thailand).
• Bank Indonesia stipulates that it has the authority to exercise a lien on a
participant’s proprietary assets (if required).

2.4.3 Credit risk

The membership criteria established by the depository
Participant eligibility criteria may vary from market to market. In all the countries analysed,
however, applicants are subject to stringent rules and regulations outlined in the depository
membership rules and/or imposed by the applicant’s specific regulator (eg banks must be
approved by the central bank). In addition, in certain markets, the depository’s parent
company (Austraclear), the exchange or the securities market regulator (CSDCC) must also
approve all new membership applications.
In general, the legal framework governing the relationship between the depository and a
participant comprises:
• a standard participation contract;
• terms and conditions of participation;
• rules and by-laws of the depository;
• relevant domestic laws and regulations.

The compliance monitoring process

• Usually, the depository monitors its participants’ adherence to the membership
criteria.
• In China (CSDCC), India (NSDL), Korea (KSD) and Taiwan (TSCD), the securities
regulators assist the depository in the monitoring process.

Disciplinary actions available to the depositories when rules are breached

It is relatively rare that central securities depositories have to resort to disciplinary action to
address non-compliance with the rules and regulations governing the CSD/participant
relationship. However, in cases where action was deemed necessary, appropriate action was
taken. No details about the specific cases have been disclosed. In cases of non-compliance,
the central depositories can:
• impose fines;
• restrict, suspend or terminate the participant’s membership.

BIS Papers No 30 303

In addition, the depositories in Japan (BOJ), India (RBI and NSDL), Korea (KSD) and Taiwan
have also issued rules that specifically address disciplinary or remedial action in case of
participant defaults or insolvency.
During the past three years:
• CSDCC Shanghai has taken action to address payment defaults;
• RBI (India) took action against participants that did not meet their settlement
obligations;
• the NSDL (India) terminated or suspended seven participants that no longer met the
net worth criteria or were disbarred from the exchange;
• BNM (Malaysia) resorted to disciplinary action and fines (no further details have
been disclosed);
• all other depositories confirmed that no disciplinary or remedial actions have been
taken against any of their participants.

The credit facilities extended by the depositories to their participants

Credit facilities and overdraft lines are important tools in mitigating liquidity and counterparty
risk as they prevent the settlement activity from being disrupted even if there are not
sufficient funds available. In most cases, the CSD does not directly extend credit facilities to
its participants, who must enter into overdraft and credit arrangements with the local
commercial banks.
• Austraclear (Australia), CSDCC (China), CMU (Hong Kong), BOJ (Japan), BTR
(Philippines), MAS (Singapore), KSD (Korea) and TSCD (Taiwan) specifically
indicated that they do not extend any credit or overdraft facilities to their participants.
• Only BNM (Malaysia) stipulated that it does extend intraday credit to its participants,
provided sufficient collateral has been posted.

2.4.4 Operational/custody risk

Establishment of guarantee funds and insurance policies
Guarantee funds
Most central depositories do not guarantee settlement (as a central counterparty or
otherwise), but act as agents in the settlement and safekeeping process. As such, they have
not created any guarantee funds, but mitigate settlement risks through the adoption of true
delivery versus payment settlement methods.
• Austraclear (Australia), CMU (Hong Kong), RBI and NSDL (India), BI (Indonesia),
BNM (Malaysia), BTR (Philippines - even though settlement of matched transactions
confirmed by both parties is guaranteed), MAS (Singapore) and the KSD (Korea)
have not established guarantee funds to cover their daily activities.
• CSDCC (China) has set up both a settlement risk (systemic failure) and a settlement
guarantee fund (participant default).
• BOJ (Japan) maintains an investor protection trust (participant default).
• TSCD (Taiwan) contributes 5% of its operating income to a “default damaged
reserve fund” to cover losses of or damage to share certificates under custody.
• TSD (Thailand) acts as a central counterparty to trades between clearing members
of the exchange only, and has established a credit line with the settlement banks to
cover settlement default. Settlement between custodian banks and their
counterparties is not covered.

304 BIS Papers No 30

Guarantee funds are more common in a trading and clearing environment. If a guarantee
fund has been created, it is important to evaluate the fund’s size, payout criteria and
replenishment schedules to ensure the available resources are adequate to mitigate the risks
they cover.
• In India, CCIL (Clearing Corporation of India Ltd) and the respective clearing houses
related to the respective stock exchange assist the depositories in the settlement
process and maintain guarantee funds to mitigate associated risks.
• In Korea, Taiwan and Thailand, the stock exchanges maintain guarantee funds to
compensate losses resulting from a member’s default.

Insurance policies
If guarantee funds are not available, it is important to review the insurance policies a CSD
has arranged to protect itself and its participants against adverse events. When analysing the
available policies, it is important to focus on the policy coverage and carrier, the amounts
insured and the applicable deductible amounts to assess the risk-mitigating capacity of the
specific policies.
• The central banks or linked depositories in our review (CMU (Hong Kong), RBI
(India), BI (Indonesia), BOJ (Japan), BNM (Malaysia), BTR (Philippines) and MAS
(Singapore)) typically do not maintain third-party insurance policies.
• Of the non-central banks, only CSDCC (China) does not maintain insurance.
• NSDL (India), KSD (Korea), TSCD (Taiwan) and TSD (Thailand) maintain insurance
policies to meet claims arising from their depository services and performance.

Performance and functionality

Functionality
For participants and underlying investors, it is important to understand how a CSD is
organised from a functional point of view, and whether risk is concentrated within the CSD or
is further disseminated to third-party providers through outside vendor/contractor
relationships.
• Austraclear (Australia), CSDCC (China), NSDL (India), BNM (Malaysia), BTR
(Philippines), MAS (Singapore), KSD (Korea), TSCD (Taiwan) and the TSD
(Thailand) handle all functions related to depository responsibilities in-house and do
not outsource any activities to third party providers.
• CMU (Hong Kong) provides all depository functions in-house except for the
safekeeping of physical certificates (private subcustodians) and computer
processing (HK Interbank Clearing Limited).
• RBI (India) provides services internally except for netting services and the
computing of participant obligations of certain trades, which is outsourced to the
clearing house CCIL.
• BOJ (Japan) performs all functions internally except for communications, for which it
employs NTT, the national telephone carrier.

Communication procedures
Most depositories communicate with their participants through either:
• proprietary technology;
• secure dial-up or leased connections;
• a combination thereof.

BIS Papers No 30 305

Control procedures and performance history
All depositories have implemented stringent internal control procedures, mitigating
operational risk. During the past three years, no adverse performance has been reported that
resulted in system disruptions of significant proportions.

Audits
When reviewing the performance, functionality and overall stability of market intermediaries
and providers, it is important to complement proprietary due diligence results with feedback
provided in both internal and external (independent) audit reviews:
• Central banks, as self-regulated entities, are generally subject to statutory audit
reviews by the government and to internal audit reviews.
• Some central banks (eg RBI - India) reported only a statutory audit requirement.
• Austraclear (Australia), CSDCC (China) and CMU (Hong Kong) are subject to both
internal and external audits.
• The NSDL (India) is subject to regular external and ad hoc regulatory (SEBI) audits.
• The KSD (Korea) is subject to annual internal/external and regulatory (FSS) audits.
• The TSCD (Taiwan) is subject to four levels of audit (internal, external and
regulatory (SFC), as well as an operational audit on its computer systems).
• The TSD (Thailand) is audited annually by the SET internal auditor (operational),
twice a year by external auditors (financial), and is subject to an occasional audit by
the SEC.
In addition, risk management policies of the CSDCC (China), CMU (Hong Kong), NSDL and
RBI (India) and the BOJ (Japan) are reviewed separately by specifically appointed risk
review committees.
Even though audit reports were not always made available for review, the depositories stated
that during the last audit no material exceptions were found. Only the NSDL (India) reported
that minor exceptions were found (no further details are available).

Data security
All depositories have implemented rigorous safeguards to ensure data security and
protection, such as unique passwords and user IDs (subject to regular change) and lockout
facilities, and depositories holding physical certificates employ guards to protect the vaults.

Availability of delivery versus payment settlement

The settlement process is the source of a variety of important risks and should therefore be
carefully analysed when assessing the soundness and safety of a central securities
depository. In view of the importance of this section, a detailed description of the settlement
process and some of the risk mitigators addressing certain aspects of the operational
processes for each country/depository in the analysis has been included in Annex 5.

2.4.5 Systemic risk

The establishment of business recovery plans
All depositories have detailed business recovery plans in place covering physical equipment,
software and data security as well as organisational structure. Most depositories also have
backup locations where business can be resumed if normal activity has been disrupted.
Exceptions to this rule are the KSD (Korea; planned for 2004) and the RBI (India), where
both the main and the backup centre are located at the same premises. Generally, the full

306 BIS Papers No 30

details of the business continuity and recovery plans are not divulged to the public for
security reasons.
In most instances, data are backed up electronically and stored off-site. Most depositories
have a documentation retention policy ranging between five (one case) and 20 years (one
case). The most common retention periods are either seven or 10 years. RBI (India) and BI
(Indonesia) did not disclose their retention policy guidelines.
Operational procedures are generally tested on a regular basis (at least once a year).
Results are usually not disclosed, or disclosed only in broad terms or in local language
(CSDCC (China), BOJ (Japan)). In some cases, testing frequencies are not divulged
(CSDCC Shanghai). MAS (Singapore) publishes full test results.
In case of a failure, the depositories have comprehensive communication policies to inform
the public and the authorities as well as the regulators.
The backup facility can generally be activated within one to four hours. KSD (Korea)
confirmed backup facilities can be made operational within three hours, NSDL (India)
reported a 24-hour required lead time and the TSCD (Taiwan) a four-to eight-hour required
lead time.
None of the depositories has had to activate the emergency plans during the past three
years, except for the TSCD (Taiwan; no details have been provided).

2.4.6 Conclusion
The analysis of Asian central securities depositories reveals, in our opinion, that the
depositories reviewed comply with the criteria outlined in Rule 17f-7 of the US Investment Act
of 1940. The structural and operational framework appears to be on a level consistent with
that of other central securities depositories serving comparable securities markets.
As outlined, the broad criteria that depositories must comply with to be considered an eligible
central securities depository under rule 17f-7 are:
• acting as a system of central handling of securities;
• being regulated by a financial regulatory authority;
• holding assets of all participants on equivalent terms;
• identifying and segregating participant assets;
• reporting periodically to participants;
• being examined at regular intervals by a regulator or independent accountant.
Some of the most noteworthy criteria we should retain from the above analysis are that
depositories should:
• consolidate the depository functions to service all securities in a domestic market;
• create real-time RTGS processing models, as these effectively mitigate risks, even
though they tend to be more costly due to the liquidity requirements;
• set up effective support systems, such as securities lending and access to credit
facilities, to ensure the system’s efficiency and effectiveness;
• confirm trade details early on in the process and harmonise settlement cycles to
allow further efficiencies;
• establish integrated and real-time links with the payment systems;
• remove physical certificates from the financial system and keep securities in
immobilised or dematerialised form (where legally possible);

BIS Papers No 30 307

• safeguard the soundness of the system by segregating proprietary and client assets
and by adopting enhanced risk management tools, such as insurance or guarantee
funds;
• subject the infrastructure and its external links to periodic independent reviews.
Payment systems should:
• be centralised, with a single payment system responsible for settlement-related
payments;
• be fully integrated with the securities depository;
• still be owned, operated or closely linked with the central bank to allow payments to
be made in central bank funds;
• be structured as RTGS or continuous net settlement systems;
• allow for electronic transfers using secure systems achieving same day value and
finality.
Other important observations we made relate to:
• Validation of a depository’s financial strength: This validation is made either through
the review of financial statements, ownership structure or risk management tools.
Asia’s government bond settlement infrastructure is still largely concentrated with
central banks or affiliates, for which this information may not be readily available.
Central banks remain fundamental to the servicing of government bonds and for the
operation of the payment systems because they:
– act as neutral participants at the centre of the system;
– receive a seal of confidence from the market;
– retain important oversight functions in order to:
• maintain the safety of payments and the payment system;
• maintain the safety of the clearing and settlement system;
• promote the clearing and settlement system’s efficiency.
The corporate bond market has evolved towards the use of non-central bank
intermediaries, in line with the practices identified in the equity markets, for which
financial statements and ownership information can be more easily obtained.
• Mitigation of systemic risk: A critical element in an environment where CSDs
become more widespread and the technology advances rapidly. Default in one part
of the payment, clearing, settlement and safekeeping system may cause a
contagion effect and affect the entire clearing and settlement infrastructure.
The soundness of the overall operating and regulatory framework and the
depositories’ business recovery plans are important elements in all due diligence
reviews. We have not found any major shortcomings with regard to regulatory or
business recovery issues in our analysis. Only the fact that the KSD (Korea) and
RBI (India) have their main and backup centres on the same premises is perceived
as a potential source of risk.
• Internal and external audit reviews: these reviews are important in assessing the
depositories’ operating processes. Traditionally, central banks have not been
subject to the same rigorous audit requirements as privately operated
intermediaries. However, the assumption that they enjoy the backing of their
respective governments in case of adverse events alleviates some of these
concerns.

