S332A315FR en

FINANCE
AND
THE REAL ECONOMY
ISSUES AND CASE STUDIES
IN DEVELOPING COUNTRIES
editors:
Yilm az Akyüz
Giinther Held
United Nations University/W orld Institute fo r Development

E conom ics Research
E conom ic Com m ission fo r Latin Am erica and the Caribbean
United Nations Conference on Trade and Development
Responsibility for the papers rest solely with the authors.
Collection: Social and Political Studies
Inscription N9 88.481
I.S.B.N.: 956 - 7363 - 02 - 1
© S.R.V. Impresos S.A., Tocomal 2052, Santiago - Chile.
Printed in Chile
Compositor: Patricio Velasco G.
Santiago, September 1993

INDEX
INDEX
INTRODUCTION ................................................................................................. 11
FINANCIAL LIBERALIZATION: THE KEY ISSUES....................................................... 19 \/

Introduction ....................................................................................................... 23
I. Interest rates and savings ................................................................................... 24
II. Financial liberalization and deepening................................................................ 27
III. Allocative efficiency............................................................................................ 30
IV. Productive efficiency and cost of finance............................................................. 35
V. Regulation of finance and financial stability ........................................................ 37
VI. Options in financial organizations ..................................................................... 42
VII. External liberalization and financial openness ..................................................... 49
VIII. Conclusions........................................................................................................ 60
EAST ASIAN FINANCIAL MARKETS: WHY SO MUCH (AND

FAIRLY EFFECTIVE) GOVERNMENT INTERVENTION? ............................................. 69
Introduction ....................................................................................................... 73
I. Why so much government intervention?............................................................. 74
II. Repression in financial m arkets........................................................................... 76
III. Discipline of business ........................................................................................ 78
IV. The government's ability to impose performance standards ................................ 79
V. Keeping the costs of investment finance low: the
case of the Republic of Korea's financial refo rm .................................................... 83
VI. Conclusions........................................................................................................ 88
SAVINGS AND FINANCIAL POLICY ISSUES IN SUB-SAHARAN A FR IC A ................... 103

Introduction ....................................................................................................... 107
I. Fragmentation of savings mobilization and financial
intermediation in Sub-Saharan Africa ................................................................. 108
II. Financial dualism and market fragmentation ....................................................... 124
III. Policy required to overcome market fragmentation.............................................. 128
THE INDONESIAN EXPERIENCE WITH FINANCIAL-SECTOR REFORM ..................... 149

Introduction ....................................................................................................... 153
I. The 1983 financial reforms: background and content............................................ 154
II. The 1988-91 financial reforms: background and content....................................... 164
III. Effects of the financial reform s............................................................................. 168
IV. Conclusions........................................................................................................ 194
J PENSION SYSTEM REFORMS AND SAVINGS IN LATIN AMERICAN AND
CARIBBEAN COUNTRIES WITH SPECIAL REFERENCE TO CHILE ............................ 201
Introduction ....................................................................................................... 205
I. Current status and challenges of pension systems ............................................... 206
II. Population, employment and returns to pension system funds............................. 215
III. Development of capital markets and returns on pension system fu n d s................. 219
IV. The Chilean experience with an individual capitalization sch em e ......................... 224
V. Conclusions........................................................................................................ 237
j BANK REGULATION, LIBERALIZATION AND FINANCIAL

J INSTABILITY IN LATIN AMERICAN AND CARIBBEAN COUNTRIES........................ 241
Introduction ....................................................................................................... 245
I. Bank regulation and supervision......................................................................... 247
II. Bank regulation in the Chilean financial liberalization experience ......................... 259
III. Overview of recent financial instabilities in
Latin American and Caribbean countries.............................................................. 273
IV. Conclusions........................................................................................................ 286
NOTE ON THE AUTHORS ........................................................................................... 291

INTRODUCTION
Yilmaz Akyüz
Giinther Held
Introduction 13
This book is a product of a joint work programme on "Financing for Deve

lopment" between the Economic Commission for Latin America and Caribbean
(ECLAC) and the United Nations Conference on Trade and Development
(UNCTAD), designed to study national experiences in financial-policy reforms
and mobilization of domestic resources with a view to drawing policy conclu
sions for developing countries. ECLAC and UNCTAD gratefully acknowledge
the financial support provided by the World Institute for Development
Economics of the United Nations University (UNU-WIDER), without which
this work would not have been possible.
The present volume contains six papers on financial-policy issues and
country experiences, presented at a joint ECLAC/UNU-WIDER/UNCTAD
seminar held at ECLAC headquarters in Santiago, Chile, on 5-6 October
1992. In the first, Yilmaz Akyuz discusses key issues in financial liberalization.
Next, Alice Amsden analyses East Asian financial policies, pondering the
reasons why there is so much effective government intervention. Machiko
Nissanke then addresses savings and financial-policy issues in Sub-Saharan
Africa. Afterwards, Donald Hanna assesses the Indonesian experience with
finandal-sector reform, followed by Andras Uthoff's study of pension-system
reforms and Giinther Held's study of bank regulations, liberalization and
financial instability in Latin America and Caribbean countries.
Akyüz examines a number of issues in financial liberalization, drawing
on the experience of industrial and developing countries in the 1980s. He
argues that the focus of financial policies in developing countries should
be industrialization and stability, and that a common feature of all modem
examples of industrialization is that they have succeeded in making finance
serve industry and trade, not vice versa. He rejects as fallacy the proposition
that financial liberalization stimulates savings, arguing that there is no simple
14 Y ilmaz A kyüz / G ünther H eld
relation between the two. Financial liberalization often gives rise to deepening,
but the latter is not always associated with a better use of resources since,
inter alia, it tends to increase financial instability and fragility which, in
turn, underminebothproductive and allocative efficiency. Prudentialregulations,
while necessary, may not always be sufficient to prevent instability, particularly
in developing countries, where financial liberalization often leads to escalation
of interest rates and excessive risk taking. It may be necessary not only
to act directly on the asset portfolios of banks but also to use interest
ceilings. Efficiency also depends on the way finance is organized, since
it influences the degree of risk, uncertainty and instability. The German-
Japanese system of bank-based finance tends to be more stable and efficient
than the Anglo-American system of capital-market-based finance, provided
that the required fiscal and monetary discipline and prudential regulations
are in place. Financial instability may be increased, and efficiency reduced,
also by excessive financial openness. The speculative element that dominates
international capital flows is capable of generating gyrations in exchange
rates, giving rise to considerable uncertainty regarding prospective investment
yields with respect of traded goods. This can be particularly damaging at
a time when developing countries place greater reliance on trade as an
engine of growth.
Amsden discusses the rationale for government intervention in financial
markets in late industrializers and the factors that account for successful
intervention in the Republic of Korea and Taiwan Province of China. Challenging
the conventional wisdom that latecomers industrialize by "getting prices
right" on the basis of low wages that procure them a comparative advantage
in labor-intensive industries, she argues that low productivity even in such
industries cannot always be offset by low wages because of political and
social constraints. Nor does foreign capital help to overcome the productivity
drawback: foreign capital typically lags behind, rather than leads, industrial
development. Thus, late industrialization involves a process of using subsidies
to "get prices wrong" so as to overcome the productivity handicap. What
distinguishes the countries that are successfully "sneaking up" to the world
technological frontier from those that are "stumbling back" or "staying behind"
is not less State intervention but rather a different set of principles governing
subsidy allocation. While subsidies in the slow-growing late industrializers
have tended to be allocated according to the principle of "giveaway", in
Introduction 15
the fast-growing ones they have tended to be allocated according to the

principle of reciprocity. In both cases the government has disciplined labor,
but in the East Asian countries it has also disciplined capital. Amsden goes
on to examine the factors underlying the ability o f the East Asian States
to impose performance standards on business, and the ways and means
by which control has been exerted. In the Republic of Korea before the
1980s, everything was prohibited except what the government specifically
allowed; since then, all is allowed except what is specifically prohibited.
The focus of financial policies has been to keep the cost of investment
finance low: despite liberalization in the 1980s, the government has continued
to provide subsidized credit to favored activities. As interest rates tended
to rise, it has concentrated on the promotion of the stock market in order
to provide cheap finance for investors.
Nissanke examines the factors impeding financial intermediation and
mobilization of household savings in Sub-Saharan Africa (SSA), drawing
on the experience of Ghana, Kenya, Malawi and Zambia. She notes that
the potential of households to generate savings is much greater than is
typically believed, particularly in rural areas. Most of these savings, however,
are held in the informal sector, not because of low deposit rates, but primarily
because of lack of access of households and small producers to bank credits.
Thus she stresses the importance of credit in the mobilization of savings,
maintaining that this is a task performed more successfully by the informal
sector in SSA as regards households and small producers. Again, access
to bank credit is impeded not so much by financial repression (i.e. direct
credit allocation favoring the public sector, credit restrictions and the like)
as by credit rationing by banks themselves, arising from imperfect and
asymmetric information and costly contract enforcement. Banks in Africa
are indeed extremely risk-averse and conservative in their lending policy.
This, together with the transaction costs involved in tapping small savings,
explains why they are also often unwilling to broaden their deposit base.
This behavior is closely linked with non-performing loans that pervade the
banking system in SSA and is reflected by an excess liquidity syndrome,
whereby banks permanently hold liquid assets well in excess of legal reserve
requirements. The unsatisfied demand is met by heterogeneous informal
lenders, who have considerable comparative advantage in obtaining information
and enforcing contracts. As Asian experience shows, segmented markets
16 Y ilmaz A kyüz / G ünther H eld
can work efficiently provided that there are adequate linkages, but these
are missing in SSA. Liberalization per se cannot overcome the impediments
to mobilization of resources arising from market segmentation; nor is the
capital market a panacea. More active policies would be needed to ease
the constraints facing banks in financial intermediation and to reduce financial
dualism.
Hanna begins by examining the theoretical links between finance, savings
and economic growth. He then focuses on the financial and real effects
of the financial reforms in Indonesia, which began with the removal of
ceilings on interest rates and credit expansion, followed by the lowering
of barriers to entry and the reduction of the direct role of the Central
Bank in credit allocation. Liberalization led to a large and sustained increase
in financial depth and breadth in the economy. Econometric work on real
private savings shows that they were not hampered by increases in financial
savings, independently of effects stemming from higher income, lower inflation
and real interest rates. The maturity of bank loans also extended over the
period of the reforms, better meshing with the investment needs of the
economy. Cost efficiency improved as well, with a reduction in overheads;
spreads between deposit and lending rates came down at all banks and
m oved closer to each other at different types of banks. However, given
the difficulties in allocating credit because of informational and incentive
problems, efficiency gains were more elusive. The author concludes that
financial reforms have been successful in Indonesia despite several factors
that differ from the standard prescription. First, the capital account was
open at the time of reforms, and second, reform began at a time of important
macroeconomic adjustment.
Uthoff examines the role of financial markets and pension funds
management, and their regulation and supervision, in the context of current
pension-system reforms in Latin American and Caribbean countries. Reforms
today seek to achieve better pension benefits without generating financial
and actuarial deficits, higher returns from portfolio investment of pension
funds, and increased savings and the development of capital markets. Pension-
system reforms require taking basic decisions regarding (i) the structure
of financing (ii) the structure of administration; and (iii) the relation between
contributions and benefits. One major concern of every reform taking place
in the region has been the proper protection and capitalization of reserve
Introduction 17
pension funds. One of the reforms which has attracted wide attention was
implemented in Chile in the early 1980s. It replaced a pay-as-you-go and
publically-administered system with one based on individual capitalization,
predefined contributions and benefits and private administration of funds.
The author emphasizes the following aspects with a view to providing good
pension benefits: (i) the role of the State in closely regulating and supervising
the portfolio allocation of pension funds, and in the provision of clear,
timely and simple information to affiliates regarding their individual accounts;
(ii) an efficient administration of reserve funds geared capitalizing them
at real annual rates of return of at least 5 per cent on average, based
on contributions equal to 10 per cent of a person's lifetime earnings profile;
(iii) the preclusion of oligopolistic behavior on the part of pension-funds,
and (iv) the provision of subsidies for participants unable to obtain minimum
pension benefits on the basis of their own contributions. The system's success
also depends on fiscal and monetary discipline.
Held looks at the outcome of different financial-policy experiences in
Latin American and Caribbean countries in the last two decades on the
basis of institutional arrangements of bank regulation and supervision. He
argues that the emergence of bank solvency problems in a sample of nine
countries in the region shares the common condition of severe flaws in
prudential regulation and supervision. These flaws have quite commonly
led to financial crises under financial liberalization, but they have also
engendered solvency problems under financial repression. Since prudential
regulation and supervision were not in place while credit allocation, interest
rates and other financial variables were liberalized, these experiences could
be better described as uncontrolled financial policies rather than financial
liberalization. In many of the countries studied, highly unstable macroeconomic
conditions were also responsible for severe solvency problems. The author
concludes that of the countries in the sample, only the Chilean and perhaps
the Costa-Rican experiences after the mid-1980s may be regarded as a fair
testing ground for the effects of financial liberalization on savings and exports.
Yilmaz A kyiiz
Giinther Held
Santiago and Geneva, August 1993.

FINANCIAL LIBERALIZATION:
THE KEY ISSUES
Yilmaz Akyüz1
1 I have greatly benefited from comments and suggestions made by various people, including
the participants of a workshop on Financial Globalization and Systemic Risk at the Center
on International Economic Relations, University of Campinas, Sao Paulo, 15-16 June 1992;
the participants of an ECLAC/UNU-WIDER/UNCTAD seminar on Savings and Financial
Policy Issues in African, Asian, Latin American and Caribbean Countries, ECLAC, Santiago,
5-6 October 1992, particularly Carlos Massad and Gunther Held; and my colleagues in
UNCTAD, Shahen Abrahamian, Andrew Comford, Detlef Kotte and Cem Somel. The
examination of financial efficiency in terms of various concepts of risk in section 5 owes a
great deal to discussions with Jan Kregel. None of the persons mentioned are, of course,
responsible for any errors. The opinions expressed in this paper do not necessarily reflect
the views of UNCTAD, and the designations and terminology used are those of the author.
CONTENTS
INTRODUCTION .............................................................................................. 23
I. INTEREST RATES AND SAVINGS ..................................................................... 24
II. FINANCIAL LIBERALIZATION AND DEEPENING............................................ 27
III. ALLOCATIVE EFFICIENCY ................................................................................ 30

1. Market failu re................................................................................................ 30
2. Successful intervention .................................................................................. 32
3. Measuring efficiency ...................................................................................... 33
IV. PRODUCTIVE EFFICIENCY AND COST OF FINANCE ....................................... 35

1. Risk, uncertainty and interest ra te s................................................................. 35
2. Intermediation margin .................................................................................. 36
V. REGULATION OF FINANCE AND FINANCIAL STABILITY .............................. 37

1. Risk-taking by b an k s...................................................................................... 37
2. Prudential regulations.................................................................................... 38
3. Interest ceilings.............................................................................................. 40
VI. OPTIONS IN FINANCIAL ORGANIZATIONS.................................................... 42

1. Bank-oriented and market-oriented finance ................................................... 42
2. Efficiency of alternative systems .................................................................... 43
3. Requirements for an efficient bank-oriented system ........................................ 46
4. Control and regulation of stock-markets ......................................................... 47
VII. EXTERNAL LIBERALIZATION AND FINANCIAL OPENNESS........................... 49

1. The concept of financial openness................................................................... 49
2. The extent of financial openness in developing countries................................. 50
3. Nature of capital flows .................................................................................. 50
4. Recent capital flows to Latin America ............................................................ 53
5. Opening stock-markets to non-residents ......................................................... 55
6. Effects of volatile capital flows on investment and trade ................................. 56
7. Controlling capital flow s................................................................................ 57
VIII. CONCLUSIONS.................................................................................................. 60
REFERENCES............................................................................................................... 63
Financial liberalization: the key issues 23
INTRODUCTION
In recent years financial policies in both industrial and developing countries

have put increased emphasis on the market mechanism. Liberalization was
partly a response to developments in the financial markets themselves: as
these markets innovated to get round the restrictions placed on them, govern
ments chose to throw in the towel. More important, however, governments
embraced liberalization as a doctrine.
In developing countries, the main impulse behind liberalization has
been the belief, based on the notion that interventionist financial policies
were one of the main causes of the crisis of the 1980s, that liberalization
would help to restore growth and stability by raising savings and improving
overall economic efficiency; greater reliance on domestic savings was neces
sary in view of increased external financial stringency. However, these
expectations have not generally been realized. In many developing countries,
instead of lifting the level of domestic savings and investment, financial
liberalization has, rather, increased financial instability. Financial activity has
increased and financial deepening occurred, but without benefiting industry
and commerce.
In many industrial countries the financial excesses of the 1980s account
for much of the sharp slowdown of economic activity in the 1990s. Financial
deregulation eased access to finance and allowed financial institutions to
take greater risks. The private sector accumulated large amounts of debt
at very high interest rates on the expectation that economic expansion would
continue to raise debt servicing capacity while asset price inflation would
compensate for high interest rates. Thus, when the cyclical downturn came,
24 Y ilmaz A kyüz
borrowers and lenders found themselves overcommitted: debtors tried to

sell assets and cut down activity in order to retire debt, and banks cut
lending to restore balance sheets. Thus, the asset price inflation was replaced
by debt deflation and credit crunch.
The recent experience with financial liberalization in both industrial
and developing countries holds a number of useful lessons. This paper draws
on this experience to discuss some crucial issues in financial reform in develop
ing countries. The focus is on how to improve the contribution of finance
to growth and industrialization; developing the financial sector and pro
moting financial activity is not synonymous with economic development.
I. INTEREST RATES AND SAVINGS
One o f the most contentious issues in financial policy is the effect o f interest
rates on savings. There can be little doubt that short-term, temporary swings
in interest rates have little effect on private savings behaviour since that
is largely governed by expectations and plans regarding current and future
incomes and expenditures: they alter the level of savings primarily by affect
ing the levels of investment and income. However, when there is a rise
in interest rates that is expected to be permanent (for instance, because
it is the result of a change in the underlying philosophy in the determination
of interest rates), will consumer behaviour remain the same, or will the
propensity to save rise? The orthodox theory expects the latter to occur,
and thus argues that removing "financial repression" will have a strong,
positive effect on savings (Shaw, 1973, p. 73).
Empirical studies of savings behaviour typically do not distinguish
permanent from temporary changes in interest rates. Recent evidence on
savings behaviour in a number of developing countries that changed their
interest-rate policy regimes shows no simple relation between interest rates
and private savings. This is true for a wide range of countries in Asia
and the Middle East (Indonesia, Malaysia, Philippines, Sri Lanka, Republic
of Korea and Turkey -Cho and Khatkhate, 1989; Amsden and Euh, 1990;
Lim, 1991; Akyüz, 1990), Africa (Ghana, Kenya, Malawi, United Republic
of Tanzania and Zambia- Nissanke, 1990), and Latin America (Massad and
Eyzaguirre, 1990) that undertook financial liberalization, albeit to different
degrees and under different circumstances.

But this should come as no surprise:
• Even according to the conventional theory, the personal propensity
to save from current income depends on the relative strength of two forces
pulling in opposite directions, namely the income and substitution effects.
Moreover, if current income falls relative to expected future income, a rise
in interest rates can be associated with a fall in savings. This often happens
when interest rate deregulation occurs during rapid inflation and is accompa
nied by a macroeconomic tightening that results in a sharp decline in employ
ment and income.
• A large swing in interest rates can lead to consumption of wealth,
especially when non-interest income is declining. This is true especially for
small savers who can react to increases in interest rates by liquidating real
assets and foreign exchange holdings in order to invest in bank deposits
in an effort to maintain their standard of living, consuming not only the
real component of interest income but also part of its nominal component
corresponding to inflation. This tendency is often reinforced by "money illu
sion" or the inability to distinguish between nominal and real interest inco
mes, something that tends to be pervasive in the early stages of deregulation.
Thus, the initial outcome of deregulation can be to lower household savings,
particularly if it is introduced at a time of rapid inflation. For instance
in Turkey high deposit rates in the early 1980s allowed a large number
of small wealth-holders to dissave.
• The behaviour of households may be quite different from that assu
med in conventional theory. For instance, they may be targeting a certain
level of future income or wealth. Higher interest rates may then lower
household savings by making it possible to attain the target with fewer
current savings. For instance, in the Republic of Korea and Japan low interest
rates combined with high real estate prices have tended to raise household
savings (Amsden and Eu, 1990).
• Financial liberalization can lower household savings by allowing
easier access to credit and relaxing the income constraint on consumption
spending. In many countries financial liberalization has, indeed, given rise
to a massive growth in consumer loans (such as instalment credits for cars
and other durables, credit card lending, etc.). This appears to have been
one reason why the household savings rate declined and the debt/income
26 Y ilmaz A kyüz
ratio rose in the 1980s in the United States -something which is at the
heart of the current debt-deflation process (UNCTAD 1991, part two, chaps.
I-II; 1992, part two, chap. E). An inverse correlation between household
borrowing and savings ratios has also been observed in most other OECD
countries since the early and mid-1980s (Blundell-Wignall and Browne, 1991).
• Even if financial liberalization and higher interest rates do not lower
personal savings, they can reduce total private savings and aggregate domes
tic savings by redistributing income away from debtors -a category which
typically includes corporations and the government. In many developing
countries undistributed corporate profits are an important part of private
savings and the most important source of business investment. Generally,
the savings rate is higher than for households: corporate retentions are high,
ranging between 60 to 80 per cent of after-tax profits, because ownership
is usually concentrated in the hands of families and there is no outside
pressure to pay out dividends (Honohan and Atiyas, 1989; Akyüz, 1991).
The redistribution of income from corporations to households through higher
interest rates can thus reduce total private savings even if it raises household
savings. In developing countries this effect can be particularly strong because
firms operate with high leverage, loan maturities are short and corporate
debt usually carries variable rates. Thus, a rise in interest rates not only
raises the cost of new borrowing but also the cost of servicing existing
debt. Evidence from the studies already mentioned suggests that in a num
ber of countries (e.g. Philippines, Turkey, Yugoslavia), sharp increases in
interest rates were a major factor in the collapse of corporate profits and
savings that took place particularly in the early phases of financial libera
lization.
Such adverse effects are especially marked when interest rates are freed
under rapid inflation. There is widespread agreement that financial liberali
zation undertaken in an unstable environment may make things worse, and
that such reforms should be undertaken only after macroeconomic balances
are attained (World Bank 1989; Edwards, 1989). Nevertheless, many countries
have resorted to liberalization as part of shock therapy against stagflation.
Thus, interest rate increases are not a reliable instrument for raising
domestic savings, but can damage macroeconomic stability and investment.
The crucial question is how to design interest rate policies compatible with
sustained stability and growth.
The historical experience of major industrial countries holds some use

ful lessons. Until the 1980s, real short-term interest rates in these countries
were slightly negative and real long-term bond rates slightly positive; i.e.
about one to two per cent below and above inflation respectively. Until
the oil shocks of the 1970s, there was sustained growth and price stability.
But since the beginning of the 1980s (for reasons to be discussed later),
real interest rates have been, on average, more than twice their historical
levels. Nevertheless, these countries enjoyed one of the longest periods of
economic expansion in the postwar period with low inflation. This genera
ted a widespread perception that high real interest rates do not impede
investment and growth, but help price stability. However, the subsequent
debt-deflation-cum-recession has clearly shown that economic expansion
attained at very high real interest rates eventually depresses income, invest
ment and growth.
II. FINANCIAL LIBERALIZATION AND DEEPENING
It is generally agreed that financial liberalization raises financial activity

relative to the production of goods and non-finandal services. However,
there is much less consensus on the causes and effects of this "financial
deepening". According to the financial repression theory (McKinnon, 1973;
Shaw, 1973) financial deepening represents increased intermediation between
savers and investment because higher interest rates raise savings and shift
them from unproductive assets towards financial assets, thereby raising the
volume of productive investment.
While it is true that financial liberalization can shift existing savings
toward financial assets, reallocation is not the only or even the most impor
tant reason for financial deepening. Financial liberalization can also lead
to deepening by redistributing savings and investment among various sec
tors, and by creating greater opportunities for speculation. Since these can
worsen the use of savings, financial deepening is not necessarily a positive
development.
The prime role of the financial system in the savings/investment process
is to intermediate between deficit and surplus sectors rather than to transfer
aggregate savings into aggregate investment. Deficit sectors (typically the
28 Y ilmaz A kyüz
corporate sector and the government) save as well as invest, while surplus
sectors (households) invest as well as save. Thus, redistribution of savings
and investment among sectors can, by changing sectoral surpluses and defi
cits, result in financial deepening without any change in aggregate savings
and investment - for instance, as already noted, when higher interest rates
redistribute income and savings from debtors to creditors. Even when this
does not alter the volume o f aggregate savings (i.e. lower savings of debtors
are compensated by higher savings of creditors), it increases deficits and
surpluses and, hence, the amount of financial intermediation. Indeed, finan
cial intermediation can increase while aggregate savings and investment
fall (Akyüz, 1991). This can happen even under the orthodox assumptions
that saving rates are positively related to the interest rate and that savings
determine investment and growth (Molho, 1986, p. 112).
In such cases financial deepening is a symptom of a deterioration
of the finances of the corporate and public sectors, reflecting an accumula
tion of debt in order to finance the increased interest bill rather than new
investment. Financial deepening driven by such Ponzi financing has been
observed in a number of countries (e.g. Turkey, Yugoslavia and New Zea
land) where financial liberalization redistributed income in favour of cre
ditors and encouraged distress borrowing.
Similarly financial deepening can be the result of a redistribution of
a given volume of aggregate investment, when, for instance, higher inte
rest rates induce households to reduce investment in housing and shift
to bank deposits. Then, the increase in the household surplus and in the
volume of deposits represents a decline in household investment, not a
rise in savings.
Financial liberalization often raises holdings of both financial assets
and liabilities by firms and individuals at any given level of income, investment
and savings. This tendency to borrow in order to purchase assets is driven
by the increased scope for capital gains generated by financial liberaliza
tion. Liberalization increases the instability of interest rates and asset prices,
thereby raising prospects for quick profits through speculation on changes
in the market valuation of financial assets. It also allows greater freedom
for banks and other financial institutions to lend to finance activities unrela
ted to production and investment, and to firms and individuals to issue
debt in order to finance speculation. These can generate considerable fi-
F inancial liberalization: the key issues 29
nandal activity unrelated to the real economy, and lead to financial dee
pening -as in the United States in recent years through leverage takeovers,
mergers, acquisitions and so on (UNCTAD, 1992, part two, chap. II).
Deepening can also result from the impact of changes in interest rates
on the form in which savings are held. Indeed, one of the main reasons
why savings do not in practice strongly respond to increases in real interest
rates is the existence of a range of assets with different degrees of protection
against inflation; returns on such assets also influence savings decisions.
The greater the influence of interest rates on the allocation of savings among
alternative assets, the smaller the influence on the volume of savings.
Whether shifts of savings into financial assets improve the use of resour
ces depends on where they come from and how efficiently the financial
system is operating. Clearly, a switch from commodity holdings can improve
the use to which savings are put. But, contrary to widespread perception,
there is very little evidence of extensive commodity holding in developing
countries as a form of savings. Such holdings entail substantial storage and
transaction costs, making their own real rate of return typically negative.
Moreover, there is considerable uncertainty regarding the movement of pri
ces of individual commodities even when the general price level is rising
rapidly. These factors, together with the existence of more liquid, less costly
inflation hedges (such as foreign currency or gold) reduce the demand for
commodities as a store of value. The large commodity holdings that exist
in African countries typically reflect the nature of production and non
monetization of the rural economy. Consequently, increases in deposit rates
are often unable to induce liquidation of commodity stocks (Aryeetey et
al., 1990; Mwega, 1990; Nissanke, 1990).
An increase in domestic interest rates can induce a shift from foreign
currency holdings to domestic assets, and repatriation of flight capital. Many
governments, however, have found it necessary to legalize foreign currency
holdings and introduce foreign currency deposits for residents and to offer
very high interest rates in order to attract foreign currency holdings to
the banking system. Certainly, in both cases the portfolio shifts can increase
the resources available for investment and deepen finance. However, as
discussed in section VII, capital flows and dollarization resulting from such
policies often prove troublesome for macroeconomic stability, investment
and competitiveness.
30 Y ilmaz A kyüz
Freeing interest rates in the formal sector can also trigger a shift away
from informal markets. However, the scope of such shifts may be limited
since the reason for informal markets is not always interest rate controls
and credit rationing. They often provide services to small and medium pro
ducers w ho do not have access to bank credits. Since financial liberalization
does not always improve their access to banks, informal markets continue
to operate after the deregulation of interest rates. As savings placed in
the informal sector assure these producers some access to credit, they are
not always willing to shift to banks when deposit rates are raised (Chipeta,
1990; Aryeteey et al., 1990; Mwega, 1990). On the other hand, when funds
are shifted to banks, the cost of finance for informal market borrowers
can rise considerably. Moreover, such shifts can result in financial "sha
llowing" because informal markets provide more financial intermediation
due to the absence of liquidity and reserve requirements (Wijnbergen, 1983;
Owen and Solis-Fallas, 1989).
It can thus be concluded that financial deepening brought about by
liberalization is not necessarily associated with a higher level a n d /or better
use of savings. Indeed, the empirical evidence does not support the claim
that financial deepening is associated with faster growth (Dombusch and
Reynoso, 1989). The degree of financial deepening is therefore not a good
measure of the contribution of finance to growth and development.
The relevant issue in financial reform is efficiency rather than deepen
ing. There are various concepts of efficiency of financial markets and insti
tutions (Tobin, 1984), but from the point of view of the role of finance
in economic growth and development, the conventional notions of allo
cative and productive (i.e. cost) efficiency are the most relevant ones.
III. ALLOCATIVE EFFICIENCY
1. Market failure
Financial markets and institutions can be said to be allocatively efficient

if they direct resources to their more socially productive use, i.e. if they
finance investment with the highest social rates of return. This concept broad
ly corresponds to what Tobin (1984, p. 3) calls functional efficiency and
provides a rationale for devoting resources to financial activity.

Allocative efficiency is closely related to the extent of "the accuracy
with which market valuations reflect fundamentals" ("fundamental-valuation
efficiency", Tobin, 1984, p. 5). Prices of financial assets provide market
signals for resource allocation. Speculative bubbles in securities markets
influence investment and consumption decisions as well as financing plans
of corporations, while exchange rate misalignments cause misallocation of
resources between traded and non-traded goods sectors.
There is ample evidence that in industrial countries financial liberali
zation has resulted in a considerable increase in the volatility of interest
rates, equity prices, exchange rates and the prices of real estate, gold, silver
and collectable assets, and caused large and sustained deviation of these
from their fundamental values (e.g. Cutler, Poterba and Summers, 1990;
Miller and Weller, 1991; Kupiec, 1991). Similarly, "takeover mania, motivated
by egregious undervaluations, is testimony to the failure of the market on
this fundamental-valuation criterion efficiency" (Tobin, 1984, p. 6). These
deviations reflect pervasiveness of speculative forces: "the similarity of pa
tterns in a wide range of asset markets suggests the possibility that they
are best explicable as a consequence of the speculative process itself." (Cutler,
Poterba and Summers, 1990, p. 36).
Quite apart from the distorting effects of speculation on asset prices
and resource allocation, financial markets also fail to allocate resources effi
ciently because of a number of imperfections not attributable to government
intervention. These include missing markets, asymmetric and incomplete
information, and various externalities not mediated by markets (Stiglitz and
Weiss, 1981; Greenwald and Stiglitz, 1986; Stiglitz, 1989a; Datta-Chaudhuri,
1990). Such market failures are more serious in developing than in developed
countries and tend to obstruct the learning process which plays a key role
in m odem industrialization. "Learning ... means that it will not be optimal
to pursue myopic policies; one cannot use current comparative advantage
as the only basis for judgements of how to allocate resources. Moreover,
it may be optimal to initially incur a loss; the imperfections of capital mar
kets thus may impose a more serious impediment on LDCs taking advan
tage of potentials for learning" (Stiglitz, 1989a, p. 199).
32 Y ilmaz A kyüz
2. Successful intervention
Governments in many countries have therefore acted to influence the allo

cation and pricing of finance as part of their industrial policy. Indeed, al
most all modem examples of industrialization have been accompanied by
such intervention. Directed and preferential credits have been the most im
portant instruments of some successful industrializes in East Asia (Ams
den, 1989; Bradford, 1986; Cho and Khatkhate, 1989; Hanson and Neal,
1985; Kato et al., 1993; Westphal, 1990). As noted by a recent report, in
Japan an important instrument of intervention was policy-based finance,
used through the Japan Development Bank "to induce the private sector
to achieve specified policy objectives". It was based on the recognition that
"if the private financial market were perfect (in terms of competition, infor
mation and freedom of transactions) policy-based finance would be unne
cessary. In reality, however, there are limits to the perfect fulfillment of
these conditions in the financial market. Thus, one can understand the
significance of policy-based finance as one means of compensating for these
market limitations" (Kato et al., 1993, p. 28).
In the Republic of Korea "government intervention was necessary not
just to steer credit in the right direction but to underwrite production during
the learning process that was far more involved than what is commonly
meant by 'infant industry protection'. Subsidized credit meant the diffe
rence between establishing new industries or not, rather than the difference
between high and low profits" (Amsden and Euh, 1990, p. 31). Thus, "exten
sive intervention by the government with South Korea's financial system
can be viewed as an internal capital market and, consequently, it could
have led to a more efficient allocation of credit than possible in a free-
market financial system" (Lee, 1992, p. 187).
But many countries have directed credit with much less success. The
differences between successful and unsuccessful intervention have been par
tly due to skill in "picking winners". While it is true that governments
are not necessarily better equipped to do this than markets, the experience
strongly suggests that whether a firm (or an industry) is a winner depends
on how it is managed. A number of factors seems to separate success from
failure in this respect:
• The ability of governments to prevent interventionist finance from
degenerating into inflationary finance, to resist excessive credit expansion

and to ensure fiscal discipline: macroeconomic stability appears to have
been a more important factor in growth than financial liberalization and
deepening (Dombusch and Reynoso, 1989);
• Making provision of support conditional upon good performance,
and seeing that government support and protection are actually used for
the purposes intended rather than simply as a handout;
• Designing objective, well-defined and market-based performance
indicators -namely, competitiveness in world markets and export perfor
mance- in order to assess the nature and extent of the support needed,
and whether it is being used effectively (Amsden, 1989; Westphal, 1990);
• Attaining social consensus on the purpose and modalities of govern
ment intervention. As noted by a recent report, this was particularly im
portant in the success of the policy-based finance in Japan: "when the gover
nment does intervene in private economic activities, or carries out economic
activities itself in place of private actors, it must not merely give some
abstract reason, but rather clearly explain the concrete need for and obtain
social agreement on those activities" (Kato et al., 1993, p. 28). This has
been achieved through extensive participation of the private sector in the
policy-formation process based on the "public-private cooperative system",
i.e. in the advisory councils, including "industry leaders and general citi
zens", as well as bureaucrats, which is still widely used (Kato et al., 1993,
p. 85).
3. Measuring efficiency
As noted above, the main impulse to financial liberalization in developing

countries has come from the frustration with ineffective and wasteful
interventm t and the belief that liberalization would raise allocative efficiency.
Thus, m ^ty countries have chosen to liberalize finance rather than reform
their industrial policies and state intervention. However, this has not always
resulted in a better allocation of credits.
In the orthodox theory, better allocation means a tendency towards
equalization of rates of return on investment in different sectors. Similarly,
a more efficient credit allocation is expected to reduce the variation of the
cost of finance across borrowers on the assumption that profit maximiza-
34 Y ilmaz A kyüz
tion requires equalization of marginal cost of borrowing and marginal rate

of return on investment (Cho, 1988).
These measures, however, are highly inappropriate. First of all, as dis
cussed below, one important determinant of the rate of return and cost
of capital is risk. When projects carry different risks, an optimal allocation
must reflect these differences in rates of return and borrowing costs. More
important, when capital markets are short-sighted, equalization of profit
rates typically means discriminating against those firms and industries with
dynamic comparative advantages and learning potentials that have to incur
initial losses. Since financial liberalization is often associated with a shorten
ing of time horizons, a tendency towards equalization of rates of profit
and cost of capital could worsen allocation.
Financial liberalization normally reduces or eliminates credits on pre
ferential terms and hence diminishes variations in cost of capital across
sectors. Therefore, measuring the effect of financial liberalization on alloca
tive efficiency in terms of reduced variations in cost of capital is tautological.
On the other hand, a successful industrial policy could reduce variance
in borrowing cost by diminishing the number of industries requiring spe
cial treatment. For instance, it has been argued that the decline in the inter
industry variance of borrowing costs in the Republic of Korea in the 1980s
compared to the 1970s reflects the success not of financial liberalization
as suggested by some authors (e.g. Cho, 1988), but of industrialization policies
(Amsden and Euh, 1990, pp. 43-44).
Financial liberalization in developing countries often changes signifi
cantly the sectoral allocation of credit. Evidence suggests that typically the
shares of service sectors, consumer loans and property-related credits tend
to increase at the expense of industry. This may result from a reduction
in directed credit allocation, which often favours industry and does not
necessarily indicate a deterioration of resource allocation. However, it is
important to note that these changes are often associated with shortening
of maturities and declines in demand for manufacturing investment credits,
when liberalization takes place in an unstable environment and results in
very high and volatile interest rates.
Perhaps more important indicators of the effects of financial libera
lization on allocative efficiency are the number of non-performing loans,
loan default rates and bank failures. Evidence from a number of countries
(e.g. the Southern Cone countries, Indonesia, the Philippines, Turkey and
Yugoslavia) indicates that deregulation of interest rates and elimination of
restrictions on financial activities have almost always been followed by in
creases in the proportion of non-performing loans in bank portfolios and
in bank failures. Again, resort to liberalization to cure instability and stagna
tion has often played an important role. These, together with external shocks,
had already greatly weakened the balance sheets of the corporate sector
and financial institutions. Deregulation of interest rates, often accompanied
by monetary tightening, further disrupted the financial position of the highly-
leveraged corporate sector, leading to increased loan default rates and even
tually to bank failures.
IV. PRODUCTIVE EFFICIENCY AND COST OF FINANCE
The traditional concept of productive efficiency refers to microeconomic

efficiency of firms in producing goods and services with given prices for
their inputs. When applied to the financial system, this concept would sim
ply be translated into intermediation cost or interest spread. However, one
must approach productive efficiency from a broader perspective and define
it as the ability of the financial system to provide finance at the lowest
possible cost. This depends not only on the extent to which financial inter
mediaries minimize the cost of intermediation between the ultimate lender
and the ultimate borrower, but also on the ability of the entire financial
system to minimize the interest paid to the ultimate lender (the lender's
interest rate).
1. Risk, uncertainty and interest rates
The Keynesian notions of lender's and borrower's risks provide an appro

priate framework for discussing the determinants of cost of finance and
the effects of financial liberalization on productive efficiency (Keynes, 1936,
p. 144). An important determinant of the lender's interest rate is the risk
due to the possibility of default by the borrower, i.e. the lender's risk. First,
there is the risk of voluntary default, or what Keynes calls the moral risk:
the lender must make an allowance for the possibility of dishonesty of
36 Y ilmaz A kyüz
the borrower. Second, there is the risk of involuntary default arising from
imperfect foresight, i.e. from uncertainties over factors outside the control
of the borrower which affect profitability. This risk, called the borrower's
risk or the pure risk, is inherent in all investment decisions and cannot
be eliminated. However, it can be reduced by access of the borrower to
better information and stabler economic conditions. The pure risk is closely
related to allocative efficiency. When finance is not efficiently allocated, the
probability of involuntary default increases. This raises the lender's risk and
the cost of finance: allocative inefficiency thus aggravates cost inefficiency.
The lender also runs a risk regarding the capital value of his assets
due to uncertainties over future interest rates and asset prices (as well as
the price level). The capital-value uncertainty increases with the volatility
of asset prices and interest rates, as well as with the increased frequency
of bank failures. These raise liquidity preference and lower the demand
for capital-uncertain assets, thereby shortening the maturities of financial
assets and pushing up interest rates, especially long-term. The degree of
productive efficiency of the financial system therefore depends in part on
its ability to attain stability and reduce capital-value uncertainty.
The search for greater allocative efficiency through financial libera
lization can greatly reduce the productive efficiency of the financial system
by giving rise to increased financial instability and raising the cost of finan
ce to investors. This is a systemic influence, quite independent of any rise
in interest rates that may result from eliminating ceilings. Indeed, the fi
nancial instability and bank failures stemming from financial liberalization
in the major industrial countries, especially the United States, in the 1980s
played a major role in considerably raising long-term interest rates and
reducing their sensitivity to changes in short-term rates (Akyüz, 1992, pp.
59-60).
2. Intermediation margin
The second component of the cost of finance, namely the intermediation