308 BIS Papers No 30

 Possible general areas for improvement on the audit requirement are to:
– include audit reviews of the interlinked operating systems, such as payment
systems or clearing houses;
– publish the audit results to make them freely available for external review.
• Asset exposure risk: These risks are mainly contained through the implementation
of true DVP and RTGS systems. Austraclear (Australia), CMU (Hong Kong; real-
time settlement), BOJ (Japan; DVP settlement), BNM (Malaysia), MAS (Singapore)
and TSD (Thailand) have implemented real-time gross settlement systems.
In countries where no true DVP systems are available, additional measures have
been put in place to address the liquidity and asset exposure risk.
– In Hong Kong (CMU), under the batch settlement mode, securities are put on
hold until payments have been confirmed.
– In the Philippines (BTR) and Thailand (BOT), investors face intraday exposure
due to the timing differences of the cash and securities settlement process.
– In China (CSDCC) and Taiwan (TSCD), securities, transferred on settlement
date, become available to the investor the next day, when the payment has
been confirmed.
– In India at the NSDL, an intraday exposure exists due to the timing differences
between the pay-in and payout schedules.
– At the Reserve Bank of India, if settlement occurs through RBI accounts, no
exposure exists. However, if cash settlement occurs outside the RBI, timing
lags and exposures exist.
• The applicable settlement cycles: Most countries have adopted a TD+3 (at the
latest) rolling settlement cycle for fixed income instruments, which corresponds to
the generally accepted G30 time frame.
Some authorities are considering shortening the settlement cycle to limit the market
and foreign exchange risks created by possible price fluctuations. Even though this
risk reduction effect is highly desirable, in order to attract foreign investors, it is
important to keep settlement cycles aligned with the foreign exchange contract
timing to allow timely and proper funding of the cash accounts.
• Liquidity risk: Gross and net settlement systems are subject to different risks. The
mechanics of a gross RTGS settlement system, for example, create larger liquidity
exposures, which call for proper risk management tools.
– In most countries, credit facilities are not available through the central
depositories directly, but through the domestic commercial banks.
– In Malaysia, lending facilities are not available to foreign investors.
– In India (NSDL), there is only limited securities lending activity (not allowed for
foreign investors) and the access to credit is uncertain.
– In China (CSDCC), liquidity risk is mitigated through securities lending
programmes and the existence of guarantee funds.
– In Japan (BOJ), liquidity risk may arise under the non-DVP settlement method.
These risks are addressed by making intraday credit limits available.
– Both in Thailand (BOT), for BAHTNET users, and Malaysia (BNM), the
depositories themselves extend credit to their participants.
– In Indonesia, access to credit facilities is restricted, even through commercial
banks.

BIS Papers No 30 309

 – In Taiwan (TSCD), some risk elements are present, as participants only have
limited access to securities lending facilities.
– In Korea (KSD), the fail-related lending programmes maintain stability in the
market.
– In the Philippines and India, the use of cheques to make settlement-related
payments introduces a risk element in the settlement process.
– The introduction of a repo market in China is considered as a positive
development.
• Counterparty risk:
– Through their design, RTGS settlement systems eliminate counterparty risk.
– Central depositories typically mitigate counterparty risk by implementing
rigorous participant membership criteria and adopting stringent membership
monitoring tools.
– If settlement is postponed due to insufficient balances of securities or cash,
principal risk is taken out of the system but is replaced by market risk.
– Failed settlements may attract market risk and expose counterparties to
consequential losses. In Hong Kong (CMU), for example, failed trades are
automatically cancelled from the system at the end of the day and must be
re-entered.
– Central depositories generally do not act as a central counterparty, except for
TSCD (Taiwan), CSDCC (China) and BOT (Thailand; payments only). As
such, counterparty exposure is reduced and guarantee funds have generally
not been put in place.
– Where guarantee funds have been implemented to safeguard financial
stability, it is important to analyse the size, payout criteria and loss-sharing
provisions of the fund.
– Central banks typically do not enter into settlement assurance provisions or
guarantee fund arrangements.
– In both India and the Philippines, cheque use creates counterparty risk on
payment banks.
In summary, we believe it is fair to conclude from the analysis that most countries are
continuously looking at enhancing the clearing and settlement infrastructure to mitigate risks
and reduce or contain costs. An illustration of this trend can be found in the implementation
of the BI-SSSS in Indonesia. With the introduction of this scripless settlement system, the
DVP, liquidity and operational risks inherent in the previous manual processes have been
effectively mitigated.

2.5 Possible future central securities depository models for Asia

2.5.1 Introduction
After reviewing a selected number of central depositories in the region and the conclusions
we have drawn, it is important to look at how the infrastructure could evolve over time.
Based on the observations about the current Asian models and taking into account trends
around the world, we would like to review the following three options:
• interlinked depositories supported by open access to different domestic systems;

310 BIS Papers No 30

• a full integration/consolidation model;
• a private, central institution linked to the domestic market infrastructure.

2.5.2 Interlinked depositories with open access

In this model, links are established between the different providers, allowing for partial
consolidation and a high degree of interoperability. This model can be seen as a first step
towards full integration (please refer to the next paragraph).

Advantages
• Compared to a highly fragmented infrastructure, this model allows for partial risk
mitigation (systemic, market, operational and liquidity risks) and potential cost
efficiencies.
• Competitive element is retained through the coexistence of several providers.
• Implementation of increased transparency.

Disadvantages
• Investors/participants are required to establish memberships at various systems.
• Investment in interoperability is required - interoperability refers to technical
compatibility of systems but also includes standardised communication/messaging,
fees, contracts and procedures.
• Each component of the system must be efficient to ensure stability and robustness
of the entire system.
• The ongoing costs of existing separate entities (maintenance and innovation) remain
- there are few economies of scale.
• The costs associated with linking multiple back office systems are important.
• The integration of the component systems may be complex.

2.5.3 Full consolidation

A fully consolidated model, either within a given market or regionally (United States/Europe)
requires close oversight by the authorities and the regulators to ensure a smooth functioning
of the system. The integrated entity must be sound, backed with the necessary financial
strength, and have stringent risk management tools and a strong and wide acceptance from
participants. Generally, standardised and integrated processes reduce some of the major
risks created in a fragmented environment.

Advantages
• Full integration provides important economies of scale.
• Economies of scale promote cost efficiencies. The absence of duplication of
processes and investment reduce the required fixed and ongoing maintenance
costs.
• Largest incentive to innovate and provide wider range of services (eg portfolio
services).

Disadvantages
• Most expensive/complex structure to develop (system and processing
re-engineering).

BIS Papers No 30 311

• Political/regulatory support required. Need for complex regulatory/taxation change.
• Important trade-off between full integration and increased systemic risk due to over-
reliance on a single system and increased risks related to a lack of competition.
• Oversight is required to promote and ensure continued improvements to the system.
The monopolistic situation may result in a lack or absence of incentives to innovate.
• On a pan-Asian basis, harmonised or streamlined processes within a given asset
class (eg settlement cycles) are preferred as this facilitates the consolidation
process and achieves the highest savings for the investors.

2.5.4 Private, central institution linked to domestic market infrastructure

The option of a central institution linked to the local market infrastructure (CSD, CCP,
clearing house) is most valuable in a multicurrency environment (cross-border), and mimics
the setup of an international central depository.
The utility should establish links through its local agents and its users, should be industry-
owned and should either provide all services in-house or contract with various third-party
entities to provide selected services. In order to provide all services in-house, this utility has
to be organised as a bank to offer cash and credit-related services.
Where traditionally domestic central depositories do not have a banking licence and settle
cash through central bank accounts, the central banks’ involvement in a multicurrency
environment becomes impractical, due to differences in time zone, timing and cutoff time.
The utility could address this by offering settlement services using commercial bank money,
through the recently launched Continuous Linked Settlement (CLS) infrastructure.
The utility model requires complex regulatory arrangements, stringent and multiple
risk/concentration/ business recovery measures and acceptance from the participant and
user communities.

2.5.5 Conclusion
When reviewing the outlook for Asia, it is important to evaluate risks in an integrated
environment and in the current context. Today, Asia is a very fragmented region with large
differences between the infrastructures servicing specific instrument types.
The domestic markets are continuing to evolve, as witnessed by the introduction of certain
functionality or initiatives in various markets:
• establishment of rating agencies for corporate bonds (India and Indonesia);
• implementation of the BI-SSSS (Indonesia).
Future enhancements are being prepared to further enhance the development of the bond
markets:
• STP processing (Australia);
• the expansion of international links and cooperation agreements (Hong Kong);
• introduction of the central counterparty concept for bond clearing (Japan, Thailand);
• a proposal to develop a long-term bond market and benchmark yield curve (Korea);
• introduction of a local fixed income exchange (Philippines);
• launch of 10-year government bond futures contract in January 2004 (Taiwan);
• a plan to move the TSD towards a T+2 settlement cycle (Thailand);
• TSD’s development of a Post Trade Integration Project, aimed at integrating all
system functions and centralising systems linked to market participants (Thailand);

312 BIS Papers No 30

• TSD’s plan to extend its service offering to include bonds and to focus on STP
processing initiatives (Thailand).
These initiatives streamline the domestic markets and offer a more robust and efficient
infrastructure, focused on risk reduction and cost efficiencies.
The choice of the future path for the clearing and settlement infrastructure will be based on a
trade-off between risk management, efficiency and costs, while at the same time
implementing a system best suited for potential growth of the market.
Where domestic markets become more sophisticated, depositories could start playing a
bigger role in the servicing of foreign investors’ portfolios.
To be successful, Asian markets must overcome certain barriers, such as:
• technology interfaces: standardised communication and automation are critical in
the development process and proprietary systems should be phased out.
• need for intraday finality: this will eliminate the need for expensive collateral or
maintaining idle cash balances, reducing both funding cost and liquidity risks.
• systems linkage: links must be established allowing information transfers on a risk-
free and efficient basis.
• differences in market mechanics: eliminate differences in settlement periods
between instruments and adopt common market practices.
• regulatory divergence: harmonise the legislation applicable to all relevant market
segments and encourage cooperation between the public and private sectors.
Rationalising the depository functionality may lead to integration or consolidation, but could
potentially give rise to monopolistic issues, especially if the remaining entity is organised as a
private, for-profit entity. Authorities must therefore implement appropriate oversight
procedures to prevent upward cost pressures and to encourage continued innovation.
Taking into account the highly fragmented environment, we believe it is fair to say that the
clearing and settlement environment is most likely to develop gradually, either by linking the
local infrastructures or by introducing a “central utility”. A steady move towards closer
consolidation may be a next logical next step, but based on the important differences
between the regulatory, operational and taxation laws in the respective markets, it is unlikely
that a full regional consolidation/integration is possible.

BIS Papers No 30 313

 Annex 1:
European clearing and settlement environment

Current situation in Europe

France
United Holland
Kingdom Belgium Germany Switzerland Italy Scandinavia
Portugal

Trading LSE Euronext Deutsche SWX/ Borsa Norex

Börse virtx Italiana HEX

Clearing LCH/Clearnet Eurex X-CLEAR CCG Future

Norex
ECP

Depository Euroclear + Crest Clearstream SIS Monte VP

Titoli VPC
VPS
APK

Future outlook of the European marketplace

Two or three Exchange Exchange Exchange

stock exchanges A B C

One or two Clearing house Clearing house

clearing houses 1 2

Two or three Depository Depository Depository

depositories α β γ

314 BIS Papers No 30

 Annex 2:
G30 recommendations

Table 3
G30 recommendations (1)

Recommendation 1a Trade comparisons between direct market participants by T + 0

Recommendation 1b Matched trade details should be linked to the settlement system

Recommendation 2 Indirect market participants to achieve affirmation by T + 1

Recommendation 3a Central depository, broadest possible participation

Recommendation 3b Widest possible range of depository eligible instruments

Recommendation 3c Immobilisation/dematerialisation to the utmost extent possible

Recommendation 3d Compatible rules and practices in case of multiple CSDs

Recommendation 4a Real-time gross settlement system

Recommendation 4b Trade netting system as per “Lamfalussy recommendations”

Recommendation 5 Delivery versus payment as defined by ISSA

Recommendation 6a Same day funds for securities settlement

Recommendation 6b Same day funds for the servicing of securities portfolios

Recommendation 7a A rolling settlement system should be adopted by all markets

Recommendation 7b Final settlement for all trades by T+ 3

Recommendation 8a Securities lending and borrowing should be encouraged

Recommendation 8b Existing regulatory and taxation barriers should be removed

Recommendation 9a ISO Standard 7775 (securities messages)

Recommendation 9b ISO Standard 6166 (ISIN numbering system)

BIS Papers No 30 315

 Table 4
G30 recommendations (2)

Recommendation 1 Eliminate paper and automate communication, data capture and

enrichment

Recommendation 2 Harmonise messaging standards and communication protocols

Recommendation 3 Develop and implement reference data standards (SWIFT, ISIN, BIC)

Recommendation 4 Synchronise timing between different clearing and settlement systems

and associated payment and foreign exchange systems

Recommendation 5 Automate and standardise institutional trade matching

Recommendation 6 Expand the use of central counterparties

Recommendation 7 Permit securities lending and borrowing to expedite settlement

Recommendation 8 Automate and standardise asset servicing processes, including

corporate actions, tax relief arrangements and restrictions on foreign
ownership

Recommendation 9 Ensure financial integrity of providers of clearing and settlement

services

Recommendation 10 Reinforce the risk management practices of users of clearing and

settlement service providers

Recommendation 11 Ensure final, simultaneous transfer and availability of assets

Recommendation 12 Ensure effective business continuity and disaster recovery planning

Recommendation 13 Address the possibility of failure of a systematically important institution

Recommendation 14 Strengthen assessment of the enforceability of contracts

Recommendation 15 Advance legal certainty over rights to securities, cash or collateral

Recommendation 16 Recognise and support improved valuation and closeout netting

arrangements

Recommendation 17 Ensure appointment of appropriately experienced and senior board

members

Recommendation 18 Promote fair access to securities clearing and settlement networks

Recommendation 19 Ensure equitable and effective attention to stakeholder interest

Recommendation 20 Encourage consistent regulation and oversight of clearing and

settlement service providers

316 BIS Papers No 30

 Annex 3:
CPSS/IOSCO recommendations

Table 5
CPSS/IOSCO recommendations

Standard 1 Legal framework

Securities settlement systems should have a well founded, clear and
transparent legal basis in the relevant jurisdictions.

Standard 2 Trade confirmations and settlement matching

Confirmation of trades between direct market participants should occur as
soon as possible after trade execution, but no later than trade date (T + 0).
Where confirmation of trades by indirect market participants (such as
institutional investors) is required, it should occur as soon as possible after
trade execution, preferably on T +0, but no later than T +1.

Standard 3 Settlement cycles

Rolling settlement should be adopted in all securities markets. Final
settlement should occur no later than T + 3. The benefits and costs of a
settlement cycle shorter than T + 3 should be evaluated.

Standard 4 Central counterparties

The benefits and costs of a CCP should be evaluated. Where such a
mechanism is introduced, the CCP should rigorously control the risks it
assumes.

Standard 5 Securities lending

Securities lending and borrowing (or repurchase agreements and other
economically equivalent transactions) should be encouraged as a method
for expediting the settlement of securities transactions. Barriers that inhibit
the practice of lending securities for this purpose should be removed.

Standard 6 Central securities depositories

Securities should be immobilised or dematerialised and transferred by book
entry in CSDs to the greatest extent possible.

Standard 7 Delivery versus payment

CSDs should eliminate principal risk by linking securities transfers to funds
transfers in a way that achieves delivery versus payment.

Standard 8 Timing of settlement finality

Final settlement should occur no later than the end of the settlement day.
Intraday or real-time finality should be provided where necessary to reduce
risks.

Standard 9 CSD risk controls to address participant defaults

CSDs that extend intraday credit to participants, including CSDs that
operate net settlement systems, should institute risk controls that, at a
minimum, ensure timely settlement in the event that the participant with the
largest payment obligation is unable to settle. The most reliable set of
controls is a combination of collateral requirements and limits.

BIS Papers No 30 317

 Table 5 (cont)
CPSS/IOSCO recommendations

Standard 10 Cash settlement assets

Assets used to settle the ultimate payment obligations arising from securities
transactions should carry little or no credit or liquidity risk. If central bank
money is not used, steps must be taken to protect CSD members from
potential losses and liquidity pressures arising from the failure of the cash
settlement agent whose assets are used for that purpose.

Standard 11 Operational reliability

Sources of operational risk arising in the clearing and settlement process
should be identified and minimised through the development of appropriate
systems, controls and procedures. Systems should be reliable and secure,
and have adequate, scalable capacity. Contingency plans and backup
facilities should be established to allow for timely recovery of operations
and completion of the settlement process.

Standard 12 Protection of customer’s securities

Entities holding securities in custody should employ accounting practices
and safekeeping procedures that fully protect customers’ securities. It is
essential that customers’ securities be protected against the claims of a
custodian’s creditors.

Standard 13 Governance
Governance arrangements for CSDs and CCPs should be designed to fulfil
public interest requirements and to promote the objectives of owners and
users.

Standard 14 Access
CSDs and CCPs should have objective and publicly disclosed criteria for
participation that permit fair and open access.

Standard 15 Efficiency
While maintaining safe and secure operations, securities settlement
systems should be cost-effective in meeting the requirements of users.

Standard 16 Communication procedures

Securities settlement systems should use or accommodate the relevant
international communication procedures and standards in order to facilitate
efficient settlement of cross-border transactions.

Standard 17 Transparency
CSDs and CCPs should provide market participants with sufficient
information for them to identify and evaluate accurately the risks and costs
associated with using the CSD or CCP services.

Standard 18 Regulation, supervision and oversight

Securities settlement systems should be subject to transparent and
effective regulation and oversight. Central banks and securities regulators
should cooperate with each other and with other relevant authorities.

Standard 19 Risks in cross-border links

CSDs that establish links to settle cross-border trades should design and
operate such links to effectively reduce the risks associated with cross-
border settlements.

318 BIS Papers No 30

BIS Papers No 30

Annex 4:
CSDs/CCPs and payment systems in Asia

Table 6
CSDs/CCPs and payment systems in Asia

Government bonds Corporate bonds

Central Central
Payment Clearing Payment Clearing
CSD clearing CSD clearing
system house system house
counterparty counterparty

Australia Austraclear RTGS n/a Austraclear Austraclear RTGS n/a Austraclear

China The China Two payment CSDCC CSDCC Same as Same as Same as Same as
Securities systems are Shanghai/ Shanghai/ government government government government
Depository and used to settle Shenzhen Shenzhen bonds bonds bonds bonds
Clearing Corp cash: a paper-
(CSDCC), based credit
Shanghai and advice
Shenzhen collected by
branches the central
bank (PBOC),
and an
electronic real-
time payment
system, China
National
Advance
Payment
System
(CNAPS)
319
320

Table 6 (cont)
CSDs/CCPs and payment systems in Asia

Government bonds Corporate bonds

Central Central
Payment Clearing Payment Clearing
CSD clearing CSD clearing
system house system house
counterparty counterparty

Hong Kong Central Clearing House None for those n/a CMU Clearing House n/a n/a
SAR Moneymarkets Automatic settled via CMU Automatic
Unit (CMU)/ Transfer HKSCC Transfer
Settlement via clearing house
Central System System
CCASS: under for securities
Clearing and (CHATS), (CHATS),
continuous net traded on the
Settlement operated by settlement sys- Stock operated by
System Hong Kong tem, HKSCC Hong Kong
(CCASS) Interbank Exchange of Interbank
settlement Hong Kong
(selected Clearing counterparty to Clearing
Exchange Fund Limited both buying and Limited
Notes are (HKICL) selling broker (HKICL)
traded on through
CCASS) novation

India Reserve Bank Cheque, n/a Clearing National Cheque, n/a BOI Shareholding
of India - Public pay-order, Corporation of Securities pay-order, Ltd (BOISL) is
Debt Office banker’s India Limited Depository Ltd banker’s clearing house for
cheque or (CCIL) (NSDL) and cheque or trades on the
Reserve Bank Central Reserve Bank Stock Exchange,
of India cheque Depository of India cheque Mumbai (BSE),
Services (India) National
Interbank Interbank Securities
Ltd (CDSL)
BIS Papers No 30

clearing - RTGS clearing - RTGS Clearing Corpo-

(expected 2004) (expected 2004) ration Limited
(NSCCL) is
clearing corpo-
ration for trades
on National Stock
Exchange (NSE)
BIS Papers No 30

Table 6 (cont)
CSDs/CCPs and payment systems in Asia

Government bonds Corporate bonds

Central Central
Payment Clearing Payment Clearing
CSD clearing CSD clearing
system house system house
counterparty counterparty

Indonesia Bank Bank n/a Bank KSEI C-BEST KPEI KSEI

Indonesia Indonesia Indonesia (scripless) (scripless) (scripless) (scripless)
through
interbank
clearing - real-
time gross
settlement
(RTGS)
system

Japan Bank of Japan (a) BOJNet n/a n/a CBs: JASDEC (a) FXYCS On-market: On-market:
(Japan (Foreign JSCC (Japan JSCC (Japan
(b) FXYCS Future 2005: Future 2005:
Securities Exchange Securities Securities
(Foreign JGBCC (JGB JGBCC (JGB
Depository Yen Clearing Clearing
Exchange Clearing Clearing
Center, Inc) Clearing Corporation) Corporation)
Yen Corporation) Corporation)
System)
Clearing Others: n/a Off-market: n/a Off-market: n/a
System) (b) BOJNet
(c) Zengin (c) Zengin
System System
(online (online
domestic domestic
yen fund yen fund
transfer and transfer and
remittance remittance
system) system)
321
322

Table 6 (cont)
CSDs/CCPs and payment systems in Asia

Government bonds Corporate bonds

Central Central
Payment Clearing Payment Clearing
CSD clearing CSD clearing
system house system house
counterparty counterparty

Korea KSD BOK-Wire KSE KSE KSD BOK-Wire KSE KSE

(Korea (Bank of (Korea Stock
Securities Korea-Wire) Exchange)
Depository)

Malaysia Bank Negara RENTAS (real- n/a n/a (a) KLSE (a) RENTAS (a-i) Securities (a) Securities
Malaysia time gross eligible Clearing Clearing
(b) RENTAS
(BNM) settlement bonds - Automated Automated
system Malaysian Network Network
managed by Central Sdn Bhd Sdn Bhd
BNM) Depository (SCANS) (SCANS)
Sdn Bhd only for
(b) n/a
(MCD) trades
settling
(b) SSTS
via Institu-
eligible
tional
bonds -
Settlement
Bank
Services
Negara
(ISS).
Malaysia
(BNM) (a-ii) n/a for
trades
BIS Papers No 30

settling
via MCD
transfer
mecha-
nism
(b) n/a
BIS Papers No 30

Table 6 (cont)
CSDs/CCPs and payment systems in Asia

Government bonds Corporate bonds

Central Central
Payment Clearing Payment Clearing
CSD clearing CSD clearing
system house system house
counterparty counterparty

Philippines Registry of Payment n/a n/a n/a Over the n/a n/a
Scripless settled counter
Securities between
(RoSS) counterparties
via cheques
Optional:
Bangko
Sentral ng
Pilipinas (BSP
- central bank)
RTGS (Q1
2004)

Singapore MAS MEPS n/a MAS CDP MEPS n/a CDP

(scripless) (scripless) (scripless)

Thailand Bank of BAHTNET Bank of Bank of Thailand BAHTNET Thailand Thailand

Thailand *(1) (RTGS) Thailand Thailand Securities (RTGS) *(2) Securities Securities
Depository *(2) Depository Depository
323
 Annex 5:
Settlement process in Asian countries

BIS settlement models:

Model 1: Securities and funds are transferred on a simultaneous, irrevocable and real-time
gross settlement (RTGS) basis.
Model 2: Securities are settled on a gross basis and cash is settled on a net basis.
Model 3: Both securities and cash are exchanged simultaneously on a net basis once a day.

Australia: Austraclear
Settlement process: BIS Model 1
• Settlement cycle: negotiable but generally on a rolling TD + 3 basis.
• Availability of securities is checked in seller’s account and position earmarked.
• If sufficient funds are available, settlement occurs in Austraclear.
• Transactions which would create a securities shortfall are rejected.
• Payment: via a feeder system to the central bank’s RTGS system.
Risk mitigators
• Matching: required but not legally binding. Matched but unsettled trades are
automatically deleted at the end of the settlement day.
• Securities lending: allowed in the market but not offered by Austraclear.
• Buy-in: there are no established buy-in procedures for debt instruments.

China: China Securities Depository and Clearing Corporation Ltd (CSDCC)

Settlement process: similar to BIS Model 2. Note: net fund movement occurs on the day after
the final movement of securities.
• Settlement cycle: securities are settled on TD on a gross trade-by-trade basis; cash
is paid on a net basis on TD +1.
• On the evening of TD, CSDCC transfers shares on a final and irrevocable basis.
• On TD +1, funds are transferred to and from the clearing bank designated by the
CSDCC.
Risk mitigators
• Matching: unclear whether a prematching or affirmation process is active.
• Brokers check the availability of securities/cash with the QFII’s appointed custodian
bank before executing trades. Trades are rejected in the event of insufficient
holding.
• Once executed, the trades are binding on the brokers.
• Securities lending: short-selling, securities lending and borrowing are not allowed.
• Buy-in: no buy-in/sell-out rules. In case of overdraft, CSDCC imposes interest
penalties based on the amount of the overdraft and holds securities as collateral.

324 BIS Papers No 30

• The depository acts as the central counterparty for all participants and guarantees
both securities and cash settlements.

Hong Kong: Central Monetary Unit (CMU)

Settlement process
• Settlement cycle: negotiable. Generally, trades executed in the morning are settled
on the same day, while those executed in the afternoon are usually settled the next
day.
• Settlement can be effected using the real-time or the batch settlement method.
Under both methods, cash and securities move simultaneously. Transfers are final
and irrevocable.
• Payment happens through CMU interface with the real-time interbank payment
system, known as the Clearing House Automatic Transfer System (CHATS).
Settlement process: real-time settlement - BIS Model 1
• Instructions are automatically matched by CMU and securities are put on hold.
• Once payment is confirmed to the CMU, the securities are released.
• In the event of insufficient securities, trades are transferred to the end-of-day batch
settlement process.
• If insufficient funds are available, transactions remain pending until funds are
available.
Settlement process: batch settlement - BIS Model 3
• At 3.30 pm, the end-of-day batch run starts.
• Net settlement obligations (securities and cash) calculated and available balances
checked.
• Settlement occurs if sufficient funds and securities are available.
• Failed transactions are cancelled at the end of each day.
Risk mitigators
• Matching: automatically upon receipt of settlement instruction.
• The CMU monitors participants’ cash and securities accounts on a continuous basis.
• Securities lending: in December 1997, HKMA introduced a securities lending
programme for eligible private debt instruments held through the CMU.
• Buy-in: no established buy-in procedures. Generally, counterparties renegotiate the
settlement of any failed trades.

India: National Securities Depository Limited (NSDL)

Settlement process
• The inter-broker settlement model is similar to BIS Model 3, whereas the broker-
custodian settlement process resembles BIS Model 2.
• Settlement does not occur on a simultaneous basis.
Settlement process: clearing house trades
• Settlement cycle: TD + 2 rolling settlement basis.

BIS Papers No 30 325

• Custodian banks advise the clearing house on TD +1 of its intent to settle on TD + 2.
• Once a custodian confirms the trade to the clearing house, settlement is binding.
• To effect settlement, the custodian delivers cash or securities to the clearing house
during pay-in and receives securities or cash from the clearing house during payout.
• The period between pay-in and completion of payout is four to eight hours.
• On pay-in date, securities are transferred to a “pool” account at the NSDL for further
credit to the clearing house’s account during the day.
• The net cash amount is paid by cheque to the clearing house on TD + 2 (at 10.30
am).
• On TD + 2, the clearing house transfers securities through the NSDL (at 2 pm).
Settlement process: hand delivery trades
• Settlement occurs outside of the clearing house.
• To settle a trade, both delivery and receipt instruction have to be put in NSDL
system.
• On settlement date, brokers initiate payment either by account transfer or by
cheque.
• Cheques (over INR 100,000 (about USD 2,050)) submitted prior to 10.30 am on SD
or cheques (under INR 100,000), presented on SD – 1 before noon receive same
day value.
• As soon as the investor receives cleared funds or securities, delivery of securities or
payment to the broker is initiated (no fixed cutoff time).
• Transfer of securities is final and irrevocable.
Risk mitigators
• Matching: not mandatory and outside the depository. Generally, only matched
instructions are entered in the depository system for settlement.
• Securities lending: While securities lending exists, foreign investors cannot borrow
securities in accordance with local regulations.
• Buy-in: Short positions between brokers are resolved through auctions on SD + 1 for
settlement on SD + 2. If securities cannot be bought in the auction, the trade is
closed out as per SEBI guidelines. Special rules apply for failed trades near ex-date.

India: Reserve Bank of India (RBI)

Settlement process
• For larger trades (above NIR 200 million), the counterparties must indicate whether
settlement will occur through the clearing house, CCIL or directly with RBI.
• As all participants maintain cash accounts with the RBI, payments are settled
through the central bank accounts.
Settlement process: settlement via CCIL - BIS Model 2
• The CCIL generates a securities and cash obligations report for each participant.
• CCIL sends details of the participants’ gross securities and net cash obligations to
RBI.