margin, reflects the microeconomic efficiency in the use of resources devo
ted to financial activity. This is particularly important in bank finance, even
though mark-ups of intermediaries in stock exchanges are not negligible
(Tobin, 1984, p. 4). The spread between lending and deposit rates is influen
ced by operating expenses, legal reserve and liquidity requirements, and

the pressure of competition on profit mark-ups. Reserve and liquidity
requirements are typically lowered as part of financial liberalization. Simi
larly, operating expenses and profit mark-ups tend to fall as entry barriers
are dismantled.
However, financial liberalization also tends to increase the spread by
raising the rate of default on loans, since banks often pass the cost of
bad loans onto other borrowers. Therefore, erroneous investment and fi
nancing decisions and allocative inefficiency can lead to cost inefficiency
by raising not only the lender's interest rate, but also the spread. The increa
se in the cost of finance, in turn, can push sound borrowers into insolvency,
thereby increasing loan default rates and pushing up the lender's risk and
the lender's interest rate further. This often leads to Ponzi financing where
by banks increasingly lend to high-risk, speculative business at very high
interest rates in order to cover high deposit rates and defaults. Such a
process is unsustainable, but it can nevertheless cause considerable waste.
V. REGULATION OF FINANCE AND FINANCIAL STABILITY
The preceding discussion has shown that stability is an essential attribute

of an efficient financial system. After many episodes of turmoil in financial
markets in both developing and developed countries, there now appears
to exist a consensus on the need for prudential regulations in order to
attain stability. But, can such regulations and supervision prevent financial
instability when interest rates are allowed to fluctuate freely and banks
are left free to compete for deposits by bidding up interest rates?
1. Risk-taking by banks
The theory of finance suggests that because information is imperfect and

asymmetric (the borrower knows more about his investment than the len
der) and contracts are incomplete (lenders cannot control all aspects of the
borrower's behaviour), banks implement their own quantity rationing by
imposing credit ceilings, and restrict deposit and loan rates in order to
avoid excessive risk-taking (Stiglitz and Weiss, 1981; Davis, 1993, pp. 13-
38 Y ilmaz A kyüz
16). Since higher interest rates tend to reduce the average quality of loans
through adverse selection (lending to high-risk borrowers willing to pay
high interest rates) and moral hazard (inducing "good" borrowers to invest
in riskier projects), the expected rate of return net of default will decline
once the loan rate has reached a certain level. This implies that even in
the absence of prudential regulations, there will be limits to price compe
tition and risk-taking in the banking sector.
However, self-restraint cannot always be relied on to prevent financial
instability, particularly in developing countries. Banks tend to engage in
speculative financing and excessive risk-taking provided that failure does
not have serious consequences for their shareholders and managers. This
happens when they can easily acquire deposit insurance, enjoy implicit or
explicit guarantees for bail-out and have easy access to the lender-of-last-
resort facility, and when sanctions and penalties for failing bank managers
are inadequate. This is often the case in developing countries where govern
ments are often all too ready to rescue banks in trouble. The moral hazard
that results is made worse by the existence of deposit insurance schemes
designed to give protection to depositors and attract funds into banks. Banks
often have to pay very little for the insurance coverage while having all
the incentives to raise deposit rates tomobilize funds to invest in high-
return, high-risk, and often speculative projects.
Furthermore, in developing countries large non-finandal corporations
are often able to exert strong influence over banks, causing bank lending
to be concentrated on a small number of firms, at the cost of increasing
their own vulnerability. Corporate distress borrowing and Ponzi financing
tend to be much more common in developing countries, and these become
particularly visible and problematic during episodes of financial .liberali
zation. The intense competition that banks in many developing countries
face from unregulated financial markets can also lead to higher interest
rates and greater risk-taking.
2. Prudential regulations
Evidence from both developed and developing countries shows that a ju
dicious combination of effective prudential and protective regulations is
necessary to prevent financial instability. In many developing countries,
however, regulations restricting excessive risk-taking and/or covering such

risks are absent. In some countries government restrictions on lending to
a single firm and the acquisition of real estate or shares in non-financial
corporations are strict but not implemented. Legal provisions against bad
assets are either absent or ignored, and capital requirements are inade
quate, non-existent or unimplemented. There is widespread non-compliance
even with legal reserve requirements, not always because they are especia
lly high, but because the monetary authorities are unable to impose su
fficient penalties.
However, prudential regulations, while necessary, may not always be
sufficient to prevent financial instability. With the freeing of deposit rates,
considerable competition can build up between the newly deregulated and
unregulated financial sectors, giving rise to sharp increases in deposit rates,
thereby raising the loan rates and deteriorating the quality of bank assets
as high-yield, high-risk lending replaces safer but lower-yielding portfolios.
It is not always possible to check this process through prudential regula
tions on the asset side of banks' balance sheets. Pressures can develop to
allow banks to enter into new lines of business in order to restore their
profitability and viability in the face of higher deposit cost. Such pressu
res will often find favour with the liberalist view underlying interest rate
deregulation, and hence result in the relaxation of constraints on types of
bank lending and investment.
The experience of the United States in the 1980s illustrates how easily
such a process can develop (UNCTAD, 1992, part two, chap. E). As the
Fed m oved away from targeting interest rates to monetarism in order to
reduce inflation and the Regulation Q ceilings on deposit rates were lifted,
banks with long-term portfolios with fixed interest rates (particularly mu
tual savings banks and Savings and Loan Associations, S&Ls) experienced
serious difficulties. Considerable pressure developed for the introduction
of legislation to attract deposits to these institutions (e.g. raising deposit
insurance limits) and to allow them to invest in high-yield, high-risk assets.
Thus, these institutions, and subsequently commercial banks, increasingly
financed consumer and credit card loans, high-yield non-investment grade
(junk) bonds, leverage buy-outs, real estate acquisition, and development
and construction loans. A large amount of debt was accumulated by house
holds and firms while banks acquired high-risk assets. This process ended
40 Y ilmaz A kyüz
with the collapse of the S&Ls with an estimated cost of about $200 billion,
and was replaced by the debt-deflation process already mentioned.
Stricter capital adequacy requirements of the type recently introduced
by BIS (UNCTAD, 1992, part two, Annex I) could have helped to slow
down this process but would probably not have prevented it. As there
was simultaneously a speculative bubble in the stock-market, banks would
have had no difficulty in raising capital on very favourable terms to cover
their high-risk investment, but would have remained exposed to risks on
both sides of their balance sheets. Indeed, this is exactly what happened
in Japan where banks can account as capital almost half of accrued but
unrealized capital gains on equities and use them to offset potential loan
losses. As the stock-market was rising rapidly in the 1980s, banks counted
on these gains instead of setting aside reserves against potential losses on
high-risk, property-related lending. The subsequent decline in stock prices,
together with the fall in property prices, thus created difficulties for banks
from both sides of their balance sheets.
There are also other instances of boom and bust where rapid expan
sion of some banks through high-risk, high-return lending increased their
stock prices sharply and allowed them to raise capital at costs lower than
the prudent banks. "In such cases neither public scrutiny of bank balan
ce sheets, nor capital ratios would have prevented the propagation of the
crisis" (Kregel, 1993, p. 10).
3. Interest ceilings
In short, competition among financial institutions can easily result in esca

lation of interest rates an d/or excessive risk-taking either because pruden
tial capital requirements become ineffective or pressures build up to relax
controls over bank asset portfolios. Such risks are greater in developing
countries. This, together with the fact that stability of interest rates and
asset prices is essential for an efficient financial system, constitutes a strong
case in favour of controlling interest rates as well as bank lending.
An effective way of doing this is to impose statutory ceilings on deposit
a n d /or loan rates. Such ceilings were widely used in industrial countries
until recent years. In Japan, for instance, interest rate regulations played
a crucial role as a "policy-based framework established throughout the high
growth period" (Kato et al., 1993, p. 122), and have not yet been abolished
totally. Again, the recent legislation in the United States regarding the de
pository institutions (the Federal Deposit Insurance Corporation Improve
ment Act of 1991, Jones and King, 1992) stipulates mandatory restrictions
on deposit interest rates for undercapitalized banks in the context of capital-
based policy of prompt corrective action. Since undercapitalization is wides
pread among banks in developing countries, the scope for the application
of such restrictions must be much greater.
Regulation of short-term interest rates through intervention in inter
bank markets is also essential for attaining greater financial stability and
preventing frequent bank failures, particularly when there is considerable
maturity mismatching between banks' assets and liabilities. Under such con
ditions, large swings in interest rates can create serious dilemmas for banks.
If banks respond to an unexpected increase in market interest rates by
raising deposit rates, their profits can be sharply reduced and their solvency
threatened. If they do not, or if they are prevented from doing so by deposit
ceilings, they may suffer a considerable deposit drain. Banks can respond
to increased swings in short-term rates with variable-rate loans or by shor
tening the maturities of their assets, as they have indeed done in many coun
tries, but when done on a laige enough scale this simply transfers the interest
rate risk onto the borrower and replaces it with greater credit risk.
It should be kept in mind that control over interest rates through
ceilings and intervention does not eliminate the need for certain types of
prudential regulations to reduce financial fragility, i.e. vulnerability to de
fault in the corporate and household sectors (Minsky, 1982, 1986; Davis,
1993). This is particularly true in developing countries where the level of
economic activity is much more variable. When activity is buoyant, banks
tend to lend increasingly against assets which carry considerable capital
risk, including not only illiquid assets such as property but also securities;
they also expand consumer credits and invest directly in securities and
property. But when the expansion comes to an end and incomes and asset
prices start to fall, the quality of bank assets can deteriorate rapidly, and
even set off a debt-deflation process and credit crunch. Reducing the fragi
lity of the financial system thus calls for prudential regulations designed
to prevent excessive investment and lending with considerable capital risk
arising from their susceptibility to changes in the pace of economic activity.
42 Y ilmaz A kyüz
VI. OPTIONS IN FINANCIAL ORGANIZATIONS
The discussions above suggest that the efficiency of the financial system
crucially depends on the way it is organized, because that influences the
nature and the degree of risk, uncertainty and instability. On the other
hand, the experience of industrial countries shows that there is no single
way of organizing finance. Consequently, an important issue in financial
reform in developing (and Eastern European) countries is what types of
financial institutions and markets need to be promoted.
1. Bank-oriented and market-oriented finance
It is possible to distinguish between two broad types of financial arran

gements according to whether or not banks and capital markets serve dis
tinct functions. In an ideal-type differentiated system, banks act primarily
within the monetary system, arranging payments and extending short-term
commercial credits. Corporations obtain investment finance in the capital
market by direct security issues, often via the intermediation of investment
banks for underwriting and brokerage. Ownership of companies is highly
fragmented: an important part of corporate securities is held by households
and institutional investors in diversified portfolios. Such a segmentation is
the essence of the Anglo-American system which we will call, for brevity,
the market-oriented system2.
In the German-type of universal banking (the bank-oriented system),
on the other hand, commercial banks play a much greater role at all stages
in the process of corporate investment. They provide investment finance
and function also, like investment banks. They also have considerable con
trol over firms both through their own equity holding and proxy votes
for private investors, and by appointing representatives on the boards of
firms. They lend primarily to firms in which they hold equity interest. Hou
2 The description of various systems here draws largely on Corbett and Mayer (1991), Kregel
(1991) and Somel (1992). For a summary account of the structural aspects of these systems,
see also Davis (1993, pp. 23-26). We do not examine here how these different systems
evolved, but there can be little doubt that government policies and regulations played a
major role.
sehold financial wealth tends to be held in banks rather than direct secu
rities, and bank credits account for a larger proportion of external financ
ing of corporate investment.
There are certainly a number of variants combining elements of both
systems. In the United Kingdom, commercial banks do not have much con
trol over corporations, but there is no legal separation between commercial
and investment banking. In Japan, commercial banks hold corporate equi
ties but are prevented from playing a major role in the underwriting of
corporate securities. Individual ownership of stocks is much smaller than
ownership by financial and non-finandal corporations, and corporate equi
ty is controlled through interlocking shareholding within industrial groups
where banks play a central role. Banks also control other financial insti
tutions (e.g. pension funds) that invest in equity. Lending by banks and
insurance and pension funds usually takes place within the same groups
and involves purchase of company bonds as well as loans. In Japan, bank
credits have played a much more important role in financing business growth
than in the other countries discussed here, although recently there has been
a shift to securities markets.
2. Efficiency o f alternative systems
In recent years many developing countries have been seeking to institute

and promote capital markets, often as part of the structural adjustment
programmes. One of the main reasons for privatizing public enterprises
has been precisely to promote capital markets. Similarly, access to equity
markets has been granted to non-residents in order to boost demand.
There are a number of arguments in favour of developing capital mar
kets as a way of overcoming the paucity of investment finance in deve
loping countries. The bank-oriented system of investment finance has tra
ditionally been viewed as inherently problematic because of the risks asso
ciated with maturity transformation in a volatile economic environment.
Such a system increases the vulnerability of firms to financial shocks since
the cost and availability of bank credit often undergo sharp and unexpected
changes. By contrast, capital markets are expected to provide firms with
more predictable, longer-term finance, while secondary markets in securities
accord savers liquidity. It is also often argued that they would exert better
44 Y ilmaz A kyüz
financial discipline over firms through shareholder action and the threat
of being taken over by other firms.
While there are often serious problems and weaknesses with a bank-
oriented system in developing countries, the benefits claimed for a market-
oriented system are unsubstantiated. It is often overlooked that the finan
cial systems in Germany and Japan have not only proved to be remar
kably stable, but also in the major respects discussed in sections El and
IV are more efficient than the Anglo-American system. Historically, finan
cial asset prices and interest rates in Germany and Japan have been less
volatile than in the United States, bank deposits more stable, and financial
disruptions and bank failures less frequent. Moreover, the cost of finance
to industry has been much lower in Germany and, more particularly, in
Japan. Evidence suggests that high capital costs have contributed to de
clining competitiveness in both industry and international banking in the
United States. Lower capital costs and a more predictable supply of fi
nance appear to have enabled Japanese firms to undertake longer-term pro
jects, including investment in research and development, whereas United
States firms have been deterred (McCauley and Zimmer, 1989; Poterba, 1991;
Zimmer and McCauley, 1991)3.
One of the main reasons for the greater stability and efficiency of
the financial systems in Germany and Japan is their ability to overcome
the dilemma posed by m odem capital markets. As noted by Keynes (1936,
chap. 12), m odem capital markets reconcile the social need for investment
with the preference of individual investors for liquidity. This is a necessity
since "if individual purchases of investment were rendered illiquid, this
might seriously impede new investment". However, while secondary mar
kets in securities accord savers liquidity, they also open up prospects for
speculation whereby most of the players "are, in fact, largely concerned,
not with making superior long-term forecasts of the probable yield of an
investment over its whole life, but with foreseeing changes in the conven
3 In conformity with contemporary trends and in response to outside pressure, Japan has
been undergoing a transition towards a market-based and open financial system, which is
not easy to reconcile with the policy of cheap finance. There have been severe fluctuations in
share prices, interest and exchange rates, and a tendency for the cost of finance to rise
(Martin, 1992).
tional basis of valuation a short time ahead of the general public". Thus,
these markets tend to operate like "casinos" where players speculate on
the speculations of other players.
The pattern of shareholding and ownership that characterizes the
German and the Japanese systems has allowed them largely to overcome
this dilemma. The fact that banks and business groups with a long-term
stake in the corporations hold the controlling interest means not only that
secondary markets tend to be less active and volatile, but also that the
managers do not need to pay much attention to how the market values
their assets from day to day, and can concentrate instead on the long term.
This also helps reduce liquidity preference and short-termism on the part
of individual investors and portfolio managers.
The bank-oriented system can exert a different and more efficient fi
nancial discipline over enterprises than the market-oriented system. Banks
in Germany and Japan (and banking groups) are often in a position to
monitor the performance of management by direct access to information
through their close and long-term relations with firms as shareholders and
creditors, and to intervene when needed in order to prevent failure. By
contrast, in the Anglo-American system of fragmented shareholding, indi
vidual investors have neither the means nor the incentive to monitor and
control corporate management. In extreme cases, market discipline is exer
cised through hostile takeovers, but these are often disruptive and waste
ful. More important, since markets tend to value the enterprise largely on
the basis of short-term financial performance, the takeover threat creates
pressures and incentives for the management to think short-term.
Furthermore, the internal capital market organized within banks and
firms connected by cross shareholding also improves enterprise performan
ce by reducing the borrower's risk by permitting economies of scale in
collecting, processing, evaluating and disseminating information. For the sa
me reason, there is less credit rationing in a bank-based system (Fama,
1985; Driscoll, 1991). Such a system also makes it possible to reduce consi
derably the lender's risk and the rate of interest, since it gives deposit
holders the liquidity they seek with a smaller risk of capital uncertainty
by pooling and institutionalizing the risk associated with individual invest
ment projects, and by reducing erroneous investment decisions.
A financial system with a close interface between banks and corpora
46 Y ilmaz A kyüz
tions tends to lower the rate of return required by investors to underta

ke investment. The expected rate of return on investment must be high
enough to cover both the borrower's risk and the rate of interest received
by the lender. However, as noted above, the borrower's risk is an impor
tant determinant of the lender's risk. The effect of this duplication of the
borrower's risk on the rate of interest can be reduced by increasing the
degree of the lender's involvement in the borrower's investment and other
managerial decisions, since the lender is then better able to assure himself
that pure risk is being properly weighed by the borrower: indeed both
components of the lender's risk (i.e. the moral risk and the pure risk) would
disappear if the lender and the borrower were the same person. The bank-
oriented system thus reduces the extent to which the borrower's risk is
duplicated in the lender's risk and the interest rate, and, hence, lowers
the cost of investment.
3. Requirements for an efficient bank-oriented system
These are particularly important considerations to be taken into account

in reforming the financial system in developing countries, where the cost
of finance needs to be kept low and firms must be able to take the long
view in order to succeed in "learning by doing". However, the experience
of many developing countries shows that the concentration of ownership
in the hands of inside investors and close relations between banks and
corporations are not necessarily conducive to good enterprise performan
ce and financial stability. Indeed, in many developing countries the equity
control of corporations is in the hands of families or business groups, and
interlocking ownership between corporations and banks is widespread. Such
arrangements have often resulted in corruption, collusive behaviour, specu
lation and inefficiency. Moreover, financial instability and short-termism in
bank and corporate behaviour are common features of these countries be
cause a number of conditions essential for an efficient bank-oriented sys
tem are not always met.
First of all, for the reasons already explained, price stability is essential
for a bank-oriented system. This calls for, above all, fiscal and monetary
discipline and a viable and relatively stable external payments position.
Prudential regulations and effective supervision are also essential in a bank-
oriented system. In particular, firms should not be allowed to own and

control banking organizations, since this will transfer the elements of the
safety net to them and burden the monetary authorities with tasks they
cannot undertake (Corrigan, 1991). In the German system prudential limits
on long-term lending and individual loans, capital adequacy requirements,
and effective supervision of banks’ risk exposure by an agency separate
from the central bank play a central role.
One argument against market-oriented finance is that "competition in
ownership is no substitute for competition in product markets" (Corbett
and Mayer, 1991, p. 20). This is also true for the bank-oriented system;
namely, it does not make up for lack of competition in the markets for
goods and services. Thus, such a system too needs to be combined with
policies depending on competition as a spur to efficiency. In those deve
loping countries where the bank-oriented system with widespread inter
locking ownership has failed, the markets for goods and services were ge
nerally highly oligopolistic and protected from competition. In contrast, in
countries where corporations were encouraged and forced to compete in
export markets, a similar financial organization made a major contribution
to industrialization.
Finally, new firms should have access to finance, and entry into new
lines of financial activity should not be impeded. This calls for some com
petition in the banking sector. However, competition policies should be de
signed to prevent monopoly power rather than to allow completely free
entry into the banking sector and unlimited price competition among banks
-practices that have often led to financial instability in both developed and
developing countries. Furthermore, specialized banks for industrial deve
lopment and controls over credit allocation can play an important role in
providing finance to new entrants.
4. Control and regulation o f stock-markets
While reform efforts need to concentrate in these areas, it is also true that
capital markets are a reality in a number of countries and they also need
to be improved. While most developing countries regulate primary issues
and stipulate a number of conditions regarding their size, maturity and
redemption and disclosure of information, there is very little effective con
48 Y ilmaz A kyüz
trol over secondary markets. Irregularities such as insider trading and fraud
are widespread and administrative capacity to undertake effective super
vision weak.
Stock prices in many of the so-called emerging markets have been
extremely erratic and subject to very large swings. By removing credit cons
traints, financial liberalization has often triggered an increase in speculative
activity by institutions and individuals. In many such countries, increased
speculative activity in the secondary market caused stock prices to rise be
fore 1987 even faster than in most of the world's major stock-markets, and
to fall, again far more than elsewhere, after October 1987 (Singh, 1992).
Most of these markets have again shown large swings over the last few
years.
Since the size of these markets is relatively small, the direct effects
of sharp falls in stock prices on the economy are negligible. However, the
state of expectations in the equities market influences the exchange rate
and capital flows since, as discussed in the next section, these markets
are open to foreigners an d/or provide alternative investment for holders
of foreign currency assets. Greater stability is thus essential to prevent
destabilizing feedbacks between equity and currency markets.
One way of reducing volatility is through the so-called "circuit bra-
kers" introduced in the United States after the October 1987 crash (Kupiec,
1991). These consist of predetermined price floors: when prices fall to the
floor, trading is suspended for a predetermined period. Such measures can
be particularly helpful in reducing intra-day bandwagon-type declines in
stock prices. Another is through the financial transactions tax long advocated
by Keynes (1936, pp. 160-161). Such a tax may help reduce speculative
instability by deterring short-term trading, improve the efficiency of the
stock-market and lengthen the time horizon of corporate managers (Stiglitz,
1989b; Summers and Summers, 1989).
Public or semi-public agencies with large holdings of securities can
also play an important role in bringing greater stability to stock prices.
For instance, in Turkey the agency dealing with privatization has operated
both as a buyer and a seller in the market for the shares of privatized
public companies, exerting a significant influence on prices, even though
its objective has not been to stabilize the market. Institutional investors
and particularly provident funds can both provide the Japanese/German
Financial uberauzation : the key issues 49
type of shareholding and control over enterprises, and help to attain grea
ter stability.
VII. EXTERNAL LIBERALIZATION AND FINANCIAL OPENNESS
Recent years have witnessed the increased integration of developing coun

tries into the international financial system in large part due to widespread
external financial liberalization. Most of these countries have also liberali
zed imports and increasingly relied on exports for growth, but the degree
of internationalization of finance has gone much further than trade. Indeed
in many countries the share of transactions with international characteris
tics in the financial sector is far greater than the share of trade in GDP.
1. The concept o f financial openness
By external financial liberalization we mean policy actions that increase the

degree of ease with which residents can acquire assets and liabilities deno
minated in foreign currencies and non-residents can operate in national fi
nancial markets, i.e. financial openness. Three broad types of transaction
can be distinguished in this respect. First, inward transactions: allowing
residents to borrow freely in international financial markets, and non-resi
dents to invest freely in domestic financial markets. Second, outward tran
sactions: allowing residents to transfer capital and to hold financial assets
abroad, and non-residents to issue liabilities and to borrow in domestic
financial markets. Third, domestic transactions in foreign currencies: allow
ing debtor-creditor relations among residents in foreign currencies such as
bank deposits and lending in foreign currencies.
Our definition of financial openness is wider than capital account
liberalization because it includes financial transactions among residents de
nominated in foreign currencies. These are an important part of banking
and finance, and affect the national economy in much the same way as
cross-border financial transactions (Bryant, 1987, chap. 3).
50 Y ilmaz A kyüz
2. The extent o f financial openness in developing countries
Widespread liberalization has occurred on all three fronts. Inward transac

tions are virtually free in a large number of countries, particularly in Latin
America where external borrowing by the private sector, often via the in
termediation of resident banks, is not subject to approval, except for ca
pital market issues. Similarly, there are few restrictions on the access of
non-resident investors to domestic capital markets. The debt crisis has pla
yed an important role in this respect: the "market-based menu" has gene
rated new prospects for arbitrage and windfall profits and significantly rai
sed the amount of equities and domestic-currency debt assets held by non
residents (UNCTAD, 1989, pp. 105-107). More recently, access of non-resi
dents to national equity markets has been encouraged in the context of
privatization programmes.
As for outward transactions, an increasing number o f developing coun
tries have adopted capital account convertibility in recent years -some to
an extent not found in most industrialized countries. Liberalization of
transactions among residents in foreign currency, however, has gone much
further. Indeed, there has been a tendency to encourage residents to hold
foreign exchange deposits with banks at home, increasing the importance
of foreign currency in the economy, i.e. dollarization. The share of foreign
currency in total deposits in recent years has reached 50 per cent in a
number of developing countries in Latin America as well as in Asia (e.g.
Philippines), the Middle East and Europe (e.g. Turkey and Yugoslavia). This
figure is well above the levels found in some international financial cen
ters such as London where the share of total bank claims (including inter
bank claims) on residents in foreign currencies barely exceeds 20 per cent
(Bryant, 1987, chap. 3; Akyüz, 1992).
3. Nature o f capital flows
The consequences of financial openness in developing countries have not

been adequately treated in the literature primarily because this is a very
recent phenomenon. Mainstream thinking is largely an extrapolation of
"open economy macroeconomics" and treats the issue in the context of
"sequencing of economic reforms". This literature emerged in large part from
F inancial uberauzation : the key issues 51
an ex post attempt to explain why the Southern Cone liberalization expe

riment failed (Corbo et al, 1986; Corbo and de Melo, 1987; Diaz-Alejandro,
1985; Dombusch, 1983; Frankel, 1983; McKinnon, 1982). It takes it for gran
ted that external financial liberalization is desirable on efficiency grounds:
it is said to have positive effects on the level and allocation of invest
ment, and these efficiency gains more than compensate for the loss of po
licy autonomy, i.e. reduced ability of governments to achieve national
objectives by using the policy instruments at their disposal (Bryant, 1980,
chap. 12).
According to this view, external financial liberalization may give rise
to perverse results only if there are problems elsewhere in the economy
-e.g. budget deficits, monetary instability, and distortions and imperfections
in goods and labour markets. On the other hand, since it is not possible
to correct these at once, external financial liberalization must be properly
sequenced. Although it is sometimes argued (e.g. Krueger, 1984) that it
may be difficult to control inflation without liberalizing the economy, the
majority view is that domestic financial markets and the current account
should be liberalized before the capital account, and that fiscal balance and
monetary stability should be attained before any liberalization (Dombusch,
1983; Edwards, 1984,1987 and 1989; Fischer and Reisen, 1992; Frankel, 1983;
McKinnon, 1982).
The benefits claimed for financial openness are generally based on
the assumption that the internationalization of finance allows savings to
be pooled and allocated globally through movement of capital across coun
tries in response to opportunities for real investment, thereby improving
the allocation of resources internationally and equalizing rates of return
on investment everywhere. Accordingly, external financial liberalization in
developing countries is expected to give rise to capital inflows, provided
that it comes after domestic financial markets have been liberalized and
interest rates raised. This is seen as a one-off phenomenon of adjustment
of domestic interest rates to world levels as capital scarcity is reduced through
an increase in the underlying capital flows.
However, the evidence strongly suggests that international capital
flows do not in practice improve the international allocation of savings.
There has been no narrowing of differences in rates of return on capital
investment in the major industrial countries, or in real long-term interest
52 Y ilmaz A kyüz
rates; nor has the link between the levels of savings and investment in
individual countries been considerably weakened (UNCTAD, 1987; Kasman
and Pigott, 1988; McCauley and Zimmer, 1989; Akyüz, 1992). The main
reason is that most international financial transactions are portfolio deci
sions, largely by rentiers, rather than business decisions by entrepreneurs.
The bulk of capital movements is motivated primarily by the prospect of
short-term capital gains, rather than by real investment opportunities and
considerations of long-term risk and return. The speculative element is ca
pable of generating gyrations in exchange rates and financial asset prices
by causing sudden reversals in capital flows for reasons unrelated to policies
an d /or the underlying fundamentals. Rather than penalizing inappropriate
policies, capital flows can help to sustain them, as has been the case in
recent years in the United States and Italy where inflows have helped to
run chronic fiscal deficits.
Thus, financial openness tends to create systemic problems regardless
of the order in which various markets are liberalized and distortions re
moved. The exposure to short-term, speculative capital flows is much grea
ter for developing than for developed countries because their instability
provides greater opportunities for quick, windfall profits on short-term capital
movements while their ability to influence capital flows through monetary
policy is much more limited.
While internal financial liberalization strengthens the link between in
flation and interest rates, external financial liberalization (unlike trade
liberalization) weakens that between inflation and the exchange rate, bringing
the latter under the domination of capital flows instead of trade balances
and the relative purchasing power of currencies: inflation differentials are
more readily reflected in nominal interest rate differentials than in the mo
vement of the nominal exchange rate. Thus, although short-term capital
inflows motivated by the lure of quick, windfall profits are often associated
with positive real interest rate differentials in favour of the recipient, such
a differential is neither necessary nor sufficient in all cases. Capital inflows
usually occur in response to a nominal interest rate differential that mar
kets do not expect to be fully matched by a nominal exchange rate depre
ciation. Such differentials often emerge when domestic inflation is much
higher than abroad and domestic financial markets have been liberalized.
Since in many developing countries inflation rates close to those prevailing
in the major OECD countries are very difficult to attain, the scope for big
arbitrage opportunities to emerge is much greater. Similarly, an expectation
that equity prices will rise faster than domestic currency depreciation can
prompt an inflow of capital. Both types of expectation can be self-fulfilling
since the inflow of funds, if large enough, can itself maintain the value
of the currency and boost equity prices.
Such inflows are typically initially a response to a favourable shift
in market sentiment regarding the recipient country. This shift may result
from external causes such as a sudden rise in export prices, or from internal
ones such as reduced inflation, better growth prospects, and greater politi
cal stability and confidence in the government's policies. After the initial
shift in market sentiment, a bandwagon develops and creates a speculative
bubble where people are lending or investing simply because everybody
else is doing so. The boom does not necessarily peter out smoothly: a re
cently liberalized, well-performing economy can suddenly find favour with
foreign capital of all sorts, but if things go wrong for some reason, the
capital can disappear just as rapidly. When the bubble bursts and the cu
rrency comes under pressure, even a very large positive real interest rate
differential may be unable to check the capital outflow.
4. Recent capital flows to Latin America
That was the story of the liberalization episodes in the Southern Cone in
Latin America in the 1970s, when high domestic interest rates, overvalued
exchange rates, freedom to borrow abroad and plentiful international li
quidity combined to induce capital inflows. But there are strong signs that
a similar process is again under way in a number of Latin American coun
tries. It is estimated that the region as a whole received about US$40 billion
in 1991, three times the level of 1990, the main recipients being Mexico,
Brazil, Argentina, Venezuela and Chile. Not all the capital inflows have
been for short-term uses, but much of them do appear to have been, particu
larly in Argentina and Brazil (Griffith-Jones et al, 1992, tables 4 and 5;
UNCTAD 1992, part two, Annex A). In the majority of these countries capital
inflows continued at an accelerated pace in 1992. In Chile where "the mone
tary authorities adopted a cautious approach based on the assumption that
the oversupply of foreign exchange was only temporary and was due to
54 Y ilmaz A kyüz
the unusually high price of copper and the low international interest ra
tes" (ECLAC 1991, p. 41), short-term capital inflows slowed down conside
rably in 1992 thanks to various measures designed to reduce the arbitrage
margin.
What is remarkable about recent capital inflows to Latin America is
not only that the recipient countries are in very different positions com
pared to the 1970s, but that they differ widely among themselves with
respect to inflation, fiscal posture, and exchange rate and trade policies.
Argentina, Chile and Mexico have liberal trade regimes whereas Brazil has
tight controls. While Brazil has had a large fiscal deficit and very high
inflation, others, particularly Chile and Mexico, have had balanced budgets
or fiscal surpluses, and moderate inflation. Capital has been attracted by
a combination of currency appreciation and high real interest rates in Chile,
Mexico and particularly Argentina (and also a booming stock-market in
the latter two), but not in Brazil where the underlying factor has been
very high real interest rates (about 4 to 5 per cent per month). Currency
appreciation is due to exchange rate policy in Argentina (which uses the
exchange rate as a nominal anchor to reduce inflation), but not in Chile
and Mexico where it is market-generated. It has led to a considerable dete
rioration of the trade balance, especially in Argentina and Mexico.
The ideal response to such capital inflows is a corresponding increase
in domestic investment in traded goods sectors. This not only prevents
a sharp appreciation of the currency by raising capital goods imports, but
can also enhance export capacity- something that may be needed especially
when capital flows dry up or are reversed. But higher investment is not
always possible when domestic interest rates are prohibitive and long-term
investment with funds borrowed abroad at lower rates carries considerable
exchange rate risk. In other words the high interest rates an d /or currency
appreciation that attract short-term capital also deter investment. In Latin
America, capital inflows resulted in a sharp swing in the transfer of resources
abroad by about 4 per cent of the region's GDP during 1990-1991, but
investment remained depressed: in Brazil and Argentina the investment ratio
remained below the levels of the 1980s when these countries had been
making large transfers on debt servicing.
The problems of macroeconomic management in the face of a massive
capital inflow are well known. Sterilizing them by issuing domestic debt
can impose a serious burden on the public sector, particularly when the
arbitrage margin is large. In Brazil, for instance, the cost of carrying the
extra US$5 billion of reserves purchased in this way amounted to about
US$2 billion during 1991-1992, adding considerably to domestic public debt
(UNCTAD, 1992, part two, Annex II; Junior, 1992). Furthermore, by incre
asing the stock of government debt, sterilization itself tends to raise domes
tic interest rates and, hence, the arbitrage margin. If, on the other hand,
the currency is allowed to appreciate, it can undermine the competitiveness
of the domestic industry, possibly eventually triggering a sharp reversal
in short-term capital flows.
5. Opening stock-markets to non-residents
Instability in short-term capital flows combined with the inherent volatility

of investment in company equity exposes the economy to even greater risks.
Since opening up domestic capital markets requires some form of currency
convertibility for non-resident equity investors, a close link can develop
between stock and currency markets even in countries where the capital
account is not fully open. This may prove to be a serious problem in Latin
America because of the increased presence of non-residents in capital mar
kets. In Mexico, for instance, equity holding by non-residents is estimated
to have amounted to more than $25 billion, or about a quarter of the mar
ket's capitalization in the second quarter of 1992 (Latin American Economy
and Business, May 1992, p.4), compared to about 5 per cent in the major
capital markets such as New York and Tokyo. The link between these two
inherently unstable markets can be further strengthened by dollarization
of the economy, when that occurs.
This link increases the potential for the emergence of foreign exchange
an d/or stock-market crises. Since the return on investment to the foreign
investor depends largely on the movement of the exchange rate, a serious
shock (e.g. a terms-of-trade deterioration) that makes a devaluation appear
inevitable can trigger both a sharp decline in equity prices and an outflow
of capital. Similarly, the m ood in equity markets can exert a strong in
fluence on the exchange rate -e.g. bullish expectations can trigger capital
inflow, leading to appreciation. By contrast, a bearish m ood in the capital
market an d/or massive profit-taking in dollars by non-residents can not
56 Y ilmaz A kyüz
only prick the speculative bubble in the stockmarket, but also lead to a
currency crisis. Recent evidence suggests that chaotic feedbacks between
financial and currency markets can easily develop: for instance, when the
bubble burst in the Tokyo stock exchange at the beginning of 1990, there
was a massive shift out of yen-denominated assets, causing also considera
ble drops in the government bond index and the currency (Akyüz, 1992).
6. Effects of volatile capital flows on investment and trade
One important consequence of sharp swings in the direction of capital flows

and greater instability of exchange rates is to increase borrower's risk. For
investors in traded goods sectors, the real exchange rate is the single most
important relative price affecting profits. But firms in non-traded goods sec
tors are also affected, depending on the imported inputs they use. Exchan
ge rate gyrations produce considerable uncertainty regarding prospective
yields of investment. By raising the average rate of return required by in
vestors to undertake investment, particularly in the traded goods sectors,
this will depress the level of investment corresponding to any given rate
of interest.
The influence of the exchange rate on investment decisions increases
with the share of foreign trade in the economy. It is thus of growing im
portance in the developing world because of widespread import liberali
zation and emphasis on export-led growth. It is therefore ironic that the
exchange rate is increasingly determined by purely financial forces delinked
from trade and investment. Exchange-rate instability can thus undermine
"outward oriented" strategies by depressing investment in exports. The evi
dence suggests that such adverse effects have occurred even in industrial
countries, where firms are better equipped to hedge against unexpected
swings in exchange rates, and that exchange rate stability has been characte
ristic of countries with sustained export growth (UNCTAD, 1987; UNCTAD,
1989, part one, chap. V).
The second systemic effect of volatile capital flows is through interest
rates. As already noted, capital-value uncertainty and interest rates both
rise as a result of increased borrower's risk as well as greater instability
in interest rates and prices of financial assets, including equities, associated
with volatile capital flows. More important, increased competition between
F inancial uberauzation : the key issues 57
domestic currency and foreign currency assets also tends to raise the cost
of finance because of the greater risk and uncertainty in developing coun
tries. The fact that most developing countries are economically and po
litically less stable than developed countries, with financial and legal sys
tems that are less able to ensure enforcement of contracts, increases the
hazards of financial investment. In a financially closed economy the safety
premium on foreign currency assets is counter-balanced by the high tran
saction costs of shifting into them, at least for most small savers, but fi
nancial openness reduces these costs considerably. Consequently, domestic
assets need to carry much higher rates of return than external assets. This
can reduce investment and impair competitiveness.
7. Controlling capital flows
Complete isolation of the financial system in a developing country from

the rest of world is neither feasible nor desirable. Successful export perfor
mance requires close interaction of banks at home with world financial
markets in order to provide trade-related credits and facilitate international
payments. The ability to borrow in international capital markets allows di
versification in corporate finance, while foreign investment in capital mar
kets can help broaden their equity base and reduce their leverage. Foreign
banks can bring greater competition in the provision of banking services,
thereby reducing the intermediation margin and the cost of finance.
Nevertheless, most developing countries need to exercise considerable
control over external capital flows in order to minimize their disruptive
effects and gain greater policy autonomy to attain growth and stability.
There are a number of techniques to control capital flows with different
degrees of restrictions and effects that were widely used in industrial coun
tries in the 1960s and 1970s (OECD, 1972, pp. 71-77; Fleming, 1973; Swi-
drowski, 1975; Swoboda, 1976; OECD, 1981; OECD, 1982). Quantitative mea
sures to limit short-term capital inflows through banks include reserve
requirements on foreign liabilities, limits on their net external or foreign-
currency positions, or on gross external or foreign currency liabilities, and
minimum holding periods and blocking of foreign deposits for such pe
riods. Similarly a number of measures may be applied to restrict external
borrowings by non-banks, including reserve requirements on their foreign
58 Y ilmaz A kyüz
liabilities, and exchange controls such as prohibition of borrowing other