326 BIS Papers No 30

• Upon verification that the securities and cash are available, the RBI debits and
credits securities on a gross basis and subsequently processes the net cash
settlement.
• CCIL receives settlement confirmation from RBI and informs its members.
Settlement process: settlement with RBI directly - BIS Model 1
• On SD (no specific time), RBI initiates simultaneous transfer of securities and cash if
sufficient positions are available.
Risk mitigators
• Matching: between counterparties. The matching is binding and the trade is
designated “ready for settlement” on the system, viewable by both the RBI and the
CCIL.
• Securities lending: Repos exist, foreign investors are not permitted to borrow
securities.
• Buy-in: under both methods, if sufficient balances are not available, settlement is
postponed until the counterparties resolve the discrepancies and/or shortages (this
eliminates principal risk, but not market risk).
• RBI levies strict penalties on participants for settlement failures.

Indonesia: Bank Indonesia (BI)

Settlement process: in February 2004, BI introduced its BI-SSSS (Bank Indonesia - Scripless
Securities Settlement System). Even though it is still too early to assess the performance of
this new system accurately, a few general observations can be made:
• The scripless system is a big step forwards from the previous manual process.
• The new scripless settlement model conforms with BIS Model 1.
• With the implementation of SSSS, government bonds and SBIs are settled on a
DVP basis, through the securities and payment systems, SSSS and RTGS.
• The scripless system is expected to reduce operational risk and improve the
efficiency of the government bond settlement infrastructure.
• Paper forms are eliminated and securities module linked to the RTGS payment
system.

Japan: Bank of Japan (BOJ)

• Settlement at the BOJ occurs in book entry form either on DVP or non-DVP basis.
Settlement process: DVP basis - BIS Model 1
• Settlement cycle: rolling TD + 3 settlement basis via the new JGB book entry system.
• Settlement occurs on a real-time gross DVP basis between 9 am and 3 pm.
• Due to the real-time nature of the process, sufficient securities/cash balances are
required.
• Pending securities receipts, pending receipts of funds or pre-advised funds are not
considered good securities or funds until received.

BIS Papers No 30 327

• Prearranged credit line can be arranged to cover cash obligations, restricted to JGB
settlement only. The use of the credit facility must be fully collateralised and will
incur an intraday funding costs.
Settlement process: non-DVP basis - no BIS model applies - cash settled on a net basis on
SD by 2.30 pm and securities on an RTGS basis by 12 pm.
• Securities are settled on an RTGS basis at the BOJ. Cash settlement usually occurs
on a net basis via the Foreign Exchange Yen Clearing System (FXYCS).
• If the cash account is not properly funded, an intraday overdraft limit will be used.
• Cash is settled during a batch run at the end of the day, therefore, no intraday
funding cost or intraday line usage charges apply.
Risk mitigators
• Matching: either by phone or through the NTT matching system on TD + 1. Not
binding.
• Securities lending: lending is allowed in the Japan market, subject to stringent
conditions with regards to tenure and collateral requirements.
• Buy-in: no formal buy-in practices. JGB fails are generally resolved directly by the
counterparties through mutually negotiated interest claims. If a trade fails and
remains outstanding for more than 10 business days, a buy-in may be initiated.

Korea: Korea Securities Depository (KSD)

Settlement process: BIS Model 3
• Settlement cycle: negotiable but generally on a rolling TD + 2 basis.
• Securities are settled on a gross basis through the KSD whilst the cash transfers
occur through BOKWire (amounts above KRW 1 billion) or manager’s cheque
(amounts less than KRW 1 billion).
• Under the Institutional Affirmation & Settlement System (INAS), KSD acts as an
intermediary to all payment obligations through its account with the Bank of Korea,
but does not assume participants’ default risk.
• Brokers use proprietary holdings to settle if investors hold insufficient positions.
Risk mitigators
• Matching: indirect participants prematch trades against the electronic trade report
received from KSD. If discrepancies occur, amended details are sent to the KSD
and new trade reports are generated.
• Unmatched transactions which were not rectified are not included in the net position
report and can only be settled directly between the broker and the underlying
investor.
• Buy-in: as per local regulations, buy-in/sell-out on SD + 2.
• If an investor continues to fail his obligations, the regulator may impose sanctions,
including the suspension of the Investment Registration Card (IRC).
• Securities lending/borrowing: KSD established detailed procedures/acts as
intermediary.

328 BIS Papers No 30

Malaysia: Bank Negara Malaysia (BNM)
Settlement process: BIS Model 1
• Settlement cycle: negotiable but generally occurs on TD + 1.
• Once a trade is completed, details are uploaded and validated in SSTS.
• SSTS transmits an unconfirmed sale/delivery advice to the receiving party.
• On SD, the receiving party confirms the transaction in SSTS.
• SSTS initiates the transfer in book entry form in the accounts with BNM.
• Securities are transferred within SSTS; cash is transferred via BNM’s RTGS
payment system, RENTAS, by debiting or crediting the participants’ cash accounts
held at BNM.
Risk mitigators
• Matching: no matching procedures.
• Buy-in: in the event a settlement fails, BNM initiates a buy-in. If the securities cannot
be bought in by 6 pm on settlement date, BNM may reverse the original transaction.
• Trades not confirmed by the receiving party remain in the system until cancelled.
• Securities lending: lending and borrowing have currently been suspended.

Philippines: Bureau of Treasury (BTR)

• Settlement cycle: same day settlement, ie on trade date (TD).
• Cash payments are generally made outside the Registry of Scripless Securities
(RoSS) and are settled directly between the counterparties. Cash can be settled by
the BTR but this is very uncommon.
Settlement process: cash settlement outside RoSS
• Upon trade execution, cash is paid on a gross basis by manager’s cheque or
cashier order.
• Cheque or cashier orders deposited before 12 noon are given same day value.
• Once payment is made, settlement details are uploaded and confirmed into RoSS
and transactions are matched by the BTR.
• Matched trades are settled on a gross basis before the end of the day (no specified
time).
Settlement process: cash settlement through the central bank - BIS Model 2
• For matched trades, the BTR prepares payment obligations and instructs the central
bank to make the appropriate cash transfers.
• The BTR transfers securities on a gross settlement basis in RoSS.
• Although lines of credit are available to cover insufficient cash balances, generally,
instructions are not uploaded into RoSS if funds are not available.
Risk mitigators
• Matching: through RoSS on trade date. Unmatched trades are considered as failed
and are automatically deleted from the RoSS system.
• Buy-in: no established procedures.

BIS Papers No 30 329

• Securities lending/borrowing: no established procedures.

Singapore: Monetary Authority of Singapore (MAS)

Settlement process: BIS Model 1
• Settlement cycle: (TD + 1) for regular trades and on trade date (TD) for cash trades.
• Regular trades, affirmed on TD, settle at 9 am on TD + 1 on a true DVP basis. In the
event of disagreement, buyer can resolve issues with seller up to 4.30 pm on TD + 1.
• Cash trades, once confirmed, settle on a true DVP basis.
• Settlement (securities and cash) occurs on a gross basis.
• Settlement occurs through MAS Electronic Payment System (MEPS), an RTGS
system for high value interbank fund transfers (IFT) and scripless Singapore
Government Securities (SGS).
Risk mitigators
• Matching: trade affirmation/confirmation via MEPS. Amendments subject to mutual
agreement, provided settlement has not occurred. If a dispute is not resolved, the
matter is referred to the SGS Market committee for a final decision.
• Buy-in: no established buy-in procedures.
• Securities lending: a repo facility is available for primary dealers, allowing them to
borrow benchmark issues.

Taiwan: Taiwan Securities Central Depository (TSCD)

Settlement process: BIS Model 3
• Settlement cycle: TD +1.
• Matched transaction reports are sent to the TSCD, the direct and the indirect
participants.
• TSCD prepares summary reports with matched trades and net securities/cash
obligations for every broker. Discrepancies must be reported to TSE/TSCD before
the end of TD +1.
• On SD (TD +1), participants verify and confirm settlement details. If verification is
not possible before the settlement cutoff time, brokers may cancel the trade, or
settle it with an undertaking to resolve any disputes with the investor at a later date.
• Securities are automatically transferred into the buyer’s account at the TSCD during
the overnight batch-run, unless a failed trade is reported by the broker to the TSE.
• The securities are available only if payment has been confirmed on SD +1. Any
delivery failure should be reported to the TSE in the evening of SD.
• Cash is settled via an electronic interbank wire transfer system on SD +1 based on
the net cash obligations indicated on the summary report for on-exchange trades.
• The TSE reports any payment defaults to the TSCD, which puts a hold on the
securities in the defaulting broker/ investor’s account at the TSCD.
Risk mitigators
• Matching: automated and binding between brokers through FAST trading system.
Indirect participants prematch trades based on the brokers’ trade execution report.

330 BIS Papers No 30

• Buy-in: failed trades are prohibited and may cause the investor’s investment license
to be revoked for three years.
• Securities lending: TSCD does not have any securities lending and borrowing
programme. However, foreign investors are allowed to lend stocks to local
borrowers (ie brokers) who need to cover settlement shortfalls due to failed trades or
transaction errors.

Thailand: Bank of Thailand (BOT)

Settlement process: BIS Model 1
• Settlement cycle: rolling TD + 2 settlement basis.
• On SD, for matched instructions, BAHTNET/2 checks the availability of securities
and cash and settles securities only upon receipt of funds.
• Government securities are cleared and settled electronically on a gross delivery
versus payment (DVP) basis through BOT’s BAHTNET/2 system, the country’s
RTGS system.
• Matched instructions that were not settled due to insufficient cash or securities are
automatically deleted at the end of the day.
Risk mitigators
• Matching: via phone. Matched instructions can be deleted only upon mutual
agreement.
• Buy-in: no established procedures.

Thailand: Thailand Securities Depository Co Limited (TSCD)

Settlement process: BIS Model 3
• Settlement cycle: rolling TD + 3 settlement basis.
• On SD, the TSD nets securities and cash obligations. The TSD effects net cash
transfers from the TSD BAHTNET account at BOT about 45 minutes after the
transfer of securities and only upon ascertaining the availability of the securities.
• If the event shares or funds are not available, the TSD must be informed and a new
net report is generated before settlement.
• If there are insufficient funds, the TSD uses the Equity Clearing Fund to settle the
trade. On SD +1, the TSD will sell the securities to reimburse the Clearing Fund and
related costs and impose a fine on the defaulting broker.
Risk mitigators
• Matching: through the TSD Net Clearing system. The matching process is not
binding for custodian banks.
• Buy-in: initiated on TD + 5 for settlement on TD + 8.
• Securities lending: mandatory borrowing programme. TSD acts as a principal, for
the account of the broker. This programme is subject to stringent rules and
penalties.

BIS Papers No 30 331

Selected bibliography
Bank for International Settlements (1997): Real-time gross settlement systems, Basel,
March.
Bank for International Settlements/International Organization of Securities Commissions
(2002): Assessment methodology for “Recommendations for securities settlement systems”,
Basel, November.
Bank of New York (2003): internal research.
Blattner, N (2003): The financial system and regulatory challenges in securities clearing and
settlement system, Lucerne, June.
CESR/ECB (2004): Summary of responses on the European System of Central Banks
(ESCB) and Committee of European Securities Regulators (CESR) joint consultation on
standards for securities clearing and settlement systems in the European Union, January.
Committee on Payment and Settlement Systems/International Organization of Securities
Commissions (1997): Disclosure framework for securities settlement systems, Basel,
February.
The Giovannini Group (2001): Cross-border clearing and settlement arrangements in the
European Union, Brussels, November.
——— (2003): Second report on EU clearing and settlement arrangements, Brussels, April.
Group of Thirty (2003): Global clearing and settlement: a plan of action, Washington DC.
HSBC (2003): internal research.
Terol, I R and D Russo (2000): The euro area securities clearing and settlement
infrastructure: recent changes and issues for debate, April.
US Securities and Exchange Commission (2002): Interagency white paper on structural
changes in the settlement of government securities: issues and options, October.

332 BIS Papers No 30

 Comments on Francis Braeckevelt’s paper
“Infrastructure in the Asian bond markets:
clearing, settlement and depository issues”

Aaron Low

This paper provides an excellent and comprehensive survey of the differences in

mechanisms of Asia’s bond settlement and clearing systems. More importantly, the author
goes on to propose various models for achieving a coherent structure that could alleviate
some of the problems plaguing the current infrastructure.
It is easy to agree that the successful evolution of an efficient Asian bond market depends on
demand and supply side dynamics. And it is also well acknowledged that poor liquidity
conditions and investor (in many cases in Asia, offshore investor) restrictions are
constraining the ability of regional and international players to transact in Asian bonds,
especially local currency issues. Slow acceptance of Asian bond benchmarks is also a major
stumbling block. And on the supply side, the lack of issues and the competition with excess
liquidity in loan markets limit the playing field.
Naturally, good infrastructure, like settlement and clearing systems, facilitates transactions,
increases trading efficiency and minimises costs and risks. Although these issues are
secondary, they are also important in minimising overall development costs in an industry
where yields are at historical lows and in a region where standards are extremely
heterogeneous.

1. Differences in Asian clearing and settlement systems

Francis begins by highlighting the historical evolution of clearing and settlement systems in
the United States, Europe and Asia. While the benefits from homogeneity are evident in the
United States, Europe is still going through a rationalisation and consolidation process that is
aided by capital market deregulation and public sector oversight. While Europe provides the
best learning curve for Asia, the regions have different starting points, although the initial
market structures have similarities. First, Europe is ahead in its convergence, with financial
market integration being an intermediate stage towards a true union. Asia undeniably shows
diversity in standards and regimes. And while the European standardisation process is
proving very costly in terms of efficiency and distortions, the costs for Asia are likely to be at
least comparable if not higher.
The author concludes that domestic systems would probably emerge, with limitations on
convergence in standards. I would be more hesitant to make that prediction. But I would
agree with the author that some form of integration is likely to evolve. With more global
institutional custodians expanding into the region and increasing global capital flows, it may
well be that domestic systems will eventually converge to a global or regional standard. It just
makes market sense for them to do so simply because the successful development of an
embryonic asset class is extremely sensitive to costs in terms of both transaction and
spreads. And with yields at historically low levels, cost consideration is too important a factor
to ignore. An example is the prevalence of multiple layers of custodian services in some
sectors due to divergence in expertise, sophistication or breadth of services offered. It is
costly both for clients and for investment managers to contend with both local and global
custodians, in terms of not only fees, but also risk management, efficiency, time, and
instruction flows. One issue we have found to be pervasive in the region is the lack of

BIS Papers No 30 333

understanding and maturity among institutional investors in the asset class that still prevails.
Naturally, this can only be addressed in time, but it does pose a challenge that cannot be
underestimated.