than commercial or supplier credits received by importers, control on do
mestic foreign currency credits to domestic importers and exporters, and
regulations regarding the timing of export and import settlements. Of these,
limits on banks' net external or foreign currency positions and exchange
controls regarding non-banks can also be applied to restrict outflows. Res
trictions on interest payments on non-resident deposits and negative inte
rest rates are also among the measures that can be used to deter capital
inflows.
Taxes may also be used to reduce the arbitrage margin and discoura
ge speculative capital flows. A tax designed to reduce interest differentials
(like the interest-equalization tax used in the United States in the past to
check outflows) can also be especially effective in checking capital inflows
in developing countries where inflation and interest rate differentials with
developed countries tend to be large. The tax rate can be used flexibly
according to the behaviour of capital flows and the objective pursued. Si
milarly Keynes' proposal for a financial transactions tax may be extended to
apply to international financial transactions in order to "throw some sand in
the wheels" and "deter short-term financial round-trip excursions" (Tobin, 1978).
Finally, various restrictions may be introduced on the access of non
residents to capital markets. One common measure is to limit foreign owner
ship to approved country funds and allow transactions on such funds only
among non-residents in order to control the flow of foreign funds in and
out of the country via capital markets. This can be combined with the
requirement that such funds be managed by local managers who are gene
rally more amenable to "moral suasion" by the authorities.
It should be kept in mind that in several industrialized countries capital
markets have been opened to non-residents only very recently. In Japan,
for instance, they were largely closed until the 1984 agreement with the
United States, and even in Europe, where an integrated financial market
is seen as an important step in the completion of a single EEC market,
restrictions on entry into capital markets still remain in a number of countries
(e.g. France and Italy). Again, the Republic of Korea only recently opened
up its capital market to non-residents, but restricted foreign acquisition to
10 per cent of total equity capital, and to 2 per cent in some strategic
industries.
Some of these techniques have recently been used in Latin America

in order to slow down short-term capital inflows. These include reserve
requirements for foreign currency liabilities (Chile and Mexico), compul
sory liquidity requirements on the short-term forex liabilities of commercial
banks (Mexico), minimum holding periods (Chile), extension of the fiscal
stamp tax to foreign credits (Chile), restrictions on company borrowing
abroad through stock and bond issues (Brazil), and limits on the dollar
amounts that banks can raise in deposits abroad as a proportion of their
total deposits (Mexico). However, such measures have generally had only
limited success. Governments are often very shy in applying effective con
trols for fear of fending off genuine, long-term capital and investment. This
is certainly a legitimate concern, particularly in Latin America, after a de
cade-long foreign exchange strangulation. However, experience shows that
capital controls might have to be introduced anyway if the process develops
into a payments crisis and capital flight. It may be easier to restrict short
term inflows and prevent debt accumulation early on than to check capital
flight in a crisis.
Controls on capital flows are not always effective when there are lar
ge arbitrage opportunities. It is thus important to bear in mind that price
stability is vital for a financially open economy, since high inflation and
wide interest rate differentials with reserve-currency areas often lead to large
arbitrage opportunities and encourage unsustainable capital flows. Further
more, exchange rate management plays an important role. Explicit or im
plicit exchange rate guarantees tend to reduce the risk involved in arbitrage
and encourage capital flows. As noted above, this has been an important
factor in attracting short-term capital to Argentina. In Chile, by contrast,
"the monetary authorities moved to resist revaluation of the peso by intro
ducing changes to create uncertainty concerning yields on short-term capital
flows" (ECLAC, 1992, p. 40). These measures included the ending of the
practice of advance announcement of devaluation of the peso, widening
of the currency band, and linking the peso to a basket of currencies instead
of the US dollar. They appear to have played an important role in slowing
down short-term capital inflows and securing greater real exchange rate
stability by introducing uncertainty regarding the movement of the exchan
ge rate. In Mexico too the authorities widened the differential points for
the peso-US dollar exchange rate to allow larger fluctuations, although its
60 Y ilmaz A kyüz
effects on capital flows seem to have been limited (Banco de Mexico, 1992,
p. 144).
Historical experience clearly shows that capital controls are no ans
wer when the underlying policies are not sustainable. For instance, mea
sures to control capital inflows are generally ineffective against capital flight
stemming from economic and political instability. Moreover, it is important
to bear in mind that capital controls are needed not in order to pursue
inappropriate policies and exchange rates, but to minimize the disruptive
effects of short-term capital flows and gain greater policy autonomy to attain
growth and stability.
VIII. CONCLUSIONS
The focus of financial policies in developing countries should be industria

lization and stability. A common feature of all modem examples of indus
trialization is that they have all succeeded in maldng finance serve industry
and trade, not the other way round. This has often necessitated a considera
ble amount of intervention and control over financial activities. On the other
hand, despite widespread claims for efficiency of financial markets, financial
liberalization in many countries in recent years has generated more costs
than benefits. These have included persistent misalignment of prices of fi
nancial assets, resulting in inefficiencies in the allocation of resources; sharply
increased short-term volatility of asset prices, resulting in greater uncer
tainty, shorter maturities and higher interest rates; excessive borrowing to
finance speculative asset purchases and consumption, resulting in unsus
tainable stocks of debt, increased financial fragility and reduced household
savings; and loss of autonomy in pursuing interest-rate and exchange-rate
policies in accordance with the needs of trade and industry.
It is equally true that government intervention in finance has often
been misguided, giving rise to inefficiency and waste. However, the appro
priate response should be to reform the State and rationalize intervention
rather than throw in the towel and simply "unleash market forces". The
main challenge is to determine where and how governments should inter
vene and to make sure that the intervention achieves its aims. The dis
cussions so far suggest the following:
• Macroeconomic stability is of cardinal importance for the stability

and efficiency of the financial system, since excessive volatility of prices
and economic activity tends to increase financial fragility, create uncertain
ty, raise interest rates and shorten the time horizon. While macroeconomic
stability itself is influenced by financial policies, monetary and fiscal dis
cipline is crucial.
• In cases where directed credits and financial subsidies are success
fully used as part of industrial policy, winners are not picked by "bureau
crats", but through a process based on a close interaction between the go
vernment and the business and the use of market signals to assess risks
and opportunities. Success also depends on ensuring reciprocity between
support and performance; use of controls, regulations and subsidies for the
intended purposes; and readiness to revise them as necessary.
• Financial policies must take account of the dual nature of interest
rates: the return aspect, which primarily influences the distribution of asset
holdings in different forms, and the cost aspect, which determines the ca
pacity of the corporate sector to generate internal funds, to undertake in
vestment and to compete in world markets. It is important to bear in mind
that while high interest rates are not necessary to increase savings, low
and stable capital cost is crucially important for investment and compe
titiveness.
• There is often a need for deposit ceilings and intervention in the
money market in order to stabilize interest rates and asset prices and pre
vent excessive risk-taking and price competition in the financial sector. Such
controls should be applied with flexibility and discretion, taking into account
macroeconomic conditions as well as the needs of trade and investment.
Rigid rules regarding the level of real interest rates are no more sensible
than those about the rate of growth of money supply in the conduct of
monetary policy.
• Prudential regulations and strong bank supervision are also essential
to prevent excessive risk-taking and financing of speculative activities by
banks. Measures such as capital requirements are not always enough to
reduce fragility: it may also be necessary to act directly on the asset port
folios of banks and restrict lending against or investment in highly capital-
uncertain assets such as securities and property, and exposure to a single
firm. Firms should not be allowed to own and control banks. Protective
62 Y ilmaz A kyüz
regulations such as deposit insurance and lender-of-last-resort facilities

should only be introduced in combination with prudential regulations.
• Most developing countries need to concentrate their energies on
strengthening their existing bank-based financial systems rather than pin
their hope on transplanting Wall Street. They also need to promote long
term equity holding via institutional investors such as provident funds, and
permit banks to hold equities within prudential limits. Transfer taxes and
"circuit breakers" may be used to deter short-term trading and reduce vola
tility in stock markets. Easy access to stock-markets and readily available
short-term financial instruments paying market returns tend to increase fi
nancial instability.
• Particular care needs to be given to the design of external financial
policies since mistakes in this area tend to be very costly and difficult
to reverse. Allowing residents uncontrolled access to international capital
markets has proved damaging in many developing countries, and short
term speculative capital flows have proved extremely troublesome. Deve
loping countries need to exercise a considerable degree of control over ex
ternal capital flows through taxes, quantitative restrictions and exchange
controls in order to minimize their adverse effects on macroeconomic equi
librium, exchange rates and trade; to control the pace of accumulation of
external debt; and to gain greater autonomy in monetary policy. Access
of non-residents to domestic capital markets should be restricted since close
links between the two inherently volatile markets can be very dangerous.
It is also important to resist the temptation to dollarize the economy in
order to keep capital at home: policies should address the root cause of
the problem and eliminate the reasons for extensive demand for foreign
currency.
• A pragmatic not a doctrinaire approach is needed towards financial
control and liberalization in developing countries. Restrictions on financial
flows and interest rates may be removed over time as they fulfil their
functions. Financial liberalization undertaken as a result of a successfully
implemented industrial policy is very different from liberalization as a reac
tion to misguided and failed intervention. Has financial liberalization ever
remedied stagnation and instability?
REFERENCES
AKYUZ, Y. (1990), "Financial System and Policies in Turkey in the 1980s",

in T. Aricanli and D. Rodrik (eds.), The Political Economy of Turkey.
Debt, Adjustment and Sustainability, London, Macmillan.
AKYUZ, Y. (1991), "Financial Liberalization in Developing Countries: A Neo-
Keynesian Approach", UNCTAD Discussion Paper, N2 36, March.
AKYUZ, Y. (1992), "Financial Globalization and Instability", in Uner Kirdar
(ed.), Change: Threat or Opportunity?, Vol. 3, Market Change, UNDP,
New York.
AKYUZ, Y. (1993), "On Financial Openness in Developing Countries", in
International Monetary and Financial Issues for the 1990s, Volume
H, UNCTAD, Geneva.
AKYUZ, Y. (forthcoming), "Does Financial Liberalization Improve Trade
Performance?", in D. Tussie and M. Agosin (eds.), Trade and Growth:
New Dilemmas in Trade Policy
AKYUZ, Y. (forthcoming), "Financial Policies for Industrialization and
Development: Reflections on Financial 'Deepening' and 'Efficiency'", in
M. Nissanke and A. Hewitt (eds.), Economic Crisis in Developing
Countries: New Perspective on Commodities, Trade and Finance.
AMSDEN, A. (1989), Asia's Next Giant: South Korea and late Industrializa
tion, OUP, New York.
AMSDEN, A. and EUH, Yoon-Dae (1990), "Republic of Korea's Financial
Reform: What are the Lessons?", UNCTAD Discussion Paper, N2 30,
April.
ARYEETEY, E., ASANTE, Y., GOCKEL, F. and KYEI, A. (1990), 'Mobilizing
domestic savings for African development and industrialization: A
Ghanaian case study", mimeo, Queen Elizabeth House, Oxford.
BANCO DE MEXICO (1992), The Mexican Economy 1992, Mexico City.
BLUNDELL-WIGNALL, A. and BROWNE, F. (1991), "Macroeconomic
Consequences of Financial Liberalization: A summary Report", Working
Paper, N2 98, OECD Dept, of Economics and Statistics, February.
BRADFORD, C.I. (1986), "East Asian 'Models': Myths and Lessons", in J.P.
Lewis and V. Kallab (eds.), Development Strategies Reconsidered, Over
seas Development Council, N.J.
64 Y ilmaz A kyüz
BRYANT, R. (1980), Money and Monetary Policy in Interdependent Nations,

The Brookings Institution, Washington, D.C.
BRYANT, R. (1987), International Financial Intermediation, The Brookings Ins
titution, Washington, D.C.
CHIPETA, C. (1990), "Mobilizing domestic savings for African development
and industrialization: A case study of Malawi", mimeo, Queen Eliza
beth House, Oxford.
CHO, Yoon-Je (1988), "The Effects of Financial Liberalization on the Efficiency
of Credit Allocation: Some Evidence from Korea", Journal of Develop
ment Economics, Vol. 29.
CHO, Yoon-Je and KHATKHATE, D. (1989), "Lessons of Financial Libera
lization in Asia: A Comparative Study", World Bank Discussion Paper,
Na 50.
CORBETT, J. and MAYER, C.P. (1991), "Financial Reform in Eastern Europe:
Progress with the Wrong Model", CEPR Discussion Paper, N2 603, Sep
tember.
CORBO, V. and de MELO, J. (1987), "Lessons from the Southern Cone Policy
Reforms", The World Bank Research Observer, Vol. 2, N 2 2.
CORBO, V., de MELO, J. and TYBOUT, J. (1986), "What Went Wrong with
the Recent Reforms in the Southern Cone?", Economic Development and
Cultural Change, Vol. 34, N 2 3.
CORRIGAN, E.G. (1991), "The Banking-Commerce Controversy Revisited",
Federal Reserve Bank of New York Quarterly Bulletin, Spring.
CUTLER, D.M., POTERBA, J.M. and SUMMERS, L.H. (1990), "Speculative
Dynamics", NBER Working Paper, N2 3242, January.
DATTA-CHAUDHURI, M. (1990), "Market Failure and Government Failu
re", Economic Perspectives, Summer.
DAVIS, E.P. (1993), Debt, Financial Fragility and Systemic Risk Clarendon
Press, Oxford.
DIAZ-ALEJANDRO, C. (1985), "Good-Bye Financial Repression, Hello Fi
nancial Crash", Journal of Development Economics, Vol. 19(1-2).
DORNBUSCH, R. (1983), "Panel Discussion on the Southern Cone", IMF
Staff Papers, Vol. 30, N 2 1.
DORNBUSCH, R and REYNOSO, A. (1989), "Financial Factors in Economic
Development", American Economic Review - Papers and Proceedings, May.
DRISCOLL, M. (1991), "Deregulation, Credit Rationing, Financial Fragility
and Economic Performance", Working Paper, N B 98, OECD Dept, of

Economics and Statistics, February.
ECLAC (1991-1992), Economic Panorama of Latin America, United Nations,
Santiago.
EDWARDS, S. (1984), ’The Order of Liberalization of the Balance of Pay
ments", World Bank Staff Working Paper, Ns 710.
EDWARDS, S. (1987), "Sequencing Economic Liberalization in Developing
Countries", Finance and Development, March.
EDWARDS, S. (1989), "On the Sequencing of Structural Reforms", OECD
Working Paper, N2 70, September.
FAMA, E.F. (1985), "What's Different About Banks?", Journal of Monetary
Economics, Vol. 15, January.
FISCHER, B. and REISEN, H. (1992), Towards Capital Account Convertibility,
Policy Brief N2 4, OECD Development Centre.
FLEMING, J.M. (1973), "Problems of Disruptive International Capital M o
vements", in A.K. Swoboda (ed.), Europe and the Evolution of the In
ternational Monetary System, Sijthoff Leiden, Institut Universitaire de
Hautes Etudes Internationales, Geneva.
FRANKEL, J.A. (1983), "Panel Discussion on the Southern Cone", IMF Staff
Papers, Vol. 30, N 2 1.
GREENWALD, B. and STIGLITZ, J.E. (1986), "Externalities in Economies
with Imperfect Information and Incomplete Markets", Quarterly Jour
nal of Economics, 101(2).
GRIFFITH-JONES, S., MARR, A. and RODRIGUEZ, A. (1992), "The Return
of Private Capital to Latin America: The Facts, an Analytical Frame
work, and some Policy Issues", mimeo, June.
HANHANSON, J.A. and NEAL, C.R. (1985), "Interest Rate Policies in Selec
ted Developing Countries, 1970-1982", World Bank Staff Working Papers,
NB 753.
HONOHAN, P. and ATIYAS, I. (1989),"Intersectoral Financial Flows in De
veloping Countries", World Bank Working Paper, N2 164, March.
JONES, D. and KING, K.K. (1992), "Implementation of Prompt Collective
Action", paper presented to the meeting of technicians of Central Banks
of the American Continent, Barbados, 16-20 November.
JUNIOR, P.A. (1992), "Política Monetária E Ingresso de Divisas: O Caso
Brasileiro Recente", paper presented to the meeting of the technicians
66 Y il m a z A k y ü z
of Central Banks of the American Continent, Barbados, 16-20 Novem

ber.
KASMAN, B. and PIGOTT, C. (1988), "Interest rate Divergence among the
Major Industrial Nations", Federal Reserve Bank of New York Quarterly
Bulletin, Autumn.
KATO, K. et a l (1993), "Policy-Based Finance: The Experience of Postwar
Japan", Final Report to the World Bank, The Japan Development Bank,
January.
KEYNES, J.M. (1936), The General Theory of Employment, Interest and Money,
London, Macmillan.
KREGEL, J.A. (1991), "Markets and Institutions in the Financing of Busi
ness", paper prepared for the Jerome Levy Institute Conference: Res
tructuring the Financial System for Economic Growth, 21-23 Novem
ber, University of Bologna.
KREGEL, J.A. (1993), "Financial Fragility and the Structure of Financial
Markets", paper presented at the Allied Social Science Associations
Meetings, Anaheim, California, 7 January.
KRUEGER, A.O. (1984), "Problems of Liberalization", in A. Harberger (ed.),
World Economic Growth, ISC Press, San Fransisco.
KUPIEC, P. (1991), "Financial Liberalisation and International Trends in
Stock, Corporate Bond and Foreign Exchange Market Volatilities", Working
Paper, N2 94, OECD Dept, of Economics and Statistics, February.
LEE, CHUNG, H. (1992), "The Government, Financial System, and Large
Private Enterprises in the Economic Development of South Korea",
World Development, Vol. 20, N2 2.
LIM, J.Y. (1991), "The Philippine Financial Sector in the 1980s", UNCTAD
Discussion Paper, N2 35, March.
MASSAD C. and EYZAGUIRRE, N. (1990), Ahorro y formación de capital:
Experiencias latinamericanos, CEPAL, Santiago.
MARTIN, P. (1992), "Not the time to gloat over Japan's shocks", Financial
Times, 26 May.
McCAULEY, R.N. and ZIMMER, S.A. (1989), "Explaining International Di
fferences in the Cost of Capital", Federal Reserve Bank of New York Quar
terly Bulletin, Summer.
McKINNON, R.I. (1973), Money and Capital in Economic Development, Wash
ington, D.C., The Brookings Institution.
F in a n c ia l l ib e r a l iz a t io n : t h e k e y issu e s 67
McKINNON, R.I. (1982), "The Order of Economic Liberalization: Lessons

from Chile and Argentina", in K. Brunner and A. Meltzer (eds), Econo
mic Polio/ in a World of Change, Netherlands.
MILLER, M. and WELLER, P. (1991), "Financial Liberalisation, Asset Prices
and Exchange Rates", Working Paper, N2 94, OECD Dept, of Economics
and Statistics, February.
MINSKY, H. (1982), Inflation, Recession and Economic Policy, Wheatsheaf, Ar-
monk, New York.
MINSKY, H. (1986), Stabilizing an Unstable Economy, Yale University Press,
New Haven.
MOLHO, L.E. (1986), "Interest Rates, Saving, and Investment in Developing
Countries", IMF Staff Papers, Vol. 33, N2 1, March.
MWEGA, F.M. (1990), "Mobilizing domestic savings for African development
and industrialization: A case study of Kenya", mimeo, Queen Elizabeth
House, Oxford.
NISSANKE, M. (1990), 'Mobilizing domestic resources for African develop
ment and diversification: Structural impediments in the formal finan
cial system", mimeo, Queen Elizabeth House, Oxford.
OECD (1972), Economic Outlook, N2 12, December.
OECD (1981), Controls on International Capital Movements: Experience with
Controls on International Portfolio Operations in Shares and Bonds, Paris.
OECD (1982), Controls on International Capital Movements: The Experience with
Controls on International Financial Credits, Loans and Deposits, Paris.
OWEN, P.D. and SOLIS-FALLAS, 0 . (1989), "Unorganized Money Markets
and 'Unproductive' Assets in the New Structuralist Critique of Finan
cial Liberalization", Journal of Development Economics, Vol. 31.
POTERBA, J.M. (1991), "Comparing the Cost of Capital in the United States
and Japan: A Survey of Methods", Federal Reserve Bank of New York
Quarterly Review, Winter.
SHAW, E.S. (1973), Financial Deepening in Economic Development, New York,
Oxford University Press.
SINGH, A. (1992), "The Stock Market and Economic Development: Should
Developing Countries Encourage Stock Markets?", UNCTAD Discussion
Paper, NQ 49, October.
SOMEL, C. (1992), "Finance for Growth: Lessons From Japan", UNCTAD
Discussion Paper, N2 44, July.
68 Y il m a z A kyüz
STIGLUZ, J.E. and WEISS, A. (1981), "Credit Rationing in Markets with

Imperfect Information", American Economic Review, June.
STIGLUZ, J.E. (1989a), "Markets, Market Failures, and Development", Ame
rican Economic Review, Papers and Proceedings, May.
STIGLUZ, J.E. (1989b), "Using Tax Policy To Curb Speculative Short-Term
Trading", Journal of Financial Services Research, Vol. 3, N 2 3, Decem
ber.
SUMMERS, L.H. and SUMMERS, V.P. (1989), "When Financial Markets
Work Too Well: A Cautious Case For a Securities Transactions Tax",
Journal of Financial Services Research, Vol. 3, NQ 3, December.
SWIDROWSKI, J. (1975), Exchange and Trade Controls, Gower Press, Essex.
SWOBODA, A.K. (ed.) (1976), Capital Movements and their Control, Sijthoff
Leiden, Institut Universitaire de Hautes Etudes Internationales, Ge
neva.
TOBIN, J. (1978), "A Proposal for International Monetary Reform", The
Eastern Economic Journal, July/October.
TOBIN, J. (1984), "On the Efficiency of the Financial System", Lloyds Bank
Review, July.
UNCTAD (1987), "The Exchange Rate System", in International Monetary and
Financial Issues for the Developing Countries, United Nations, Geneva.
UNCTAD (1989-1992), Trade and Development Report, United Nations, Ge
neva.
WESTPHAL, L.E. (1990), "Industrial Policy in an Export-Propelled Econo
my: Lessons from South Korea's Experience", Economic Perspectives,
Summer.
WIJNBERGEN, S. van (1983), "Interest Rate Management in LDCs", Journal
of Monetary Economics, Vol. 12, September.
WORLD BANK (1989), World Development Report, Washington, D.C.
ZIMMER, S.A. and McCAULEY, R.N. (1991), "Bank Cost of Capital and
International Competitiveness", Federal Reserve bank of New York Quar
terly Bulletin, Winter.
EAST ASIAN FINANCIAL MARKETS:
WHY SO MUCH (AND FAIRLY EFFECTIVE)
GOVERNMENT INTERVENTION?
Alice H. Amsden
CONTENTS
INTRODUCTION .......................................................................................................................... 73
I. WHY SO MUCH GOVERNMENT INTERVENTION? ................................................. 74
II. REPRESSION IN FINANCIAL MARKETS ..................................................................... 76
III. DISCIPLINE OF BUSINESS ............................................................................................... 78
IV. THE GOVERNMENTS ABILITY TO IMPOSE

PERFORMANCE STA N D A R D S...................................................................................... 79
1. The state autonom y ..................................................................................................... 80
2. Financial system characteristics: The case of the Republic of Korea ................... 81
V. KEEPING THE COSTS OF INVESTMENT FINANCE LOW:

THE CASE OF THE REPUBLIC OF KOREA'S FINANCIAL REFORM .................... 83
1. Interest R a te s ................................................................................................................. 83
2. Developing the stock m arket ..................................................................................... 84
3. Interest rates and savings ............................................................................................ 86
VI. CONCLUSIONS.................................................................................................................... 88
ANNEX OF T A B L ES...................................................................................................................... 89
REFERENCES 99
E a st A s ia n f in a n c ia l m a r r é is : w h y s o m u c h (and f a ir l y e f f e c t iv e ). 73
INTRODUCTION
Because the economies of the Republic of Korea and Taiwan, Province of

China, have grown so fast, their financial systems, and the reforms they
underwent in the 1980s, have been objects of curiosity in other developing
countries. This is especially the case since the financial systems of both
countries have exhibited high degrees of repression. Although both systems
were reformed in the 1980s they still deviate substantially from the free
market model.
This paper has four purposes. The first is to explain why govern
m ent intervention in financial markets has been so extensive, not just in
the Republic of Korea and Taiwan, Province of China, but in the general
case of countries that have industrialized "late". Late industrialization is
defined here as a process of industrial development in countries without
the competitive asset of pioneering technology, which was the hallmark
of the First and Second Industrial Revolutions. The second purpose is to
analyse why government intervention in financial markets in the Republic
of Korea and Taiwan, Province of China, has worked relatively well. The
third purpose is to review briefly the major characteristics of the financial
system of the Republic of Korea. Finally, the paper makes some observa
tions about Korean financial reform in the 1980s.
74 A l i c e H. A m sd e n
I. WHY SO MUCH GOVERNMENT INTERVENTION?1
The basic Heckscher-Ohlin trade model assumes identical production func

tions in the same industry in all countries, but in reality more advanced
economies may be more cost effective, even in industries with a high la
bour content, due to superior infrastructure in their operating setting, better
management and workforce skills, as well as their store of tadt, nontransfe-
rable productivity and quality improvements. At least in the short run,
the standard form of technology transfer -designs, blueprints and production
equipment- or even a turnkey transfer, are usually insuffident to overcome
the productivity gap. Under these conditions, relatively low wages do not
translate into relatively low unit labour costs, and do not serve as an entrée
into world markets.
Exchange rate devaluations may lower real wage costs in international
markets. But currency depredations are usually constrained by political and
social conditions, workers' physiological intake requirements, and the need
to import production inputs, including wage goods. Even after real currency
devaluations in the 1960s in the Republic of Korea and Taiwan, Province
of China, low wages were found to be an insufficient competitive advanta
ge against Japanese competition notwithstanding the labour intensity of a
leading sector like cotton textiles (see, for example, Amsden, 1989; Clark,
1987). Latin America in the 1980s demonstrated that it was impossible to
lower real wages without triggering a wage response that led to serious
price instability and general macroeconomic disorder in the form of volatile
nominal and real interest rates, unpredictable foreign exchange rates, variable
and large fiscal defidts, and stop-go growth (UNCTAD, 1989, chap. 4; Taylor,
1988).
In contrast with conventional arguments favouring specialization in
labour-intensive industries, the doyen of institutional theories of late indus
trialization, Alexander Gerschenkron, views catching up as a process of 're
volutionary', 'eruptive' spurts, with backward countries promoting "those
branches of industrial activities in which recent technological progress has
been particularly rapid" (1962: 9-10). Leading American and German enter
1 Sections 2-5 (a) rely heavily on Hikino and Amsden (1993).

E a s t A s ia n f in a n c ia l m a r k e t s : w h y s o m u c h (and f a ir l y e f f e c t iv e ) . 75
prises could and did leap-frog ahead of Britain in the most dynamic sectors
such as chemicals and steel because British firms could not establish impe
netrable international entry barriers in the nineteenth century.
By the twentieth century this strategy had become impractical for two
reasons. First, over time the gap between the most and least advanced
countries (or even the average developed and developing country) had grown
far greater. The distance to the world economic frontier, measured as the
ratio of GDP per capita of the most developed countries to the appropriate
comparison, ranged from only 1.8 to 3.3 for the average backward European
country at the end of the nineteenth century compared with 11.9 for the
average developing country after the Second World War and 25.7 for the
typical least developed country in the 1970s (Hikino and Amsden, 1993).
Second, with the rise of global enterprises possessing "organizational
capabilities" based on a core technology (Chandler, 1990), Gerschenkron's
idea of leaping to the world technological frontier was no longer feasible.
The institutionalization of R&D in such enterprises allowed them to erect
entry barriers around their proprietary technology, which kept newcomers
out. The only economy in the twentieth century to attempt to leap-frog
to the world technological frontier ended in failure, namely the former
USSR, which was Gerschenkron's primary concern.
Gerschenkron conceived domestic enterprises in Europe as the agents
of industrialization, but increasingly after the Second World War, the
multinational firm came to be viewed in certain development theories as
the agent capable of transferring technology to backward countries. By rais
ing productivity there, the multinational supposedly precluded the need
for government intervention. Whether in the role of exporter of labour-
intensive manufactures (as in Taiwan, Province of China, and Puerto Rico),
or developer of import substitution industries (as in Mexico and Brazil),
the multinational firm was credited with nudging backward countries closer
to world productivity standards.
Nevertheless, casting the multinational firm at the heart of the deve
lopment drama has proved to be problematic in practice. Few developing
countries after the Second World War have managed individually to attract
significant amounts of foreign investment. Even in those Asian countries
which the multinationals have patronized, investments from abroad have
been found to amount to only a small fraction of aggregate capital require
76 A uce H. A m sd e n
ments (Amsden, 1992). Foreign capital typically lags rather than leads indus
trial development and tends to flow to backward countries where industria
lization has already started, and only then accelerates it (Herman, 1991).
Moreover, simply because a foreign investor is a multinational firm no longer
necessarily implies that it operates at the world frontier. The production
and design problems that afflict it at home may merely be transferred abroad
-as in the case of the South American operations of some North American
automobile companies (Shapiro, forthcoming). Thus, the multinational firm
cannot be counted on automatically to overcome the diseconomies of back
wardness mentioned above.
In sum, conventional development theory recommends that latecomers
industrialize by "getting the prices right" (allowing supply and demand to
determine prices), and typically by using low wages to gain a comparative
advantage in labour-intensive industries. But, in fact, successful late
industrialization has been a process of using subsidies to lower production
costs, such as capital, to get the prices "wrong" (preventing market forces
from determining prices, such as interest rates) in order to overcome the
handicap of an absence of proprietary technology, the inability to leap
frog over more advanced countries, and the inadequacy of a low wage
advantage (Amsden, 1989; 1992). Therefore, a necessary condition for
industrializing in the twentieth century is systematic and well-coordinated
government intervention to promote manufacturing investment. As sugges
ted below, what seems to distinguish developing countries that are "sneak
ing up" to the world technological frontier from those that are "stumbling
back" (such as the Philippines and Argentina) or "staying behind" (say,
Bangladesh and Pakistan) is not less state intervention but rather a different
set of principles governing subsidy allocation.
II. REPRESSION IN FINANCIAL MARKETS
Even in the classic economic liberalism of the First Industrial Revolution

in England, the Government played a positive and significant role (Coats,
1971; Taylor, 1972). Starting from the late eighteenth century, governments
were actively involved in developing transportation, communications, and
all sorts of education as well as stable banking systems and legal and
E a s t A s ia n f in a n c ia l m a r k e t s : w h y s o m u c h ( and f a ir l y e f f e c t iv e ) . . 77
administrative frameworks generally. Over time, more direct microecono

mic intervention that affected price competition increased everywhere to
the extent that tariff protection to infant and other industries became wides
pread for a host of reasons related not just to industrial development but
also to revenue and politics (Nye, 1991). Ideology aside, laissez-faire simply
never existed (lively, 1955; Goodrich, 1967; Hughes, 1991). As Maddison
has observed: "Before 1913, government intervention had a similar flavour
almost everywhere" (1989: 108).
Nevertheless, given a lack of competitive assets, late-industrializing
states in the twentieth century did all this and much more, examples being
Brazil, Turkey, India, the Republic of Korea, Taiwan, Province of China,
and Japan, but because Japan was relatively less developed, it could do
less than later industrializers. A case in point relates to financial markets,
where state intervention went much further than even Alexander Gerschen-
kron envisioned. In the absence of highly developed banking institutions,
Gerschenkron recognized the need for governments to arrange finance for
manufacturing investments. But late-industrializing governments have not
only made finance available, they have also targeted capital to specific firms
as well as selected industries on highly concessionary terms.
In Taiwan, Province of China, for example, a market-determined inte
rest rate could be said to have been approximated in the 1960s, 1970s and
1980s by the "curb market" interest rate. The curb rate was not determined
in perfect competition because there were large, wholesale lenders, but it
was still quite competitively established. Below the equilibrium curb market
rate was the rate set by the government-owned commercial banks. These
banks were habitually getting the interest rate "wrong", as evidenced by
the fact that the commercial bank rate was consistently below the curb
rate, often for one and the same borrower. Moreover, the nominal interest
rate paid by the big firms that received commercial bank credit was higher
than their effective interest rate because they on-lent to smaller firms at
higher prices (Biggs, 1988).
After the Republic of Korea's financial "liberalization" of the 1980s,
the "right" price of capital may be said to have been approximated by
the interest rate in the secondary short-term government bond market. The
misnomer of "liberalization" is indicated by the fact that, whereas in May
1989 the interest rate in this market was 18.9 per cent, the loan interest
78 A u c e H . A m sd e n
rate of government-controlled commercial banks was only 12.5 per cent

(Amsden and Euh, 1993). Obviously, commercial bank credit was still being
subsidized even after "liberalization". The Republic of Korea had a three-
tier financial structure for the first 25 years of its development. It was
characterized by a curb market interest rate, a commercial bank rate, and
a rate on foreign loans. Due to inflation and the relative constancy of the
exchange rate, the real interest rate on foreign loans was negative through
out most of this period (Park, 1985). Not all three different prices that exis
ted side by side in the Republic of Korea's capital market could have been
right, and the negative real interest rate on foreign loans was fundamentally
"wrong" in a capital-scarce country.
Even Thailand, with its reputation for economic liberalism, had positive
real interest rates for only 24 out of .52 quarters between 1970 and 1982.
A World Bank study called this performance "quite respectable in compari
son with most developing countries" (Hanson and Neal, 1984: annex 6,3).
III. DISCIPLINE OF BUSINESS
In comparing the behaviour of more and less economically- successful la

te industrializers that are "sneaking up" closer to the world frontier with
those that are "stumbling back" or just "staying behind", an important di
fference between them lies in their management of the subsidy allocation
process. Slow-growing late industrializers have tended to blanket business
with subsidies without being willing or able to extract concrete performan
ce standards in exchange. By contrast, fast-growing late industrializers have
generally succeeded in disciplining subsidy recipients, imposing strict and
monitorable performance standards on them (Amsden, 1989; 1991a; 1992).
In imposing performance standards on business, the government also sub
jected itself to evaluation by objective criteria.
In the case of Taiwan, Province of China, subsidies to exporters in
the 1960s were tied to targets administered by industry associations that
were overseen by government agencies. These associations acted as cartels.
They collected dues from members out of which bonuses to exporters were
paid. Firms were allocated export targets and penalized if they fell short
of their targets (Haggard, 1990; Wade, 1990). Loan officers of Taiwan, Pro-
E a s t A s ia n f in a n c ia l m a r k e t s : w h y s o m u c h ( and f a ir l y e f f b c t iv e ). 79
vince of China's State-owned banks were also held personally responsible

(in terms of pay and promotion) for the credit they allocated. Consequen
tly, they were both conservative in their lending policies -lending only to
relatively large firms- and careful in their monitoring of how effectively
borrowers used their credit (Biggs, 1988).
While "infant industry protection", as conceived in the nineteenth cen
tury, was in theory a one-shot deal, designed to enable a new enterprise
to reach a minimum efficient scale of operation, subsidization in late
industrialization has in practice been multi-stage. It has operated not just
at the start but also at later points in an economy's catch-up trajectory.
For instance, the Taiwan, Province of China's machine tool industry re
ceived little government support in its early growth phase, but was sub
sidized subsequently to help it acquire financially troubled United States
machine tool companies and move up into a higher quality market niche
(Amsden, 1977; OECD, 1990). Ironically, the United States machine tool in
dustry was also receiving government subsidies at the same time as the
Taiwan, Province of China's machine tool industry was receiving them, but
with no performance standards attached (Amsden, 1991b). By the 1990s the
Taiwan, Province of China's Government was making preferential treatment
of business depend on whether the firm in question met conditions related
to R&D spending, personnel training, and even environmental protection
standards (Dahlman and Sananikone, 1990).
Thus, subsidies in the slow-growing late-industrializing countries have
tended to be allocated according to the principle of "giveaway" whereas
in the fast-growing ones they have tended to be allocated according to
the principle of reciprocity. In both cases the government has disciplined
labour. What distinguishes the East Asian countries is that the government
has also disciplined capital.
IV. THE GOVERNMENT'S ABILITY TO IMPOSE

PERFORMANCE STANDARDS
The factors underlying the ability of the East Asian States to impose per
formance standards on business have operated at different layers of com
plexity.
80 A uce H. A m sd en
1. The state autonomy
Imposing performance standards on business in exchange for subsidies in

the early phase of industrialization requires a critical degree of "autono
my" on the state's part. This autonomy is influenced by the political power
of private economic interest groups, whose strength and cohesiveness, in
turn, seem to depend upon: (a) the level of development of the manufac
turing sector; and (b) income distribution. What made it possible for the
East Asian states to discipline business was the relative weakness of both
manufacturing and agrarian interest groups at the outset of postwar in
dustrial development.
Asia's manufacturing sector, except for Japan's, was extremely un
derdeveloped after the w ar even by the standards of other backward coun
tries. According to two indicators of manufacturing development presented
in table 1 -the ratio of manufacturing to agricultural net product and the
net value of manufacturing per capita- East Asia's manufacturing sector
in 1955 was much less advanced than Latin America's (the backwardness
of India's manufacturing sector is exaggerated due to India's huge agrarian
population). Owing to their weakness, East Asian manufacturing enterpri
ses became dependent on state support to achieve a growth spurt, and
had fewer of their own institutions than Latin American business to shield
them from state interference.
Second, East Asian income tended to be more equally distributed than
Latin American income (see table 2). Assuming that unequal income dis
tribution implies access to a disproportionate share of resources by small
groups in agriculture or industry, the more concentrated economic power
is, the more these groups can both by-pass state sources of investment
finance and buy government favours. Japan, the Republic of Korea and
Taiwan, Province of China, all underwent a land reform in the late 1940s
during which time their agrarian aristocracies were expropriated. With a
manufacturing sector only in its infancy, and an agrarian sector devoid
of powerful interest groups, the challenge to state authority remained weak.
The Philippines represents the outstanding Asian exception to sustained
growth, and it has regressed economically, bearing the burden of a highly
unequal income distribution.
E a st A s ia n f in a n c ia l m a r k e t s : w h y s o m u c h (and f a ir l y e f f e c t iv e ). 81
2. Financial system characteristics: The case of the Republic of Korea2
Another set of historically-spedfic factors influencing the East Asian state's

ability to discipline business relates to the character of financial institu
tions. All major financial institutions under Japanese colonialism (which
emerged in the Republic of Korea and Taiwan, Province of China, before
the turn of the century and endured until the end of the Second World
War) were state-owned, and commercial banks continued to be state-owned
after independence. The ownership and control of commercial banks gave
the postwar state in the Republic of Korea and Taiwan, Province of China
control over investment finance (including foreign borrowing) and hence
enormous leverage over borrowers.
Over time, all institutions of the Republic of Korea's financial system
have been unified under one regulatory umbrella, the Ministry of Finance,
so that the parts and the whole of the system can act in concert. Political
leaders have come and gone in the Republic of Korea since the 1950s, and
Korean students have returned from the United States brandishing first
Keynesian and now monetarist theories, but there has been some continuity
in the way the Korean "State" has managed financial activities. Such consis
tency has arisen from the professionalism and high calibre of the bureau
cracy in the Ministry of Finance, which recruits its officials from the elite
universities on the basis of a gruelling civil service exam3. Up until the
present, virtually no bureaucrats in the Ministry of Finance (except poli
tically-appointed Ministers) had been educated outside the Republic of Ko
rea (aside from short-term courses). Such bureaucrats consult and possibly
respect the importance of the free market theories of United States-trained
economists, but they do not necessarily implement them, which is one rea
son for the yawning gap between what the Republic of Korea's financial
system is said to be and what it actually is.
2 Sections 5(b)-6(c) rely heavily on Amsden and Euh (1990 and 1993).
3 Tokyo National University, Seoul National University and Taiwan National University (the
last tw o in emulation of the first), were all founded with the purpose of supplying educated
governm ent officials, all of whom had to pass a "high" civil service exam. In general, all
three East Asian countries have invested heavily in education, and this has undoubtedly
contributed to the efficiency of the bureaucracy.
82 A l ic e H . A m sd en
The financial system has incorporated an intangible set of dependen

cies, power relationships, and informal understandings that lends flexibi
lity to the management of political and economic contingencies and variabi
lities in the environment. Informality has been instrumental in providing
government officials with levers to discipline big business and other reci
pients of state support.
Before the 1980s, the government prohibited a myriad of financial prac
tices; since the 1980s, it is more likely to prevent them. At one time, eve
rything was prohibited except w hat the government allowed. Since then,
all is allowed that is not prohibited. In addition to political pressures, this
change reflects the demands which a more complex economy places on
the financial system in terms of flexibility: the need for greater flexibility
to meet unforeseen contingencies is best handled by informal understand
ings and mutual dependencies rather than formal rules. The shift from
prohibition to prevention hinges more on informal than formal methods
of government intervention.
Since 1988 the monetary authorities have prevented interest rates from
rising too high -in both commercial banks and non-bank financial institu
tions (NBFIs)- by means of "window guidance": for instance, a Bureau Chief
in the Ministry of Finance will call the President of a bank, finance or
trust company, a n d /o r officials of associations of financial institutions and
provide "advice" on both deposit and loan rates.
The Ministry of Finance has several means at its disposal to make
"window guidance" binding. In the case of commercial banks, which were
privatized in the early 1980s, the MOF (and not shareholders) still influen
ces the appointment of officers, privatization notwithstanding. The appoint
m ent of directors is enforced a fortiori in the case of specialized banks,
which have not been privatized and which account for roughly 20 per cent
of total financial assets (see table 3). In the case of non-bank financial ins
titutions, which have always been private, the MOFs leverage over inte
rest rates hinges on its power: (i) to regulate these institution's financial
portfolios by setting limits on the amount of funds they can invest in,
say, real estate, stocks, and bonds; (ii) to regulate their size, by deciding
whether they may increase their paid-in capital (previously the MOF had
control over the ability of NBFIs to create new branches); (iii) to order
the Superintendent of Banks or another regulatory agency comparable to
E a s t A s ia n f in a n c ia l m a r k b is : w h y s o m u c h (and f a ir l y e f f e c t iv e ) . 83
the United States Securities and Exchange Commission to investigate a fi

nancial institution for "irregularities"; and (iv) to control the ability of
NBFIs to refinance through the Bank of Korea, which the MOF governs.
V. KEEPING THE COSTS OF INVESTMENT FINANCE LOW: THE

CASE OF THE REPUBLIC OF KOREA’S FINANCIAL REFORMS
1. Interest rates
Whether before or after reform, Korea's Ministry of Finance (MOF) has had
as one of its objectives lowering the level of financial costs for an ever-
increasing circle of firms. One consideration that has guided the Ministry
of Finance is that low interest rates are necessary to stimulate investment,
while high interest rates are not as necessary to mobilize savings.
In December 1988 the monetary authorities became serious about re
form and, as part of a sweeping financial reform, deregulated interest rates
on bank loans and discounts, and on the rates of both short-term and long
term financial markets. After December 1988 the MOF was no longer em
powered to set ceilings on interest rates (which, in a repressed financial
market, had become actual rates) or to earmark subsidized credit for spe
cial borrowers.
A brief example suffices to substantiate the continuation of the MOF's
informal interventions. In May 1989, the interest rate in the secondary short
term government bond market (special law bonds), which is a close appro
ximation to a free market rate, was 18.9 percent. The loan interest rate
of commercial banks was only 12.5 percent. This represents a big gap bet
ween the two markets and is therefore evidence against the liberalization
of interest rates which, in theory, occurred from 5 December 1988 onwards.
The only interference by government in loan allocation in the Republic
of Korea that is supposed to continue to exist is in relation to small and
medium-sized firms: all financial institutions, even branches of foreign banks
operating in the Republic of Korea, are required to set aside a certain portion
of their loans for small and medium-sized enterprises, since even foreign
banks, which were always free to choose their clients, tended to slight such
firms in the past. In practice, the MOF determines to which firms financial
84 A l ic e H. A m sd en
institutions may lend long term. Targeted industries receive preferential

credit in the form of access to bank loans which, as noted above, carry
below- market interest rates, if not the super preferential rates they once
carried. The government has also acted to help targeted industries by other
means, for instance, by developing plant sites for them near Seoul, whe
re land is expensive but labour prefers to live; by providing tax breaks
for R&D; and by financing joint ventures in R&D between business and
government.
Thus, despite liberalization in the 1980s, the government of the Repu
blic of Korea continues to provide subsidized credit to special customers.
Nevertheless, the principal means by which the government has tried to
reduce the costs of finance have come from outside the banking system.
2. Developing the stock market
Instead of liberalizing (raising) the interest rates of commercial banks to

mobilize domestic savings (and thereby reduce reliance on foreign loans),
the government has also attempted to provide both attractive outlets for
savers and cheap finance for investors by expanding the stock market. The
government has also attempted to expand the stock market to diffuse the
wealth of the "chaebol" (big diversified business groups) through greater
public ownership.
In 1980, price-eamings ratios in the Republic of Korea averaged 3
(compared to 8 in the United States and 20 in Japan). The cost of finance
through the issuance of stocks was therefore 33 per cent. In 1989, such
ratios averaged 14, which meant that the cost of finance had fallen to 7.1
per cent. Euh and Baker (1990) estimate that equity financing in 1989 cost
only 3 per cent, in terms of cash flow, even considering tax effects, as
dividend yields were only 2 per cent versus 12.5 per cent for preferential
commercial bank loans.
Price-eamings ratios have risen in the Republic of Korea in response
to general prosperity -a current account surplus provided unprecedented
liquidity- and government plans to develop the capital market. Prices also
responded to the performance of the Korea Fund, and other open-end type
investment trusts sanctioned by the government, which enabled foreigners
to invest indirectly in the Korea Stock Exchange (KSE). The Fund, amount
ing to US $60 million, was established in May 1984. It aimed at long

term capital appreciation through investment primarily in equity securities
of Korean companies. It soon traded at a 100 per cent premium, driving
up the price of a wide range of shares traded on the KSE.
The government intervened on both the supply and demand sides
to deepen the stock exchange. The Ministry of Finance drew big companies
into the capital market as suppliers of stocks and bonds by preventing
them from borrowing overseas. Ceilings on debt-equity ratios were also
enforced by closing a loophole whereby business groups tried to evade
debt-equity ceilings by cross-holding stocks of affiliated companies. Such
ceilings have represented a far-reaching form of government intervention
in the affairs of business by international standards. The MOF began to
enforce debt-equity ceilings and thereby reduce the financial instability of
big business and increase the supply of corporate stocks and bonds for
sale. A regulation that required share issues to be offered at par was res
cinded (with a notable exception, which will be discussed below). By the
late 1980s, real interest rates in the Republic of Korea were roughly three
times higher than international rates, once an appreciation of the Korean
won is taken into account (see table 4). This made it even more advanta
geous for firms to acquire finance on the capital market by issuing bonds
or stocks than by borrowing from the banks.
The number of companies Usted on the Korean Stock Exchange rose
from 342 in 1985 to 669 at the end of 1990 (see table 5), and the number
of shares traded rose from 0.8 bilhon to 3.2 bilHon. On a cumulative basis,
indirect bank finance remained more important in the RepubUc of Korea
than direct equity and bond finance, as in almost every developed and
developing country except the United States. On an annual basis, howe
ver, indirect finance became less important than direct finance in the late
1980s (see table 6).
On the demand side, government measures to deepen the stock m ar
ket had a dual objective: to mobilize savings and speculative funds by
raising demand for securities, and to spread stock ownership among the
lower income earners of the population. In addition to providing tax breaks
and opening investment trusts for foreigners, demand-side measures have
included the following: first, the MOF has aimed to enhance the integrity
and trustworthiness of the capital market. Government guarantees of cor
86 A u c e H . A m sd en
porate bonds, introduced in 1972, have continued. Stocks have been divi
ded into two tiers, blue chips and more risky issues. A country-wide drive
has attempted to educate the ordinary investor. Measures have been con
sidered by the government to reduce insider trading and improve disclo
sure.
Second, stock market institutions themselves have been strengthened,
partly in anticipation of the opening of the Republic of Korea's capital m ar
ket to foreign investors and securities companies. The MOF allowed Korean
securities companies to increase their paid-in capital by a multiple of 7
in a four-year period. An attempt has also been made to increase participa
tion on the KSE of institutional investors by allowing non-bank financial
institutions to hold securities in their portfolios. The ownership structures
of life insurance companies and securities dealers themselves have been
amended to encompass ownership of corporate equities and bonds. At the
end of 1990, institutional investors accounted for 46 per cent of outstanding
shares, a lower proportion than in most advanced countries but a far higher
one than five years earlier (Securities Supervisory Board, 1992).
Third, to spread "people's capitalism", the government allowed 5 per
cent of a company's new issues of stocks or debentures to be reserved
for its employees for sale at par rather than market value. Directors and
executives were not allowed to be included in the scheme. Some finance
companies were successful in establishing such a programme, and since
they had a relatively high level of paid-in capital and few employees, the
gains to any individual employee sometimes reached US$20,000. This sti
mulated demands by workers in other companies for the establishment of
similar schemes, demands that were sometimes supported by strikes. By
1990 Employee Stock Ownership Associations covered some 945,600 wor
kers in 784 companies. The total number of shareholders was roughly 10.4
million people out of a population of about 42 million.
3. Interest rates and savings
In the 1980s, credit continued to be allocated preferentially, but real de

posit rates rose as did savings. Whereas real savings deposit interest rates
were negative through most of the 1970s, they became positive in the 1980s,
reaching a peak of 7.3 per cent in 1986 (see table 7). Nevertheless, there
E a s t A s ia n f in a n c ia l m a r k e t s : w h y s o m u c h (and f a ir l y e f f e c t iv e ). 87
does not appear to be any simple relationship in the Republic of Korea

between interest rates and savings.
Table 8 provides data on national and household savings rates, expre
ssed as percentages of GNP. What is evident is that the rising trend of
rates of saving began well before the onset of liberalization and low in
flation rates. In 1978-1979, when inflation was high and real savings deposit
interest rates were negative during the big push into heavy industry, the
ratio of national saving to GNP reached 28 per cent. This almost equalled
the savings rate of 1984-1985 (29 per cent), when inflation was down and
the era of positive real interest rates had dawned (see table 8). Savings
rates jumped in the late 1980s, but this coincided with a surplus in the
current account of the balance of payments. By definition, therefore, sav
ings rose.
Any postulation of a simple relationship between real interest rates
and savings is further discredited by econometric exercises suggesting that
the two variables are weakly and insignificantly correlated with one ano
ther in the Republic of Korea, whether savings are defined at the national
or household level (Han, 1988; Yu, 1988). If simple explanations are preferred,
then income is a better explanator of savings behavior than real interest
rates, a finding for the Republic of Korea that is in keeping with more
general observations for developing countries. Khatkhate (1988) reports that
in a wide range of developing countries, "the level of the real interest rate
by itself has little or no impact on the selected macroeconomic variables"
(p. 577, emphasis added).
The message seems to be that savings behaviour must be understood
in a wider policy-making and socio-economic context than merely interest
rate determination. The econometric findings in the Republic of Korea of
a weak relationship between savings and real interest rates open the door
to the possibility that savings in the 1980s were influenced by opportuni
ties for capital gains in the stock market, and by an increased demand
for housing, both induced by rising incomes. Housing as a share in total
urban household expenditure rose from 14 per cent in 1965 to 17 per cent
in 1975 and to 27 per cent in 1985 (Song, 1991). The demand for housing
may be expected to raise the savings rate insofar as the Republic of Korea's
financial sector still does not provide mortgage loans.
88 A l ic e H . A m sd e n
VI. CONCLUSIONS
Without the competitive asset of new products and processes, late-indus

trializing countries have found it difficult to propel themselves into the
league of international competitors on the exclusive basis of a low-wage
advantage. Even in their leading sector in the 1960s (cotton spinning and
weaving), and notwithstanding their relatively modem physical and hu
man infrastructure achieved by dint of United States aid, the Republic of
Korea and Taiwan, Province of China, still could not compete on quality
and price using market-determined production costs against the higher
productivity of Japanese textile manufacturers. Government intervention in
financial markets, therefore, has been oriented towards lowering costs, es
pecially of capital, by subsidizing interest rates.
The Governments of both the Republic of Korea and Taiwan, Provin
ce of China, have intervened fairly effectively, as measured by aggregate
rates of productivity and output growth, because they instituted a principle
to guide subsidy allocation that differed from that of slower-growing late-
industrializing countries. In the two countries, subsidies have tended to
be allocated to business according to the principle of reciprocity, in exchan
ge for concrete and monitorable performance standards, whereas in slower-
growing countries they have tended to be allocated without strings attached.
In the 1980s, both countries (the focus of this paper) reformed their
financial systems by making them more competitive, but the Ministry of
Finance continued to coordinate financial market institutions and to intro
duce measures to keep credit articifially cheap for selected firms. Interest
rates in the 1980s were not deregulated, but the major method employed
by the government to create cheap credit was largely institutional: it in
troduced a series of measures to strengthen the stock market.
The Republic of Korea has achieved its objective of modernizing its
financial sector, therefore, principally by creating institutions or remode
lling old ones, not by relying exclusively on market forces to achieve de
sired goals. Sometimes new institutions have liberated market forces, but
at other times they have repressed them. To achieve similar goals, other
countries wishing to follow the Republic of Korea's experience would have
to create equivalent, if different, institutions, but not assume that goals are
achievable merely by letting the forces of supply and demand operate freely.
ANNEX OF TABLES
E a s t A s ia n f in a n c ia l m a r k e t s : w h y s o m u c h ( and f a ir l y e f f e c t iv e ). 91
Table 1
RATIO OF MANUFACTURING TO AGRICULTURAL NET PRODUCT
AND NET VALUE OF MANUFACTURING PER CAPITA,
LATIN AMERICA AND ASIA, L955
Ratio of m anufacturing N et value of m anu

to net agricultural facturing per head
Country product (US$)
Latin America
Argentina 1.32 145
Brazil 0.72 50
Chile 1.35 75
Colombia 0.42 45
Mexico 1.00 60
Peru 0.52 25
Venezuela 1.43 95
Asia
India 0.30 7
Indonesia 0.20 10
Republic of Korea 0.20 8
Philippines 0.32 13
Thailand 0.28 10
Source: Alfred Maizels, Industrial G row th and W orld Trade, (Cambridge: At the Uni
versity Press, 1953), as cited in Christopher Freeman, "Catching up in world growth
and world trade," (mimeo), Science Policy Research Unit, Sussex University, Uni
ted Kingdom
92 A l i c e H. A m sd e n
Table 2
INCOME DISTRIBUTION (THE RATIO BY W HICH THE INCOME
OF THE TOP FIFTH OF THE POPULATION EXCEEDS THAT OF
THE BOTTOM FIFTH)
Country/territory Year Ratio

Asia
East Asia
Hong Kong 1981 12.1
Japan 1979 4.0
Korea, Republic of (a) 1981 4.9
Southeast Asia
Taiwan, Province of China (b) - 4.3
Indonesia (c) 1983 11.9
Philippines 1971 16.1
Singapore 1977-78 7.5
South Asia
Thailand 1975-76 11.2
Bangladesh 1976-77 7.6
India 1975-76 10.1
Latin America
South America
Brazil 1982 27.7
Central America
Dominican Republic (d) 1976-77 12.5
H Salvador (d) 1976-77 8.6
Guatemala 1979-81 10.6
Mexico (d) 1977 15.4
(a) Urban only (data for other countries are national).
(b) Statistic reported in Kuo-Ting Li, The Evolution o f Polio/ Behind Taiw an's
D evelopm ent Success (New Haven: Yale University Press, 1988).
(c) Rural only, as reported in Alan Gelb and associates, Oil W indfalls: Bless
ing or C urse?, New York, Oxford University Press, 1988.
(d) Based on available rather than total households, which tends to bias
estimates of inequality downwards.
Source: (all countries and territories except Taiwan, Province of China, and
Indonesia): National survey data reported in United Nations, "Special study”,
N ational A ccounts Statistics, (New York: 1985).
Table 3
DEPOSITS BY TYPES OF INSTITUTIONS
(end of year - billion won, %) (a)
1975 1980 1985 1990 1991

Banking Institutions 2,779.2 12,421.8 31,022.7 84,054.1 98,507.9
(Deposit money banks) (b) (78.9) (70.4) (53.7) (41.2) (40.1)
Commercial banks 1,923.3 7,752.1 18,157.0 51,174.6 59,494.9
(54.6) (43.8) (31.4) (25.1) (24.2)
Specialized banks 855.9 4,669.7 12,865.7 32,879.5 39,013.0
(24.3) (26.4) (22.3) (16.1) (15.9)
N onbank financial institutions 743.5 5,281.4 26,780.2 120,078.1 146,918.6
(21.1) (29.8) (46.3) (58.8) (59.9)
Development institutions 28.8 144.8 227.2 1,773.8 2,281.1
(0.8) (0.8) (0.4) (1.5) (0.9)
- Korea Development Bank 28.8 135.8 198.5 846.0 860.3
- Korea Long-Term G e d it Bank - 9.0 28.7 927.8 1,420.8
Investment companies 195.3 1,771.9 9,552.6 33,964.0 25,120.3
(5.5) (10.0) (16.5) (28.3) (14.3)
- Investment and Finance Co. 180.5 902.0 3,238.6 8,575.3 7,012.6
- M erchant Banking Corp. - 213.2 835.4 1,209.0 1,443.1
- Securities Investment Trust Co. 14.8 633.1 5,399.6 24,179.7 26,664.6
Saving institutions 382.6 2,445.8 10,273.4 55,961.2 72,675.0
(10.9) (13.8) (17.8) (46.6) (29.6)
- Trust accounts at banks 186.5 1,042.7 3,928.0 29,174.6 36,619.4
- M utual Savings Finance Co. 50.4 419.5 2,765.4 8,940.3 12,021.2
- Credit unions 7.4 121.9 556.5 2,657.7 3,827.3
- M utual credit 138.3 861.7 2,814.4 13,823.5 18,317.9
- Postal savings - - 209.1 1,365.1 1,889.2
Insurance companies 136.8 918.9 6,727.0 28,379.1 36,842.2
(3.9) (5.2) (11.6) (23.6) (15.0)
Life Insurance Co. 136.8 918.9 6,580.7 27,165.8 35,316.0
Postal Life Insurance - - 146.3 1,213.3 1,526.2
Total 3,522.7 17,703.2 57,802.9 204,132.2 242,426.5
(100.0) (100.0) (100.0) (100.0) (100.0)
(a) Deposits in won currency on the balance sheet of institutions.
(b) Figures in ( ) indicate shares of deposits at each type of institution.
Source: The Bank of Korea, Monthly Bulletin (various issues), as cited in Amsden and Euh (1993).
Table 4
A COM PARISON OF INTEREST RATES IN
THE UNITED STATES AND KOREA
A. Real prim e rate in Korea adjusted for changes in exchange Rate

1979 1981 1983 1985 1987 1989 1991
(1) Korean prime rate (a) 18.5 19.0 10.0 11.5 11.5 12.5 12.5
(2) Exchange rate changes (b) 0 -6.2 -6.2 -7.6 8.0 0.7 -6.2
(3) (1) in terms of dollars 18.5 12.8 3.8 3.9 19.5 13.2 6.3
B. Real (dolar adjusted) Korean prim e rate on US prim e rate.

1979 1981 1983 1985 1987 1989 1991
(1) Korean prim e rate 18.5 12.8 3.8 3.9 19.5 13.2 6.3
(2) US Prime Rate 15.3 15.8 11.0 9.5 8.8 10.5 6.5
(3) (l)/(2 ) (%) 120.9 81.0 34.5 41.0 221.0 125.7 96.9
(a) At the end of the year.
(b) Negative numbers indicate won depredation.
Source: The Bank of Korea, Monthy Bulletin (various issues), as cited in Amsden and Euh (1993).
E a st A s ia n f in a n c ia l m a r k e t s : w h y s o m u c h (and f a ir l y e f f b c t iv e ) . 95
Table 5
SOUTH KOREA’S FINANCIAL REFORMS
STOCK MARKET INDICATORS
(end of year -100 million won)
1980 1985 1990 1991
Listed companies 352 342 669 686

N um ber of stocks listed 437 414 1,115 1,013
Total capital of listed companies (a) 24,214 46,654 239,816 255,096
Total m arket value (a) 25,266 65,704 790,197 731,178
(as % of GNP) 6.9% 9.0% 60.6% 51.6%
Total trading volum e (a) 11,340 36,206 534,545 625,649
Stock population (b) 753 772 1,731 1,432
(as % of total population) 2.0% 1.9% 4.0% 33%
Composite stock price index 106.9 163.4 696.1 611.0
Dividend yield 19.6% 6.0% 2.6% 3.1%
P /E ratio 2.6 5.2 12.8 10.4
(a) 100 million won.
(b) Thousand.
Source: Ministry of Finance, Capital Market Statistics (January 1992), as cited in Amsden and Euh (1993).
96 A u c e H . A m sd en
Tableó
SOUTH KOREA'S FINANCIAL REFORMS
SOURCES OF FUNDS BY THE CORPORATE SECTOR
(in percentage)
63-65 66-71 72-76 77-81 82-86 87-91

Internal funds 47.7 25.4 32.9 23.3 33.5 26.4
External funds 52.3 74.6 67.1 76.7 66.5 73.6
Total 100.0 100.0 100.0 100.0 100.0 100.0
External funds 100.0 100.0 100.0 100.0 100.0 100.0
Borrowings from
M onetary
institutions 48.4 41.8 51.1 53.7 41.8 36.0
Banks 33.5 32.8 34.3 32.6 22.6 17.0
N onbanks 15.0 9.0 16.8 21.1 19.2 19.0
Securities
(direct finance) 27.6 14.3 21.8 24.8 27.5 37.4
Debts 1.2 0.7 2.5 4.2 11.0 14.5
Stocks 21.4 11.8 18.1 14.4 16.5(a) 22.9(a)
Capital paid in 5.0 2.7 1.3 1.9
C orporate bills - - 1.8 5.5 3.9 3.3
G overnm ent and
curb market loans 8.5 7.8 -0.3 0.8 24.9(b) 20.2(b)
Borrowing
from abroad 15.4 36.2 26.6 15.2 1.9 3.1
(a) Stocks and capital paid in.
(b) Others included.
Source: The Bank of Korea, Annual Report, (various issues), as cited in Amsden and Euh (1993).
E a st A sia n f in a n c ia l m a r k e t s : w h y s o m u c h ( and f a ir l y e f f e c t iv e ) . 97
Table 7
TREND OF NOM INAL AND REAL INTEREST RATES
(in percentage) (a)
Nom inal interest rate G NP deflator Real in terest rate Change

Savings Curb (Rate of Savings Curb in the
deposit m arket change) deposit m arket price
rate (b) rate (c) (d) rate rate of land
1963 15.0 52.4 29.3 -14.3 23.1 -
1964 15.0 61.4 30.0 -15.0 31.4 47.0

1965 18.8 58.8 6.2 12.6 52.6 32.0
1966 30.0 58.7 14.5 15.5 44.6 -
1967 30.0 56.4 15.6 14.4 40.8 101.0

1968 27.6 55.9 16.1 11.5 39.8 51.3
1969 24.0 51.2 14.8 9.2 36.4 122.5
1970 22.8 50.8 15.6 7.2 35.2 2.8
1971 22.2 46.3 12.5 9.7 17.7 51.3
1972 15.7 38.9 16.7 -1.0 16.1 4.0
1973 12.6 39.2 13.6 -1.0 25.6 8.1
1974 14.8 37.6 30.5 -15.7 7.1 25.5
1975 15.0 41.3 25.2 -10.5 16.1 21.9
1976 15.5 40.5 21.2 -5.7 20.3 21.0
1977 16.2 38.1 16.6 -0.4 23.3 46.7
1978 16.7 41.7 22.8 -6.1 18.9 79.1
1979 18.6 42.4 19.6 -1.0 22.8 22.0
1980 22.4 45.0 24.0 -1.6 21.0 17.0
1981 19.2 35.3 16.9 23 18.4 7.1
1982 10.9 30.6 7.1 2.1 23.5 5.6
1983 8.0 25.8 5.0 3.0 20.8 31.7
1984 8.9 24.9 3.9 5.0 21.0 21.6
1985 10.0 24.0 4.2 5.8 19.8 7.0
1986 10.0 24.3 2.7 7.3 21.6 7.3
1987 10.0 23.2 3.4 6.6 19.8 14.7
1988 10.0 22.7 5.9 4.1 18.4 27.4
1989 10.0 23.1 5.2 4.8 17.9 32.0
1990 10.0 19.9 10.6 -0.6 9.3 20.6
1991 10.0 23.4 10.9 -0.9 12.5 12.8
(a) Bank of Korea survey data.
(b) Market rate from December 5,1988.
(c) Unofficial rate surveyed by the BOK.
(d) Until 1970 based on 1975 constant price; from 1971, based on 1985 constant price.
Source: The Bank of Korea, Monthly Bulletin, (various issues), as cited in Amsden and Euh (1993).
98 A l ic e H. A m sd en
Table 8
GROW TH OF THE FINANCIAL SECTOR
(unit: %)
1975 1978 1980 1985 1990 '91 (a)
M l/G N P 11.7 11.3 10.4 10.4 9.5 10.6

M 2/G N P 31.3 33.1 34.2 39.2 40.8 40.7
M 3/G N P 36.0 42.7 48.6 75.2 114.8 118.4
Corporate bond/G N P - 2.5 4.5 10.0 7.0 15.2
Domestic credit/G N P 39.5 36.4 45.8 58.4 56.1 57.4
Deposit share (%)
Banks 82.1 76.6 69.9 53.7 41.2 40.1
NBFls 17.9 23.4 30.1 46.3 58.8 59.9
National saving
(as % of GNP) 18.2 29.7 23.1 29.1 35.5 36.1
Stock market
Capitalization/G N P _ -
6.9 9.0 60.6 51.6
(a) Estimated data.
Source: Ministry of Finance, Fiscal and Financial Statistics, Pt. 1 (1992), as cited in Amsden and Euh
(1993).
E a s t A s ia n f in a n c ia l m a r k e t s : w h y s o m u c h ( and f a ir l y e f f e c t iv e ) . . 99
REFERENCES
Amsden, Alice H. (1992). "A theory of government intervention in late in

dustrialization", in Louis Putterman and Dietrich Rueschemeyer (eds.),
The State and the Market in Development: Synergy Or Rivalry?: (Boulder:
Lynne Rienner), pp. 53-84.
(1991a). "The diffusion of development: The late-industrializing model
and greater East Asia", American Economic Review, 81: 282-286.
(1991b). "Innovation and decline: The American machine tool industry"
(mimeo), Graduate Faculty, New School for Social Research.
(1989). Asia’s Next Giant: South Korea and Late Industrialization (New
York and Oxford: Oxford University Press).
(1977). "The division of labor is limited by the "type" of market: The
Taiwanese machine tool industry". World Development, 5: 217-234.
and Yoon-Dae Euh (1990). 'Republic of Korea's Financial Reform: What
Are the Lessons?", UNCTAD Discussion Paper Ns 30, (April, Geneva,
Switzerland).
(1993). "South Korea's 1980s financial reform: Good-bye financial re
pression (maybe), hello new institutional restraints", World Development,
21: (forthcoming).
Biggs, T.S. (1988). "Financing the emergence of small and medium enterprise
in Taiwan: Financial mobilization and the flow of domestic credit to
the private sector", Employment and Enterprise Policy Analysis Discussion
Papers Na 15 (Washington, D.C.: Employment and Enterprise Develop
ment Division, Office of Rural and Institutional Development, Bureau
of Science and Technology, United States Agency for International De
velopment).
Chandler, Alfred Dupont, Jr. (1990). Scale and Scope: The Dynamics of Indus
trial Capitalism (Cambridge, Ma.: Harvard University Press).
Clark, Gregory (1987). "Why isn't the whole world developed?: lessons from
the cotton mills", Journal of Economic History, 47: 141-173.
Coats, A. W. (ed.) (1971). The Classical Economists and Economic Policy (Lon
don: Methuen).
Dahlman, Carl J. and Ousa Sananikone (1990). "Technology strategy in Tai
wan: Exploiting foreign linkages and investing in local capability" (mi
meo), (Washington, D.C.: World Bank).
Euh, Yoon-Dae and James Baker (1990). The Korean Banking System and Foreign
Influence. London: Routledge.
Gerschenkron, A. (1962). Economic Backwardness in Historical Perspective (Cam
bridge, Ma.: Harvard University Press).
Goodrich, Carter (ed.) (1967). The Government and the Economy: 1783-1861
(Indianapolis: Bobbs-Merrill Co., Inc.).
Haggard, Stephan (1990). Pathways From the Periphery: The Politics of Growth
in Newly Industrializing Countries (Ithaca, N.Y.: Cornell University Press).
Han, Dae-Woo (1988). "Liberalization of interest rates in Korea: its impact
and policy issues", Monthly Economic Review (in Korean), 397 (December):
1- 21 .
Hanson, J. and C. Neal (1984). "A review of interest rate policies in selected
developing countries" (mimeo), (Washington, D.C.: World Bank, Finan
cial Unit, Industrial Department (Sept.)).
Herman, B. (1991). "International finance of developing Asia and the Paci
fic in the 1990s" (mimeo), Department of International Economic and
Social Affairs, United Nations, New York.
Hikino, Takashi and Alice H. Amsden (1993). "Staying behind, stumbling
back, sneaking up, soaring ahead: Late industrialization in historical
perspective", in William J. Baumol, Richard R. Nelson, and Edward
N. Wolff (eds.), Convergence in Productivity, With Some International
Comparisons and Historical Examples (New York: Oxford University Press).
Hughes, Jonathan R.T. (1991). The Governmental Habit Redux: Economic Con
trols From Colonial Times to the Present (Princeton, N.J.: Princeton Univer
sity Press).
Khatkhate, Deena R. (1988). "Assessing the impact of interest rates in less
developed countries", World Development, 16 (May): 577-588.
Lively, Robert (1955). "The American system: A review article", Business
History Review, 29:81-97.
Maddison, A. (1989). The World Economy in the 20th Century (Paris: Develop
m ent Centre Studies, OECD).
Nye, John V. (1991). "The myth of free trade Britain and fortress France:
Tariffs and trade in the nineteenth century", journal of Economic Histo
ry, 51:
Organisation for Economic Co-operation and Development (OECD) (1990).
Industrial Policy in OECD Countries: Annual Review (Paris: OECD).
E a s t A s ia n f in a n c ia l m a r k e t s : w h y s o m u c h (and f a ir l y e f f e c t iv e ). 101
Park, Yung Chul (1985). "Korea's experience with debt management", in

G. Smith and J. Cuddington (eds.), International Debt and the Develop
ing Countries (Baltimore: Johns Hopkins Press). Securities Supervisory
Board (various years), Monthly Review (Seoul: Securities Supervisory
Board).
Shapiro, Helen (forthcoming). "Automobiles: Trade and investment flows
in Brazil and Mexico", in D. Yoffie (ed.), From Comparative Advantage
to Globally Regulated Competition (Boston, Ma.: Harvard Business School
Press).
Song, Byung Nak (1991). The Korean Economy (Hong Kong: Oxford Universi
ty Press).
Taylor, A. J. (1972). Laissez-Faire and State Intervention in Britain (London:
Macmillan (for Economic History Society)).
Taylor, Lance (1988). Varieties of Stabilization Experience (Oxford: Clarendon
Press).
United Nations Conference on Trade and Developoment (UNCTAD) (1989).
Trade and Development Report, 1989 (UNCTAD/TDR/9), United Nations
publication, Sales NQ: E.89.II.D.14 (New York: United Nations).
Wade, Robert (1990). Governing the Market: Economic Theory and the Role of
Government in East Asian Industrialization (Princeton: Princeton Univer
sity Press).
Yu, Byung-Ha (1988). "Household savings in Korea", Monthly Bulletin (in
Korean), 478 (September): 31-44.
SAVINGS AND FINANCIAL POLICY
ISSUES IN SUB-SAHARAN AFRICA
Machiko Nissanke
CONTENTS
INTRODUCTION .......................................................................................................................... 107
I. FRAGMENTATION OF SAVINGS MOBILIZATION AND

FINANCIAL INTERMEDIATION IN SUB-SAHARAN A F R IC A .............................. 108
1. Characteristics of household savings in Sub-Saharan Africa ................................ 108
2. Fragmentation of financial sy ste m s............................................................................ Ill
a) Informal fin a n c e ...................................................................................................... Ill
b) Formal financial in stitu tio n s.................................................................................. 113
i. Excess liquidity in the banking sy s te m .......................................................... 114
ii. N on performing l o a n s ...................................................................................... 121
II. FINANCIAL DUALISM AND MARKET FRAGM ENTATION.................................... 124
III. POLICY REQUIRED TO OVERCOME MARKET FRAGM ENTATION..................... 128
ANNEX OF FIG U R E S.................................................................................................................... 137
REFERENCES.................................................................................................................................. 145
S a v in g s a n d f in a n c ia l p o l ic y is s u e s in S ub- S a h a r a n A f r ic a 107
INTRODUCTION
It has been increasingly recognized in recent years that the financial system
and its intermediation function could play a vital role in economic deve
lopment by effectively mobilizing and allocating resources and facilitating
transformation and distribution of risks and maturities among savers and
investors (World Bank 1989).
The Structural Adjustment Programmes (SAPs) in Sub-Saharan Africa
adopted by m any countries in response to the worsening economic crisis
of the 1980s have placed high priority on financial sector reform. Financial
liberalization policy, based largely on the arguments advanced by the Me
Kinnon-Shaw school1, has been implemented subsequently in these countries
as part of economic reform packages. Special Financial Sector Adjustment
loans (FINSAP loans) have been taken up to uphold reform measures and
to restructure and strengthen distressed financial systems in several coun
tries (e. g. Ghana, Tanzania and Uganda). However, the FINSAP has focused
mainly on putting distressed formal institutions on a sound financial foot
ing through restructuring and refinancing of balance sheets. Emphasis is
placed on prudence, an improved regulatory environment and supervision
of the banks' operations. However, these reform measures, together with
the financial liberalization policy, have at best had limited impact on econo
mic development so far.
1 See, Mckinnon (1973), Shaw (1973) and Fry (1982, 1988). For a critical literature survey of
this school, see Nissanke (1991).
108 M a c h ik o N is s a n k e
Drawing on the lessons of liberalization experiences of a number of

Latin American and Asian countries in the 1970s and 1980s, recent literature
on financial policies emphasizes the need for careful design of the sequence,
pace and timing of financial liberalization and the importance of its coor
dination with changing macroeconomic conditions2. In the case of low-inco-
me economies, however, it is particularly critical to design the financial
sector reform with a view to the speed of deepening of markets. Extensive
efforts at institution-building are needed to achieve sufficient market depth
with adequate market-support structure.
Contrary to the main thrust of financial reform within structural ad
justment programmes as originally devised, interest rate adjustments alone
are neither effective nor sufficient to stimulate improved resource mobi
lisation and allocation as they do not directly address the core systemic
or structural problems as perceived either by individuals or by the financial
institutions.
This paper examines, drawing largely on case studies of Ghana, Kenya,
Malawi and Zambia (Nissanke et al. 1991a), first, the current fragmented
state of savings mobilization and intermediation in Sub-Saharan Africa (sec
tion I) and then presents various theoretical explanations for this prevailing
condition (section H). The concluding section attempts to address the issue
of how best to overcome the existing situation.
I. FRAGMENTATION OF SAVINGS MOBILIZATION AND