2. Applicable models for Asia

The paper then proceeds to propose three possible models for Asian bond markets, taking
into account regional differences in existing infrastructure, current transformation trends in
the region, and the basic compatibility of most systems with Rule 17f-7 of the US Investment
Act. The models differ in the extent of integration across the system.
The first proposal, which the author terms the “Interlinked Depository” model, essentially
retains domestic differences while allowing for some degree of integration across the region
between sovereign entities. Absent regional leadership, I would agree that this is the most
realistic and achievable option. While the attraction of this model would be its limited “direct”
costs, its “indirect” costs from the structural inefficiencies that may arise in the “interlinking”
process and its limited scalability in the system are ultimately less attractive.
The second model, the “Full Consolidation” approach, provides more complete integration
and scalability but incurs more direct or developmental costs. The requirement would be the
presence of an oversight body that may be difficult for sovereign countries to agree on.
Third, and perhaps most improbable at this stage, would be the “Private Central Institution”
proposal, which operates seamlessly as if the region were one bloc. The author rejects this
model due to the difficult arrangements that are required from various countries.
While I would agree that the first model presents the most feasible and realistic option, I
would postulate that the overall costs (direct and indirect) might prove higher. Clearly, with
the relatively short period that local economies have had to incubate their bond markets, an
important question that arises is whether this short history constitutes an advantage for the
region in making it possible to bypass the standard domestic development in favour of a
more standardised platform. Global standards are evolving and generally converging to a
commoditised structure, but it does take time to gravitate to an “end equilibrium”. If this were
a process governed by pure market forces, then the speed and direction of that gravitation
would be obvious. But given the structural differences in the region in terms of regulation, tax
and operational issues, it would be hard to disagree with the paper.

3. Conclusions
This paper points to a “first-best” solution that could be reached if Asia were a homogenous
region. But it highlights issues that suggest extremely challenging work to seriously address
the limitations of the current financial infrastructure. A comforting point is the speed at which
some countries have addressed these issues and restructured their systems. The hope is
that while we have seen the impressive speed in the convergence of the real Asian
economies, the day may not be too far away when we finally witness the convergence of the
Asian financial sector.

334 BIS Papers No 30

 Bond market regulation and supervision in Asia

Bernhard Eschweiler 1

Executive summary
Modern economies need efficient financial markets. In Asia, financial market development
has primarily been centred around banking and to some degree equity markets. Bond
markets always played a smaller role and some people believe that the absence of large and
robust local bond markets may have helped to cause the Asian crisis. True or not, Asian
policymakers have focused on local bond market development since the crisis. Indeed, local
bond markets have grown in size, but are still viewed as underdeveloped.
This has raised the question, on the one hand, of whether there are too many restrictions
that hamper market development. On the other hand, there is also the question of whether
current prudential regulatory standards are sufficiently sound. The purpose of this paper is to
analyse and compare the different degrees of liberalisation of Asian local bond markets and
their prudential standards. The key findings are as follows.
• The extent of bond market liberalisation and prudential regulation varies
substantially within the region. Only Hong Kong SAR and Singapore are on a par
with global standards and best practices.
• There is ample room for liberalisation and deregulation in most countries to promote
the development of local bond markets.
• Prudential standards are not grossly out of line with the respective degrees of bond
market liberalisation, but that is no reason for complacency since there is a need for
regulators not to fall behind the changes and developments in the marketplace.
• The basic structure and content of securities regulation in Asia looks increasingly
similar to the model adopted in most other parts of the world, but there are notable
deficiencies in some countries concerning enforcement. On the one hand,
supervisors are often too bureaucratic. On the other, they often lack the ability or
even the will to enforce basic standards.
• A key factor undermining the effectiveness of prudential regulation in some
countries is the general weakness of the legal and accounting infrastructure, which
is partly a function of the prevailing attitude towards common rights versus special
interests.
• The two areas with the biggest weaknesses are issuer disclosure and the prevention
of systemic risks. Disclosure standards for new issues are largely observed, but
regular reporting is weak. Supervisors’ understanding of market positions and
related risks, with regard to both individual investors and intermediaries, is often not
sufficient, largely due to resource issues.
In summary, there is clearly a need to strengthen prudential standards in most countries.
However, this may be better achieved and have a greater impact on bond market

1
The views expressed in this report are those of the author and not necessarily those of JPMorgan.

BIS Papers No 30 335

development if it takes the form of adoption and implementation of global best practices
rather than attempts to harmonise bond market rules and regulations within the region.

Defining regulation
Any meaningful discussion of financial market regulation first requires a description of the
basic regulatory framework. Not so long ago, financial market regulation in Asia consisted
primarily of a set of rules and restrictions that were mostly aimed at ensuring market (and
broader macroeconomic) stability and protecting onshore financial institutions from offshore
competition. In the late 1980s and early 1990s, some of these rules and restrictions,
especially capital controls, were eased. This led to a surge in offshore borrowing and what
followed is well known history.
One of the key lessons of the Asian crisis is that financial liberalisation should not occur in
isolation. Critical for success is the existence of adequate prudential regulation and
supervision that protects investors, ensures that markets are fair and transparent and
reduces systemic risks. Equally important is that the policy regime, especially the exchange
rate regime, is sufficiently flexible to cope with increased capital mobility.
In the years since the Asian crisis, there has been sustained effort in every country to
improve the prudential regulation and supervision of the financial sector and progress has
been made everywhere, although to different degrees. Some countries have also adjusted
their monetary policy frameworks to cope better with the rise in capital mobility, but the
preference for tight monetary policy control, especially over the exchange rate, has remained
strong. As a result, many countries have been slow to ease the foreign exchange and capital
controls they imposed during the Asian crisis.
Another popular conclusion from the Asian crisis is that financial intermediation relied too
much on traditional commercial banking and that most countries lacked strong local capital
markets. In response, every country in the region has made efforts to promote its local bond
market. These markets have grown substantially compared with the years before the Asian
crisis (on average, more than tripling in size), but most people would still view them as
underdeveloped. Much of the market growth has been driven by government issuance, which
in turn was largely a legacy of the Asian crisis. Buy-and-hold investors, mostly commercial
banks, still dominate markets and liquidity is low, while corporates still struggle to raise funds
in the domestic bond market and rely on bank loans or offshore borrowing.
The recognition that local bond markets remain underdeveloped has more recently led to
several regional efforts to promote their development, including the Asian Bond Fund (ABF)
initiative by the 11 EMEAP central banks. A number of issues have been identified as
hampering the development of local bond markets, including access barriers, especially for
foreigners, lack of funding and hedging instruments, inadequate clearing, settlement and
trading systems, lack of liquid benchmark curves, and last but not least insufficient prudential
regulation and supervision.
This paper will focus on the regulatory aspects and compare the degree of financial
liberalisation of local bond markets with the relevant prudential regulatory and supervisory
conditions. The study focuses on eight local bond markets in non-Japan Asia which are in
EMEAP economies, namely China, Hong Kong SAR, Indonesia, Korea, Malaysia, the
Philippines, Singapore and Thailand.
First, however, some more clarification of the term “regulation” is required. The usage of this
term still creates some confusion and often has a negative connotation. On the one hand,
people talk about regulatory restrictions that hamper market development and the need for
deregulation and market liberalisation. On the other hand, there is growing demand for more
sound regulation and supervision of market practices, especially following the Asian crisis

336 BIS Papers No 30

and the corporate governance scandals of recent years. To provide a framework, this paper
differentiates between two types of regulation, namely economic regulation and prudential
regulation.

Economic regulation
The motivation behind economic rules and restrictions often varies, but the end effect is that
they undermine the free operation of market forces by prohibiting certain business activities
or making them difficult. Good examples are market entry restrictions, capital controls, price
controls and certain taxes. Often, economic regulations are used to support macroeconomic
policy objectives, like financial market or foreign exchange stability. But while such
regulations may help governments achieve their policy objectives, they are typically
inefficient and lead to a misallocation of resources. Moreover, market participants often seek
loopholes to circumvent these restrictions, which leads to a whole new set of problems.
Another motive behind economic regulations, especially entry barriers, is to protect domestic
financial institutions from foreign competition. However, such protection typically leads to
inefficiencies and preserves poor market practices.
The aim of any developing economy should be to gradually reduce economic regulations and
open up its markets. The only caveat is that such liberalisation should not run ahead of other
economic, policy and market reforms, including the establishment of strong prudential
regulation. Thus, when comparing the different degrees of financial liberalisation in the
region, one needs to be mindful of the circumstances and the feasible extent of deregulation
against the background of economic, policy and market conditions. In other words, the
current degree of financial liberalisation in Hong Kong or Singapore is not a realistic near-
term goal for countries like China or Indonesia.

Prudential regulation
Prudential regulation and supervision are meant to protect investors, ensure that financial
markets are fair, transparent and efficient and reduce systemic risks. A strong prudential
regulatory environment is the key to a successful financial centre. Contrary to popular
wisdom, most financial institutions do not want lax regulation. The customers of financial
firms need the assurance that the institutions with which they do business have high ethical
standards, are prudently run, have high-quality staff, and adhere to the highest business
standards. Unless they have confidence in these factors, they will simply take their business
elsewhere.
Most financial institutions recognise that good regulation is a valuable asset which raises the
value of their services in the eyes of their customers. It is no accident that the most
successful financial centres, New York, London, Hong Kong and Singapore, all have rigorous
supervision. Even so, detailed mechanical rules and ratios enforced by frequent checking are
undoubtedly burdensome. When these rules constrain otherwise desirable transactions, they
can contribute to driving business away.
Good prudential regulation works with the grain of market forces and should provide
incentives to reinforce prudent instincts. This is why the trend of regulation and supervision is
towards encouraging high-quality risk management processes, and away from detailed
monitoring of balance sheet ratios. It stresses transparency, market discipline and self-
regulation, and not just compliance with formal rules. In this spirit, the relationship between
the regulatory agency and the regulated entity should not be adversarial.
The financial sector should regard the supervisor as a partner and counsellor, rather than as
a policeman enforcing rules. The supervisor should be able to offer guidance to the financial
institution when it falls short of best practice elsewhere in the industry, or when the business
model being followed seems to have under-appreciated risks. Only when the institution

BIS Papers No 30 337

concerned has taken on unacceptable risks, misled investors or violated the law in any other
way should the supervisor forcefully intervene in the public interest.
Having said that, it is also critical for effective regulation that the supervisory institutions are
empowered to do their job, have the right people and resources, and are supported by sound
legal and accounting standards. This means that objectives and responsibilities must be well
defined, especially when there is more than one supervisory authority. Supervisors must
have the training and background to deal with their private sector counterparts on equal
terms and their compensation should not lag too far behind private sector standards. Legal,
accounting and other financial infrastructure elements are often outside the scope of financial
regulation, but supervisors need to impress on other public bodies the need to establish
robust standards.

Economic regulation and local bond markets

There are many types of economic regulations in the region that have some restrictive
impact on the development of local bond markets. This study focuses only on those that are
the most common and have the most disruptive impact. In no particular order, these
economic regulations are:
• Rules that limit foreign participation in the local bond market;
• Bond issuance restrictions;
• Price and interest rate controls;
• Rules that limit the use of hedging instruments;
• Taxation;
• Custody, settlement and clearing restrictions.

Foreign access
There are still substantial restrictions in several countries on access by foreign investors,
issuers and intermediaries who want to participate in the local bond markets (Table 1). At
one extreme, China is currently the most closed market. Investors can only enter the local
bond market if they apply for a Qualified Foreign Institutional Investor (QFII) licence, which is
a laborious process. On the issuance side, the government now seems willing to open the
local market to multilateral agencies, but no bonds have been issued so far. And for
intermediaries, access is currently only available through joint ventures. Even so, the
schedule of China’s WTO agreements and the current reform drive promise more opening of
China’s capital markets to foreign participation in the next few years. Only the move to full
capital account convertibility probably remains many years (if not decades) away, given the
poor health of many domestic financial institutions, especially the state-owned commercial
banks.
At the other extreme, Hong Kong and Singapore are the most open financial centres in the
region. The main difference between the two is the non-internationalisation policy of the
Singapore dollar, which in practice only means that foreign bond issuers must swap the bond
proceeds that are not used for domestic investment purposes into another foreign currency.
Getting an intermediary licence is also a bit more difficult in Singapore compared to
Hong Kong. In between these extremes, Korea, and to a slightly lesser extent Thailand, are
more accessible, especially for investors. Indonesia, Malaysia and the Philippines are more
on the closed side, in particular in terms of issuer access, which, as with China, is limited to
multilateral agencies on a case by case approval basis.

338 BIS Papers No 30

 Table 1
Foreign access

China Hong Kong SAR Indonesia Korea

Investor access Very limited Very open Limited Open

Issuer access Very limited Very open Very limited Limited
Intermediary access Very limited Very open Limited Limited
Foreign exchange/
capital controls Heavy None Some Some

Malaysia Philippines Singapore Thailand

Investor access Limited Limited Very open Open

Issuer access Very limited Very limited Open Very limited
Intermediary access Limited Limited Open Limited
Foreign exchange/
capital controls Some Some None Some

Issuance restrictions
Bond issuance restrictions not only affect foreign issuers, but domestic issuers as well.
Protection of investor interests is often the motivation behind these rules, but this may come
at the expense of unnecessarily constraining an issuer’s ability to go to the market. There are
two types of issuance models: disclosure-based and merit-based.

• In the disclosure-based model, which is increasingly becoming the global standard,

the issuer is required to disclose all relevant information, but investors have to
decide themselves whether the bond is fairly valued.