FINANCIAL INTERMEDIATION IN SUB-SAHARAN AFRICA
1. Characteristics of household savings in Sub-Saharan Africa
Even in the low-income economies of Sub-Saharan Africa the potential of the

household sector for generating savings is significant. Our study suggests that
savings capacity and potential for growth in this sector is by no means negli
gible. However, there are several characteristics of household savings to note.
2 See, for example, Cho and Khatkhate (1989), Corbo and DeMelo (1985,1987) and the W orld
Bank (1989).
S a v in g s a n d f in a n c ia l p o l ic y is s u e s in S ub-S aharan A f r ic a 109
First, a critical difference in the size, form and pattern of savings be
tween rural and urban households may be noted. Rural households tend
to have higher average and marginal propensities to save. Voluntary savings
capacities in rural areas may be substantial, with far more liquidity than
is usually assumed. Though liquidity generated does not necessarily consti
tute a surplus as such3, the agricultural production cycle and the seasonal
and highly variable income creates an absolute need for savings. The pa
tterns are distinctly "short-run", governed by the seasonality of agricultu
ral production and that of liquidity flows of rural household income.
Secondly, to the extent that rural households dominate the household
sector, financial savings have a modest share in total household savings.
According to the theory, non-financial forms of asset-holding as stores of
value indicate large-scale 'efficiency loss' to the economy, where the finan
cial system can potentially perform an efficient intermediation function. Un
productive commodity holdings could entail substantial storage and tran
saction costs. In the countries studied, the insignificant share of financial
assets in total savings undoubtedly also reflects the insufficient degree of
monetization of the economy at large. Indeed, a reverse process of de
monetization has taken place in the countries where economic conditions
had deteriorated sharply in the 1980s4. The asset composition of household
savings is generally determined by the nature of the economic activities
which a particular household is engaged in as well as the degree of liqui
dity, risk, return structure of different assets and their storage and trans
portation costs. In this latter sense, it is also affected by the perception
of the households as to their liquidity constraints and the ease with which
they can switch between different sets of assets.
The third characteristic of household savings to note is that the finan
cial savings of the household sector are small in size per unit of transaction
and short-term in frequency. As noted by Deaton (1989), savings by the
household sector in developing countries are of a "high frequency" natu
re, as opposed to the "low frequency" life-cycle savings in developed eco
3 As Von Pischke notes, low levels of surplus do not necessarily accompany low levels of
liquidity (Von Pischke (1991) p. 81).
4 For the degree of monetization and the factors explaining the process of de-monetization in
the sample countries, see Nissanke et. al. (op. cit.).
nomies: savings, in essence, provide a buffer between uncertain and unpre

dictable income and an already low level of consumption. High volatility
and seasonality in income necessitate savings of this nature primarily to
protect one's livelihood in high-risk environments. Under these conditions,
financial savings of household display strong preference for more liquid
assets and maxi-mum flexibility. Even contractual savings, conducted within
the informal sector, e.g. for insurance purposes, tend to be for a very short
term time horizon and on a limited scale.
At the same time the small size and high-frequency of savings imply
high transaction costs to financial intermediaries in tapping the household/
non-corporate sector. Nevertheless, small but regular savings by large num
bers of householders/traders have been the basis of operations for many
informal associations, such as ROSCAs (rotating savings and credit associa
tions) and susu-collectors. This is closely related to the fourth characteristic
of household savings: most of the financial savings of the household sector
have been mobilized through informal financial agents or institutions. The
country-estimates suggest that savings mobilized and credit extended by
the informal financial sector appear to exceed by a large multiple those
of formal institutions. These savings are used for social and private consump
tion, as well as productive purposes, including farming activities such as
purchasing fertilizer or small-scale investment, mostly in retail trade and
commercial activities.
Numerous factors may be advanced as reasons inhibiting a large part
of the population from using formal institutions, even where such facilities
exist in rural areas. In our case studies, the high transaction cost of borrow
ing, lack of marketable collateral and low probability of acquiring credit
in addition to mistrust ranked high on the list; low interest rates were
only a marginal factor affecting borrowers' attitude to using savings facili
ties. From the lenders point of view, the combination of high risks (arising
from lack of marketable collateral and default risk) and high transaction
costs associated with small short-term loans and deposits have undermined
efforts to bring financial services to low-income groups, the micro and small-
scale enterprise sector and rural small holders. It is commonly believed
that an insufficient branch network, particularly in the rural areas, is the
inhibiting factor. However there is only weak statistical evidence connect
ing the rate of diffusion of banks (and real interest rates), on the one hand,
S a v in g s a n d f in a n c ia l p o l ic y iss u e s in S u b - S a h a r a n A f r ic a 111
and the growth of financial savings, on the other. Unless transactions costs
can be lowered and risks management improved, neither the expansion
of the number of branches, mobile facilities and specialized rural banks
nor interest rate adjustments will rectify the current inefficient and passive
attitude of banks towards savings mobilization in the household sector.
The persistent financial dualism between the formal and informal sec
tors and the limited progress in financial "widening"5 by the formal sector
can thus be attributed specifically to the incentive to interact of both bo
rrow ers/savers and lenders/intermediaries. As mentioned above, the lack
of access to credit facilities has hindered the majority of the population
in using the savings facilities of formal institutions. Our country studies
point to the critical importance of linking savings mobilization with credit
provision. There is also ample evidence that credit unrelated to savings
has undermined many targeted programmes (Adams et al. (1984), Seibel
(1989), Meyer (1989) among others).
At the same time, commercial banks, the dominant financial institu
tion in these countries, often do not have the incentive to actively mobilize
savings, due largely to the excess liquidity syndrome. Thus, one of the prime
factors explaining the relative size of the informal financial sector and its
vibrant presence within the household sector lies in the striking degree
of constrained behaviour of formal institutions. The characteristics of the
financial system will be examined below.
2. Fragmentation of financial systems
a) Informal finance
While many of the traditional forms of savings and informal financial ac
tivities have existed for a long time in these countries, a dynamism has
begun to gain momentum in recent years, as the formal sector has failed
to stimulate economic growth or respond to the need of real sector deve
lopment, in particular, that of the private sector. Rather financial distress
5 The process of financial widening can be viewed as extending the frontier/boundary between
formal and informal activities (Von Pischke (1991)).
112 M a c h ik o N iss a n k e
permeated the formal system in many of Sub-Saharan African countries

as the financial position of governments and many parastatals had sharply
deteriorated in the 1980s. Presently, thriving forces are operating on a visible
scale within parts of the informal financial sector. In Ghana, the informal
sector, SHSM-collectors in particular, performs a very dynamic role in savings
mobilization (Aryeetey et a l 1991a). In Malawi and Zambia the informal
sector plays a more active role in lending operations, enjoying considerable
comparative advantages over the formal sector in terms of accessibility,
flexibility, transaction costs, and low default rates (Chipeta and Mkandawire
(1991a) and Mutukwa (1991a). Alternative methods of risk assessment,
monitoring and the application of social peer pressure for repayment has
achieved results unattained by the formal sector.
The emerging dynamism of the informal sector has not been transmi
tted to the formal financial sector through complementary relationships or
links on any significant scale. Indeed, the informal market has continued
to thrive out of a need to fill gaps in the financial system. Despite the
growing size of the informal sector, the process of forming an integrated
financial system through developing linkages of these activities to more
efficient forms of mobilization and intermediation of savings has hardly
begun.
However, in emphasizing the strengths of the informal system, one
should also be aware of its limitations in financial intermediation on a larger
scale. In the absence of sufficient information and market signals, funds
mobilized through the informal sector are at present rarely channelled into
economic activities beyond a confined sphere. For example, most of savings
mobilized by s«sw-collectors from small-scale trading activity are channelled
into the expansion of that sector only. The money is not used for lending
to other groups in informal activity, such as small-scale manufacturing, re
pair shops or handiwork on a significant scale. Informal lenders in Malawi
and Zambia do lend to a wide range of productive activities. Yet, the high
risk of these activities and the extreme short-run nature of lending impe
de the full financial intermediation function to be performed by these single
agents. Mutukwa (1991) notes that group-based associations are subject to
highly co-variant risks, since their members tend to engage in the same
economic activity in the same community and belong to the same income
group. This factor and the seasonality of most rural income make it hard
S a v in g s a n d f in a n c ia l p o l ic y is s u e s in S ub-S aharan A f r ic a 113
for these associations to expand and perform a dynamic role in financial

intermediation.
The localized nature of the informal activities, combined with high
risks and transaction costs beyond the limited vicinity act as a constraint
on viable large-scale financial intermediation by the informal financial sec
tor on its own. For this reason, the potential use of savings mobilized through
the informal sector for economy-wide diversification remains unrealized.
Consequently, a process of industrialization and development based on in
formal small production units has not yet got off the ground. Savings mo
bilization is also currently fragmented into a large number of operations
by multifarious informal activities.
Moreover, there is little intermediation between saving and investing
units within the informal sector. Intermediation usually takes place only
over time for each individual unit. Informal savers often earn no interest
or may even incur a negative nominal interest rate (as in the swsM-collector
system), even though investing units in the economy would be willing to
pay positive real rates for funds. Few informal financial institutions provide
term finance (either directly or through linkages with the formal system).
b) Formal financial institutions
The paucity of institutions in terms of both structure and numbers is not

necessarily the kernel of the problem: a large number of institutions have
been established. However, despite their institutional diversity and efforts
to extend the branch networks, the formal financial institutions have failed
to meet the need of the real sector development in Ghana, Zambia and
Malawi, particularly. Nor have they succeeded in meeting the credit needs
of the wider groups of population in all four countries studied. The pro
blem of poor performance of formal institutions arises from the severity
of their operational constraints and fragmentation within the formal sys
tem, which is most clearly demonstrated by the coexistence of "excess li
quidity syndrome" and "non-performing term loans"6.
6 See Nissanke (1992a) for a fuller discussion of the brief description of the two phenomena
sum m arized here.
i. Excess liquidity in the banking system
For a number of Sub-Saharan countries, the disequilibrium resulting from

the "excess liquidity" syndrome has long been a wide-spread phenomenon
where banks have been allowed to operate largely on a commercial basis7.
The term is applied here to the situation where the banking institutions,
in particular those specializing in the short end of the market such as com
mercial banks, hold liquidity in excess well above the required statutory
level. Excess liquidity is defined as net excess of actual voluntary hoard
ings of liquid assets (in cash and short-term government paper) over the
minimum reserve requirements8. The levels of minimum reserve require
ments in developing countries are usually set as high as 40 to 50 per cent
of total bank deposits compared to the typical levels of around 10 to 15
per cent in industrialized countries (Kitchen 1986). This implies that total
assets held in liquid assets can often exceed the level of 70 per cent of
total deposit liabilities of the banks. These banks have held their surplus
7 Some commercial banks were established by expatriate banks primarily for financing external
-often colonial- trade in preindependence years and later nationalized to become majority-
govem m ent ow ned or rem ain foreign controlled. Others were newly established in the post-
independence era as publicly ow ned. Operational decisions of m ost of these banks are in
principle guided by commercial considerations, though they were often required to take on
a developm ental function in advancing loans to priority sectors. In contrast, banking
operations of some state-owned commercial banks have been governed prim arily by the
need to finance fiscal deficits and to cover operational loss of public enterprises and marketing
boards (eg. Tanzania and Uganda, see Collier and G unning (1991) for the Tanzanian case).
1 The term "excess liquidity" here refers to banks' voluntary hoardings of liquid assets as
opposed to "monetary overhang" when nonbank public accumulates liquidity as a result of
shortages in the goods markets. The former phenomenon becomes prevalent, when perceived
default risks increase and banks resort to a sharp retrenchm ent in lending, exacerbating the
recessionary tendencies resulting from reduced aggregate dem and. Caprio and H onohan
(1991) report that in the United States of America this was observed during the Great
Recession of the 1930s as well as in the "credit crunch" of 1990-1991. In the earlier episode,
the share of liquid assets in total deposits of the US banking institutions was reported to
have nearly doubled from 35 per cent in 1929-1930 to 65 per cent in 1935, then further to 70
per cent in 1940. The deep recession that seized major industrialized economies in 1992 has
also been characterized by debt deflation on a global scale, w here portfolios of banking
institutions have been rapidly deteriorating with increasing non-perform ing loans and
debtors-consumers have been m aking adjustm ents to their over-leveraged positions from
excess borrowing in the previous period (UNCTAD 1992).
S a v in g s a n d f in a n c ia l p o l ic y is s u e s in S ub- S a h a ran A fr ic a 115
assets on their own volition in liquid form, in cash with no return or in

treasury bills or government stocks which, until very recently, yielded low
returns9.
An explanation of this phenomenon can be manifold: First, it is a
reflection of the early stage of real sector development and a generally
poor, often unstable economic environment. Further, part of excess liqui
dity in the banks' portfolio may simply reflect a degree of inefficiency of
banking operations, i.e. their disproportionately high operational/interme
diation costs, which are not covered by the prevailing spreads and thus
require holding a large proportion of liquid assets. A high proportion of
non-performing loans with inadequate provisions for bad debts also forces
banks to hold more liquid assets than would appear necessary on the basis
of a balance sheet analysis.
Furthermore, liquidity in excess could contain transitory components
as well as permanent ones. The former would in fact reflect some time
lags in bankers' responses to temporary external shocks (sudden changes
in the terms of trade or a large influx of capital) or policy shocks such
as devaluation or uncontrolled money supply. Accordingly they should be
viewed as short-term disequilibrium phenomena. The accumulation of liqui
dity in these cases is temporary, reflecting the slow speed of adjustment
by banking institutions to shocks and thus also a degree of operational
inefficiency.
The main component of excess liquidity holdings of the commercial
banks in Kenya belongs to this category; the Kenyan banks have indeed
been dynamic in savings mobilization and credit provision (Mwega 1991).
As shown in Fig. 1, the minimum statutory required liquidity ratio in Ke
nya has been set at the very low level of 14 to 18 per cent, a level not
far from that observed in industrial countries. Indeed, the actual liquidity
9 N aturally Treasury Bills and Government Paper held as secondary reserves are not nearly
as liquid as cash prim ary reserves. Risk-adjusted returns on these holdings have been,
how ever, high enough to justify investm ent in them com pared to lending to activities
perceived to be highly risky, even when nominal returns on them were low. The recent
u p w ard adjustm ents of rates of returns on these paper as a part of financial liberalization
and monetary stabilization policies im plem ented in a num ber of con tries have m ade m ost of
lending operations unattractive to banks, except very low cost and highly rem unerative
lending (Nissanke 1992a).
ratio maintained by the commercial banks has been far below those ob
served in the other case studies throughout the 1970s and 1980s, ranging
from 18 per cent in 1974 to 1975 and 1983 to 28 per cent in 1977 and
1986, with fluctuating margins from year to year. A preliminary analysis
of annual changes in the commercial bank liquidity ratio (Mwega 1991)
suggests that there are the considerable adjustment lags involved between
an increase in commercial banks deposit liabilities and their increased lend
ing to borrowers10. In contrast to the other three countries studied, excess
liquidity in Kenya is of more transitory nature. Nevertheless, the excess
in liquidity ratio has still averaged more than 5 per cent of deposit liabi
lities throughout the 1970s and 1980s, exceeding 10 per cent in some years.
The sheer scale and persistence of the syndrome in the other countries
studied (Ghana, Malawi and Zambia) points to the need for analytical ex
planations of the permanent components of excess liquidity. The nature
and scale of the phenomenon of excess liquidity in these three countries
are illustrated in the following statistics:
In Ghana, the required minimum reserve requirements have histo
rically been set very high (Fig. 2): over 80 per cent of deposits in 1979
and 1980. The statutory level had been gradually brought down to 40 per
cent in 1985. The actual reserves held in liquid assets by the banks were
over 70 per cent in 1980-1983, indicating the extreme degree of 'low lend
ing trap1 in those years. In 1988, the year when the liberalization policy
was already well in place, the commercial banks held about 15 per cent
of their total deposit liabilities as excess reserves. Of these, about 5 per
cent was held in cash (in addition to the statutory requirement for 20 per
cent cash reserves) and 10 per cent was in the form of government paper
(over and above the statutory requirement for 25 per cent secondary re
serves). This brought the share of liquid assets to 60 per cent of total de
posit liabilities for the entire banking system in that year. The level of
liquid assets actually held by the banks subsequently increased, reaching
over 65 per cent in March 1991, as shown in Fig. 3.
10 The Kenyan country study suggests that the commercial banks liquidity decisions are
influenced by several factors including the maximun lending rates, the structure of deposit
liabilities, the required minim um ratios, and the changes in the commercial banks m one
tary base.
S a v in g s a n d f in a n c ia l p o l o is s u e s in S ub- S ah a ran A fr ic a 117
In Malawi, the excess liquid assets held by the commercial banks

amounted to from 20 to 25 per cent of total deposits in the period 1973
to 1985, while the minimum reserve ratio required was set at 25 per cent,
making 45 to 50 per cent of their total assets cash or government papers11.
There was a sharp run-down of liquid assets in the late 1970s, when the
commercial banks had accumulated non-performing loans through their lar
ge exposure towards the estate tobacco sector, as tobacco prices collapsed.
This experience has made the commercial banks, which had taken some
years to recuperate their loan portfolio, much more conservative in asset
management. Despite the rise in the minimum liquidity ratio to 30 per
cent in 1980, the scale of excess liquidity began to escalate further after
198312. Voluntary holding of excess liquid assets were over 20 per cent
in 1985; this persisted throughout the liberalization period, further increas
ing to 40 per cent above the minimum reserve ratio, bringing the share
of liquid assets to 70 per cent of total deposits in 1987 (see Fig. 4)13.
The scale of the problem has been much greater in Zambia than in
the other countries in the sample (Fig. 5). The commercial banks are
known to have persistently carried considerable amounts of excess liqui
dity throughout the 1970s and 1980s, excepting 1974. First, the statutory
required ratio had been set in a range of 25 to 40 per cent most years.
To mop up liquidity in the system, seen to affect the monetary stability,
the central bank raised the minimum required level to 40 per cent in re
cent years. The actual liquidity ratio has been above 60 per cent for most
years, reaching 82 per cent in 1988, exceeding the already high statutory
level by 25 to 45 per cent.
As explanations for the excess liquidity syndrome illustrated by the
se statistics, two theoretical propositions may be advanced14:
n In the 1970s, the largest component of liquid assets was held in governm ent stocks rather
than in Treasury Bills.
12 During the 1980s, the major part of liquid assets was held as deposits at the central bank,
rather than in governm ent securities, as in the 1970s. Some of these deposits at the central
bank reflected the bank's liabilities for outstanding im port paym ents awaiting extemali-
zation because of inadequate foreign exchange.
13 It is reported that the trade liberalization measures taken in 1989 subsequently eased the
banks' excess liquidity problems to some extent.
14 For more detailed discussion of the two paradigms, see Section II below.
First, using the Financial Repression Hypothesis, McKinnon (1973),

Shaw (1973), Fry (1982, 1988), it may be argued that the problem is asso
ciated with the governments' intervention in credit markets through a poli
cy of low interest rate, a government credit ceiling and a restrictive sec
toral credit- allocation policy.
Indeed, there is indication that the policies of low interest rates and
directed credit allocation have had, to a varying degree, negative impacts
on the operations of commercial banks. Credit guidelines imposing m an
datory sectoral allocation have undoubtedly served to undermine incenti
ves for banks to mobilize savings, as is most apparent in the Ghanaian
case. Once banks had enough deposits to meet the set credit targets of
the central bank, savings mobilization was no longer encouraged. If the
banks were not favourably disposed towards those priority sectors, loans
were withheld. Consequently they did not see an urgent need to mobilize
additional savings.
Furthermore, lending decisions by banks have not been based strictly
on commercial assessments of risks against expected rates of return but
rather have often been influenced by political clout and connections. In
deed, notable conflict between government sectoral credit allocation and
the banks' commercial considerations has been with sectors designated as
the priority, as these are identified by the bankers as high-risk areas. Hence,
such conditions may have contributed to passive banking practice; a com-
po-nent of excess liquidity may have thus been related to 'involuntary cre
dit rationing' as a result of enforced patterns of credit allocation.
An alternative hypothesis can be drawn from the Imperfect Informa
tion Paradigm (Stiglitz and Weiss (1981), Stiglitz (1989)), to explain the persis
tence of the syndrome, even in the-post liberalization years, as resulting
from 'voluntary credit rationing' and self-imposed caution on the part of
bankers operating in a high risk environment with imperfect and costly
information.
In less developed economies, successful applicants who are conside
red low-risk tend to represent established large-scale enterprises or estate
owners, either foreign or publicly owned. Many, particularly the state enter
prises, are known to have chronically suffered from poor management, low
capacity utilization and acute financial problems, prompting the current
privatization debate (Adam et al. 1992). Potentially dynamic productive small-
S a v in g s a n d f in a n c ia l p o l ic y is s u e s in S u b-S aharan A f r ic a 119
scale entities have little chance of obtaining bank credit due to high tran
saction costs and perceived risks in dealing with them. High transactions
costs can prevent banks from undertaking adequate screening, monitoring
and credit-enforcement procedure against default risks. The perceived high
risks of lending make banks extremely risk-averse in extending credit to
small borrowers, particularly those without collateral assets.
Bankers often explain their overly conservative lending policy by cit
ing the absence of viable projects to which to lend. It can be argued that
it is not necessarily the absence of viable projects per se but the lack of
mechanisms by which risks and viability can be adequately assessed, and
new ventures selected, monitored and repayment enforced. The adoption
of liberalization policies have had limited impact on this component of ex
cess liquidity. While latent demand for credit by solvent borrowers may
remain unsatisfied, the banking sector accumulate excess liquidity, withdraw
ing support from productive investment. Excess liquidity can thus give rise
to 'a low lending trap' (Caprio and Honohan (1991)) and can impede the
emergence of vigorous entrepreneurship.
Several implications may be derived from the Excess Liquidity Syndrome:
First, faced with limited lending opportunities worthy deemed and
excess liquidity hoarding, the banks lack incentive to mobilize additional
savings. Banks have even regarded the shift to deposits with longer matu
rity as raising the average costs of keeping funds, given the absence of
good investment opportunities promising a high rate of return which could
justify the wide margin between deposit and lending rates to cover high
intermediation costs. Therefore, commercial banks in Malawi and Ghana, for
example, are reported to have been turning depositors away, in particular
discouraging them from opening accounts bearing high rates of intrest15.
15 The adoption of an interest rate liberalization policy has h ad very limited im pact on the
resolution of this phenom enon, even where higher financial savings at the formal financial
institutions followed, as illustrated by the reaction of M alawian banks to the upw ard
adjustm ents in interest rates in 1987-88 (Chipeta and M kandaw ire (1991a) and Nissanke
(1992a). The banks are, first of all, constrained to raise the lending rate because of the
adverse selection incentive and moral hazard effects (Stiglitz and Weiss (1981). Even if the
lending rate is raised to cover the intermediation cost after the u p w ard adjustm ent in the
deposit rate, the resulted reduction in dem and for loanable funds squeezes their profit
margin and m akes it expensive to keep interest-bearing deposits.
Secondly, the 'low lending trap' engendered by the excess liquidity

syndrome implies a serious deficiency in the system of financial interme
diation and it obstructs a steady flow in the savings-investment circuit.
Given the short-term liability structure, and the weak capital base in most
cases, the portfolio management of the commercial banks is geared to match
ing the maturity structure of assets with the existing maturities of liabi
lities. Very little term transformation has taken place. Thus, the commercial
banks in Ghana, Malawi and Zambia have been very selective and discrimi
natory, exhibiting extreme conservatism in their portfolio-management.
Thirdly, there is an interesting link between excess liquidity and the
mechanism of crowding-out of private finance by public finance requirements
in these economies. From the commercial banks' point of view, there is
little effective demand for credit from the private sector and hence the pu
blic sector automatically assumes an unrivalled position as a receiver of
formal credit and as a vendor of financial paper. Yet, existing literature
suggests that limited access to credit is one of principal constraints for
hindering an expansion of the productive capacity of micro enterprises and
the SSE sector. This points to the considerable gaps between effective secu
ritized demand and notional unsecuritized demand for credit16. Conven
tional asset-based lending which requires tangible collateral has p ut a ma
jor restriction on lending.
Finally, the prevalence of the excess liquidity syndrome casts serious
doubts on the efficacy of monetary instruments for stabilization. Monetary
tightening -a standard policy response to the emergence and growth of
excess liquidity in these economies- would not get to the crux of the pro
blem. Credit ceilings imposed under stabilization programmes are singu
larly ineffective in an environment in which the portfolio management of
banks is conservative and reactive rather than dynamic. At the same time,
it is difficult to regulate the money supply through manipulating bank li
quidity, when the relationship between bank liquidity and credit is unsta-
16 Von Pischke (1991) argues, however, that the concept of credit needs and the perception of
unsatisfied credit dem and should not be adopted as a prime criterion for lending. In order
to achieve long-run viability and sustainability of financial institutions, he instead advoca
tes using the risk-adjusted debt capacity, which could dynamically evolve over time, as a
basis for credit policy.
S a v in g s a n d f in a n c ia l p o l ic y issu e s in S u b- S a h a r a n A f r ic a 121
ble and unpredictable due to excess liquidity. Furthermore, upward adjust

ments to returns on treasury bills necessiated for monetary stabilization
can render most private sector lending unattractive, resulting in a further
intensification of the "crowding-out" effect.
ii. Non-performing loans
Excess liquidity holdings are observed alongside a growing number of non

performing loans and low recovery rates, particularly among publicly
owned financial institutions, where political interference in lending deci
sions is prevalent. The World Bank (1989) has suggested that in developing
countries well over 20 per cent of total loans of banks are non-perform
ing. This may be an under-estimate for Africa, where loan repayment rates
are often reported as low as 2040 per cent.
In general, the causes of financial distress manifested in large pro
portions of non-performing loans in the banks' portfolio can be attributed
to numerous factors, both internally and externally generated. In the low-
income, often commodity-dependent economies, the financial institutions
are frequently exposed to large external shocks with the subsequently re
quired policy adjustments far exceeding the capacity of the economies or
the financial institutions to absorb these volatile forces and manage the
associated aggregate risks. The portfolio structure of these institutions is
little diversified. This makes it hard to offset financial loss of one activity
against gains from others. Hence a wave of loan defaults can ripple through
out the financial system, due to high risk covariance17 characteristic of the
se economies. Financial distress ensues.
The difficulties of financial institutions to cope with these external
forces are undoubtedly exacerbated by their internal structural weakness,
such as the low capital adequacy ratio, poor asset quality as a result of
inadequate loan appraisal and poor contract enforcement procedures, not
to mention political patronage and corruption. The inadequate bank super
vision by the central bank makes it hard to arrest an accumulation of non
performing loans at an early stage.
17 W hen economies are narrowly based, credit risks are highly correlated across loans exten
ded for similar economic activities.
The problem of non-performing loans is, however, most pronounced

among the development finance institutions (DFIs) or the targeted credit
schemes such as agricultural credit schemes. These were established at a
time when the supply-leading approach was prominent; the former were
set up in recognition of the shortages of medium- and long-term credit
to the priority sectors; the latter were expected to encourage flows of
concessionary formal credit specifically to the rural areas. Many of the
DFIs were also created so as to achieve developmental and social objec
tives, such as diversification and enterprise/employment creation and thus
to provide combined financial and technical services to local enterprises.
For this the DFIs were typically funded on term-loans/equity from external
sources which involved exchange and interest rate risks. They have conti
nued to be financed either by external finance or through the government,
often through refinancing schemes of the central bank when they failed
to meet the capital adequacy requirements.
Most DFIs or special schemes have not been involved in mobilization
of local savings to any significant extent and loan recovery was not given
high priority, with inevitable consequences for their funds. Their portfolio
position has been further weakened by taking on large financial risks, and
specializing in long-term lending to high risk areas such as agricultural
and industrial finance. Their excessive exposure in high risk sectors is
compounded by generally poor risk analysis or inadequate assessment of
debt capacity, excessive political pressure on lending decisions, and limited
opportunities for asset divestiture.
Frequent recapitalization has not solved the underlying problem fac
ing the DFIs. This points to a failure of the standard 'supply-leading'
approach, which creates financial institutions to meet demand for term ca
pital without simultaneously addressing other fundamental bottlenecks in
hibiting the viability of these banking operations -high transaction costs
and inadequate risk analysis and management- indispensable for their
spedalized lending operation. Conventional term risk management schemes
appear to have been inadequate, and in some cases inappropriate, to cope
with the degree of risk facing the DFIs.
It is important to recognize the need to resolve, in practical terms,
the tension that exists between the developmental and social objectives of
the DFIs (and other special lending schemes) and their financial and ope
rational liability. Yet, there are several examples of DFIs which have achie
ved better performance by adopting a more flexible, pragmatic and com
mercial approach, while at the same time managing to meet development
objectives. Good performers can be found not only in more dynamic Asian
economies but also in Sub-Saharan Africa (e.g. in Zimbabwe and Botswana).
The recommendation that DFIs should reorientate their priorities to
commercial goals, however, leaves open the critical question of how to reach
the social objectives of loan provisions to the priority sectors and meet
the credit need of micro enterprises and low-income people, thus follow
ing a broad-based and more equitable development path. Adams and his
associates (1984) suggest that policies affecting income levels through input
and output pricing policies are more appropriate in stimulating economic
development than special credit programmes.
However, serious reservations have been cast on the notion that cre
dit will "trickle down" to where it is needed through market forces and
without policy intervention. Can a viable alternative system to the conven
tional DFIs be created to ensure equitable development, through a combi
nation of financial institution-building, market deepening, linkage develop
m ent and new innovative schemes and instruments? The example of NGO
and donor involvement in innovative schemes suggests a possible way,
although it is impossible to deduce that these will proliferate to the point
where widespread self-sustained development is achieved unless it is foun
ded on active mobilization of local savings18.
The phenomena of excess liquidity and non-performing term loans
suggest that banking institutions taking the traditional approach have fai
led to make any significant inroads on expansion at the "frontier". Further
more, the co-existence of these two phenomena not only highlights a de
gree of fragmentation within the formal system, but also the serious mis
match of liquidity positions and the asset maturity structure of the sys
tem: While short-term money abounds in the form of excess liquidity in
one segment of the banking system, within the economy as a whole the
re is an endemic shortage of capital, particularly long-term loan provisions
18 For these "grassroots" banking programmes, m odelled largely on the Grameen Bank of
Bangladesh, see below.
124 M ach uco N is s a n k e
for productive investment and diversification. The formal system lacks the
abi-Hty to take on the risks associated with the maturity term transformation19.
The absence of functional inter-bank money and viable capital markets li
mits the ability of financial institutions to reduce or hedge against these
risks. Resources mobilized by financial institutions are used more by go
vernment and the public sector, with no real choices being made in terms
of risk assessment or prudent resource allocation to private users of credit.
The intermediation function of the formal system is thus seriously cir
cumscribed.
Meanwhile, dualism of the formal and informal sectors has been in
creasing and little progress has been made in financial widening by the
formal institutions in Sub-Saharan Africa in the last three decades. Any
practical long-term financial sector development policy for these economies
has to take into account predominant characteristics of their financial sys
tem and operations; namely, financial market fragmentation and pronounced
dualism between the formal and informal financial sectors. The next section,
explores theoretical explanations for this phenomenon.
II. FINANCIAL DUALISM AND MARKET FRAGMENTATION
The conventional theory of "financial dualism" suggests that the extent and
degree of segmentation of financial markets between formal and informal
sectors reflects the inherent dualism of economic and social structures,
rooted firmly in the population's traditional values and practice. It is sug
gested that as development proceeds, the informal sector would wither
away as it would be integrated into the formal system.
Empirical evidence emerging from developing countries indicates, ho
wever, that the actual process of interaction between the financial struc-
19 M any argue that this weakness in m aturity transformation and long-term loan provision
m ay not present a problem, so long as the formal banking institutions are able to supply
sufficient working capital, while retained profits and internally generated savings could be
used for term investm ent as Mckinnon's complementary hypothesis suggests. It should be
noted, however, the growth of enterprises may be slower if fixed investm ent is solely
dependent on retained earnings.
S a v in g s a n d f in a n c ia l p o l ic y iss u e s in S ub- S a h a ran A fr ic a 125
ture and real sector development is more intricate than this simple view
suggests. In some dynamic economies of Asia, integration of the financial
system has indeed taken place; the intermediation efficiency of the sys
tem as a whole has increased over time. Neverhteless, a heterogenous and
dynamic informal financial sector continues to exist. Indeed the formal and
informal sectors form a continuum of the integrated and well-functioning
financial system, with specialization in financial services by each sector
(Biggs 1991, Ghate 1990).
In contrast, in other developing countries, where economic performan
ce has been poor or stagnant, financial dualism appears to have deepened
over time; the two sectors persistently form almost discrete financial en
claves with little interaction between them. In such circumstances, market
segmentation can be detrimental to efficient functioning of the financial
system. A theoretical explanation is required to understand this phenome
non20.
The 'financial repression' hypothesis attributes the prime cause for the
dualism and fragmentation to repressive financial policies. With selective
or sectoral credit policies, a segmented and fragmented credit market emer
ges in which some borrowers obtain (rationed) credits at highly negative
real interest rates, while non-favoured borrowers m ust seek funds in ex
pensive and unreliable informal credit markets21.
In contrast, the imperfect information paradigm (Stiglits and Weiss
(1981), Stiglitz (1989)) explains the fragmentation of financial markets in
terms of the nature of credit transactions. According to this paradigm, in
the presence of imperfect information and costly contract enforcement,
market failures result from adverse selection and incentive effects and moral
hazard, which undermine the operation of financial markets. As the interest
rate charged increases, borrowers with worthwhile investments may be
discouraged from seeking loans and the 'quality' of the mix of applicants
changes adversely. Further, borrowers have an incentive to adopt projects
that promise higher returns but with greater risks; this increases the pro
20 Explanations for market segmentation are not necessarily limited to the two hypotheses
discussed below, nor are these m utually exclusive competing hypotheses. Both Hypotheses
may well be valid in combination, as suggested by Roemer and Jones (1991), (see below).
21 See Nissanke (1991) for a critical review of this thesis.
bability of defaults. There is also a moral hazard with high interest rates,
when applicants borrow to pay interest on previous loans or simply to
stave off bankruptcy rather than to invest or to finance working capital.
With these possibilities, lenders may choose not to raise interest rate
to clear the market, when faced with excess demand for loans, since the
lenders' expected return on a loan could deteriorate as the interest rate
increases. Instead, they opt to select borrowers by adopting their own ra
tioning procedure. Thus, the interest rate has generally to take on an addi
tional function of regulating the risk composition of the portfolio of finan
cial intermediaries; it then often fails to perform its market clearing role;
hence market equilibrium is frequently characterized by credit rationing,
even in the absence of credit controls and direct credit allocation. In such
a case, by this paradigm, free markets do not ensure Pareto efficient allo
cation, nor do policies that move the economy closer to free market so
lutions enhance welfare.
While these features are universally applicable to all credit transac
tions, the problems arising therefrom are likely to be most pronounced in
low-income countries, where the economy-wide information flows are ex
tremely limited, access to information asymmetric and information gather
ing extremely costly. Where there is a generally low level of income and
asset accumulation, with widespread poverty and highly skewed income
and asset distribution, strict collateral requirement restricts access to credit;
moreover the applicability of various colla teral-substitutes, such as reputa
tion and group responsibility, cannot be well tested. There is limited scope
for legal enforcement owing to the inadequate infrastructure of legal ins
titutions.
With these deficiencies in markets and market-support infrastructure,
the transaction costs for financial intermediaries -which in addition must
minimize depositors' risks- are prohibitively high. As discussed above, banks
tend to become overly conservative and risk-averse in extending loans to
non-established borrowers. This sets economic constraints on the growth
of the formal financial institutions in these countries, unless there is some
impetus to break through the stalemate. If the formal institutions are push
ed against these operational constraints, their commercial viability is threa
tened. It is this particular aspect of credit transactions that largely explains
both the failure of the subsidized credit programmes administered through
formal institutions to reach the targeted group and the high loan concen
tration in the relatively well-off groups which have collateral or reputa
tion at their disposal.
The recent study by Hoff and Stiglitz (1990) advances an explanation
for the extreme segmentation of markets in terms of asymmetric access
to information on borrowers and the differences in costs of screening, mo
nitoring and contract enforcement across lenders. The formal institutions,
even those specifically established for this purpose, faced with high tran
saction costs, ration out smaller borrowers, farmers and the poor, so as
to safeguard their operational viability. The unsatisfied demand for credit
of those left out by the formal sector is satisfied by heterogenous informal
lenders. Informal lenders have access at reasonable costs to local informa
tion on their borrowers with whom they have social interpersonal rela
tions, thus reducing screening and monitoring costs. Furthermore, tangible
assets are not necessarily the only form of collateral since informal len
ders are flexible in using other available m eans/forms for collateral subs
titute. They could, for example, rely on interlinked credit contracts with
other markets (land, labour and product), social group pressures and per
sonal knowledge of borrowers for their risk management and assessment
of debt capacity. Their service is far more flexible in loan terms and speed
compared with formal institutions. Hence, informal lenders have sizeable
comparative advantages over the formal lenders in engaging in small-scale
credit transactions because of their lower transaction costs and different
ways of handling risk. Informal associations and agents have a competitive
edge in small and short-term deposit mobilization. The result is market
segmentation and there is no automatic mechanism to integrate these seg
mented markets.
In examining the applicability of these two hypotheses explaining mar
ket segmentation by formal and informal sectors in Sub-Saharan Africa,
one can note an interesting conceptual distinction made by Roemer and
Jones (1991) between 'fragmented market’ and 'parallel market'; parallel mar
kets arise principally to evade government controls and regulations; mar
kets can be fragmented owing to some inherent characteristics of particular
market operations, even in the absence of government controls. They sug
gest that "credit markets in developing countries display characteristics of
both parallelism and fragmentation" (p. 8).
128 M achiko N issanke
Evaluated in this light, it can be said that while parallelism stems

mainly from the adoption of repressive policies, fragmentation is more effec
tively explained by the imperfect information paradigm. If the informal fi
nancial sector were strictly a parallel market, its activities would have been
reduced by liberalization measures. General evidence from the 1970s and
1980s suggests strongly that liberalization per se cannot overcome the
fragmentation of the markets, leading to financial integration.
III. POLICY REQUIRED TO OVERCOME MARKET

FRAGMENTATION
As part of Stabilization-cum-Structural Adjustment Programmes, many

Governments in Sub-Saharan Africa initiated large-scale restructuring of the
financial system in the 1980s. Emphasis in these programmes was placed
on the need (i) to adopt financial liberalization measures, i.e. de-controlling
interest rates and credit allocation to provide savers and intermediaries suffi
cient incentives, and (ii) to enhance regulatory and supervisory functions
so as to ensure prudence of the financial institutions.
Indeed, alongside a stable macroeconomic environment, adequate pru
dential supervision and regulation over financial institutions and transac
tions are now widely recognized as among the prerequisites to successful
financial liberalization. In the absence of these conditions and lack of
uncoordination with other macroeconomic and trade reform measures, a
hasty implementation of financial liberalization can further destabilize the
economies, as witnessed by the Southern Cone countries.
However, financial policies in Sub-Saharan Africa should also encom
pass measures which look into the phase that extends beyond the 'repair'
of the financial system and 'restoration' of the solvency of financial institu
tions. Our studies suggest that liberalization per se has failed to improve
the performance of banking institutions in savings mobilization and finan
cial intermediation in any sustainable manner. Difficulties in performing
effective intermediation continue to impede further advancement in savings
mobilization. In advancing liberalization measures, little attention has been
paid to the fact that motivation for increased financial savings at the for
mal financial institutions is closely related to accessibility to credit facilities
S avings and financial policy issues in S ub-S aharan A frica 129
offered. At the same time, the limited lending opportunities considerably

dampen efforts for active savings mobilization by the formal institutions.
This interrelationship between savings and credit facilities/lending condi
tions from the perspectives of lenders and savers/borrowers has been ne
glected in the financial repression hypothesis, where lending opportunities
are assumed to abound.
Indeed, financial liberalization policies have hardly addressed the issue
of high credit risks and transaction costs -the key constraint facing the bank
ing institutions in their lending operations. This remains the prime cause
for the extreme risk aversion and conservatism of many commercial banks
as well as the financial distress and illiquidity of many DFIs s p e c ializin g
in development term loans.
As in many other cases of policy reform measures, getting 'prices
right' is not sufficient to promote savings mobilization and financial inter
mediation, unless efficient markets and financial institutions are in place
(Collier and Mayer (1989)). Price signals transmitted from shallow markets
are inherently unstable and volatile. Furthermore, literature suggests that
financial liberalization could result in higher overall intermediation costs
in economies where the informal sector has been playing an important in
termediation function.
Among the measures addressing institution building, capital market
development has become one of the focal points in the recent discussions
on finandal-sector development. Indeed, the chronic mismatch of liquidity
and portfolio positions of different financial institutions calls for the crea
tion and development of inter-bank money markets by which to facilitate
rapid and easy realignment of liquidity positions. There are few well-func
tioning inter-bank money markets in the developing countries. Where they
do exist, their operations are dominated by treasury bills and other govern
ment short-dated stock, issued as government's short-term deficit financ
ing instruments. Government stocks and bills are treated by financial ins
titutions as almost 'risk-free' assets and often constitute a major compo
nents of their liquidity holdings. Other instruments such as CD, commercial
papers and inter-bank lines exist in a few countries but are used much
less.
However, our research suggests that the gap in the liquidity posi
tion of different banking institutions is of a structural nature. It cannot
be eliminated simply by activating short-term inter-bank transactions. Ge

nerally, the capacity to generate long-term capital does not emerge spon
taneously. In the demise of DFIs, capital markets are seen increasingly as
a credible alternative source and mechanism of supporting long-term in
vestment financing. Particularly, in the light of the structural gap in the
maturity structure of financial assets in the economies and the shortage
of long-term loan provisions, many policy-makers have turned to develop
ment of the capital market as a potential conduit for channelling long
term funds into the productive sectors while the banking system provides
short-term finance to investors.
Though potentially useful, capital markets take time to develop and
mature; they often fail to perform many of the functions ascribed to them
in the short-term. The paucity of financial instruments and potential parti
cipants inhibits deepening the market. Presently, in many countries, govern
ment long-dated securities and bonds tend to be placed outside the mar
ket with quasi-public institutions and held to maturity. In most cases, the
market in corporate bonds or other private share stocks is at a very inci
pient stage of development in the economies where the public sector has
hitherto dominated. Many family-owned indigenous enterprises are unders
tandably reluctant to go public. Tax policy in favour of deposit income
over dividends is also seen as impeding a transfer from the bank-based
financial system to the capital market based one.
In terms of sequencing the financial sector development, many argue
that the capital market could play its full potential only at the later stage
of economic development. For many developing countries in Sub-Saharan
Africa, it is unrealistic to expect discernable benefits from capital market
development in the immediate future. As policies are introduced to encou
rage capital markets, the improvement in banking institutions' operation
should be given due attention so that the two policies are mutually rein
forcing; the economies could eventually benefit from the advantages of both
bank-based finance and capital-market-based finance. The capital market
development is not a panacea. It should not be seen as a substitute for
policies of strengthening banking and other financial institutions through
lifting their operational constraints, namely addressing the issue of high
credit risks and transaction costs in dealing with small borrowers and sa
vers. This calls for more proactive policies to ease constraints facing bank
ing institutions in financial intermediation. Our discussion shows that such

steps could make these institutions more innovative and active in savings
mobilization. One possible way forward may be to address directly the cu
rrent state of market fragmentation.
In this regard, it is interesting to note the argument advanced by
Biggs (1991), suggesting that market fragmentation per se does not neces
sarily entail intermediation inefficiency and that market fragmentation
could work. This stands in sharp contrast to the popular view that finan
cial dualism can have an adverse impact on economic development, equity
and the allocative efficiency of resources22. Biggs takes the opposite view:
that the benefits of financial dualism could outweigh the cost associated
with it.
Gting the experience of Taiwan Province of China, Biggs argues that
the decentralization of lending and fragmentation of borrowing groups, to
optimize screening and monitoring of loans, can reduce an economy's ove
rall intermediation costs and increase investment efficiency in developing
countries with information-imperfect financial markets. The high opportu
nity cost of investable funds, determined by the market rates prevailing
in the curb market, kept the average efficiency of aggregate investment
high, and deterred entrepreneurs from undertaking lower yielding invest
ments, even when they had access to cheap bank funds. He concludes that
the development of a dualistic financial system -with the formal sector serv
ing 'full-information' borrowers, the informal lenders serving 'information
intensive' borrowers- "helped credit intermediaries allocate funds to 'informa
tion-intensive' borrowers at a lower cost and more efficiently than would
have been possible if all investable resources were channelled through for
mal sector banks" (p. 168).
A key to success in Taiwan Province of China in achieving interme
diation efficiency in the fragmented markets appears to be dependent on
creating conditions both for individual segments to evolve and grow and
for effective linkages to develop between them. First the government adop
ted a deliberate policy to encourage the development of an active informal
22 Qiandavakar (1985) address the issue of the impact of the large presence of the informal
financial sector on the efficacy of monetary control.
sector (the curb market) as "an efficient adjunct to regulated credit institu
tions". Secondly, the informal lenders have been adaptive to develop new
and innovative instruments for risky projects, and hence, they have ex
tended loans to borrowers without demanding real assets as collateral.
Furthermore, while the financial market exhibited a clear dualism into for
mal and informal segments, demarcated for borrowers depending on the
size of their activities, formal credits were available indirectly to smaller
borrowers through an extensive subcontracting system. Larger firms offered
on-lending facilities such as trade credit or suppliers credit to their sub
contractors and suppliers, acting as "de-facto intermediaries". The use of
"market interlinkage" and "credit layering" has been instrumental to their
success and supportive to industrial growth.
In other Asian countries, "interlinking" of contracts across financial,
trade and production transactions and "credit layering" have been used by
informal lenders for risk management in rural areas, wherever informal
sectors do perform an intermediation role effectively. These techniques help
to improve information gathering and contract enforcement. Yotopoulos and
Floro (op. cit. 1991) report that the informal lending operations with two-
tier credit layering, involving traders-lenders (commodity-based) and far-
mers-lenders (land-based), are extensively used in the rural Philippines.
These experiences in Asia suggest that market segmentation may not
be necessarily a sign of inefficiency, or a cause of inefficient intermediation.
Market segmentation, however, constitutes a problem when there is total
lack of communication or links between the different segments, constraining
severely the transmission of price and policy signals across the system.
This points to one key condition for a segmented system to work efficiently;
namely, there must be linkages, indirect if not direct, between segments.
In low-income African economies, this condition appears to be absent
or under-developed. Fig. 6. illustrates the flow of funds (savings and credit)
among different economic agents and sectors in these economies. Following
the conventional classification adopted in the literature on "dualism", econo
mic activities (the real sector) and financial services ( the financial sector)
are sub-grouped into the formal (modern) and the informal (traditional)
sectors.
The formal financial institutions predominantly provide services to
established large-scale "formal" sector activities, many of which, (in parti
Savings and financial policy issues in S ub-S aharan A frica 133
cular, State-owned) have had very low returns and shown poor financial
performance. The fiscal record of the governments -beneficiaries of most
of the credit from the central banks and whose papers other formal insti
tutions have largely invested in- has been generally dismal, exposed to
large external and internal shocks. The performance of many formal ins
titutions has been correspondingly poor. Past attempts to provide credit
and loans to the small-scale sectors, through the DFIs and the targeted
credit schemes, have largely failed, as little attention was given to the build
up of viable financial institutions (Meyer, 1989, 1991). Thus, the formal fi
nancial sector has suffered from a legacy of either conservatism or finan
cial distress. There is at present very little financial flow to the small-scale
sector from formal financial institutions (diagonal flows from the upper-
left box to the lower-right one in Fig. 6).
The financial needs of 'peripheral' activities of micro enterprises, SMEs
and small-scale farmers -upon whom the livelihood of a majority of the
population is dependent- are met by heterogenous informal financial orga
nizations and agents. Multifarious informal financial activities have deve
loped a dynamism of their own, demonstrating flexibility and adaptability
to the needs of local communities in which they operate. However, infor
mal financial arrangements are usually organized in a confined local vici
nity; hence their intermediation function is often constrained by their limi
ted scale of contacts, subject to highly seasonal funds and highly covariant
risks affecting the whole local community.
Consequently, markets and financial flows are largely segmented.
Critically, horizontal linkages are weak in both financial and real activities:
linkages between the formal and informal financial sectors (between the
two upper boxes in Fig. 6) are insubstantial. Since backward -and forward-
linkages of economic activities also tend to be weak (between the lower
boxes), formal credit fails to reach smaller borrowers through indirect rou
tes such as "credit layering" or subcontracting systems. Thus there is little
effective linkage whereby information and funds can flow between segments.
Presently, there is neither a dynamic market force for workable spe
cialization of financial intermediation by sector, nor policies to recognize
efficient functioning of the informal sector as intermediaries, i.e. those who
could sort borrowers according to 'information intensity' and credit risk.
There is no automatic mechanism to integrate the segmented markets into
a coherent system through specialization. Flows of funds take place mostly

within an enclave sphere in each sector, i.e. flows are restricted to vertical
movements.
In the face of fragmentation -where functional specialization by each
sector has not materialized as part of an integrated system to perform effi
cient mobilization and intermediation- the required policy would be not
to deregulate markets in haste, but to work first on closer integration of
the informal and formal markets (Roemer and Jones 1991).
Given the positive role played by the informal sector, many today
recognize that integration through simple institutionalization (formaliza
tion) of informal activities is not appropriate. As suggested by Yotopoulos
and Floro (1991), the necessary integration may be achieved several other
ways; (i) infusing into the formal institutions some of the flexibility of in
formal operations; (ii) strengthening the structure and performance of
informal market operations; (iii) developing linkages between the formal
and informal financial sectors so that the latter becomes a conduit of loana
ble funds of the formal institutions to reach borrowers otherwise rationed
out. Such schemes can lead to improved repayment rates, as informal len
ders screen and monitor a smaller group of borrowers on the basis of per
sonal relationships. The development of better linkages could be seen as
leading to a financial system where the formal and informal markets could
specialize and function complementarily, mutually reinforcing each others'
strengths.
This line of thinking has been echoed in several recent studies. For
example, the study by Germidis et a l (1991) recommends policies of both
integration of the informal sector into the formal system so as to reduce
financial dualism, through in-depth institutional and operational reforms
of the latter, and interlinkage between the informal and formal sectors, by
maximizing the positive and minimizing the negative aspects of each. The
objective of developing linkages is defined as to "reduce the gap between
the two sectors by promoting closer links between formal and informal
operators on a more systematic scale" (p. 214). The informal sector is seen
as "a means of retailing formal financial services to areas, sectors or popu
lation groups that are difficult to reach". They further propose that inte
gration and interlinkage should be undertaken not only concurrently but
somewhat sequentially, arguing that linkage in the short term would en-
train integration in the long term. Development of linkage is viewed as

a strategy of "soft integration", a step towards achieving the goal of an
integrated and competitive financial system. Indeed, better linkages are al
so desirable, and required, among institutions and agents within each seg
ment for better intermediation.
This new thinking has begun to replace the traditional approach as
a basis for developing innovative financial schemes to provide credit to
the SSE and microenterprise sector as well as the small-scale agricultural
sector (Levitsky et a l 1989, Meyer 1991). In them, credit delivery tends
to be linked to savings mobilization. Seibel (1989) proposed a model of
linking savings and credit: "as savings without credit lead to demotivation,
resource deflection and the inadequate financing of small enterprise, and
as credit without preceding savings leads to haphazard loan spending, to
risky business ventures and poor repayment morale." (p. 103). Empirical
evidence confirms that loan recovery is higher if credit provision originates
at least partially in mobilized savings. This basic philosophy has been adop
ted by many donor agencies and NGOs for instituting grassroots banking
(for poverty alleviation) and income-generating programmes (Von Pischke
1991, Holt 1991 and Yaron 1991).
The new experimental schemes tend to adopt other features of infor
mal financial activities such as group formation, replacing traditional tangi
ble collateral requirements with group guarantees and using peer pressure
and assessment of reliability based on knowledge of the participants. Many
of them also charge commercial-based interest rates to recover, at least par
tially, operational costs and thus create conditions for self-sustainability of
the programmes in the long-run.
The applicability, replicability and long-term viability of these schemes
are yet to be tested in Sub-Saharan Africa; several schemes, based on the
experience of the Grameen Bank of Bangladesh, have been instituted by
NGOs and donor agencies in a few countries (e.g. the Mudzi Fund in Mala
wi and K-REP in Kenya). However, the schemes are still at an incipient
stage and require laborious efforts of committed staff for initiating then
developing them into programmes and institutions that are self-sustainable.
Many regard these as too specialized to be a part of normal banking opera
tions. Furthermore, the majority of these schemes are at present heavily
dependent on financial and technical assistance from external agencies and
NGOs. They do not perform a function of financial intermediation with

locally generated savings. It may take some time before they can form part
of countries' integrated financial system.
In many parts of Sub-Saharan Africa, few positive steps have yet been
taken in this direction; even a nascent form of linkages in financial inter
mediation is yet to emerge. Consequently, the economies persistently lack
a system of effectively intermediating funds between surplus and deficits
units. While integration of the financial system could also be achieved
through dynamic growth of the informal sector, a policy of developing fi
nancial interlinkage might accelerate financial intermediation and stimulate
linkage development between the financial and real sectors as well as among
informal and formal financial intermediaries. Whichever policy measures
are taken, they should strengthen both existing formal institutions and in
formal operators by easing their respective operational constraints. This may
allow formal financial institutions to engage, without jeopardizing their
commercial liability, in broad-based development, meeting the latent de
mand for credit from the small-scale entrepreneurs and small-holders.
ANNEX OF FIGURES
KENYA
Liquidity Ratio
Source: Bank of Kenya, Annual Report

GHANA
Liquidity Ratio
% of deposit liabilities
Source: Bank of Ghana, Quarterly Economic Bulletin

Savings
G H AN A
and financial policy issues in

Liquidity Ratio
O)
T3
Sub-Saharan A frica
ZI
erg
co
O
<u
Oh
Source: Bank of Ghana

j—*
MALAWI
31
oq’
4*.
M achiko N issanke
Source: Reserve Bank of Malawi, Financial & Economic Review
ZAMBIA
Liquidity Ratio
cfi
<U
rÛ
(C
cn
0
Oh
01
T3
Source: Bank of Zambia, Annual Report

L
Fig. 6
FINANCIAL IN TERM EDIATION : FLOW OF SA V IN G S AND CREDIT
Financial flows (w eak:------------- , strong:-------------- )
REAL SEC TO R FINANCIAL SECTOR

S avings an d financial policy issues in S ub-S aharan A frica 145
REFERENCES
Adam, Christopher. S, William, P. Cavendish and Percy Mistry (1992), "Ad

justing Privatization", James Curry Publishing Co.
Adams, Dale W , Douglas H. Graham, and J.D. Von Pischke (1984), "Under
mining Rural Development with Cheap Credit", Westview Press
Adams, Dale W (1987), 'Taking a New Look at Loan Defaults", Economics
and Sociology Paper, Na 1375, Department of Agricultural Economics
and Rural Sociology, Ohio State University, Columbus.
Adams, Dale W (1991), "Taking a fresh Look at Informal Finance" in Philippe
Callier ed, "Financial System and Development in Africa, EDI, World
Bank.
Aryeetey, Ernest, Yaw Asante and Alexander Kyei (1991a), "Domestic Savings
Mobilization for African Development and Diversification" (mimeo)
International Development Centre, University of Oxford
Aryeetey, Ernest (1991b), 'The Relationship between the Formal and Infor
mal Sectors of the Financial Markets in Ghana," presented at the work
shop of the African Economic Research Consortium, Nairobi, May 1991.
Biggs, Tyler S. (1991). "Heterogeneous Firms and Efficient Financial Inter
mediation in Taiwan," Markets in Developing Countries, eds. Michael
Roemer and Chris Jones, ICS Press, San Francisco (forthcoming).
Caprio Gerard, Jr. and Patrick Honohan (1991), "Excess Liquidity and Mone
tary Overhangs", World Bank, mimeo.
Corbo Vittorio and Jaime De Melo (1985), "Overview and Summary", World
Development Vol. 13, NQ 8, pp. 863-866.
Corbo Vittorio and Jaime de Melo (1987), Lessons from the Southern Cone
Policy Reforms", The World Bank Research Observer, Vol. 2, Na 2.
Chandavakar, Anand G. (1985), "The Non-Institutional Financial Sector in
Developing Countries: Macroeconomic Implications for Savings Poli
cies", Savings and Development, Na 2, 1985.
Chandavakar, Anand G. (1989), "Informal Credit Market in Support of Micro
business", in Levitsky (ed.) op. cit.
Chipeta, C. and M.L.C. Mkandawire (1991a)," Domestic Savings Mobilization
for African Development and Diversification" (mimeo) International
Development Centre, University of Oxford.
Chipeta, C. and M.L.C. Mkandawire (1991), "The Informal Financial Sector
146 M achíko N issanke
and Macroeconomic Adjustment in Malawi", African Economic Res

earch Consortium, Research Paper 4.
Cho Yoon Je and Deena Khatkhate (1989), "Lessons of Financial Liberali
zation in Asia: Comparative Study", World Bank Discussion Paper,
Washington D.C.
Collier Paul and Colin Mayer (1989), "The Assessment: Financial Liberali
zation, Financial Systems, and Economic Growth", Oxford Review of
Economic Policy, Vol. 5, N Q 4
Collier Paul and Jan W. Gunning (1991), "Money création and Financial
liberalization in a Socialist Banking System: Tanzania 1983-88", World
Development, Vol. 19, N a 5, pp. 533-538.
Deaton, Angus (1989), "Saving in Developing Countries: theory and review",
paper presented at the First Annual World Bank Conference on Eco
nomic Development, Washington, April 1989.
Fry, Maxwell J. (1982), "Models of Financial Repressed Developing Econo
mies", World Development, Vol. 10, N s 9, pp. 731-750
Fry, Maxwell J.(1988), 'Money, Interest, and Banking in Economic Develop
ment", The John Hopkins University Press, Baltimore and London.
Germidis, Dimitri, Denis Kessler and Rachel Meghir (1991), "Financial Sys
tems and Development: What Role for the Formal and Informal Fi
nancial Sectors?", OECD, Development Centre Studies. OECD, Paris.
Ghate, P.B. (1990), "Interaction Between The Formal and Informal Financial
Sectors", paper presented for United Nations International Conference
on Savings and Credit for Development, 28-31 May 1990.
Hoff, Karla and Joseph E. Stiglitz (1990), "Imperfect Information and Rural
Credit Markets-Puzzles and Policy Perspectives", The World Bank Eco
nomic Review, Vol. 4, Number 3.
Holt Sharon (1991), 'Village Banking" (mimeo) World Bank, Washington
Kitchen, Richard L. (1986), "Finance for the Developing Countries", Chiches
ter, John Wiley and Sons.
Levitsky, Jacob (1989) et al, "Microenterprises in Developing Countries",
Intermediate Technology Publications, 1989, London.
McKinnon, Ronald I. (1973), "Money and Capital in Economic Development",
The Brookings Institution, Washington, D.C.
Meyer, Richard (1989), "Financial Services for Microenterprise: Programmes
or Markets?", in Levitsky (ed.) op.cit.
Meyer, Richard (1991), " Rural Finance Research: Priorities, Dissemination

and Policy Impact", Paper presented at the Vth SACRED Meetings,
FAO, Rome.
Meyer, Richard L. and Carlos E. Cuevas (1990), 'Reducing the Transaction
Costs of Financial Intermediation: Theory and Innovations", paper
presented at the International Conference on Savings and Credit for
Development, Copenhagen, 28-31 May, 1990
Mutukwa Mbelo (1991), 'Mobilizing Domestic Savings for African Develop
ment and Diversification: Case Study of Zambia" (mimeo) Internatio
nal Development Centre
Mwega, Francis (1991), "Domestic Savings Mobilization for African Deve
lopment and Diversification" (mimeo) International Development Cen
tre, University of Oxford.
Nissanke, Machiko (1991), "Theoretical Issues on Finance and Development:
A Critical Literature Survey", Queen Elizabeth House Working Papers,
June 1991
Nissanke, Machiko (1991a), 'Liberalization Experience and Structural Impe
diments to Savings Mobilization and Financial Intermediation" (mi
meo) International Development Centre, University of Oxford
Nissanke, Machiko (1992a), "Excess liquidity in the Banking System in SSA"
(mimeo) International Development Centre, Oxford University, March
1992
Roemer, Michael and Christine Jones (1991), "Markets in Developing Coun
tries: Parallel, Fragmented, and Black", International Centre for Eco
nomic Growth, HIID, ICS Press
Seibel, Hans Dieter (1989), "Linking Informal and Formal Financial Insti
tutions in Africa and Asia", in Levitsky (ed.) op. cit.
Shaw, Edward S. (1973), "Financial Deepening in Economic Development",
Oxford University Press, New York.
Stiglitz, Joseph E. (1989), "Financial Markets and Development", Oxford Re
view of Economic Polio/, Vol. 5, N s 4.
Stiglitz, Joseph E. and Andrew Weiss (1981), "Credit Rationing in Markets
with Imperfect Information", American Economic Review 71.
UNCTAD (1992), Trade and Development Report 1992.
Von Pischke, J.D. (1991), "Finance at the Frontier", the World Bank EDI
Publication, Washington
148 M achuco N issanke
World Bank (1989), "World Development Report 1989: Financial Systems

and Development", New York, Oxford University Press
Yaron, Jacob (1991), "Successful Rural Finance Institutions", World Bank
Discussion Papers N a 150.
Yotopoulos, Pan A. and Sagrario L. Floro (1991), 'Transaction Costs and
Quantity Rationing in the Informal Credit Markets: Philippine Agri
culture", (in Roemer and Jones (1991, op. rif)).
THE INDONESIAN EXPERIENCE WITH
FINANCIAL-SECTOR REFORM
Donald P. Hanna (*)
(*)The views expressed are the author's and not necessarily those of the World Bank. Inneke
Herawati provided indispensable word-processing support.
CONTENTS
INTRODUCTION .............................................................................................. 153
I. THE 1983 FINANCIAL REFORMS: BACKGROUND AND CONTENT .................... 154

1. Macroeconomic Background ............................................................................................ 154
2. Financial Sector Pre-reform Stru ctu re............................................................................. 156
3. Financial Repression ........................................................................................................... 157
4. Macroeconomic Pressures for Reform ........................................................................... 162
5. The First Financial Sector Reforms ................................................................................. 163
II. THE 1988-91 FINANCIAL REFORMS: BACKGROUND AND C O N TEN T 164

1. The Macroeconomic Background ................................................................................... 164
2. The 1988-91 Financial Reforms ....................................................................................... 165
3. Summary of R efo rm s........................................................................................................... 167
III. EFFECTS OF THE FINANCIAL REFORMS ........................................................................ 168

1. Effects on the Financial System ....................................................................................... 168
a) Financial Deepening and Interest R a t e s ................................................................... 169
b) Maturity Structure .......................................................................................................... 172
c) Bank Cost Efficiency, Allocative Efficiency & R is k ................................................ 173
i. Cost E fficien cy.......................................................................................................... 174
ii. Bank Margins and Profitability and Efficiency .............................................. 174
iii. Credit Allocation and Risk .................................................................................. 178
iv. Loan Pricing and Default Risk ............................................................................ 179
v. Efficiency and Return on Investment at F irm s................................................ 182
vi. Risks from New Operations ................................................................................ 183
vii. Liquidity Risk .......................................................................................................... 183
viii. Summary of Banking Efficiency and R is k ......................................................... 185
2. Real Effects of Financial Reform ..................................................................................... 186
a) S a v in g s................................................................................................................................ 186
b) Investm ent......................................................................................................................... 191
3. Macroeconomic Conditions and Financial R efo rm ..................................................... 192
IV. CO N CLU SIO N S........................................................................................................................... 194
BIBLIOGRAPHY ...................................................................................................................................... 196

I
I
T he I ndonesian experience with financial- sector reform 153
INTRODUCTION
Maintaining growth and development of the real economy requires a

complementary growth of the financial system capable of providing indispen
sable financial resources to the growing markets as cheaply and efficiently
as possible. For some time, Indonesia has been engaged in serious efforts
to stimulate the domestic financial system, and thereby economic growth,
through major policy packages in 1983, 1988, 1990 and 1991. This paper
begins with a discussion of the nature of the specific reforms carried out
in Indonesia, the environment in which they were undertaken and, most
importantly, their effects on the real economy.
Theories of the link between financial performance and economic
growth, particularly as advanced by McKinnon (1973), and Shaw (1973)
have been the basis for a series of financial reforms around the world,
most prominently in the Southern Cone of Latin America, but also in Tur
key, parts of Africa and East Asia, including Indonesia. These experiences
have raised concerns about the real sector effects of financial reforms, parti
cularly with regard to the freeing of interest rates and their effects on savings
and investment1. Various authors analyzing this experience have focused
1 Van Wijnbergen (1983) and Taylor (1983) present an argument based on the efficiency of the
curb market and the effects of interest rates on the supply of goods (through the cost of
working capital) whereby freeing interest rates may not lead to growth. These arguments
turn on empirical magnitudes that need to be evaluated for each case. Other prominent
debates concern the timing and sequencing of financial reforms, the content of reforms and
the nature of financial-market supervision. See for example Edwards (1984), Caprio and
Atiyas (1992).
154 D onald P. H anna
on the need to avoid severe macro-imbalances as a precondition for success

ful financial reform because of the reduced likelihood of engendering high
ex post real interest rates and the risk of currency substitution exacerbating
any fiscal imbalance (Hanson, 1992). Others have cited the examples of
the Southern Cone as warnings of the dangers of opening the capital account
before the current account (McKinnon, 1973, 1982; Edwards, 1984; Khan
and Zahler, 1983). Others have argued that the timing should be simulta
neous (Little, et al. 1970; Krueger, 1981 and Michaely, 1986, cited in Hanson
1992). After discussing the nature of the financial reforms implemented in
Indonesia over the last decade, this paper addresses their effects on the
real economy and on the financial system itself. Questions analyzed inclu
de:
• The effect on the financial system itself, including asset growth,
maturity structure, spreads, profitability and interest rates and risk;
• The effects on the level of investment and savings;
• The impact of the macroeconomic environment on financial reform;
• The sequencing of financial reform, in particular the opening of
the financial system in the presence of a closed current account
and an open capital account.
I. THE 1983 FINANCIAL REFORMS:

BACKGROUND AND CONTENT
1. Macroeconomic Background
Indonesia began the 1980s as a high growth, low-income country heavily

dependent on oil (Table l)2. Growth had averaged 7.6 per cent through
the 1970s, while inflation, low in the early part of the decade, had picked
up to levels over 40 per cent in 1974 before falling to the 10-20 per cent
range for the rest of the decade. Revenues from oil, which accounted for
up to 80 per cent of exports, kept the current account balance positive
2 For an excellent overview of the Indonesian economy, with a focus on international debt,
see Woo and Nasution (1989).
T he I ndonesian experience with financial - sector reform 155
or at a mild deficit. Oil revenues from the State-owned oil monopoly, Per-
tamina, also accounted for as much as 70 per cent of budgetary revenues,
keeping the domestic tax effort low. A period of hyperinflation in the mid-
1960s had led to the imposition of a policy that eschewed domestic-bond
financing of the government budget. The same concern over inflation led
to the adoption of an open capital account in 1970, whereby all movements
of capital by the non-finandal private sector into and out of the country
were completely unrestricted. Only banks and public enterprises faced
restrictions on the amount of foreign lending they could undertake. A debt
crisis in 1975 brought on by irresponsible borrowing by Pertamina led to
tight controls over public-enterprise borrowing and a strong desire by the
Indonesian government to limit its external debt burden. The crisis helped
Indonesia avoid the external debt build-ups that plagued many other middle-
and low-income oil producers in the late 1970s and early 1980s.
Government revenues were used throughout the 1970s to promote key
domestic industries, usually State-owned, behind a barrier of tariffs and
quantitative restrictions. Oil inflows created a Dutch disease problem during
the 1970s that Indonesia attempted to counter in 1978 by devaluating its
fixed exchange rate. Although non-oil exports initially responded positively
(Woo and Nasution, 1989), inflation in the intervening period, however,
led to continued appreciation of the rupiah.
Table 1
KEY M ACROECONOMIC IN DICA TO RS (a)
75-83 83-87 88-89 1990 1991
Average real growth rates (% p.a.)

GDP 6.5 5.0 6.6 7.1 6.6
Non-oil GDP 7.0 5.7 7.8 7.5 6.3
Non-oil exports 10.5 12.2 17.8 2.8 24.3
Non-oil imports 13.8 -8.2 12.7 26.0 9.6
Fixed investment 10.7 -3.7 11.9 14.6 6.0
Macroeconomic balances (%) (b)
Current account/GNP -7.8(d) -2.5 -1.9 -3.4 3.8
Non-interest current account/GNP -6.0(d) 2.0 2.3 0.4 0.3
Overall public sector balance/GDP -4.3(d) -3.0(e) -2.1 (e) 0.2(e) -0.4 (e)
MLT debt service/exports 16.8 (d) 34.8 35.8 29.7 31.6
MLT debt/exports 127.6 234.3 212.2 187.6 196.5
MLT debt/G N P 33.1 65.5 60.1 57.8 59.0
156 D onald P. H an n a
Table 1 (Con t.)
75-83 83-87 88-89 1990 1991
Structure of the economy (%) (b)

Non-oil manufacturing/GDP 9.9 12.8 13.9 14.9 15.4
Non-oil exports/non-oil imports 37.4 80.8 88.8 71.2 77.5
Public savings/GDP 8.9 4.8 6.4 10.2 9.4
National savings/GDP 21.0 19.1 20.9 20.9 20.3
Fixed investment/GDP 25.1 19.2 20.6 22.5 22.7
Private fixed investment/total
fixed investment 52.1 59.2 58.7 59.0 57.7
Consumption/GDP 73.9 75.6 74.2 74.2 75.0
Prices
Oil prices (US$/bbl) (b) 28.9 17.6 17.9 22.6 18.7
Non-oil terms of trade (83/84 = 100) (b) 100.0 96.3 95.3 95.3 91.5
Domestic inflation (% p.a.) (b) (c) 16.2 7.9 6.7 7.9 9.4
(a) Balance o f payments and fiscal data are for fiscal years (starting April 1). Other indicators are for
calendar years.
(b) For last year of m ulti-year periods.
(c) As m easured by the consum er price index, with an adjustment for rice prices during 1987-89; yearly
average.
(d) For 1982/83.
(e) Calendar year.
Source: Central Bureau of Statistics.
2. Financial Sector Pre-reform Structure
Indonesia's financial system at the beginning of the 1980s was typical of

most developing countries (Nasution, 1983; Balino and Sundarajan, 1986).
Finance was dominated by commercial banks, which accounted for 95 per
cent of financial assets (Table 2). Banking was dominated by five State
commercial banks which, along with the Bank of Indonesia (BI) the central
commercial bank, controlled 80 per cent of financial assets. State banks had
a number of advantages, including extensive branch networks, access to
BI, and the exclusive right to receive public-enterprise deposits.
The other 15 per cent of the banking systems assets were in the hands
of 21 banks authorized to operate in foreign exchange (11 foreign and 10
domestic), 60 private domestic banks limited to rupiah operations and 29
development banks. Private banks had grown during the 1970s after having
been outlawed in the early 1960s. They were hampered, however, by res-
frictions on branching and access to BI as well as to public enterprises.

Domestic private banks were nonetheless favoured over foreign banks which
operated without access to BI credit and were limited to a maximum of
two branches. Two national and twenty-six development banks, all but one
State-owned, provided some long-term financing with funds from the public
and multilateral banks. In addition, an extensive rural financial system ope
rated through thousands of very small village banks which were precluded
from offering demand deposits. Other institutions in the financial system
played a very minor role. Some 13 non-bank financial institutions had been
set up in the early 1970s to promote investment banking and the stock
market. Their activities were funded through the issuance of certificates
of deposit (CDs), purchased by commercial banks, and through equity, again
chiefly supplied by the state banks and BI. A domestic stock market, revi
ved in 1977, was moribund. By 1982 only 24 companies had issued shares,
mostly foreign-owned ones seeking to comply with investment laws requir
ing gradual sales of equity to Indonesians. A few bond issues had also
been floated in the market, but overall the funds raised in the capital mar
ket amounted to only 1.7 per cent of financial assets in 1982. Insurance
companies and pension funds were small, keeping what assets they had
in land or short-term deposits.
3. Financial Repression
Recycling oil revenues within Indonesia in the 1970s led to an elaborate

system of directed, subsidized credit, known in the country as liquidity
credit, that accounted for 48 per cent of all bank lending by 1982 (Table
3). Refinancing by BI at subsidized rates was plentiful so long as the loans
met the criteria for any of a myriad of directed credit schemes. Schemes
targeted small-scale firms, farmers, transmigrants, home owners and public
enterprises. Terms varied, with the portion of a loan eligible for refinancing
varying from 20 per cent to 100 per cent and the rediscount interest rates
from one-third to one-half of the subsidized rate paid by the final borrower.
The bulk of refinancing was provided through State banks which granted
loans on the credit at rates of 6 to 12 per cent. Refinancing was the chief
source of funding for State banks since, despite the open capital account,
their interest rates on deposits of over 3 months were controlled by BI
Table 2
STRUCTURE AND GROW TH OF THE INDONESIAN FINANCIAL SYSTEM
(number, billions of rupiah and %)
Number Nominal Asset Size Real Growth

1982 1988 1991 1982 1988 1991 1978-82 1982-88 1988-91 1990-91
Bank of Indonesia 1 1 1 13,707 42,445 55,220 1.4 10.8 0.5 2.4

DMB 118 111 185 17,105 65,693 159,392 8.7 14.4 21.3 6.4
State 5 5 5 12,257 39,862 70,158 1.8 11.6 10.6 -1.5
Private FX 10 12 23 1,168 10,189 45,654 23.7 28.1 41.7 10.7
Foreign 11 11 29 1,172 3,215 12,070 -2.6 8.8 35.8 11.6
Non-FX Private 60 51 96 720 4,972 12,868 22.5 24.2 23.4 7.8
Development 29 29 29 1,336 5,046 14,505 8.8 14.1 26.9 25.3
Savings Bank 3 3 3 452 2,409 4,137 57.4 19.8 9.8 20.2
Other (a) 113 209 276 1,447 6,681 N.A. N.A. 17.5 N.A. N.A.
Rural banks (b) 5,808 5,783 6,243 86 637 N.A. 19.5 25.3 N.A. N.A.
Total (c) 6,040 6,104 6,954 32,345 115,456 225,575 6.2 13.2 14.1 4.6
(a) Comprises N BFls, insurance and leasing companies. The nbfi data are as of March 1991.
(b) Village and Urban Peoples Credit Banks (BPRs).
(c) Totals reflect previous year stocks for those categories for which 1991 data is unavailable.
Source: Bank of Indonesia.
(Table 4). Private banks, on the other hand, were not subject to interest-
rate controls, neither on deposits nor on loans. As a result, private banks
offered deposit rates two to three times higher than State banks.
Table 3
LIQ U ID ITY CREDITS
(billions of rupiahs; end of period)
Mar Mar Mar M ar Mar Mar

1983 1985 1988 1990 1991 1992
State Banks
Liquidity Credit (LC) 3,876 6,426 9,176 13,873 10,807 12,048
L C /State Bank Credit 43.8 46.7 10.1 32.6 19.8 19.5
State Bank L C /A ll LC 92.2 93.6 89.0 79.5 82.7 87.5
Private National Banks
Liquidity Credit 208 298 753 2,602 1,693 1,280
LC Share (%) 15.4 9.1 9.2 11.5 4.8 3.0
LC/A11 Bank Credit 1.8 1.6 2.2 3.7 1.7 1.1
All Banks
Liquidity Credit 4,203 6,862 10,311 17,451 13,061 137,68
LC Share (%) 37.1 36.8 30.7 24.6 13.1 11.9
Source: Bank o f Indonesia.
Table 4
IN TEREST RATES A T COMMERCIAL BAN KS, 1982-91 (a)
(annual %)
Decem ber of 1982 1985 1986 1987 1988 1989 1990 1991
Nominal Deposit Rates (b)

State Banks 6.0 16.0 14.7 17.3 18.2 17.2 19.4 22.3
Private FX Banks 18.5 17.8 15.5 18.5 19.6 18.2 20.0 23.3
All Banks NA 16.9 15.4 18.4 19.0 17.7 19.6 22.7
Ex post Real Deposit Rates (c)
State Banks -3.4 9.0 6.4 10.6 9.1 6.5 11.2 16.1
Private FX Banks 8.0 10.7 7.1 11.7 10.4 7.5 11.7 17.1
All Banks N.A. 9.9 7.1 11.6 9.9 7.0 11.4 16.5
Nominal Lending Rates (d)
State Banks 12.0 15.3 18.5 20.0 20.2 19.7 21.2 25.1
Private FX Banks N.A. 24.2 23.02 3.6 23.8 21.7 25.1 28.2
All Banks N.A. 22.1 21.12 2.1 22.3 21.0 23.0 26.1
Table4 (Cont.)
D ecem ber of 1982 1985 1986 1987 1988 1989 1990 1991
Ex post Real Lending Rates (c)

State Banks 2.1 8.4 9.9 13.1 11.0 8.8 12.8 18.8
Interest Rate Spreads
Lending-Deposit Spread (d)
State Banks 5.7 -0.6 3.3 2.3 1.7 2.1 1.5 2.3
All Banks N.A. 4.4 4.9 3.1 2.8 2.8 2.8 2.8
On-shore Off-shore
Dollar Spread (e) 0.9 -0.5 -0.0 1.0 1.9 -0.0 0.3 2.5
On-Shore Rupiah On-Shore
Dollar (e) 3.4 9.0 8.1 9.3 8.5 9.2 11.0 14.6
On-shore Rupiah-LIBOR (e) 43 8.4 8.1 10.4 10.6 9.1 11.4 17.4
Memo Items:
Annualized Semester
CPI Inflation 9.7 6.4 7.8 6.1 8.3 10.0 7.4 6.2
6 Month US Dollar LIBOR 13.6 8.6 6.9 7.3 8.1 8.3 7.8 4.6
6 Month On-shore
US Dollar Rate (f) 14.6 8.1 6.8 8.4 10.2 8.3 8.1 7.2
(a) For rupiah transactions, excluding liquidity credits. Rates shown include all outstanding loans or
tim e deposits, not marginal rates.
(b) Average rate for six-m onth tim e deposits.
(c) Rate calculated using the actual annualized sem ester inflation as proxy for expected inflation.
(d) Average nominal rate on working capital loans. Because o f long credit m aturities, the average shown
responds slowly to current rates. Thus the lending rates cannot b e directly compared with the deposit
rate.
(e) Spread calculated using Private FX bank on-shore rates.
(f) Rate on offer at Private FX banks.
Source: Bank o f Indonesia and author's calculations.
In addition to the direction of credit through its extensive refinancing

program, BI also set individual limits on credit expansion for each bank,
thereby influencing the allocation of almost all credit (Table 5). Aggregate
and subsector limits on credit growth had started as a means of controlling
the money supply, since policy makers felt that relying on a fractional re
serve system in the face of an open capital account and significant cross
border flows would be ineffective. Of course a system of credit control
did not protect against the effects of loose domestic monetary policy, which
still led to periodic losses of international reserves. It did, however, insu-
late the domestic financial system from large foreign exchange inflows (the
main concern of policy makers during the period of oil wealth), since these
could not be transformed into credit.
Table 5
SECTORAL SHARES & GROW TH OF CREDIT AND GDP
(%)
Growth
1982 1988 1991 88-91
Share of Total Bank Credit

Agriculture 8.7 8.4 7.6 34.2
Mining 16.0 1.2 0.6 17.1
Manufacturing 24.9 33.3 31.2 35.1
Trade 31.4 30.8 29.5 35.9
Services 14.2 16.5 17.7 39.8
Others 4.8 9.9 13.4 47.3
Share of GDP
Agriculture 24.1 24.1 19.3 8.8
Mining 19.5 11.6 12.4 18.5
Manufacturing 12.0 18.5 21.8 21.6
Trade 15.9 17.3 17.5 16.4
Services 24.8 24.6 25.6 17.5
Others 3.7 3.8 3.3 11.2
Memo items:
Total Credit (trillions of rupiahs) 10.3 32.8 100.4
Total GDP (trillions of rupiahs) 62.5 139.5 226.5
Source: Bank of Indonesia, Central Bureau of Statistics.
The elaborate system of credit allocation ultimately thwarted govern

ment objectives and reduced efficiency. From the perspective of the go
vernment, the plethora of programs and lack of prioritization makes it diffi
cult to judge whether the allocation of credit that ensued was desirable.
The large loan subsidies encouraged a misdirection of funds, shortchanging
intended beneficiaries. From the standpoint of efficiency, the subsidized ra
tes encouraged financing of low-retum projects, or ones with levels of capital
intensity inappropriate for a low-wage, labor-surplus country. Equalization
of the marginal efficiency of investment across sectors, a partial measure
of allocative efficiency, was hindered by the targeted nature of directed
credit. The existence of high percentages of rediscount and subsidized cre

dit insurance from a State-owned insurance company weakened the incen
tive of State banks to choose viable projects or to supervise them once
funded. Woo and Nasution (1989) point out that State banks rarely met
their credit goals, a result that they hypothesize may have been due to
large bribes called for by these banks, bribes that raised the cost of State-
bank credit above that of private banks.
For the State banks, the existence of credit ceilings together with con
trols on deposit and lending rates weakened the incentive to aggressively
seek deposits. Private domestic banks, operating without interest-rate res
trictions, showed faster growth than State banks. Because of generous credit
ceilings, private banks were able to increase assets at a real rate of over
20 per cent per year between 1978 and 1981 (Table 2). State banks, though
showing real growth of less than 2 per cent, were still able to collect enough
funds to develop high levels of excess reserves. These reserves were either
converted via foreign exchange as a hedge against devaluation or lent out
to private banks in the interbank money market.
The combination of extensive State ownership, interest and credit con
trols and entry barriers led to financial markets that were highly segregated
(Harris, et al., 1992). Public enterprises banked with the State banks, whose
extensive branch network gave them advantages in raising funds. Large
conglomerates, fostered in the hot-house of trade protection, were also cus
tomers of the State banks. These banks therefore provided almost all of
industry's funding. Besides the fragmentation of internal markets, different
groups had varying degrees of access to foreign financing. Of the private
firms, the large conglomerates had the best access to foreign borrowing.
Chinese-owned firms sometimes had access to foreign funds through off
shore Chinese banking links with Hong Kong and Singapore.
4. Macroeconomic Pressures for Reform
By 1982, the worsening price of oil and world-wide recession had under
mined Indonesia's balance of payments and fiscal balance, thus prompt
ing a series of macroeconomic adjustments. The current-account deficit had
reached 7.8 per cent of GDP, while oil-tax receipts had fallen 13 per cent
in real terms during the fiscal year (April to March).
One of the most important steps taken was a devaluation of the ru
piah by 38 per cent in March 1983, which brought the real exchange rate
back to its 1978 level, when the last devaluation had occurred. This was
undertaken to spur non-oil exports, which had responded well to the 1978
devaluation, and to increase rupiah revenues in the budget.
The devaluation was coupled with a series of fiscal adjustments to
make it stick. These involved cuts in current expenditures and reductions
in domestic subsidies to oil, public enterprises and food. This was reinfor
ced by delaying or cancelling dozens of large public-sector projects, includ
ing four multi-billion-dollar petrochemical projects planned by Pertamina.
Efforts at increasing domestic tax collections were also undertaken.
5. The First Financial-Sector Reforms
The fall in oil revenues played havoc with the old structure of the financial
system that had relied on significant recycling of the government's oil reve
nues through the banking system. Policy makers began to focus on the
need to promote the mobilization of domestic savings to maintain investment
in face of tightening external constraint. There was also growing concern
about the need to provide banks with more flexibility in the allocation
of credit, with the goal of increasing the efficiency of its use.
The reform process began in August 1982 as BI cut back on the provi
sion of directed credits for some low-priority sectors. This was followed
in June 1983 by more substantial reductions in the utilization of directed
credits to 14 "priority areas" and a hike in their interest rates to 12 per
cent. BI also moved to limit its direct credit to the public, relying instead
on refinancing. At the same time, the bank announced the elimination of
ceilings on lending rates at State banks, except for directed credits. Interest
rates on time deposits longer than six months were also freed3. Further
efforts to mobilize domestic funds included the authorization of bearer CDs
and the elimination of a 20 per cent withholding tax on domestic dollar
deposits. Finally, and most importantly, BI eliminated all credit ceilings.
3 This was followed several months later by the elimination of limits on six-month time
deposits, so that only savings deposits continued to have interest-rate ceilings.
Although these bold reforms greatly increased the flexibility of exist

ing banks in pricing and allocating credit, they stopped short of lowering
entry barriers, either among banks or between banks and other financial
institutions. No new foreign banks were allowed entry, nor were new li
censes for foreign exchange operations issued. The opening of new bran
ches continued to be severely constrained, giving State banks a clear ad
vantage that was further solidified by their having sole possession of State
enterprise business.
II. THE 1988-91 FINANCIAL REFORMS:

BACKGROUND AND CONTENT
1. The Macroeconomic Background
The move to lower entry barriers was to take five years. In the intervening
time Indonesia's external environment improved in 1984-85 and then worse
ned considerably, with the 1986 plunge of the price of oil and the appre
ciation of the yen both pushing the current account into the red. As in
1983, the Indonesian authorities responded with a combination of exchange-
rate and fiscal policies to restore the balance of payments. Real current
government expenditures were cut through a salary freeze, subsidies were
reduced and capital spending slowed. Another sharp devaluation was
announced in September 1986, lowering the value of the rupiah by 50 per
cent. In response to the stabilization program, Indonesia was also able to
step up its aid program, with large amounts of balance-of-payments support
coming from the World Bank, the Asian Development Bank and Japan.
Unlike the 1983 crisis, this time stabilization efforts were coupled with
trade-reform measures. This began in 1985 with a reduction in the dispersion
of tariff rates. A duty-drawback system for exports was revamped in May
1986. More reforms accompanied the 1986 devaluation, setting a pattern
of replacing non-tariff barriers with tariffs which were subsequently lowe
red. In an effort to attract foreign investment, licensing requirements were
simplified in 1987. By 1991, a series of almost annual trade-reform packages
had succeeded in sharply lowering export bias and variance in the trade
regime and broadening the scope and ease of foreign direct investment.
Stabilization and trade reform left the Indonesian economy in a different

position in 1988 than it had been at the advent of financial reforms in
1983. Inflation had been kept under 10 per cent, helped by the trade re
form that created more effective competition from imports. Strong efforts
to reform the domestic tax system and increase collections, coupled with
austere spending, had helped keep the fiscal deficit below 4 per cent of
GDP (except for 1986) despite the decline in oil revenue (Table 1). Strong
export growth had reduced the dependence of the trade account on oil.
Nevertheless, the debt burden had increased, both because of yen appre
ciation and because of larger current account deficits (Table 1).
2. The 1988-91 Financial Reforms
Having confronted the stabilization challenge of 1986, attention turned once

again to the financial sector in October 1988. The government announced
a far-reaching set of reforms, quickly termed PAKTO (the October Package)
that dramatically lowered entry barriers to financing:
• Banking licenses were made available to new banks that could

meet new minimum capital requirements.
• The process for obtaining a foreign-exchange license was simpli
fied.
• Freer branching by domestic banks was allowed so long as stan
dards of prudential soundness were met.
• Foreign joint-venture banks were authorized, with an extended bran
ching network.
• Limitations on the activities of banks and non-banks were lessened.
• State enterprises were allowed to hold up to 50 per cent of their
assets in private banks.
• The right to issue CDs was extended to all banks and non-bank
financial institutions (NBHs).
• The burden of monetary control was lowered by a reduction in
reserve requirements from 15 per cent to 2 per cent of deposits.
Efforts to promote competition were coupled with improvements in

prudential supervision of banks. Regulations limiting lending to persons,
firms or groups to 20 per cent and 50 per cent of equity were phased
in. The new capital requirements forced greater levels of equity for banks4.
Insurance company soundness was improved by the imposition of solvency
requirements.
Having dealt with banking, the next reform package (PAKDES, in
December 1988), focused on stimulating the capital market and other fi
nancial institutions. The government issued new regulations covering the
establishment of multi-finance companies empowered to engage in leasing,
factoring, venture capital, credit-card operations and consumer credit. The
same activities were made available to banks. Another set of regulations
came out governing securities trading, including prohibitions against in
sider trading. A major limitation to the stock market was eliminated when
domestic deposits were subjected to a 15 per cent withholding tax, the
same tax levied on dividend payments. New regulations also opened the
market to foreign investors.
PAKMAR, the March 1989 package, was aimed at refining the pru
dential regulations first announced in October 1988. It contained a series
of decrees clarifying, among other matters, the development and control
of NBFIs, lending limits, joint-venture bank capital ownership and bank
mergers, the definition of bank capital, reserve requirements and bank
investment in stocks. The long-standing absolute limit on external borrow
ing was replaced with restrictions on the net open position of banks in
foreign exchange (25 per cent of equity). Furthermore, the requirement of
prior BI approval of off-shore lending was eliminated. This allowed banks
to borrow off-shore more freely so long as they lent domestically in foreign
exchange or otherwise covered their position, thereby further freeing up
the capital account.
PAKJAN, announced in January 1990, took on the directed credit pro
grams that had continued to exercise a large though diminishing role in
the banking system despite initial efforts to curtail them in 1983 (Table
3). This time priority programs were limited to four activities (mainly fi
nance for small farmers and foodstuffs). One important sector excluded
from directed credit was export finance. Interest rates were moved closer
4 This equity cushion was quickly erased when lending expanded (to be discussed in the next
section).
to the market level and the portion of credit available for refinancing was
lessened. Mandatory, subsidized credit insurance was abolished. All these
measures were additional incentives for originating banks to more carefu
lly select and monitor their borrowers. As a political compromise, the eli
mination of directed credit programs for small businesses was replaced by
a requirement that 20 per cent of a bank's loans be made to small bo
rrowers.
PAKFEB, the latest set of reforms, and announced in March 1991,
returned once again to prudential regulations. New professional standards
were set for bank directors. Loan-loss provisioning standards were over
hauled, now involving a financial analysis of customers rather than simply
a check of whether their payments were current. A new, more quantita
tive evaluation of bank soundness, based on capital, asset quality, mana
gement, equity and liquidity, was implemented. Finally, banks were obli
ged to adopt the risk-based capital adequacy standards (as stated in the
Basel Agreement) by the end of 1993 (subsequently extended to the end
of 1994).
These important reforms were followed by several important new
pieces of basic legislation covering banks, pension funds and insurance
companies, all issued in February and March of 1992. Immediately before
hand, a regulation establishing the operating rules for closed-end mutual
funds was also handed down. The Banking Law is particularly important
since it eliminates any legal distinction between private and State banks
or between NBFIs and commercial banks. The former must now choose
between operating as a security house or a bank. New, stiffer penalties
for fraud are included in the law, along with a provision for the partial
privatization of State banks.
3. Summary of Reforms
Before moving to a discussion of the effects of financial reform, we will

now briefly summarize the discussion so far. The pre-reform system was
dominated by State-owned banks whose credit decisions and interest rates
were largely controlled by the BI, itself a major commercial lender. As a
result credit decisions were based on administrative decisions at BI with
preference being given to public enterprises, protected domestic firms and
agriculture. Competition between State and private banks and between banks
and other financial intermediaries was stymied by high entry barriers and
the ceiling on credit expansion at all banks. Nonetheless, the open capital
account put limits on Indonesia's ability to manipulate domestic interest
rates by offering larger depositors and borrowers an off-shore option. The
financing of fiscal deficits through foreign borrowing put a cap on expen
ditures that helped contain inflation and domestic demand.
The pressures of adjusting to external shocks, chiefly the oil-price de
clines of the 1980s, and concerns over the implications of pervasive controls
of credit, led to the 1983 reforms which removed all credit controls and
freed interest rates on all but directed credits. The latter were streamlined
but continued to play an important role in the system until their reform
in early 1990. The second stage of reforms, begun in 1988, tackled the high
entry barriers to banking and attempted to foster competition between banks
and other financial intermediaries. This was followed in 1991 by strengthen
ing of prudential regulation, a process that had first begun with some re
forms in 1989.
Despite this generally market-based reform, there are still some areas
of finance where the market is not given full play. As mentioned above,
the most important is the ruling requiring banks to lend 20 per cent of
their portfolio to small customers. For foreign and joint-venture banks this
rule has been replaced with a requirement that 50 per cent of all lending
be to exporters. The reinstitution of quantitative limits on bank external
borrowing in November 1991 also limits the openness of the capital account.
III. EFFECTS OF THE FINANCIAL REFORMS
Having presented the macroeconomic background and thenature of the

reform process, we now turn to an evaluation of its effect both on the
financial system and the real economy.
1. Effects on the Financial System
The McKinnon-Shaw framework highlights the importance of reform of the

financial system in promoting financial savings and improving interme-
diation. We now review how the financial system has reacted in three key
areas: (a) financial deepening and interest rates (b) transforming the ma
turity of savings to match tire needs of investment projects, and (c) the
efficiency of intermediation.
a) Financial Deepening and Interest Rates
Following reform, the financial system has deepened and broadened. Re
form led to a rapid expansion of financial assets within the economy after
credit and interest rates ceilings were eliminated in 1983 (Table 2). Overall
financial asset growth in real terms more than doubled (to 13.2 per cent)
between 1982 and 1988 compared to its rate between 1978 and 1982. Private
foreign-exchange bank growth was the strongest in banking, surpassed on
ly by leasing, which had a much smaller asset base. Private banks were
particularly successful in attracting time deposits from private individuals
and firms, despite a narrowing of their interest spread over State banks
as the latter responded to their new freedom by boosting rates 6 to 8 per
centage points (Table 4). This move narrowed the differential from 8.5 per
centage points on 3-month time deposits in 1982 to 2.6 percentage points
in 1983. The differential fell to less than 2 percentage points by 1985 and
is currently closer to one percentage point. Initial gains in deposits at pri
vate banks were also made despite restrictions on the number and location
of private bank offices. Insurance companies, public enterprises and govern
ment accounts all continued to be dominated by State banks, chiefly be
cause of legal and regulatory restrictions. The freedom to set interest rates
did allow State commercial banks to boost real growth to almost 12 per
cent, after a real growth of less than 2 per cent earlier on. Interestingly,
despite the initial attempts to curtail directed credits, BI's assets expanded
at close to 11.8 per cent real per annum between 1982 and 1988, compared
to only 1.4 per cent between 1978 and 1982. This growth did not slow
down until 1988-1991.
After the lowering of entry barriers in 1988, growth expanded further,
helped by a sharp increase in the number of firms. Assets at deposit mo
ney banks grew by nearly 20 per cent in real terms between 1988 and
1991, pushed by private foreign-exchange and non-foreign-exchange banks,
whose asset growth was 41 per cent and 35 per cent, respectively. On
the deposit side of the balance sheet, private banks in only two years have
increased their share of deposits to over 40 per cent, while that of State
banks has fallen to below 50 per cent. Part of this increase is due to the
success of private banks in attracting savings accounts. Nominal growth
of savings accounts has exceeded 40 per cent since 1988.
Overall financial savings increased dramatically as a result of the 1983
and 1988 reform measures. M2 to GDP ratios increased sharply between
1982 and 1991, after an increase in the previous nine years of only 4 per
cent (Table 6). The bulk of this improvement came because of an expan
sion of quasi-money -time and savings deposit- that went mostly to pri
vate banks after the interest rate deregulation of 1983 (for time deposits)
and 1988 (for savings deposits). The link between this dramatic jump in
financial savings and domestic savings relative to national accounts is ex
plored below. In the meantime, however, it is important to note that an
important, though unknown, amount of the increase in M2 to GDP was
due to the movement of deposits out of Singapore and other off-shore
markets and into Indonesia.
An interesting impact of the reforms is evident in the figures on Ml
to GDP, where M l is composed of currency, C, and demand deposits, DD
(Table 8). As Chant and Pangestu (1992) have pointed out, the 1983 re
forms had little effect on the M l/G D P ratio, with the ratio actually falling
in 1983 and 1984. This pattern is consistent with Ml being held chiefly
for transactions and therefore only marginally affected by the higher ra
tes available on quasi-money. The 1983 increase in interest rates did lower
demand for M l relative to income somewhat, as people turned to the more
remunerative quasi-money. After the 1988 reforms, which allowed greater
branching and improved service, including interest on demand deposits,
there has been a discrete jump in the M l/G D P ratio. The effect of higher
interest rates on demand deposits is also visible in the stagnant share of
currency to GDP and the increase in DD/GDP.
Looking at the allocation of credit across institutions shows a diffe
rent pattern from that noted earlier for deposits. State banks actually ex
panded their share of total credits between 1982 and 1988 from 61.7 per
cent to 65.1 per cent. The maintenance of State bank dominance in credit
provision was due in large part to the continued importance of liquidity
credits funneled principally through State banks (Table 3) and to the decli-
ne in direct lending by BI, whose share fell from 26 per cent in 1982 to
only 4 per cent in 1988. Because of the withdrawal of BI from direct credit
creation, private banks also expanded their share of credit, which surged
from 9 per cent to 23 per cent.
Lower entry barriers created a stampede of new firms into financial
services. In banking, more than 75 new banks were opened between 1988
and 1991 (Table 2). Not only new banks, but the number of branches sur
ged as well, an increase of more than 1100 to 3700 between 1988 and 1991.
The PAKDES reforms initially sparked the equity market, leading to the
licensing of over 200 stock-brokerage firms. The market itself finally be
gan providing equity finance to complement greater amounts of debt fi
nance, though this slowed down in 1990 with the rise in domestic interest
rates and institutional growing pains (see discussion below).
Tableó
M ONETARY AGGREGATES IN GDP, 1978-1991
(%)
Currency Demand Ml Quasi M2

deposits money
1978 5.1 5.1 10.3 5.4 15.7

1982 4.7 6.7 11.4 6.3 17.7
1983 4.3 5.5 9.7 9.1 18.9
1984 4.1 5.4 9.5 10.4 20.0
1985 4.6 5.8 10.4 13.5 23.9
1986 5.2 6.2 11.4 15.6 26.9
1987 4.6 5.5 10.2 17.0 27.1
1988 4.4 5.7 10.1 19.4 29.6
1989 4.4 7.6 12.0 23.1 35.1
1990 4.6 7.5 12.1 30.9 43.0
1991 4.1 7.5 11.6 32.1 43.6
Source: Bank of Indonesia.
Not only has overall asset growth been strong, but the number of
products available to savers and borrowers has multiplied and service im
proved. Some of the fastest growth, for example, has occurred in leasing
and other newer financial activities. In banking, the range of attractive sa
vings current and time deposit accounts widened. Home mortgages and
consumer finance are now available from a number of providers. Domes

tic syndicated loans allow on-shore financing of large projects by smaller
banks.
b) Maturity structure
Accompanying the rapid expansion of bank liabilities and assets was a

lengthening of maturities (Table 7). On the liability side, this was because
of a fall in the importance of demand as compared to time deposits. Time-
deposit maturities shortened dramatically, but were still longer than those
of demand deposits. For all banks, 24-month deposits shrank to only 10.5
per cent in 1988, as compared to 45.7 per cent in 1982. The deregulation
of savings deposits has led to a dramatic increase in their share of liabi
lities that, combined with a shortening of the maturity structure of time
deposits, has led to a shortening of the average deposit maturity in the
last few years. Since 1983, credit maturities have slowly lengthened. This
means that loan maturities are somewhat better matched to the profiles
of investment projects. The lengthening of credit maturities has occurred
only at private banks. State banks and development banks have shortened
the maturity of their credits. The modest improvement in the maturity of
credits has come at some cost to bank soundness. It implies that banks
are more exposed to interest-rate risk due to a widening maturity mis
match. Some of this risk is hedged by the use of variable rate loans.
Table 7
COM M ERCIAL BAN K M A TU RITY STRUCTURE
(billions of rupiahs; end of period)
1983 1986 1988 1990 1991 M aturity W eights

months
Bank Credits
< 3 months 1,829 1,991 2,354 3,495 3,957 1.5
3-6 months 1,787 3,443 5,392 5,057 5,726 4.5
6 months-1 year 4,438 11,237 18,866 43,753 49,539 9.0
1-3 years 1,604 2,092 3,731 8,778 9,938 24.0
> 3 years 3,944 8,974 14,825 29,904 33,859 60.0
unclassified 105 56 105 191 218 1.5
Total 13,707 27,793 45,273 91,178 103,237
Table 7 (Cont.)
1983 1986 1988 1990 1991 M aturity Weights

months
Average Maturity
(mths) (a) 23.8 25.5 26.0 26.3 26.6
Bank Funds
Demand Deposits 6,031 8,157 10,350 19,254 22,013 1.0
Savings Deposits 584 1,387 2,174 9,662 10,595 1.0
Total Rupiah Time
Deposit 4,441 10,393 19,622 38,789 40,559
< 3 months 934 1,280 3,895 11,899 11,802 1.5
3-6 months 605 1,448 4,133 5,886 9,160 2.0
6 months-1 year 843 1,767 2,719 5,958 8358 9.0
1-3 years 1,948 5,898 8,350 13,910 9323 15.0
unclassified 111 132 525 1,136 1,716 1.5
Total All Deposits (b) 11,056 19,937 32,146 67,705 73,167
Average Maturity
(mths) (a) 5.8 8.6 7.8 73 5.1
(a) Calculated by weighting th e share of each m aturity in total credit (deposits) by the assumed matu
rity weight (the last colum n of the table).
(b) This total does not include FX tim e deposits.
c) Bank Cost Efficiency, Allocative Efficiency & Risk
Tobin (1984) divides banking efficiency into three categories. The most
straightforward is cost efficiency: Do banks intermediate funds at the lowest
possible cost for a given level of risk? This entails looking at overhead
costs and, more generally, at the spread between borrowing and lending.
The second measure of efficiency, allocative efficiency, tries to incorporate
whether borrowers receive credit at a price commensurate with their risk.
Alternatively, it can be thought of as how successful the financial system
is in equalizing the marginal return on investment across the economy.
The measure of efficiency disregards the cost and risk of credit to focus
on liquidity. Since banks will necessarily have some maturity mismatch on
their balance sheet, prudent management requires keeping enough liqui
dity on hand to meet unexpected demands for cash. This last item can
be seen as a way of verifying that cost efficiency has not been achieved
through increasing risk. In general one could extend the analysis to other
sorts of risks (foreign-exchange, interest-rate, etc.). We will now examine
these different measures of efficiency.
i. Cost Efficiency
Indonesian banks have become more cost efficient since 1983. Non-interest
operating expenses (NIOE), chiefly wages, rents and office supplies, have
fallen sharply as a share of average total assets (Table S f. For all banks
in 1982 NIOE chewed up 4.26 per cent of average assets while accounting
for an exorbitant 7.11 per cent in non-foreign exchange banks. Interestingly,
State banks showed lower overhead costs than private banks, implying that,
though overstaffed, State bank wage levels were quite low. By 1988 NIOE
figures had fallen to 3.23 per cent overall, a decline of 24 per cent, with
private banks successfully bringing their overheads in line with those of
State banks. As the 1988 reforms took hold, NIOE initially fell further as
competitive pressures began to build. By 1990, however, the costs of a ra
pidly expanding branch network and higher wage rates for skilled staff
pushed up NIOE at all but the foreign banks6. With the slowdown in this
network expansion, banks, faced with lower asset growth and higher loan
losses (see below), have responded by once again reducing overhead, most
noticeably at State banks.
ii. Bank Margins and Profitability and Efficiency
A key to sparking domestic financial intermediation is to lower the spread

between borrowing and lending rates. For banks to be more efficient,
though, the lowered spread should not be offset by higher risk, a point
to which we return in the next section. There are two means of measuring
spread (Chant and Pangestu 1992). The first is to look at spreads on an
ex ante basis, that is, the difference between posted deposit and lending
5 The figures in Table 8 are based on unaudited balance-sheet and income statements collected
by the Bank of Indonesia.
6 Note that foreign banks were still subject to limits on branches and had traditionally had
higher wage bills because of more extensive use of expatriate staff.
rates. This spread includes banks expectations as to future losses, expecta

tions that can go very wrong. Alternatively one can judge spreads ex post
by looking at the actual results from a bank's financial statements. Howe
ver, one should cautiously interpret the results in this case as well, since
losses embedded in the portfolio may not yet be fully realized.
Table 8
CO ST EFFICIENCY FOR BANKS
(% of average total assets)
1982 1988 1989 1990 1991
Non-Interest Operating Expense

State Banks 4.14 3.33 3.58 3.87 1.84
Private FX Banks 6.89 3.63 3.35 3.59 3.44
Private Non-FX Banks 7.11 3.68 2.87 3.87 3.57
Foreign Banks 4.01 3.76 4.18 3.97 2.63
All Banks 4.26 3.33 3.10 3.82 2.50
Source: Bank o f Indonesia and author's estimates.
As we have seen, the attractiveness of depositing in the domestic mar

ket improved with the increase in real interest rates (Table 4). Improved
cost efficiency and lower reserve requirements (after 1988) should have al
so narrowed the spread. Table 9 provides information on ex post spreads
derived from the unaudited balance sheet and income statements of Indo
nesian banks as supplied to BI. The presentation follows the OECD for
mat first used for developing countries by Hanson and Rocha (1982). It
begins with interest income and expense, the difference being the net inte
rest margin. Adding income and expenses from fee-based business, includ
ing foreign exchange, and subtracting non-interest operating expenses gives
the net operating margin. Inclusion of extraordinary income and expenses,
which include provision expenses, gives the pre-tax return on assets. The
return on equity shown in the table is net of tax.
Both interest income and interest expenses as a share of average assets
increased after the 1983 reforms. This is not surprising since both measu
res reflect interest rates (the difference being that the figures in Table 9
use assets, rather than loans as their base). As would be expected, State
banks show the sharpest rise in interest income, from 1.4 per cent in 1982
to 9.6 per cent in 1988. This sharp rise reflects not only freedom to deter-
Table 9
BANK C O STS AND M A RGIN S
(% of average total assets)
1982 1988 1989 1990 1991
Interest Income
State Banks 1.42 9.57 10.48 11.25 13.57
Private FX Banks 1.59 15.82 14.32 16.48 20.59
Foreign Banks 4.45 11.71 12.05 12.52 12.99
All Banks 1.60 11.17 11.60 13.18 16.03
Interest Expense
State Banks 4.12 7.40 7.77 8.08 11.43
Private FX Banks 7.95 11.90 10.81 12.49 16.14
Foreign Banks 9.15 7.57 7.97 6.95 7.83
All Banks 4.82 8.25 8.83 9.29 12.60
Net Operating Margin
State Banks 2.55 2.45 2.63 2.84 1.31
Private FX Banks 4.00 2.89 2.37 2.26 2.24
Foreign Banks 4.51 3.70 3.64 3.98 4.53
All Banks 2.99 2.64 2.57 2.91 1.98
Pre-Tax Return on Assets
State Banks 2.40 1.40 1.55 1.78 0.31
Private FX Banks 3.60 1.93 1.68 1.39 1.24
Foreign Banks 4.32 1.78 2.80 3.02 3.39
All Banks 2.81 1.68 1.67 1.99 0.97
Return on Equity
State Banks 67.31 44.54 47.34 65.59 16.16
Private FX Banks 32.89 29.65 18.43 15.07 12.63
Foreign Banks 119.90 48.77 18.59 17.33 25.26
All Banks 45.97 29.93 24.82 28.21 16.65
mine loan rates, but also the elimination of credit ceilings which allowed
State and other banks to move more assets into loans. State bank inte
rest expense, however, does not increase nearly as much as interest inco
me -only 3.3 percentage points, as compared to 8.2 percentage points. This
increase is smaller than that of private foreign exchange banks which were
the State banks' closest rivals. The lower increase may be due to the con
tinued presence of public-enterprise deposits in State banks, as well as the
continued access of State banks to refinancing from BI. Another part of
the difference is explained by the implicit government guarantee that State
banks enjoy.
Lower interest costs at State banks did not translate into lower interest
margins at these banks because of low loan income, which between 1982
and 1988 reflected the continued importance of liquidity credits (Table 3).
The 1983 reforms did, however, lead to sharply higher interest margins
at all banks, going from 2 per cent to 4 per cent by 1988, somewhat exceed
ing the 3.5 per cent interest margins earned by US banks in 19887. Such
margins were negative at all banks in 1982, while the figure for private
non-foreign exchange was -8.8 per cent! This is only an accounting ploy,
however, since gross operating margins were all quite high in 1982, made
possible because income was being recorded from fees, rather than from
interest. It is a mystery as to why, without restrictions on interest rates,
private and foreign banks would resort to this subterfuge.
Reductions in NIOE between 1982 and 1988 allowed net operating
margins to fall less than gross margins for all types of banks. The opening
of new private banks pushed down margins at these banks, especially in
1991 when they approached international levels. State bank net operating
margins were steady, however, until 1991 when, in the wake of reductions
in refinancing and the loss of public-enterprise deposits, interest expenses
grew 2.5 percentage points. All domestic bank returns on assets also declined
between 1988 and 1991, chiefly as a result of higher provision expenses.
This was particularly true for state banks where the combination of higher
interest and provision expenses reduced profits to only 0.1 per cent of assets.
Meanwhile, foreign banks, helped by low deposit costs and high income
from foreign exchange trading, have been able to hold pre-tax profits at
levels two to three times those of domestic banks.
Overall, the examination of ex post margins shows a banking sys
tem that has significantly reduced its spread as measured by net operat-
7 See Duca and McLaughlin (1990) for the US figures.

ing margin, chiefly through a reduction in fees and overhead that has out
stripped increases in the interest margin. Pre-tax returns on assets are
approaching levels seen in recent years in the US.
Hi. Credit Allocation and Risk
One of the central aims of financial deregulation was to improve the allo
cative efficiency of the banking system by eliminating ceilings on credit
and interest rates while limiting directed credits. These moves were designed
to allow resources to be directed to high-return sectors of the economy,
at prices that reflected the risk inherent in those sectors. As w e have seen,
the reforms were associated with lower costs of intermediation and greater
credit flows. But was the credit better allocated? To assess the effects of
the reforms in this area we look first at the sectoral allocation of credit,
and then at the results of a firm level study, following this with an analysis
of changes in the riskiness of bank lending.
The overall allocation of credit was not greatly affected by the 1983
reforms (Table 5), though the pattern has roughly followed changes in GDP.
In part this may have been due to the continued importance of liquidity
credits within overall credit until very recently. Trade continued to receive
about 30 per cent of all credit between 1982 and 1988. Manufacturing's
share of credit rose to almost one-third at the expense of lending to mining.
These changes mirrored movements in value added over the period, with
the manufacturing share in GDP increasing and the mining share falling.
Since 1988 credit growth has accelerated most rapidly in services and
"other" two areas that have not seen a corresponding increase in value
added. This is due to expanding home mortgages, car loans and other consu
mer finance. Part of the increase in 1990 was in response to the govern
ment decree mandating that 20 per cent of a bank's portfolio be lent to
small borrowers. Most consumer loans satisfy the government's criteria. Cri
tics have charged that this consumer and property-development lending
has spurred conspicuous consumption and unhealthy increases in land pri
ces. Experiences in other countries would support the view that lending
to consumers and for commercial real estate is riskier than lending to other
sectors. Indeed, the 1990/91 tight money policy, instituted in part to slow
overall credit growth, took the steam out of the property market and led
to difficulties for some banks, including the bankruptcy of a newly-formed

bank with large real-estate loans, Bank Summa.
The bankruptcy and subsequent liquidation of Bank Summa, the tenth-
largest private bank with over US$1 billion in assets, demonstrates the pro
blem not only of concentrating lending on real estate, but also of lending
to related parties. A major portion of Bank Summa's loans were made to
affiliated companies. The poor performance of these companies further
weakened the bank. Indonesia, with all its major private banks controlled
by conglomerates, is particularly exposed to tie risks of related lending.
For this reason adherence to prudential standards requiring that no more
than 30 per cent of equity be lent to any one group is critical if financial
soundness is to be achieved. The Bank Summa case does present a useful
precedent in that the owners agreed to cover all losses due to lending
to related companies, over and above the equity they held in the bank.
The case also points out the need for further improvements in Indonesia's
procedures for dealing with problem banks. Shortcomings in laws and
regulations regarding bank liquidation have complicated the process of set
tling accounts.
iv. Loan Pricing and Default Risk
Looking at the sectoral allocation of credit gives only a partial picture of

the allocative effects of banking reform. Increased lending to sectors that
have low returns, as measured by high loan default rates, is not an efficient
allocation of credit. Efficient allocation thus requires that banks be allowed
to price their loans appropriately. However, until 1990, because of the
predominance of liquidity credits with fixed interest rates and targeted markets,
there was little need for adequate loan pricing nor little leeway to do so.
Credit risks could be passed off to government insurance companies at
subsidized premiums, further lessening the incentive for banks to adequate
ly assess and price such loans. This was particularly true of State banks
owing to the importance of liquidity credits in their portfolios. The scaling
back of liquidity credits, coupled with the strong growth in non-liquidity-
credit loans, however, has allowed banks greater freedom in pricing their
loans and in choosing sectors and borrowers to lend to. One of the first
reforms undertaken in 1983 eliminated restrictions on deposit interest rates.
Altering bank cost of funds and thereby bank profitability has had pro
found effects on banks loan pricing, as can be seen in Table 9. Both these
changes should lead, a priori, to an improvement in banks' management
of default risk. Offsetting this, though, is the pressure on banks to build
market share by aggressively expanding their lending portfolios at the cost
of underpricing loans. While this has happened at some banks, margins
in general do not seem to support this conclusion. However the sheer speed
of the growth in credit of some 50 per cent per annum between 1988 and
1991 must have led to a weakening of credit quality8.
Determining which effect has predominated as regards pricing and
allocative efficiency is highly speculative. An answer requires a judgement
about changes in the level of risk in bank portfolios and the adequacy
of measures to cover that risk. One approach to this question is to look
at how loan-loss reserves and provision expenses have behaved over recent
years (Table 10). This approach, however, requires caution for several rea
sons. First, BI portfolio examinations have traditionally focused only on
the status of interest payments and collateral in classifying loans and deter
mining reserve adequacy. Hence, the level of reserves needs not reflect the
underlying financial strength of borrowers. Commercial-bank criteria have
generally been stricter in classifying their portfolios. However, accounting
standards as to when to declare a loan non-performing and how much
to provision were not tightened until 1991 with a two-year phase-in. The
standard for the treatment of accrued income when a loan becomes non
performing was not implemented until early 1991. Prior to that, interest
payments, for example, could be capitalized so that a loan would appear
current although a borrower had made no payments and could conceiva
bly be bankrupt9. Analysis is further complicated by the nature of a great
part of bank lending. Firms generally receive lines of credit that are rolled
over, the so-called "evergreen" loans. If the line is large enough, a company
can easily keep current by simply drawing down the line to meet debt
service. In a period like 1988-90, when aggregate credit growth exceeded
the interest rate for the system on the whole, no one paid off any debt
8 See CSIS, (1990).

9 Of course, banks, wishing to prop up their balance sheet can still capitalize interest illegally.
but instead simply borrowed more. Furthermore, the fastest rise in credit
came in 1989 and 1990 as the overall economy grew at over 7 per cent.
Such strong growth can reduce company loan defaults until economic
growth slows, making it appear that current levels of reserves are adequate,
based on recent experience, but potentially inadequate in the future. Indeed,
as growth slowed in 1991 and 1992 problem loans increased as a share
of bank portfolios. Finally, tax laws limit loan-loss reserve deductibility to
3 per cent of loans for private banks and 6 per cent for State banks, limit
ing banks' interest in exceeding these levels. All of these factors combine
to make drawing conclusions about default risk from loan-loss reserves
tentative.
Table 10
LOAN L O SS RESERVES AND PRO V ISIO N S (%)
1982 1988 1989 1990 1991
Loan Loss Reserve/

Total Loans
State Banks 3.2 4.2 2.3 4.4 4.5
Private FX Banks 1.7 1.2 1.0 0.8 1.1
Foreign Banks 0.6 2.0 1.6 1.4 1.9
All Banks 2.7 3.3 3.3 2.7 2.8
Provision Expense/
Total Loans
State Banks N.A. 1.2 0.6 1.3 0.9
Private FX Banks N.A. 1.0 0.7 0.8 0.7
Private Non-FX Banks N.A. 0.9 0.7 0.6 1.2
Foreign Banks N.A. 1.1 0.7 0.6 0.8
All Banks N.A. 1.1 1.0 1.0 0.8
Provision Expense/
Interest Margin (a)
State Banks N.A. 44.9 36.0 35.5 47.3
Private FX Banks N.A. 20.7 16.3 18.1 14.9
Private Non-FX Banks N.A. 17.1 16.7 13.4 27.2
Foreign Banks N.A. 17.2 12.8 8.7 14.6
All Banks N.A. 29.2 28.2 22.3 23.4
(a) The interest margin is defined as the difference between interest loans and
interest expense on deposits.
Source: Bank of Indonesia and author's calculations.
With these caveats in mind, the figures in Table 10 show that loan-
loss reserves at all banks rose as a share of loans between 1982 and 1988,
fell in 1989, and since then have increased to levels higher than in 1982.
Provision expenses at state banks measured as a share of the interest mar
gin show a similar pattern since 1988. This fall and subsequent rise in
loan loss reserves is just the opposite of the pattern of credit growth shown
in Table 2. Taken at face value it implies that banks, particularly State
banks, have experienced a deterioration in credit quality during the last
two years. So long as external factors have not dramatically affected the
conditions of borrowers between 1982 and now, this would imply that allo
cative efficiency has worsened, at least since 1990.
One striking feature of the figures on loan-loss reserves and provision
expenses is the sharp difference between the levels at State and private
banks. In 1991, State banks recorded almost four times the loan-loss reser
ves of private banks, and were placing almost half their (smaller) interest
margin into those reserves. Part of this discrepancy reflects sounder port
folios at private banks whose portfolio decisions were more often guided
by profitability. However, the level of reserves at private banks is barely
over the minimum of 1 per cent set by the Bank of Indonesia. For this
low level to be prudent, almost the entire private bank loan portfolio would
need to be sound! The low level of reserves, then, must also reflect attempts
by private banks to bolster profits by foregoing needed provision expenses.
Unfortunately the data do not allow us to compare the size of these unco
vered losses today versus the pre-reform period. One can simply note that
reserves are marginally higher today than in 1982, when the economy was
slowing down sharply, so that the problem may have diminished since
1982.
i;. Efficiency and Return on Investment at Firms
Harris, et al. (1992) tackle the question of allocative efficiency from a diffe
rent angle, focusing on whether the financial reforms led to a equalization
of the marginal efficiency of investment across different firms. This mea
sure of efficiency has shortcomings since it is an equilibrium concept, while
changes in returns in different sectors due to changing technology, prices,
etc. could alter returns unexpectedly. Using plant level data from a sample
of about 200 manufacturing firms they find that there was a more efficient
allocation of credit between 1984 and 1988; small and large firms, which
show the highest increases in the productivity of capital and a convergen
ce to the absolute levels of medium-sized firms, increased their access to
credit in the wake of the reforms. Unfortunately, the analysis by Harris,
et al. does not include post-1988 data. Since this is the period when banks
showed increasing loan-loss reserves, it will be important to engage in fur
ther analysis before forming a definitive opinion on the implications of
the financial reform for allocative efficiency.
vi. Risks from New Operations
Besides managing credit or default risk, the opening up of other areas of

financial intermediation to banks has added to the risks they face. Severe
exchange-trading losses by at least one Indonesian bank and commercial
real-estate losses at another demonstrate that opening up new areas to bank
operations can lead to inexperienced staff exposing the bank to excessive
risks. Coping with the risks and opportunities of the new environmentis
a major challenge for banks as well as for regulators10.
vii. Liquidity Risk
A more traditional component of risk to banks is the risk of becoming

temporarily illiquid. Illiquid banks jeopardize the role of the banking sys
tem in providing the means of payment for transactions in the economy
and the gains of a smoothly functioning payments mechanism. To lessen
this risk, a key element of the reforms has been aimed at strengthening
the interbank money market to provide individual banks with liquidity to
meet temporary shortfalls. Much progress has been made with the average
weekly volume of lending increasing eight-fold since 1986. Maturities have
lengthened to as long as one month, though the bulk of lending is still
overnight. Despite the deepening liquidity, though, rates in the interbank
10 See section C for a fuller discussion of the need for improved risk management and
supervision.
market remain volatile. With the tightening of monetary policy in mid-

1990, rates have ranged between 7 per cent and 60 per cent p. a., as major
suppliers of funds -the five State banks- have coped with the reduction
in liquidity credits and public-enterprise deposits.
The reduction of required reserve requirements to 2 per cent from
15 per cent in October 1988 had important implications for the liquidity
policies of commercial banks. At 2 per cent of deposits, required reserves
were in some cases lower than prudential liquidity management would
warrant. For this reason banks overall have held "excess'1 reserves since
1988, with the bulk of the excess accounted for by State banks. Despite
the tight money policy implemented in May 1990 (see discussion in section
C), banks have still maintained excess reserves, though a smaller percentage
of current liabilities. The tighter monetary policy, however, has led some
banks to access BI’s discount windows to maintain needed liquidity when
funds have become scarce in the interbank market.
Bank loan-to-deposit ratios also remained very high. Their level re
mained almost unchanged from 1982 to 1990 for most groups of banks.
Such high levels make it difficult for banks to adjust to temporary illi
quidity, in part because loans are not as marketable as other assets. Levels
are particularly high for the state commercial banks, complicating risk
management. For this reason BI in March 1991 explicitly incorporated li
quidity as an element of its quantitative evaluation of soundness, prompt
ing banks to improve their liquidity position.
Another way of looking at financial-system risks is to judge the mar
ket's view of subsequent reforms in the Indonesian financial system in terms
of the spread between US dollar deposit interest rates held on-shore and
those held off-shore. Since both assets are denominated in the same curren
cy, any difference in the rates would reflect transactions costs and, more
importantly, the market's perception of the riskiness of the Indonesian fin
ancial system11. This measure goes beyond simply judging allocation effi
ciency from looking at credit risk. It encompasses other risks such as inte
11 One might argue that foreign-exchange deposits are not sufficiently large in a domestic
financial system for their price to be a good reflection of their true value. In Indonesia these
deposits have made up 10-20 per cent of time deposits and 5-10 per cent of demand depo
sits since 1982.
rest-rate risk, maturity and liquidity risk, as well as the risk of managing
new operations12. Figures for 1982-1988 (Table 4) show that the first stage
of reforms lowered the spread, though the negative number in 1985 pro
bably reflects transaction costs and a relatively thin market. The spread
widened in 1986-1988 as the economy went through macroeconomic ad
justment. In the aftermath of the 1988 reforms, however, the spread again
fell. This decline was accompanied by large increases in capital inflow,
chiefly external borrowing. Nonetheless from 1989 on, the figures show a
steady rise in the difference between on-shore and off-shore dollar depo
sits. Thus, the market believes that risks have increased since 1989.
viii. Summary of Banking Efficiency and Risk
Summing up the evidence on efficiency and risk, the 1983 reforms have
certainly lowered intermediation margins and fostered credit expansion,
much of it going to profitable firms, including small firms that were pre
viously rationed. Increased lending to consumers, higher provisioning and
higher market perception of financial risk, however, point to a growing
risk in the financial system since 1988 as both the number of banks and
credit growth have surged.
G edit allocation is closely related to the quality and quantity of infor
mation available in the market and the options available to bankers should
a borrower default. Weak accounting standards and a scarcity of practicing
public accountants, lack of registration of collateral and strict banking se
crecy laws all limit the quantity and quality of information available. Ina
dequate bankruptcy laws limit the ability of banks to act against delinquent
debtors. Although steps have been taken to shore up these weaknesses,
their existence makes it easier to understand why financial reforms would
be less effective in improving the efficiency of credit allocation than in
increasing the quantity.
A strong supervisory agency can help to ensure a sounder allocation
of credit by limiting the actions of banks that engage in risky activities,
knowingly or not. As we have noted, Indonesia moved to improve pru-
12 The latter two issues are taken up below.

dential supervision as early as 1989. The most through-going reforms, ho

wever, were not announced until March 1991. Furthermore, improvements
in personnel as well as the regulatory framework are critical for effective
supervision. The knowledge needed by a bank examiner to effectively do
the job is only acquired over the course of five to ten years. It therefore
should not be surprising that in a era of rapid asset growth, with a great
inflex of new bankers, some poor credit decisions were made, thus increas
ing risk in the financial system.
2. Real Effects of Financial Reform
a) Savings
Sparking savings, particularly by the private sector, was one of the avo
wed goals of financial reform in Indonesia. In judging whether this goal
was achieved we first look at data from the Indonesian flow of funds and
then estimate a private consumption function for Indonesia, focusing on
the effects of financial savings on domestic savings.
The Indonesian flow of funds, available from 1984-89, provides infor
mation on savings and investment from various sectors of the economy,
including firms, households and the government (Table 11). The figures
on domestic savings rates show an 8.5 per cent increase, from 26.4 per
cent to 34.9 per cent of GDP, between 1984 and 198913. This is casual evi
dence of a positive correlation between savings and financial reform. Closer
examination of the figures shows that 7.7 per cent of the increase comes
from private firms which doubled their savings rates. None of the other
sectors show rates of increase anywhere near that size. Private firm savings,
however, include the residual left from adjusting the other sectors savings
rates to the aggregate level taken from the standard national accounts. This
number is probably overestimated for three reasons: first, because the size
of the change is not reflected in other sectors, particularly households; se
13 The figures on savings and investment used in this section rely on official estimates. Other
estimates, prepared by the World Bank, show a similar pattern of savings and investment
but at much lower shares in GDP (see Table 1 for the World Bank estimates).
T he I ndonesian experience wrm financial- sector reform 187
cond, because the aggregate rate of investment, 37 per cent, is one of the
highest in the world; and, third, because inventories make up nearly 10
per cent of the aggregate investment number.
Even if the aggregate domestic savings figures are correct, they only
show a coincidence between financial reforms and increases in savings. To
get a better understanding, we estimated a private consumption function
that controls for other effects. Private consumption was chosen rather than
private savings to reduce the impact of measurement errors on the esti
mation, while taking advantage of the accounting identity that links pri
vate disposable income, consumption and savings. Private, rather than gross
domestic savings, was estimated because of the independence of public sa
vings and financial reform14.
Table 11
STRUCTURE OF SECTORAL G R O SS SA V IN G, 1984-1989
(% of GDP)
1984 1985 1986 1987 1988 1989
Non-Finandal 25.38 25.44 23.65 26.81 30.33 33.18

Government 5.84 5.64 5.38 4.30 4.55 6.20
Firms 9.87 9.98 8.23 11.85 14.61 16.14
Public 2.18 1.96 1.49 2.07 1.41 0.71
Private 7.69 8.02 6.74 9.78 13.2 15.43
Households 9.67 9.82 10.04 10.66 11.17 10.84
Financial 1.06 1.49 0.85 3.03 0.58 1.68
Banks (a) 0.89 0.95 0.53 2.30 0.20 1.24
Non-banks 0.17 0.54 0.32 0.73 0.38 0.44
Foreign 1.04 2.31 4.96 3.10 2.22 2.13
Total 27.48 29.24 29.46 32.94 33.13 3.99
Memo Item:
GDP (billions of Rp.) 85,702 93,056 98,490 118,795 135,099 159,336
(a) Including th e Central Bank.