• In the merit-based model, a regulator decides whether an issuer is fit to launch a

bond. The regulator’s decision may be based on pure discretion, but more typically
is based on a number of criteria, like the issuer’s past financial performance,
capitalisation, size of the issue, rating and so on.
In Asia, Hong Kong, Malaysia and Singapore have adopted the disclosure-based model
(Table 2). For Malaysia, the main deviation is that issuers have to be rated before they can
launch a bond. How restrictive the merit-based models in the rest of the region are depends
largely on the regulatory requirements. The recently released new issuance rules in
Thailand, for example, have become more issuer-friendly for large and frequent borrowers,
but create a higher hurdle for small and infrequent borrowers.
Korea’s issuance requirements have also become more liberal, but arrangers still have to
commit to fully underwrite the bond. China has the most restrictive issuance requirements.
Issuers must be rated at least AA and have to get approval from two or sometimes three
regulators. In addition, the central bank sets or approves the interest rate level of the new
bond.

BIS Papers No 30 339

 Table 2
Issuance rules

China Hong Kong SAR Indonesia Korea

Issuance model Merit Disclosure Merit Merit

Mandatory rating Yes No Yes Yes
Number of domestic
rating agencies 2 None 1 3

Malaysia Philippines Singapore Thailand

Issuance model Disclosure Merit Disclosure Merit

Mandatory rating Yes Yes No Yes
Number of domestic
rating agencies 2 1 None 2

The mandatory rating requirement, which is standard in almost every economy except for
Hong Kong and Singapore (note that both these economies are also the only ones without at
least one domestic rating agency), is undoubtedly meant as a protection for investors. This
may put pressure on issuers in the less developed markets to comply with the disclosure
requirements. However, this benefit declines and the rating requirement becomes more of an
additional cost as the market develops and disclosure standards are generally met.
Furthermore, investors may become complacent and rely too much on the judgment of the
rating agencies instead of making their own assessments.

Price and interest rate controls

Only China still has direct price and interest rate controls. The People’s Bank of China
(PBoC) effectively controls the cost of borrowing. For loans and bonds, the key rate is the
central bank rate and the PBoC determines or approves the spread between this rate and
the borrowing rates. The objective is to keep overall borrowing costs low, but also to leave
enough margins for the banks between lending and borrowing rates. While interest rates are
market-determined in the rest of the region, some countries still resort to moral suasion and
other forms of indirect intervention to keep interest rates within desired ranges.

Hedging instruments
Lack of hedging instruments is repeatedly listed as one of the top obstacles to the
development of local bond markets. At the moment, only Hong Kong permits the full range of
hedging instruments. To be sure, derivatives are complex financial instruments and need to
be used with care, but the reluctance to approve their use often has more to do with the fear
that they may be used to destabilise markets than real prudential concerns. It must also be
recognised that hedging instruments can be unavailable despite a neutral stance by the
authorities, owing to a lack of liquidity in the market.
After Hong Kong, Korea, Singapore and Thailand have taken the most steps to liberalise the
use of derivatives in the local fixed income markets (Table 3). In China, the use of derivatives
is still the least developed, but a new series of guidelines is currently paving the way for the
introduction of basic interest rate derivatives. In Indonesia, Malaysia and the Philippines, the
use of derivatives is highly restricted and, unlike in China, there are no signs that this will
change soon.

340 BIS Papers No 30

 Table 3
Availability of hedging instruments

Risk type China Hong Kong SAR Indonesia Korea

Foreign exchange Limited Yes Yes Yes

Duration No Yes No Yes
Yield curve No Yes No Yes
Credit No Yes No No

Risk type Malaysia Philippines Singapore Thailand

Foreign exchange Limited Yes Yes Yes

Duration Limited No Yes Yes
Yield curve No No Yes Yes
Credit No No No No

Taxation
The issue as regards taxation is not so much one of principle (whether capital income and
gains should be taxed), but one of distortion. There are legitimate reasons why governments
want to tax capital income and capital gains. The problem is that it is difficult, if not impossible,
to tax the different forms of capital income and capital gains equally. Moreover, there is a
growing global trend not to tax foreign investors. So investors are likely to avoid those countries
that still do so. Finally, even if there are tax treaties in place, the paperwork is often so
laborious and refunding takes so long that many foreign investors stay out of the market.
In Asia, taxation is still a key factor that keeps many foreign investors away from local bond
markets. Only Hong Kong and Singapore effectively do not tax foreign investors (Table 4).
China has also done away with the withholding tax and only taxes capital gains if bonds are
not held until maturity. Korea now has tax treaties with many countries, but the high amount
of taxes initially withheld and the long time period until refunds are paid out still deter many
foreign investors.

Table 4
Taxation

Tax type China Hong Kong SAR Indonesia Korea

Withholding tax None None 20%1 27.5%1

Capital gains tax 33%2 None None 11% or 27.5%3

Tax type Malaysia Philippines Singapore Thailand

Withholding tax 15%1 20-32%1 None4 15%1

Capital gains tax None None None 15%1
1
Can be reduced or waived by tax treaty with certain other countries. 2 If bonds are not held until maturity
3
(plus 5% profit tax). The lower of 11% of gross sales proceeds or 27.5% tax on net capital
gains. 4 Originally 10% but waived for all bonds issued after 1998. Singapore also has the most tax treaties of
any country in the region.

BIS Papers No 30 341

Custody, settlement and clearing
Custody, settlement and clearing are the last areas one should highlight where restrictions
are undermining local bond market activity and development. Settlement and clearing
systems and conventions have substantially improved throughout the region (delivery versus
payment and real-time gross settlement systems are standard in most countries), but they
remain much localised.
Foreign investors that are active in local bonds have to use a local custodian and settle and
clear their trades locally. Besides entailing extra effort and cost, the often short settlement
periods in the local markets leave little time for foreign investors to process their trades (in
Hong Kong, Korea, Malaysia, the Philippines and Singapore, local currency bonds settle
either on the same day or the day after).
For the providers of custody services, another issue is the reluctance of local authorities to
allow them to outsource part of their activities, especially to processing centres outside of the
country. This also often applies to banks and securities firms that intend to move their back
and middle offices to a central location outside the country. China currently prohibits the
offshoring of any processing activities, while most other countries require approval. In some
economies, for example Hong Kong, obtaining approval is primarily a matter of proving that
the data and information are properly protected, while some other authorities tend to reject
outsourcing requests simply to keep the business onshore.

Prudential regulation and local bond markets

The three broadly accepted objectives of prudential regulation and supervision for financial
markets are the protection of investors, ensuring fair, transparent and efficient market
practices and reducing systemic risks. Financial regulation and supervision is typically
divided into three main sectors: banking, insurance and securities markets. Bond markets fall
under the scope of securities market regulation and supervision, but there is typically some
overlap with banking and insurance regulation. The regulatory structure varies from country
to country, but basically evolves around one main law that contains most parts of the
securities market regulation and one main agency that is responsible for supervision. In
some countries, one agency supervises all financial sectors (banking, insurance and
securities markets), but most countries have separate agencies overseeing each sector.
Securities regulation and supervision are broader in scope and more complicated than
banking and insurance regulation and supervision. In banking and insurance, regulation and
supervision deals almost entirely with the intermediary (ie banks and insurance companies),
while securities regulation and supervision has a much broader scope, including securities
intermediaries (eg brokers, dealers and investment advisers), exchanges, collective
investment schemes and issuer disclosure.
The roots of securities market regulation and supervision in Asia go back to well before the
Asian crisis, but have undergone substantial changes since then. Today, the basic
framework of securities regulation and supervision in the region looks very similar to the
standards in the rest of the world. In each country, there is typically one main law that
governs securities regulation and one main authority that supervises the sector, although
there are typically other relevant laws and supervisory authorities (Table 5).

342 BIS Papers No 30

 Table 5
Securities regulation - laws and authorities

China Hong Kong SAR Indonesia Korea

Main securities laws China Sec Law Sec & Fut BAPEPAM Sec & Exch Law
Comm Ord Rules, Cap Mkt
Law1
Number of other laws
relevant for securities 7 4 3 4
Main securities China Sec Sec & Fut Comm BAPEPAM Fin Supv Comm
supervisor Reg Comm
Year established 1992 1989 1976 1998
Accountable to State Council Fin Secretary MoF MoFE
2
Staffing 1,525 361 443 1,670
Funding Gov & fees Self-funded Gov & fees Gov & fees
4
Other relevant CBRC, PBoC, HKMA Bank Indonesia FSS, SFC5
regulatory authorities SAFE3
Structure of overall
financial sector
supervision6 Multiple Multiple Multiple Single

Malaysia Philippines Singapore Thailand

Main securities laws Sec Ind Act Sec Reg Code Sec Ind Act Sec & Exch Act
Number of other laws
relevant for securities 7 5 6 4
Main securities Sec Comm Sec & Exch MAS7 Sec & Exch
supervisor Comm Comm
Year established 1993 1936 1971 1992
Accountable to MoF MoF MoF MoF
Staffing2 549 361 987 391
Funding Self-funded Gov & fees Self-funded Self-funded
Other relevant
regulatory authorities BNM, CCM8 BSP9 None Bank of Thailand
Structure of overall
financial sector
supervision6 Semi Multiple Single Multiple
1 2
BAPEPAM = Badan Pangawas Pasar Modal, the capital market supervisory agency. See Central Banking
3
Publications (2004). CBRC = China Banking Regulatory Commission; PBoC = The People’s Bank of China;
SAFE = State Administration of Foreign Exchange. 4 HKMA = Hong Kong Monetary Authority. 5 FSS =
Financial Supervisory Service; SFC = Securities and Futures Commission; the FSS and SFC are executive
bodies of the Financial Supervisory Commission. 6 Integration/division in the supervision of the three main
financial sectors (banking, insurance and securities): multiple = at least one supervisor for each sector; semi =
one supervisor for two sectors; single = one supervisor for all sectors. 7 MAS = Monetary Authority of
8
Singapore. BNM = Bank Negara Malaysia (Central Bank of Malaysia); CCM = Commission Companies of
9
Malaysia. BSP = Bangko Sentral ng Pilipinas (Central Bank of the Philippines).

BIS Papers No 30 343

Legal and accounting standards
Securities regulation and supervision cannot be seen in isolation from the broader legal
framework and accounting requirements. Appropriate and effective legal and accounting
standards form the foundation for the regulatory and supervisory framework. In some places,
it is a lack of sound legal and accounting standards that undermines what on paper looks like
good regulation and supervision. In many parts of Asia, legal and accounting standards were
poor before the 1997-98 crisis and, while reform is under way, six years is not a long time in
which to make the types of structural and behavioural changes that typically require more
than one generation.
Only Hong Kong, Singapore and to some extent Malaysia had robust legal standards before
the Asian crisis. All three also practise common law. The legal code in the rest of the region
is based on forms of European civil law, which is often complicated by the inclusion of other
legal cultures, like socialist law in China, customary law in Indonesia, classical Chinese law
in Korea, or Buddhist law in Thailand. In general, the quality of the law is improving in most
parts of the region, but there are still some notable gaps, like private property law in China
and Indonesia.

Table 6
Legal and accounting standards

Hong Kong
China Indonesia Korea
SAR

Type of law Civil Common Civil Civil

Legal opacity index1 100 55 86 79
Formal compliance with Some material Closely aligned No material No material
International Accounting differences differences differences
Standards
Accounting opacity index1 86 53 68 90
1
Corruption opacity index 62 25 70 48
Corruption perceptions
index2 3.4 ... 1.9 4.3

Malaysia Philippines Singapore Thailand

Type of law Common Civil Common Civil

Legal opacity index1 ... ... 32 65
Formal compliance with No material Some material Closely aligned No material
International Accounting differences differences differences
Standards
Accounting opacity index1 ... ... 38 78
1
Corruption opacity index ... ... 13 55
Corruption perceptions
index2 5.2 2.5 9.4 3.3
1 2
See PricewaterhouseCoopers (2001); the index ranges from 0 (least opaque) to 150 (most opaque). See
Transparency International (2003); the index ranges from 10 (highly clean) to 0 (highly corrupt).

344 BIS Papers No 30

As important as the quality of the law is the quality of the legal system (eg judges,
lawyers, etc). Before the Asian crisis, the quality of the legal system was well behind the
capital market development in most countries. The crisis revealed the inadequacy of most
legal systems in the region, although one should be mindful that any existing legal
system would have been severely challenged if faced with the kind of systemic
insolvencies seen during and after the crisis.
Since the Asian crisis, there have been widespread efforts by all countries to improve
their legal systems and progress is visible. However, one fundamental problem is that the
prevailing political, social and economic power structures in some countries continue to
work against efficient legal systems. Not coincidently, this bias to resist efficient legal
systems correlates with the persistence of corrupt practices in several countries
(Table 6).
On the accounting side, most countries in the region comply largely with International
Accounting Standards (IAS) and/or US GAAP. Only China and the Philippines still have
significant gaps, such as in the measurement of fair market value, but both are working to
converge to international standards over the next few years. As with the legal standards,
the problem lies less in the quality of the accounting rules than in their implementation.
To some extent, this is a resource and training issue. But the more fundamental problem
is whether there is the political will to change old habits.

IOSCO principles of securities regulation

All countries in this study are members of the International Organization of Securities
Commissions (IOSCO) and subscribe to its principles of securities regulation. These
30 principles are grouped into eight categories (Table 7):

• Principles relating to the regulator;

• Principles for self-regulation;
• Principles for the enforcement of securities regulation;
• Principles for cooperation in regulation;
• Principles for issuers;
• Principles for collective investment schemes;
• Principles for market intermediaries;
• Principles for the secondary market.
The first four categories concern the regulatory and supervisory institutions directly, while
the remaining categories outline the regulatory principles for the four main areas of
securities regulation (ie issuers, collective investment schemes, market intermediaries
and the secondary market). It is beyond the scope of this paper to assess in detail to
what extent these principles have been implemented by each country - the IMF and the
World Bank are currently leading a project to assess the observance of the principles -
but some preliminary observations can be made.

BIS Papers No 30 345

 Table 7
IOSCO principles of securities regulation

Principles relating to the regulator

1. The responsibilities of the regulator should be clear and objectively stated.
2. The regulator should be operationally independent and accountable in the exercise of its
function and powers.
3. The regulator should have adequate powers, proper resources and the capacity to perform its
functions and exercise its powers.
4. The regulator should adopt clear and consistent regulatory processes.
5. The staff of the regulator should observe the highest professional standards, including
appropriate standards of confidentiality.
Principles for self-regulation
6. The regulatory regime should make appropriate use of self-regulatory organisations (SROs)
that exercise some direct oversight responsibility for their respective areas of competence, to
the extent appropriate to the size and complexity of the markets.
7. SROs should be subject to the oversight of the regulator and should observe standards of
fairness and confidentially when exercising powers and delegated responsibilities.
Principles for the enforcement of securities regulation
8. The regulator should have comprehensive inspection, investigation and surveillance powers.
9. The regulator should have comprehensive enforcement powers.
10. The regulatory system should ensure an effective and credible use of inspection,
investigation, surveillance and enforcement powers and implementation of an effective
compliance programme.
Principles for cooperation in regulation
11. The regulator should have authority to share both public and non-public information with
domestic and foreign counterparts.
12. Regulators should establish information sharing mechanisms that set out when and how they
will share both public and non-public information with their domestic and foreign counterparts.
13. The regulatory system should allow for assistance to be provided to foreign regulators who
need to make inquiries in the discharge of their functions and exercise of their powers.
Principles for issuers
14. There should be full, timely and accurate disclosure of financial results and other information
that is material to investors’ decisions.
15. Holders of securities in a company should be treated in a fair and equitable manner.
16. Accounting and auditing standards should be of a high and internationally acceptable quality.
Principles for collective investment schemes
17. The regulatory system should set standards for the eligibility and the regulation of those who
wish to market or operate a collective investment scheme.
18. The regulatory system should provide for rules governing the legal form and structure of
collective investment schemes and the segregation and protection of client assets.
19. Regulation should require disclosure, as set forth under the principles for issuers, which is
necessary to evaluate the suitability of a collective investment scheme for a particular
investor and the value of the investor’s interest in the scheme.
20. Regulation should ensure that there is a proper and disclosed basis for asset valuation and
pricing and the redemption of units in a collective investment scheme.

346 BIS Papers No 30

 Table 7 (cont)
IOSCO principles of securities regulation

Principles for market intermediaries

21. Regulation should provide for minimum entry standards for market intermediaries.
22. There should be initial and ongoing capital and other prudential requirements for market
intermediaries that reflect the risks that the intermediaries undertake.
23. Market intermediaries should be required to comply with standards for internal organisation
and operation conduct that aim to protect the interests of clients, ensure proper management
of risk, and under which management of the intermediary accepts primary responsibility for
these matters.
24. There should be procedures for dealing with the failure of a market intermediary in order to
minimise damage and loss to investors and to contain systemic risk.

Principles for the secondary market

25. The establishment of trading systems, including securities exchanges, should be subject to
regulatory authorisation and oversight.
26. There should be ongoing regulatory supervision of exchanges and trading systems which
should aim to ensure that the integrity of trading is maintained through fair and equitable rules
that strike an appropriate balance between the demands of different market participants.
27. Regulation should promote transparency of trading.
28. Regulation should be designed to detect and deter manipulation and other unfair trading
practices.
29. Regulation should aim to ensure the proper management of large exposures, default risk and
market disruption.
30. Systems for clearing and settlement of securities transactions should be subject to regulatory
oversight, and designed to ensure that they are fair, effective and efficient and that they
reduce systemic risk.

Source: IOSCO (2003).

Principles for the regulator, self-regulation, enforcement and cooperation

First, many of the principles relating directly to the regulatory and supervisory institutions
have generally been implemented in the region. Nevertheless, there are some notable
differences between countries. Second, implementation is often narrow and lacks
effectiveness. A good example is the spread of regulatory responsibilities across several
agencies or the lack of clarity of roles. To be clear, though, this is not to suggest that all
countries should create a single regulator for the whole financial sector as adopted by Korea
and Singapore. Whether there are multiple regulators or only one, most important is that
responsibilities are clearly defined and standards and rules are consistent across sectors and
institutions.
In that respect, the most confusion still exists in China. The patchwork of eight securities
laws leaves gaps and is not entirely consistent, while overlapping responsibilities with the
other three authorities involved in securities supervision often undermine the effectiveness of
the China Securities Regulatory Commission. Korea, on the other hand, is a good example
of the fact that even the single regulator model can have its problems. The trifurcated
structure of the Financial Supervisory Commission appears cumbersome and in some cases
gets in the way of regulatory effectiveness.
A different issue that is often raised by people familiar with securities regulation in Asia is that
the supervisory authorities are often not sufficiently independent from the government or

BIS Papers No 30 347

other public authorities (like the central bank) to fulfil their roles. The desire for operational
independence is understandable, and so is the desire to reduce the degree of political
influence over regulatory decisions. In many countries, the government often interferes with
the supervisory agency in respect of enforcing laws and regulations.
At the same time, it is only appropriate that a supervisory agency be accountable to a
government body, typically the Ministry of Finance, and that it closely coordinate with other
relevant government bodies, especially in times of crisis. More broadly, coordination and
cooperation with other regulatory authorities and government institutions, both domestic and
foreign, is an area where more progress is needed. This is even true for Hong Kong, whose
regulatory framework is generally viewed as one of the best in the world. A case in point is
the growing number of transactions and relationships between Hong Kong and the mainland.
The supervision of the Hong Kong affiliates of mainland companies is no doubt strong, but
little is typically known about the financial conditions of the parent companies.
Resources are another area where conditions differ. In terms of number of staff, supervisory
agencies in the region have sufficient manpower to do their jobs. The problem is not quantity
but quality, in terms of basic training as well as ongoing training to stay in touch with financial
developments. Another issue in this regard is staff retention, which is not only a matter of pay
but also motivation. In fact, even Hong Kong and Singapore, which have the highest quality
of staff and the best pay, have difficulty retaining their top people.
Resource constraints are not limited to people. Another issue, especially for the less
developed countries, is the technical ability to closely monitor markets and trading positions
to detect manipulation and systemic risks. Interestingly, resource constraints occur
everywhere, no matter what the type of funding. For those agencies that depend primarily on
government funding, the message is that more is needed. For those agencies that are self-
funded, the message may be that other sources may be needed, even if that may somewhat
undermine the public image of independence.
Large differences in regulatory standards are also apparent in the area of enforcement.
While most regulators have formally adequate powers of inspection, investigation and
surveillance, they are often constrained in their ability to require compliance and impose
penalties. Moreover, some regulators appear complacent and seem to deliberately overlook
violations. Supervisors in Hong Kong and Singapore have the best reputation and there are
few violations (if any) that go undetected and unpunished. Supervisors in Korea and
Malaysia have sufficient regulatory authority and recent actions point to a growing
commitment to enforcement, but there is still a sense that supervisors occasionally turn a
blind eye to apparent regulatory violations.
In terms of style and procedures, Hong Kong and Singapore have the highest standards, but
Hong Kong’s securities supervisors are viewed as more user-friendly than their Singapore
colleagues, who still have a reputation for being somewhat paternalistic. At the other end of
the spectrum, China still struggles with many basics. In general, the rule in China remains
that, if the law does not already approve a transaction, then it cannot be done. Another issue
in China is poor documentation of new procedures and guidelines, which places demands on
the experience of financial intermediaries, investors and issuers.
Korea’s securities supervisors have made good progress in issuing appropriate and timely
rules and procedures. Market perception, however, is that there are too many rules of
sometimes mixed quality, while implementation is still very bureaucratic. Obtaining broader
public comment would also help Korea’s supervisors to dispel the sense that certain groups
within the financial sector disproportionately influence some policies.
Use of self-regulatory organisations (SROs) in securities regulation is mostly limited to public
exchanges, which usually have SRO status. In practice, exchanges focus primarily on stock
market activity and less on bond markets, where secondary market trading takes place
mostly over the counter. One exception is Thailand, where the Thai Bond Dealing Centre (an
SRO) is responsible for monitoring activity in the bond market. The typical issue with an SRO

348 BIS Papers No 30

like a public exchange is that its own capability to regulate, monitor and discipline its
members or market segments is often limited, while the relevant regulator does not always
exercise full and consistent oversight over the SRO.

Principles for issuers

Disclosure standards, accounting and auditing practices and investor protection are issues
that have received much attention around the world following the series of corporate
governance scandals in recent years. In Asia, these concerns are compounded in many
countries by prevailing legal uncertainties. Most countries in the region fulfil the basic
disclosure requirements for new issues, but there are concerns regarding the timeliness and
accuracy of reporting of new material information. Even in Hong Kong and Singapore, which
have the soundest disclosure standards in the region, the release of non-periodic information
is sometimes slow.
Issuer adherence to disclosure standards is often lax because of limited civil liability. Issuers
in Malaysia, for example, face little or limited consequences if they are slow to disclose vital
information or make false or misleading statements. The other main shortcoming in many
countries relates to accounting and audit. The issues here concern both the quality of staff
and the oversight of those staff by the professional associations and their respective official
regulators. Only Hong Kong and Singapore are on a par with global best practices in this
area.
In some cases, however, disclosure requirements are too stringent. In the new Thai issuer
and disclosure guidelines, for example, the financial adviser, who is typically the underwriter,
has the same responsibility and liability to ensure the correctness of all information provided
in the prospectus as the issuer. The intention behind this is to make sure the underwriter
undertakes proper due diligence. However, this does not conform to global best practices
and may lead to higher transaction costs.

Principles for collective investment schemes

Regulation and supervision of collective investment schemes is mixed in the region, largely
as a result of the recent development of this industry in many countries. In the least
developed markets, proper product descriptions, clear definition of principal and agent, and
risk warnings fall short of global standards. In some countries, retail investors are often not
sufficiently advised about the differences between bond funds and regular bank deposits,
which leads to an underestimation of the potential risks and an overestimation of the
potential returns.
In contrast, some countries are clearly too strict in the way they regulate collective
investment schemes, which undermines their growth potential. This is, for example, the case
in Korea. Another issue concerns the growing internationalisation of fund management. Hong
Kong, for example, currently has the largest fund management industry in the region, but
many funds distributed in Hong Kong are managed offshore, which creates challenges for
local supervisors.

Principles for market intermediaries and the secondary market

The licensing process for new market intermediaries appears sound from a prudential
perspective. In fact, the issue is more that some countries are still too stringent in giving new
licences, especially for foreign intermediaries. However, other aspects of the ongoing
supervision of market intermediaries, such as the monitoring of capital requirements, risk
management, governance, failure procedures and other prudential controls, still require
strengthening in many countries.

BIS Papers No 30 349

Related to the regulation of market intermediaries is the supervision of secondary market
activities. The regulation of public exchanges in the region is generally approaching global
standards, but the detection and prosecution of manipulation and other unfair trading
practices are not consistently enforced in all countries. Partly, this is a resource problem, with
modern surveillance systems lacking. In addition, some countries need to step up
enforcement of the self-regulatory responsibilities of their exchanges.
In the local bond markets, the large majority of secondary market trading is over the counter.
Price transparency is low, given the low liquidity of most local bond markets. Not surprisingly,
supervisors struggle in their efforts to monitor the markets and prosecute in the event of
market manipulation. This is also true for many interest rate derivatives and, among other
reasons, has made many regulators reluctant to approve new products.

Conclusion
Comparing the degree of local bond market liberalisation with prudential regulatory and
supervisory standards reveals three clusters (Graph 1). Hong Kong SAR and Singapore are
the most advanced economies in the region and essentially in line with best standards in the
rest of the world. Both have superior prudential regulatory and supervisory systems, but Hong
Kong is slightly more liberal in terms of market access and product innovation. The main
challenge for Hong Kong is to improve cooperation with the mainland authorities in order to
better understand the credit quality of the Chinese parent companies of Hong Kong affiliates.

Graph 1
Liberalisation/supervision grid
100
Singapore

80
Supervision index; 100 = global best practice

Hong Kong SAR

Korea

Malaysia
Thailand
60

40 China Philippines

Indonesia

20

0
0 20 40 60 80 100
Liberalisation index, 100 = global best practice
Note: The scatter chart above represents the author’s judgment of the degree of
market liberalisation and supervision based on the analytical framework outlined in
this study and informal feedback from market participants.

350 BIS Papers No 30

The next cluster consists of Korea, Malaysia and Thailand. Of the three, Korea undoubtedly
has the most developed bond market. In fact, the bond market probably plays a more
important role in Korea’s economy than in Hong Kong’s or Singapore’s. However, there is
room to liberalise the market more and to strengthen supervision. Thailand’s bond market is
visibly smaller than Korea’s, but not much behind in terms of deregulation and supervision.
Malaysia’s prudential regulatory and supervisory standards are largely on a par with Korea
and Thailand, but the bond market is much more closed, in terms of both foreign access and
product innovation. Given Malaysia’s advanced supervision, liberalisation could accelerate
bond market development without creating immediate prudential concerns.
China, Indonesia and the Philippines are at the low end of the scale. All three countries have
improved prudential regulation and supervision, but China has come from further behind and
has made the most progress. If this trend continues, China may leave behind Indonesia and
the Philippines, which have fewer resources with which to improve prudential regulation and
supervision. The main constraint for China is the low probability that it will achieve full capital
account convertibility in the next few years.
Overall, it is probably fair to say that general prudential standards governing local bond
markets in Asia are not grossly insufficient when set against the degree of market
development and liberalisation. In fact, Malaysia and Singapore could immediately ease
restrictions without creating any prudential concerns. However, there is no cause for
complacency. First, given the development and sophistication of market practices and cross-
border flows, regulation and supervision cannot stand still. Second, the current efforts in the
region to promote local bond markets will probably lead to more liberalisation and market
activity and, thus, require stronger prudential standards.
Whether this requires extensive harmonisation of bond market regulation and supervision
within the region, as is often called for, is not clear. Different political, legal and economic
structures make it very difficult to strive for a high degree of harmonisation. More important
than efforts to harmonise all aspects of regulation and supervision within the region is that
each country aims for global best practice and unbureaucratic rules and procedures.
Having said that, there are some areas of more practical concern where harmonisation and
cooperation would undoubtedly be welcome.
• More cooperation around clearing, settlement and custody would greatly facilitate
cross-border investment.
• Integration and linking of trading platforms would also help cross-border investment
and boost price transparency.
• Concerning taxation, harmonisation towards the lowest common denominator
(ie zero withholding and capital gains taxes) would be consistent with trends in most
other parts of the world and help market development.
Finally, adopting common standards in line with global best practices could further enhance
the value and credibility of local rating agencies.

References
Asian Development Bank (2003): Harmonization of bond market rules and regulations, August.
Basel Committee on Banking Supervision (2001): The joint forum, core principles, cross-
sectoral comparison, Bank for International Settlements, November.
Central Banking Publications (2004): How countries supervise their banks, insurers and
securities markets.
China, China Securities Regulatory Commission, www.csrc.gov.cn.
Crockett, Andrew (2003): “Principles of financial regulation”, speech at the Seoul Financial
Centre Conference, October.

BIS Papers No 30 351

de Luna Martinez, José and Thomas A Rose (2003): “International survey of integrated
financial sector supervision”, World Bank Policy Research Working Paper 3096, July.
Hong Kong SAR, Securities and Futures Commission, www.hksfc.org.hk.
Indonesia, Capital Market Supervisory Agency (Bapepam), www.bapepam.go.id.
International Monetary Fund (2003): People’s Republic of China - Hong Kong Special
Administrative Region: financial system stability assessment, June.
——— (2003): Republic of Korea: financial system stability assessment, March.
International Monetary Fund and World Bank (2002): Experience with the assessments of
the IOSCO objectives and principles of securities regulation under the Financial Sector
Assessment Program, 18 April.
International Organization of Securities Commissions (2003): Objectives and principles of
securities regulation, May.
Korea, Financial Supervisory Commission/Service, www.fsc.go.kr.
Malaysia, Securities Commission, www.sc.com.my.
Organisation for Economic Co-operation and Development (2003): White paper on corporate
governance in Asia, July.
Philippines, Securities and Exchange Commission, www.sec.gov.ph.
PricewaterhouseCoopers (2001): The opacity index, January.
Republic of the Philippines (2001): Report on the observance of standards and codes,
corporate governance country assessment, September.
Singapore, Monetary Authority of Singapore, www.mas.gov.sg.
Thailand, Securities and Exchange Commission, www.sec.or.th.
The World Bank Group: Reports on the observance of standards & codes, Malaysia,
www.worldbank.org/ifa/rosc_cg_malaysia.html.
Transparency International (2003): Corruption perceptions index 2003, October.

352 BIS Papers No 30

 Panel discussion: questions to be analysed

Yung Chul Park

It is my honour and privilege to serve as the moderator of this panel of distinguished experts
on capital markets. Over the last two days, we have listened to presentations and
discussions about 13 papers analysing many issues involved in developing regional bond
markets in Asia. In order to summarise and better understand what we have learned in this
conference, we have organised this panel discussion with the expectation that the panel
members will help us chart a course of development that will establish deep and liquid
regional bond markets in Asia. For this purpose, I would like to ask the panel members for
their views on some issues that are crucial for Asian bond market development.
The first question I would like to raise is: does Asia need regional bond markets?
• Would it be easier and cheaper to issue local currency bonds in regional bond
markets than in global bond markets?
• Would these regional markets help diversify the foreign exchange reserve portfolios
of Asian countries?
• Would regional bond markets improve resource allocation and stimulate investment
in the region?
The second question refers to the ideal structure of Asian bond markets: what types of
regional bond markets would best serve the bond financing needs of Asian governments and
corporations? Possible approaches are:
• Improving and expanding existing regional bond markets such as the samurai
market in Japan
• Creating an offshore bond market (replication of the old eurobond market)
• Promoting regional financial centres through competition
Finally, I would like to ask the panel members: what could ASEAN+3 do to develop deep and
regional bond markets in Asia? Aspects to consider:
• Develop and open local bond markets - a prerequisite to the promotion of regional
bond markets in Asia
• Facilitate regulatory, tax, and other institutional harmonisation and policy
coordination
• Construct regional capital market infrastructure, including clearing and settlement,
regional rating and credit enhancement institutions

BIS Papers No 30 353

 Panel discussion

Tom Byrne

This conference addressed issues concerning the development of an Asian bond market.
Moody’s view is that governments and corporations in Asia will continue to seek access to
the global capital market, but that national markets in the region will develop and mature,
allowing firms to increase their funding in domestic debt markets. As in other regions,
Moody’s has in recent years expanded in Asia, by building up its offices in Tokyo, Hong Kong
and Singapore, as well as by buying into local rating agencies. For example, Moody’s has
majority ownership in KIS, one of the national rating agencies in the Korean market. The use
of a national rating scale in Korea - although not integrated into Moody’s global rating scale’s
default and expected loss probabilities - provides local investors with valuable distinctions of
relative creditworthiness in a credit rating convention they are more familiar with.
I think it is important to consider the role of a rating agency, which is valid in any market,
regional or global. In Moody’s view, the main and proper role of credit ratings is to enhance
transparency and efficiency in debt capital markets, by providing an independent opinion of
relative credit risk, reducing the information asymmetry between borrowers and lenders. This
function enhances investor confidence and allows creditworthy borrowers broader
marketability of their debt securities.
Moody’s believes that there is room to enhance the disclosure of its own rating processes.
Moody’s does not believe the accusation that it is a “black box” to be fair or accurate.
Moody’s rating methodologies and practices have been published periodically, and senior
officers of the firm have made speeches to professional forums and made presentations to
government regulators concerning this issue. Moody’s has codified core principles of good
rating practices; they are:
• Ratings must be independent of commercial relationships with an issuer.
• No forbearance: Moody’s shall not refrain from taking a rating action out of concern
for the potential effect it may have on the issuer or the market.
• Controlling conflicts of interest: Moody’s does not give investment advice.
• Confidential information is not disclosed, and used only internally for rating
decisions.
• Judicious consideration will be taken in assessing all the circumstances relevant to
an issuer’s creditworthiness.
• Rating committees make rating decisions that reflect the collective experience of
judgment of the organisation, not the opinion of any single person.
Furthermore, Moody’s believes that independence, objectivity and reliability have been the
heart of the rating agency’s role in credit markets for nearly a century. Moody’s would be
concerned if additional regulatory oversight were to reach into the underlying methodology
and practices of the credit rating practice - particularly if regulation were to change the nature
of the product offered by the rating agency from one based on credibility with the investor
community to one of a licensing function for the government.
Moody’s believes that innovation and competition between rating agencies better serve the
market than harmonisation or cooperation with the industry. Well functioning rating agencies,
in the manner described above, will help improve market transparency and efficiency in
allocating capital.

354 BIS Papers No 30

 Panel discussion

Aaron Low

1. Does Asia need regional bond markets?

There is no disputing the need for Asian bond markets to provide sources of financing for
Asia’s capital needs and instruments for investing Asia’s savings. The economic concept of
“market completeness” comes to mind. A more critical question is whether issuance should
be through local or hard currency. Asia’s situation is somewhat similar to some emerging
markets, where sovereigns prefer to issue in hard currency, in contrast to the G5 economies,
where sovereigns typically issue in local currency to the natural, local investors. Valuations
that influence the choice between local and hard currency issuance tend to vary across the
region, depending on domestic monetary policy as well as on demand and supply conditions.
For example, it is currently cheaper for Korean corporations to issue long-dated bonds in
dollars due to US-Korea yield curve spreads and currency swap spreads. Another important
consideration is that global demand is generally better for longer-term maturities while local
investor demand tends to be concentrated on the shorter end.
Naturally, increasing the opportunity set with Asian bonds will help diversify foreign exchange
reserve portfolios. My view is that additional return/risk benefits would accrue from credit risk
diversification rather than from currency/interest rate diversification. As long as Asian
exchange rate policy targets the US dollar with full capital account convertibility, local yields
will be highly correlated with US interest rates in terms of systematic movements.
Diversification benefits will then arise from sovereign or credit risks.

2. What types of regional bond markets will best serve the bond
financing needs of Asian governments and corporations?
Absent a common regional currency, the Asian dollar bond market is perhaps the most
attractive alternative to a regional bond market. This approach does imply an additional set of
costs that includes issues of corporate governance, disclosure, rating fees, etc. As long as
developed Asia relies on the external sector and hard currency earnings for its growth, it
would be natural to finance that growth with hard currency debt. Corporations and
governments also look closely at the cheapest form of financing. Local currency and even
offshore issuance increases the financing opportunity set and should be an important priority,
especially for firms that do not have foreign revenues or operations. Regulatory bodies can
facilitate and improve market infrastructure, but markets will gravitate to the cheapest and
most efficient alternative.

3. What could ASEAN+3 do to develop deep and regional bond

markets in Asia?
Building acceptance for regional benchmarks would be a good start. We have seen the
introduction of two recent local currency benchmarks in the region that include issues with
acceptable liquidity and size. These are in addition to a couple of dollar issue regional
benchmarks that were more in demand when the dollar was strong. If the recent weak dollar
outlook continues, local currencies will probably face more demand.

BIS Papers No 30 355

An important yardstick of success would have to be the depth of liquidity in Asian bonds,
both for local and for hard currencies. The current lack of liquidity in secondary issues poses
problems for active managers, pricing vendors, market-makers and traders. There are
naturally some exceptions, but illiquidity is the rule. The need to boost liquidity and trading is
a paramount concern, and any bond market cannot be considered a success if most bonds
are held to maturity. Fortunately, there are options available to address this matter.
Using active external fund managers, traditional or hedge fund types, would be a big boost.
Conventional fears of market volatility are overblown, in my view. First, fundamentals look
positive with the excess pool of Asian savings and improving regional fiscal discipline
providing strong support against trading volatility. The real sources of volatility will more likely
prove to be G3 interest rates and spreads. Second, the real source of volatility in Asia is
equities rather than bonds, and with excessive pools of savings channelled into real estate
and stocks, the lack of a bond market increases systematic risk.
Asian investors are also heavily invested in global bonds, using global benchmarks. These
do provide both interest rate and currency diversification, but there is also a need for a
meaningful regional component, especially if local investors are to internalise asset and
liability management practices.

356 BIS Papers No 30

 Panel discussion

Robert N McCauley 1

In his remarks, Sang Yong Park offered a counsel of despair. Only an offshore regional
market, he argued, could break the vested interests that are preventing the development of
domestic bond markets. The flaw in this prescription, however, is that governments have to
allow their currencies to be used offshore, and the same vested interests that block domestic
market development will prevent offshore market development. The US dollar was already
internationalised when the US government made the policy errors that encouraged the
development of the offshore eurodollar bond market. Offshore markets in the Deutsche mark
or Swiss franc were limited by restrictions imposed by the German and Swiss authorities,
with the effect of protecting vested interests. Offshore regional markets do not, in my view,
offer the way forward.
The real choice is between the global bond market and domestic bond markets. As debt
managers, governments need to recognise that global securities firms naturally argue for
global markets. They cannot be expected to give full weight to the direct costs, and perhaps
more importantly, the indirect costs, of going global. Debt managers need to exercise
discipline in choosing between global and national markets. Consider the example of the last
Korean sovereign issue.
Last June, the Republic of Korea sold a $1 billion dollar bond due in 2013 in the global
market. Underwriters Barclays, Citigroup and Goldman Sachs were paid to distribute the
bond to global investors. The trade press, in particular FinanceAsia (June 2003, page 8),
reported that about 75% of the bonds were placed in Asia, with less than 10% in Korea,
15% in Europe and 10% in the United States.
It is widely believed that, owing to secondary market purchases, Samsung Insurance has
become the largest holder of this dollar bond. Since the insurer has Korean won liabilities to
its policyholders, it has reportedly converted the dollar cash flows from this bond into won
with a cross-currency swap.
Consider the Rube Goldberg contraption that this circuit of transactions represents. The
Republic of Korea pays underwriters a fee to place the bond with global investors. Samsung
Insurance pays half the bid-ask spread to the bond’s market-maker and half the bid-ask
spread to the derivatives dealer who arranged the cross-currency swap. Would it not have
been easier and cheaper to sell a 10-year Korean won government bond in the Seoul bond
market?
Look at the transaction from another perspective. The Republic of Korea paid 92 basis points
over the yield on a 10-year US Treasury note on its bond. This was a vast improvement on
the 355 basis points paid in April 1998 on a 10-year bond. Following the money, the $1 billion
is added to Korean foreign exchange reserves. There, Korea’s reserve managers will try to
earn returns over US Treasury yields by investing the $1 billion proceeds. If they are able to
obtain a return of 40 basis points over US Treasury yields by buying agency paper or bonds
backed by credit cards or mortgages, they would be doing well. But this would imply a net
cost of something like 50 basis points per year, or around $50 million over the life of the

1
Views expressed are those of the author and not necessarily those of the Bank for International Settlements.

BIS Papers No 30 357

10-year bond. The good news is that, in buying the bond, Samsung Insurance recaptures
some of this sum for Korea as a country.
Again, would it have made more sense for the Korean government to have added another
$1 billion to the domestic bond market? There, could these bonds have satisfied pension
funds’ and the insurers’ need for long-duration, won-denominated assets? Could they have
added to the mass of bonds available for trading and repoing? Could they have helped
lengthen and deepen the won benchmark yield curve?
Global securities firms highlight the externality of having the sovereign bond set a benchmark
in the dollar (or euro) market for other issuers. Why? With a sovereign benchmark, Korea
Development Bank or Korea Export Import Bank bonds can be properly priced by
international investors. Indeed, FinanceAsia reported that the success of the sovereign deal
led to a repricing of the Korea Development Bank from US Treasury plus 120 to US Treasury
plus 107 basis points. Why? With a well developed agency dollar yield curve, bonds for
Korean corporations and banks can be properly priced by international investors. As
FinanceAsia noted, “the sovereign does not need the money, but was merely keen to create
a new tighter benchmark from which corporate Korea could benefit”. Such externalities come
at a price, however, which can be measured not only in terms of the direct cost, but also in
terms of the opportunity cost.
That is, the benefits of setting such benchmarks come at the missed opportunity of more
issuance in domestic currency in the domestic market. In late 2002, the Kingdom of Thailand
came close to issuing a $1 billion bond in the global market. At the last minute, the deal was
cancelled, although the Kingdom has done a pair of floating rate dollar deals. Consistent with
Fernandez and Klassen’s result that the sovereign and corporate choice of currency is
similar, Thai bond issuance in foreign currency since the crisis has been more limited than
Korean bond issuance in foreign currency. Ultimately, there are positive externalities to
corporate issuance from government issuance within limits in dollar or domestic currency.
Given limited amounts of government debt and a policy preference for domestic bond
issuance, debt managers should think twice before selling dollar bonds.

358 BIS Papers No 30