Source: Central Bureau o f Statistics, R o w o f Funds 1984-1989, p. 39.
14 One could argue that a substantial reform which increased demand for M l would increase
government seignorage and, ceteris paribus, increase public savings. As we have seen above,
M l/G D P did increase after 1988. Nonetheless, treating expenditures or tax revenues as
given is not realistic since they are both instruments of fiscal policy and therefore their
movements will reflect both policy changes and exogenous factors.
Before presenting the results we will briefly highlight the choice of

variables used in estimating private consumption by drawing on issues dis
cussed in the literature15. All work on consumption focuses on some measu
re of income as the chief determinant. Early empirical studies used current
income, later replaced by permanent income or wealth as theory began
to look at consumption in an intertemporal setting. Under the life-cycle
or permanent income approach, consumption was only increased in response
to permanent changes in income, while transitory increases were saved.
Current income continued to play a role, however, if liquidity constraints
inhibited consumers from maintaining consumption levels consistent with
their permanent income.
The role of real interest rates in private consumption is another conten
tious issue. In theory the sign of the impact is indeterminate a priori since
it depends on which predominates: the substitution effect (consume less
today to have more tomorrow when higher interest has increased your
wealth), or the income effect (increase consumption today and tomorrow
because the higher interest rate has increased your wealth). Schmidt-Hebbel
and Webb (1992) note that changes in real interest rates can cause shifts
between various assets -out of physical assets and into financial assets- which
could affect measurement of private savings, because of difficulties in treat
ing the purchase of consumer durables, for example, but without affect
ing the overall level of private savings (see discussion below). Empirical
results differ as to the sign of the real interest rate, but the effect is usually
quite small (Fry, 1988).
Ricardians include public savings as a determinant of private consump
tion, arguing that expectations of future taxes lead consumers to reduce
spending one-for-one with public dissaving (Barro, 1974). Some studies have
included inflation as an explanatory variable though its sign is indetermi
nate a priori. Inflation increases uncertainty about the value of future earn
ings and could potentially reduce savings because of the lower return, or
increase them for precautionary motives. Foreign savings are also assumed
to affect private savings, usually with the argument that access to foreign
15 For reviews of savings in developing countries see Schmidt-Hebbel and Webb (1992), Deaton
(1989) and Fry (1988).
savings reduces domestic savings since the former do not solely finance
additional investment.
Finally and most importantly for this paper, the amount of financial
wealth has been posited as a determinant of private savings. The argument
is that greater levels of financial wealth imply fewer liquidity constraints
on borrowers and therefore higher consumption levels. Note that this pre
mise runs opposite to the usual argument that financial reform, by increas
ing interest rates, will increase savings.
Using the independent variables just discussed -permanent income,
liquidity, foreign and public savings, the real interest rate and inflation-
we estimated a private consumption function using annual data from 1970
through 1991. The data were taken from a database maintained by BI. All
variables were logs of current values, except for the real interest rate and
foreign savings, the latter appearing as a share of private disposable in
come. Permanent income was calculated as a three-year moving average
of private disposable income (Schmidt-Hebbel and Webb, 1992). Public sa
vings were taken from national accounts and exclude public enterprises.
The real interest rate was calculated using the three-month average time
deposit rate and actual consumer price inflation for the year in question.
The results of the least squares regression of private consumption on
the five variables and a constant shown in Table 12 show an excellent
fit and no sign of first-order autocorrelation. Since several variables were
not stationary, an Engle-Granger test for unit roots in the estimated equa
tion was done using the first difference of private consumption. As expected,
the sign of logged private consumption was negative. However, due to
a fairly small sample size the non-stationary hypothesis cannot be rejec
ted. There is a large and significant correlation between private consumption
and permanent income, with an elasticity of .93, very dose to the theoretical
life-cyde value of one. Government savings come in with a unexpectedly
positive and significant sign. This could be the result of the high correlation
between oil prices and government savings, meaning that higher savings
did not result from increased taxes with their contractionary effect on pri
vate consumption, but rather from additional income from abroad. This
conjecture is consistent with the negative, though insignificant, coeffident
on the foreign saving rate. The real interest rate and inflation have negative
and insignificant coeffidents.
190 D o n a ld P . H anna
Table 12
DETERMINANTS OF PRIVATE CONSUMPTION, 1970-1991
Variable Coefficient STD. Error T-Stat. 2-Tail Sign.
Constant 0.403 0.127 3.181 0.006

Permanent Income 0.935 0.026 36.641 0.000
Real Interest Rate -0.001 0.002 -0.455 0.656
Inflation -0.001 0.002 -0.840 0.414
Liquidity -0.026 0.012 -2.103 0.053
Foreign Saving -0.254 0.143 -1.782 0.095
Government Saving 0.031 0.015 2.137 0.049
R-squared 0.9998 Mean of dependent var 10.028

Adjusted R-squared 0.9998 S.D. of dependent var 1.243
S.E. of regression 0.017 Sum of squared resid 0.005
Durbin-Watson stat 2.003 F-statistic 17,758.000
Log likelihood 62.066 ADF t statistic -3.830
Note: Coefficients estimated using OLS.
Source: Author's calculations with data from bi.
Rerunning the equation without the inflation and interest-rate varia

bles further confirmed the above results. The permanent income coefficient
increased to .95. Public savings were still positively correlated with con
sumption, though the significance level rose to 68 per cent and foreign
savings remained negatively corrected with private consumption.
From the standpoint of financial reform, the most interesting finding
is the negative effect of liquidity (M2) on private consumption, which im
plies that increases in financial savings increase domestic savings. Since the
permanent income coefficient is close to one, the two results indicate that
liquidity constraints on private consumption were not a significant factor.
The question remains as to why financial savings and domestic savings
would be positively correlated when real interest rates are not statistically
significant. A possible explanation is that increases in financial savings are
associated with shifts in domestic savings away from hoarding goods. If
this is associated with higher measured domestic savings, because of sta
tistical problems in the treatment of durable goods, a positive correlation
will result. Properly specified domestic savings and financial savings will
therefore not be positively correlated. Still, even if this is the explanation,
if greater financial intermediation allows more efficient, cheaper investment,
there will still be benefits to higher financial savings.
T h e I n d o n e s ia n e x p e r ie n c e w it h f in a n c ia l - s e c t o r r e f o r m 191
b) Investment
Increasing the quantity and efficiency of investment was a primary aim

of the Indonesian reform efforts. As we have just seen, the reforms are
correlated with higher domestic savings, which in turn boosts investment.
Harris, et al. (1992) present evidence that the marginal efficiency of in
vestment improved between 1984-1988, though, as noted earlier, the mea
sure they use -equalization of rates of return across sectors- has flaws. They
go on to estimate an investment equation as a function of sales, cash flow
and leverage. Cash flow has been included to capture the effects of liqui
dity constraints, while leverage has been included to test whether higher
levels reduced credit access and depressed investment. Their results show
a significant reduction in the importance of cash flow to small firms follow
ing the 1983 deregulation and a reduction in overall market segmentation.
The discussion in section IV A. of efficiency of credit allocation and hence
investment, however, means that these improvements did not necessarily
persist into the second stage of financial reform when entry barriers were
lowered, though the effect on the quantity of investment has clearly per
sisted.
Indonesian financial reform involved freeing interest rates and led,
ex post, to high real rates. This could be construed as retarding investment.
Such an argument would be dubious for two reasons. First, from a theore
tical standpoint, firms base investment decisions on ex ante rates, not ex
post ones. If expectations of inflation consistently outpaced actual inflation,
ex post rates would overstate ex ante ones. There is evidence that this
has been the case in Indonesia. Table 4 shows the spread between on
shore dollar and on-shore rupiah deposit rates. The spread has consistently
overstated the actual depreciation of the rupiah (with exception of the two
maxi-devaluations). With fairly open goods markets linking domestic and
international prices and an explicit exchange rate policy rule of maintain
ing the real exchange rate (at least in the post-1986 period), this overesti
mation on devaluation can be linked to an overestimation of internal infla
tion. Second, from an empirical standpoint, the work of Harris, et al. (1992)
shows that measured returns on capital in manufacturing far outstripped
the 10-15 per cent ex post lending interest rates prevailing since financial
reforms began. Given this fact, the increased access to credit that the fi-
192 D o n a ld P . H a n n a
nanáal reforms stimulated were probably much more important in spark

ing investment than the somewhat higher measured price. It should be
recalled that prior to the 1983 reforms credit was rationed so that for many
firms, the price was effectively infinite.
3. Macroeconomic Conditions and Financial Reform
Most prescriptions for financial reform call for stable macroeconomic con
ditions on the eve of financial reform (e.g. Caprio and Atiyàs, 1992). Prefe
rence is generally given to opening the financial system after the current
account and before the capital account (Hanson, 1992), and the references
dted therein). The Indonesian experience with finandal reform shows that
within limits neither a stable economy nor a dosed capital account is ne
cessary for successful finandal reform. As was discussed in section II, In
donesia in 1983 was in the midst of a serious stabilization program to
adjust to the effects of deteriorating external conditions. Inflation, though
relatively low at just over 10 per cent, was accelerating. A yawning current
account defidt of 7.4 per cent of GDP needed to be financed. The capital
account had been open for over a decade, while a web of trade and invest
ment regulations restricted the current account. Nonetheless, the freeing of
interest rates and elimination of credit ceilings successfully boosted finan
cial and domestic savings.
Rather than condudve external conditions, the Indonesia experience
spotlights the need for macroeconomic management that recognizes and
adjusts to the constraints imposed by the external environment and the
government's own budget limitations. Adjustment to the external environ
ment was the thrust of the 1982-83 stabilization program with its major
devaluation and extensive fiscal adjustment. The fiscal balance moved from
a deficit of 24 per cent of GDP in 1983 to a surplus of 1.4 per cent. Deva
luation provided the real depreciation needed to reach external balance,
while fiscal adjustment made the real depreciation stick. In these efforts,
the open capital account and the government's commitment to avoid fi
nancing the fiscal deficit from domestic sources played a critical role in
conditioning the response of the economy. Commitment to the open capi
tal account with a quasi-fixed exchange rate meant that attempts to soften
the external blow by loose monetary policy would lead to capital outflows
T h e I n d o n e s ia n e x p e r ie n c e w it h f in a n c ia l - s e c t o r r e fo r m 193
and a quick reversal of policy. Expansionary fiscal policies were limited

by access to official external finance, itself difficult to attract without a cre
dible adjustment program. As a result, in the presence of lower oil prices
expansionary policies were severely restricted. By releasing interest rates
in this environment, there was little risk of currency substitution. Weak
domestic demand also meant that releasing credit ceilings was unlikely to
lead to excessive increases in domestic credit.
The importance of macroeconomic management, the role of the capi
tal account and the absence of domestic debt finance is further reinforced
by Indonesia's experience with the second stage of its reforms in 1988. By
that year Indonesia had already applied a second round of adjustments
in view of the sharp decline of oil revenues in 1986. Trade and investment
restrictions had been relaxed. The economy was poised for growth a much
more propitious time for financial reform. Nonetheless, the 1988 reforms,
though successful in spurring further improvements in cost efficiency, com
petition and growth in the financial sector, also led to a growing problem
of poor credits. This was the result of a combination of circunstances that
macroeconomic policy failed to adjust to initially. Unlike in 1983, trade re
forms opened the door to new profit opportunities within the Indonesian
economy, spurring a rise in investment demand. Banks moved aggressively
to finance these investment plans, spurred by competition from new en
trants. Thus a private sector imbalance between savings and investment
developed against which the prohibition against domestic deficit finance
did not protect. This rise in demand was initially validated by the govern
ment's attempts to lower interest rates through a combination of loose m o
netary policy and tightening in late 1989 and early 1990. Only when the
open capital account led to a US$1 billion loss in reserves in April 1990
did the government move to tighten monetary policy. This immediately
dealt with the interest-rate problem, by raising rates to levels consistent
with international parity. The private-sector savings-investment gap, howe
ver, was little affected by the adjustment in monetary policy. Only in 1991,
when fiscal policy tightened, was internal balance restored. In the interim
a significant amount of new credit had been extended by banks, credits
that became more difficult to service in 1991 as interest rates remained
high and activity slowed somewhat.
194 D on a ld P . H anna
IV. CONCLUSIONS
This paper began by looking at the theoretical links between financial

growth, savings and economic growth. Discussion focused on four impor
tant aspects of the financial system and growth:
• the depth and breadth of financial savings;
• the ability of the financial system to alter maturities to suit the
needs of investors;
• the efficiency of the financial system, both in terms of cost and
in allocation of credit;
• the ability of the financial system to deal with informational asym
metries and incentive problems.
To promote these aspects of the financial system prescriptions for fi
nancial reform have frequently called for reform during periods of macro-
economic stability and under the protection of a closed capital account.
What do phased reforms in Indonesia -the removal of ceilings on interest
rates and credit expansion, then the lowering of barriers to entry, M o w e d
closely by a reduction of the direct role of the central bank in favour of
a stronger supervisory presence- show about this set of prescriptions/effects?
The reforms have led to a large and sustained increase in financial
depth and breadth in the economy, with the M2 to GDP ratio rising from
18 per cent in 1982 to over 44 per cent in 1991. What is more, econometric
work on private savings in Indonesia shows that this increase in financial
savings has not hampered domestic saving independent of any effect from
higher incomes, lower inflation or the real interest rate. The maturity of
bank loans has also been extended over the period of the reforms, better
meshing with the investment needs of the economy. Cost efficiency has
improved with a reduction in overhead. Spreads between deposit and lend
ing rates have come down at all banks and have moved closer to one
another at different types of banks.
In the area of credit allocation and financial-system risk there is less
certainty about the positive benefits of the reforms, particularly the lower
ing of entry barriers in 1988. Evidence from manufacturing firms shows
improved access to credit and a movement toward equalization of returns
on investment across firms. Data on loan losses, the broader allocation of
credit outside of manufacturing and an empirical estimate of tinandal-system
risk all point to increasing risk since 1988. Given the difficulties in allo
cating credit because of information and incentive problems, it is not
surprising that efficiency gains are more elusive. These information and
incentive problems highlight the need for improvements in the legal ac
counting and prudential systems that support efficient financial systems.
Indonesia's reform efforts were generally successful despite several
factors that differed from standard prescriptions. First, Indonesia's capital
account was open at the time of reforms; second, reform began in a time
of serious macroeconomic adjustment. Indonesia's experience highlights the
need for macroeconomic management that recognizes the limitations im
posed by the external budget constraint and that of the government rather
than macroeconomic stability per se, and responds quickly to those cons
traints. Indeed the open capital account aided macroeconomic management
by providing quick feed-back when domestic policies moved out of line
and limiting the inflationary consequences of excessive demand stimulus.
The absence of domestic finance for fiscal deficits removed an important
source of pressure on domestic financial markets. Quick response to exter
nal imbalance, backed by strong fiscal adjustment, aided the successful re
forms in 1983. Overly stimulatory monetary policy, and a private-sector
savings-investment gap which was not sufficiently offset by fiscal policy,
exacerbated the credit expansion of 1989-90. Furthermore, the slow respon
se of fiscal policy put more of the burden of adjustment on monetary po
licy, keeping interest rates high and thereby prolonging the problem of
non-performing loans within the banking system. Adroit macroeconomic
management can thus promote successful financial reform even in periods
of macroeconomic adjustment with a economy open to speculative capital
flows.
196 D on a ld P . H a n n a
BIBLIOGRAPHY
Ahmed, Sadiq and Basant K. Kapur (1990), How Indonesia's Monetary Po
licy Affects Key Variables, Policy, Research and External Affairs Papers,
WPS 349, World Bank.
Balino, Tomas J.T. and V. Sundararajan (1986). "Financial Reform in Indone
sia" in H.S. Cheng, Financial Policy and Reform in Pacific Basin Coun
tries, Singapore: Oxford University Press.
Barro, Robert (1988), "Are Government Bonds Net Wealth?", Journal of Po
litical Economy, Vol. 82.
Bemanke, Ben S. and Mark Gertler (1986). Agency Costs, Collateral and Bu
siness Fluctuations, National Bureau of Economic Research Working Pa
per #2015, September.
Caprio, Gerard (1992), Policy Uncertainty, Information Asymmetries, and Fi
nancial Intermediation, Country Economics Department, World Bank, Po
licy Research Working Paper, WPS 853.
Caprio, Gerard and Izak Atiyas (1992), "Policy Issues in Reforming Finance:
Lessons and Strategies," The Economic Development Institute (EDI)
World Bank.
Caprio, Gerard et al. (1990), "Research Proposal: The Impact of Financial
Reform", Financial Policy and Systems Division, Country Economics
Department, World Bank.
Center for Strategic and International Studies (1990), Survey of Recent De
velopments, Jakarta: CSIS, September.
Chant, John F. and Mari Pangestu (1992). "An Assessment o f Financial Re
form in Indonesia: 1983-1990," World Bank, Washington.
(1988), 'The Analysis of Efficiency of Indonesia Banking", mimeo.
Cheng, Hang-Sheng (1986), Financial Policy and Reform in Pacific Basin Coun
tries, Lexington, Lexington Books, Mass.
Cole, David C. and Betty F. Slade (1990), "Financial Development in Indo
nesia," Economic and Development Institute, W orld Bank.
de Juan, Aristóbulo (1987), "From Good Bankers to Bad Bankers: Ineffec
tive Supervision and Management Deterioration as Major elements in
Banking Crises," Country Economic Department, World Bank.
Deaton, Angus (1989), "Saving in Developing Countries: Theory and Re
view", Proceedings of the World Bank Annual Conference on Development

Economics, Washington, D.C.: World Bank.
Duca, John V. and Mary M. McLaughlin (1990), "Developments Affecting
Federal Reserve Bulletin, July.
the Profitability of Commercial Banks",
Edwards, Sebastian (1984), The Order of Liberalization of the External Sector
in Developing Countries, Princeton Essays in International Finance, N2
156.
Fama, Eugene (1985), "What's Different about Banks?," journal of Monetary
Economics, Vol. 15, January.
Fry, Maxwell J. (1988), Money, Interest and Banking, The Johns Hopkins
Press, Baltimore.
Gelb, Alan H. (1989), Financial Policies, Growth, and Efficiency, Country Eco
nomics Development, World Bank Working Paper, WPS 202.
Gertler, Mark and Andrew Rose (1991), Finance, Growth and Public Policy,
World Bank Working Paper, WPS 814.
Giovanini, Alberto (1983), "The Elasticity of Savings in Developing Coun
tries: The Existing Evidence," World Development, Vol. 117, July.
(1985), "Saving and the Real Interest Rate in LDCs", Journal of De
velopment Economics, Vol. 18: 2-3.
Hanson, James A. (1992), Opening the Capital Account: A Survey of Issues
and Results, Policy Research Working Papers, WPS 901, World Bank,
May.
Hanson, James A. and Roberto Rocha (1986), High Interest Rates, Spreads
and the Costs of Intermediation, World Bank, Industry and Finance Se
ries, #18.
Harris, John R., Fabio Schiantarelli and Miranda G. Siregar (1992). 'The
Effect of Financial Liberization on Firm's Capital Structure and Invest
ment Decisions: Evidence from a Panel of Indonesian Manufacturing
Establishments, 1981-1988", Boston University.
Haubrich, Joseph G. (1989), "Financial Intermediation Delegated Monitor
ing and Long-Term Relationships", Journal of Banking and Finance, Vol.
13.
Hill, Hal (1990), "Indonesia's Industrial Transformation, Part 1", Bulletin of
Indonesian Economic Studies, Vol. 26: 2.
(1990), "Indonesia's Industrial Transformation, Part 2", Bulletin of In
donesian Economic Studies, Vol. 26: 2.
198 D o n a ld P . H anna
Khan, Moshin and Roberto Zahler (1983), "The Macroeconomic Effects of

Oranges in Barriers to Trade and Capital Hows: A Simulation Anal
ysis", IMF Staff Papers.
King, Robert G., and Ross Levine (1992), Finandal Indicators and Growth
in a Cross Sedion of Countries, Country Economics Development, World
Bank Working Paper, WPS 819.
Krueger, Anne (1981), "Interactions Between Inflation and Trade-Regime
Objectives in Stabilization Programs", in W. Cline and S. Weintraub,
eds., Economic Stabilization in Developing Countries, Washington: Broo
kings.
Little, Ian, Tibor Skitovsky and Maurice Scott (1970), Industry and Trade
in Some Developing Countries: A Comparative Study, Oxford: Oxford Uni
versity Press.
Myths of the West: Lessons from Developed Countries for
Mayer, Colin (1980),
Development Finance, World Bank Working Paper, WPS 301.
McKinnon, Ronald I (1973), M oney and Capital in Economic Development,
Washington, D.C.: Brookings Institution.
— (1982), "The Order of Economic Liberalization: Lessons from Chile and
Argentina", in K. Brunner and A. Meltzer, eds., Economic Policy in
a World of Change, Carnegie Rochester Conference Series, Amsterdam:
North Holland.
Michaely, Michael (1986), "The Timing and Sequencing for Trade Libera
lization Policy", in A. Choksi and D. Papageorgiou, eds., Economic
Liberalization in Developing Countries, 41-59, London: Basil Blackwell.
Nasution, Anwar (1983), Financial Institutions and Policies in Indonesia, Sin
gapore: Institute of Southeast Asian Studies.
Neal, Craig R (1990), Macro-Financial Indicators for One Hundred Seventeen
Developing and Industrial Countries, Policy, Research, and External Affairs,
Development Economics, World Bank Working Paper, WPS58.
Sachs, Jeffrey D. and Susan M. Collins (1989), Developing Country Debt. Vol.
3: Country Studies - Indonesia, Philippines, Turkey, Chicago, University
o f Chicago Press.
Schmidt-Hebbel, Klaus and Stephen B. Webb. "Public Policy and Private
Saving", in V. Corbo, S. Fischer and S. Webb, eds., Adjustment Lending
Revisited: Public Policies to Restore Growth, Washington, D.C.: World
Bank.
Stiglitz, Joseph E. and Andrew Weiss (1981), "Credit Rationing in Markets

with Imperfect Information," American Economic Review 71, June.
Sundararajan, V. and Tomas J. T. Balino (1988), "Issues in Recent Banking
Crises in Developing Countries," M F Working Paper.
Taylor, Lance (1983), Structuralist Macroeconomics, New York: Basic Books.
Tobin, James (1984), "On the Efficiency of the Financial System," Lloyds Bank
Review , Vol. 153, July.
W oo, Wing Thye (1991), "Strengthening Economic Management in Indo
nesia," mimeo.
W oo, Wing Thye and Anwar Nasution (1989), "Indonesia Economic Policies
and their Relation to External Debt Management," in J. Sachs and S.
Collins, eds., Developing Country Debt, Vol. 3, Chicago: University of
Chicago Press.
World Bank (1989), World Development Report, Oxford, Oxford University
Press.
Vittas, Dimitri (1991), Measuring Commercial Bank Efficiency, Use and Misuse
of Bank Operating Ratios, Country Economics Department, World Bank
Working Paper, WPS 806.
PENSION SYSTEM REFORMS AND SAVINGS IN
LATIN AMERICAN AND CARIBBEAN COUNTRIES
WITH SPECIAL REFERENCE TO CHILE1
Andras W. U th ofP
1 This paper is based on national case studies carried out under ECLAC/UNDP projects
RLA/90/001 and RLA/192/003, "Finandal Policies for Development". These case studies
have been published by ECLAC in separate edited volumes (Uthoff and Szalachman, 1991
and 1993 (forthcoming), and Iglesias and Acuña, 1991).
2 The author wishes to thank Gunther Held for his comments on an earlier draft of this paper.
The author has also benefited from comments received at different ECLAC seminars.
3 The views expressed in this study, which has been reproduced without formal editing by
ECLAC, are those of its author and do not necessarily reflect the views of the organizations.
CONTENTS
INTRODUCTION ........................................................................................................................... 205
I. CURRENT STATUS AND CHALLENGES OF PENSION SYSTEMS ......................... 206
II. POPULATION, EMPLOYMENT AND

RETURNS TO PENSION SYSTEM FUN D S..................................................................... 215
III. DEVELOPMENT OF CAPITAL MARKETS AND

RETURNS ON PENSION SYSTEM FUNDS .................................................................. 219
1. Public sector reform s................................................................................................... 221
2. Banking regulation and supervision......................................................................... 222
3. Security and insurance market regulation and supervision ................................. 223
IV. THE CHILEAN EXPERIENCE WITH AN

INDIVIDUAL CAPITALIZATION SCHEME ................................................................ 224
1. Main characteristics of the new system ................................................................... 224
2. Prudential regulation of pension fund investments............................................... 225
3. Pension system results and financial market developments................................. 226
a) Pension system population coverage .................................................................. 227
b) Pension benefits, now and then ............................................................................ 229
c) Pension Funds and Financial market developments ........................................ 230
d) Pension Funds, savings and investment.............................................................. 234
4. Other problems pending............................................................................................. 235
a) Challenges related to financial markets .............................................................. 236
b) Other challenges...................................................................................................... 236
V. CONCLUSIONS................................................................................................................... 237
P e n s io n s y s t e m r e f o r m s a n d s a v in g s in L a t in A m e r ic a n a n d . 205
INTRODUCTION
Pension systems in Latin American and Caribbean countries have been

developed since early this century on the basis of compulsory contributions
by workers. The systems have alternative built-in mechanisms to transform
these contributions into savings for old-age security and other contingency
income benefits. Thus the performance of the system can be evaluated in
terms of the implicit returns estimated from relationships between future
benefits and current contributions. Such evaluations show that the systems
are not providing significant returns to contributors4.
In addition, the systems also face operational problems and most
governments are seriously assessing and considering feasible reform alter
natives. This is due to the unsustainability of their current financial and
actuarial deficits, especially in light of their effect on public-sector account
balances, and hence on macroeconomic policy management. This last issue
is becoming increasingly important in evaluating pension-system reform in
the region.
4 Uthoff and Szalachman 1991, Sistemas de Pensiones en America Latina. Diagnóstico y

alternativas de reforma. Costa Rica, Ecuador, Uruguay, Venezuela. CEPAL, Proyecto Regio
nal Políticas Financieras para el Desarrollo. Santiago de Chile. Uthoff and Szalachmann,
1993, Sistemas de Pensiones en América Latina. Diagnóstico y alternativas de Reforma.
Volumen 2. Bolivia, Brazil, Guatemala, México, Colombia. CEPAL, Proyecto Regional
Políticas Financieras para el Desarrollo. Santiago de Chile. Iglesias and Acuña, 1991. Sistemas
de Pensiones en América Latina. Chile: Experiencia con un Régimen de Capitalización
1981-1991. CEPAL. Proyecto Políticas Financieras para el Desarrollo. Santiago de Chile.
206 A n d r a s W. U t h o f f
Although most systems were created as mechanisms for accumulating

savings to finance pension benefits, the systems have been merged with
other components of social security and their funds used for redistributive
purposes by financing the increase in both population and other social se
curity contingencies coverage. Today, the alternative effects of the system
design on savings and the development o f capital markets are also being
considered.
The objectives of this paper are four: to briefly describe ECLAC's find
ings from case studies on the evaluation and reform alternatives to pension
systems in Latin America and the Caribbean (ECLAC5, 1992); to discuss
factors that define the return on alternative pension systems financing sche
mes; to indicate institutional and market factors needed to improve pen
sion system performance; and to assess, in light of the above, the last ten
years of Chile's experience with a large-scale and unique system based on
individual capitalization of contributions.
I. CURRENT STATUS AN D CHALLENGES OF PENSION SYSTEMS
Previous research by ECLAC indicates that it is not enough for financial

policy to rely exclusively on the response of voluntary savings to market
forces. The evidence shows that the response of personal savings to interest
rates is poor, reflecting a strong correlation between the substitution effect
(higher interest rates raises the cost of present versus future consumption)
and the income effect (with higher interest rates the same amount of sa
vings allows for greater purchasing power). In turn, private firms are net
borrowers and thus higher interest rates negatively affect their savings. The
studies undertaken conclude that there is a need to develop policy instru
ments to promote compulsory saving but within the context of regulated
and supervised institutions. The role of institutional savings through pen
5 See ECLAC, 1992, "Pension Systems in Latin America: Diagnosis and Reform Alternatives".
Case Studies in Latin America and the Caribbean. Conclusions and Recommendations.
Santiago, 3-5 December 1990 and 22-23 August 1991. ECLAC, Santiago de Chile, September
1992.
P e n sio n sy s t e m r e f o r m s a n d s a v in g s in L a t in A m e r ic a n and. 207
sion schemes is of considerable importance here (Massad and Eyzaguirre,

1990)6.
Currently, alternative mechanisms involve the financing of pension
schemes to overcome a short-sighted attitude among people towards their
own voluntary saving for old-age security purposes. These alternative pen
sion schemes have been developed within a continuous process and/or
combinations of two extreme cases. One is the "pay-as-you-go" system, whe
re an intergenerational solidarity mechanism operates, through which those per
sons now in the labour force are obliged to contribute to the financing
of defined pension benefits for those presently retired an d/or to the financ
ing of reserve funds for future contingencies. The other is an "individual
capitalization" system where an intragenerational solidarity mechanism opera
tes, by which each person's compulsory contribution is saved and invested
for future pension benefits in an individual account.
Alternative systems include the following: a mixture of a basic social
pay-as-you-go scheme, complemented by voluntary private capitalization
alternatives (e.g., Brazil); or a social pay-as-you-go scheme with capitaliza
tion of reserve funds (like most countries in the region). There are several
implications of these multiple alternatives, such as the difficulties in enfor
cing social-security legislation; the variety of programme benefits being im
plemented; the need for periodical revision of contribution rates and their
relation to defined benefits; the regulation of the pension system market
and its enforcement; the need for regulating the portfolio investment of
social pension systems reserve funds and their enforcement, and so forth.
Perhaps what best summarizes the challenges of pension systems is
their need to provide relatively high returns on contributions when the
latter are considered as savings for old-age security purposes. Table 1 illus
trates this point. Based on a hypothetical exercise, the table reflects the
amount of savings accumulated by a contributor to the system had h e /
she experienced a flat life-eamings profile equivalent to 100 pesos in real
terms, contributing with 10 per cent of net earnings to the system every
6 See Massad, C. y Eyzaguirre, N. (eds.) 1990, Ahorro y formación de capital. Experiencias

latinoamericanas. Argentina, Brazil, Chile, El Salvador México. CEPAL/PNUD Proyecto
Regional Conjunto "Políticas Financieras para el desarrollo". Grupo Editor Latinoamerica
no. CEPAL, Santiago, Chile.
month for the number of years given in each column (30, 35, 40 and 45),
with contributions capitalized at the real rates of returns given in each
row (-5, -3, 0, 3, 5).
The amount of savings obtained after this person retires (Part 1 of
the Table), and its equivalent in pensionable years with a pension equal
to constant real salary (100 pesos) (part 2 of the Table), indicates that un
less real capitalization rates are set at around 5 per cent, the system will
provide pension benefits far below the average real level of earnings (given
the 10 per cent contribution rate). Thus, with low capitalization rates the
system would need higher contribution rates to achieve the goal of pension
benefits equal to the earnings of the last working years, an d /or the exten
sion of the working lifespan (increased retirement age). Quite to the con
trary, higher capitalization rates would allow for better pension benefits
a n d /or extension of life expectancy at the age of retirement (earlier reti
rement), a n d /or lower contribution rates.
Returns on systems based on capitalization schemes depend on the
composition of the investment portfolio of the funds and capital-market
performance. In turn, returns on simple pay-as-you-go schemes are set equal
to the rate of change of real wages plus that of the number of contribu
tors minus the rate of growth of life expectancy at the age of retire
ment. Systems combining both schemes have a return which results from
a weighted average of the two above.
Table 1
SIMULATION OF INDIVIDUAL PENSION BENEFITS:
IMPORTANCE OF CAPITALIZATION IN DEALING
WITH SHORT-SIGHTED ATTITUDESf*)
1. Amount Accumulated at Age of Retirement according to

Years of Service and Capitalization Rate_______________
Years
Rate 30 35 40 45
0% 3,600 4,200 4,800 5,400
3% 5,842 7,434 9,284 11,432
5% 8,357 11,408 15,324 20,349
-5% 1,858 1,976 2,068 2,139
-3% 2,370 2,596 2,790 2,957
P e n s io n sy s t e m r e f o r m s a n d sa v in g s in L a t in A m e r ic a n a n d . 209
Table 1 (Conl.)
2. Number of Years in which Participant Can Receive a Pen

sion Equal to 100 % of monthly Earnings according to
Years of Service and Capitaliztion Rate________________
Years
Rate 30 35 40 45
0% 3 3.5 4.0 4.5
3% 5.3 6.9 8.8 11.3
5% 8.6 12.9 20.3 37.3
-5% 1.5 1.7 1.7 1.8
-3% 2.0 2.2 2.3 2.4
(*) Calculations are for an individual earning 100 pesos a m onth in
real term s throughout his complete working life and with 10
per cent contribution tax.
Despite the original design, pension systems have been merged with
other social-security contingency systems (e.g., job-related accidents and
professional sickness, non-job-related accidents and non-professional health
and maternity care and leave for the working mother, maternity and health
care for dependents, dependent allowances, national health systems or pu
blic health programmes). It has been very common practice, on the basis
of arguments in favour of redistributing income by creating a social-securi
ty net, to use pension system reserve funds to expand the population and
contingency coverage under social security. This practice has been mentio
ned as one of the main factors contributing to the financial and actuarial
deficits of the pension system. Use of reserve funds to cover temporary
fiscal deficits, which are not repaid by the State, and use of funds for
social-security purposes other than defined pensions, have been justified
on the grounds of improving equity and the result of joint management
of all social security schemes (Uthoff and Szalachman, 1991 and 1993)7.
7 See for details, Uthoff, Andras and Raquel Szalachman 1991 Sistemas de Pensiones en
America Latina. Diagnóstico y alternativas de reforma. Costa Rica, Ecuador, Uruguay,
Venezuela. CEPAL, Proyecto Regional Políticas Financieras para el Desarrollo. Santiago de
Chile, and Uthoff Andras and Raquel Szalachman 1993 Sistemas de Pensiones en América
Latina. Diagnóstico y alternativas de reforma. Volumen 2. Bolivia, Brazil, Guatemala,
México, Colombia. CEPAL, Proyecto Regional Políticas Financieras para el Desarrollo.
Santiago de Chile.
Once those systems started facing difficulties in providing reasona

ble pension benefits and in achieving financial and actuarial equilibria, the
trend in Latin America has been to develop them into a compulsory, pu
blic-sector-managed, simple "pay-as-you-go system". This trend is not a
definitive solution. Aside from the problems pointed out in the previous
paragraph, a simple pay-as-you-go system is subject to additional sources
of financial disequilibria: rapid changes in the demographic structure of
the population covered by the system; effects of the economic cycle on
the labour market, and therefore on potential contributors; evasion and
corruption; poor returns on the investment portfolio; under-declaration of
earnings while accumulating seniority in the system; and defined benefits
which bear no relation to changes in contributions and returns on invest
ment.
Table 2 summarizes the main objectives, problems and alternative ad
justments to improve current pay-as-you-go systems operating in the re
gion. It can be noted that no simple solution is at hand. These systems
require quite sophisticated adjustments involving an overall reform not on
ly of social-security schemes, but also on macroeconomic and financial
management grounds.
Based on such an assessment, pension system reforms should also
consider their implications vis-à-vis capital market developments and sa
vings investment processes. The region needs to increase its investment
rates from 16 to 22 per cent to grow at 5 per cent and improve the per
formance of regional economies in the world economy. Such investment
rates should allow for technological improvements, to redistribute income
through human capital investment opportunities, and to make development
enviromentally sustainable. In the absence of significant and permanent ca
pital inflows from abroad, domestic savings both public and private should
be increased. Institutional savings may promote these efforts by contribu
ting both directly and indirectly to savings (CEPAL, 1990)8.
8 See CEPAL, 1990 Transformación productiva con equidad. La tarea prioritaria del desa
rrollo de América Latina y el Caribe en la década del noventa. Naciones Unidas, Santiago
de Chile.
P e n s io n s y s t e m r e f o r m s a n d s a v in g s in L a t in A m e r ic a n and. 211
Table 2
PROBLEMS AND POSSIBLE REFORMS IN PAY-AS-YOU-GO SYSTEM
Objective Particular problem Proposal Instruments
Productive employ Shortage of producti Reorientation of eco Investment in human

ment ve employment. nomic policy towards resources, savings, fac
generation of produc tor productivity, ex
tive employment. panding existing mar
kets for domestic
goods, technological
development.
To improve collection Evasion and/or delays To create incentives Creation of unique list
of funds in payment of social for punctual payment of contributors and
security taxes, debt of social security con dependen ts;creation of
from the state as an tribution taxes, to crea individual accounts;
employer, lack of con te mechanisms to ne inspection of payments
tributions from inde gotiate employer debt. and indexed instru
pendent workers. ments to charge real
interestrates to debtors
to the system; indexed
government bonds;
simplification of pro
cedures.
Investment of reserve Low returns on re Regulation and super Creation of regulation
funds serve fund investment vision of investment and supervisory insti
portfolio; use of funds portfolio; development tution; risk classifica
for other government of capital markets; se tion commission; fi
executed programmes; paration of govern nancial interm edia
lack of indexed ins ment programme ad tion; periodic and pu
truments to cope with ministration. blic balances of social
inflation. security accounts; se
paration of fiscal ac
counts by social secu
rity and other govern
ment programmes.
Solidarity and wider Systems cover only a Development of mixed Use of funds in indivi
population coverage. small share of the la systems with a basic dual accounts as co
bour force and its de minimum pension and llateral for access to
pendents; there are supplements based on personal credits; stan
large population individual savings ef dardized minimum
groups facing income forts; development of pension; self-financing
restrictions. incentives to promote of additional pension
Table 2 (Cont.)
Objective Particular problem Proposal Instruments
saving among inde- benefits; separation of

pendent workers. pension programme
accounts from the ge
neral budget.
To make pension be- Individual pension be- Relate pensions to mi- Age at retirement and
nefits bear relation to nefits bear no relation nimums determined pension benefit s
contributions. to changes in life-time by law and comple- according toabalanced
contributions; exis- ment them with addi- formula taking into
tence of non-eligible tional benefits accord- account life expectancy
beneficiaries; indexa- ing to life-time addi- at retirement and pre-
tion of pension be- tional voluntary con- vious contributions;
nefits which bear no tributions and their flexibilization of age at
relation to indexation capitalization. retirement on the basis
of financing sources of of adjustment of pen-
the system. sion to similar formula
as above; implement a
basic minimum pen
sion anda self-financed
supplement; develop
ment of life insurance
schemes; development
of indexed instruments
to be used in the ope
ration of the system;
regulation and super
vision of eligible de
pendents; elimination
ofprivilegesbearingno
relation to contribu
tions.
Improvement ofsys- Excessiveadministra- Adm inistrative re- Rationalization; sepa-

tem administration five expenses and ine- form. ration of programme
fficiency administrations; legis
lation; computeriza
tion of system admi
nistration; training;
human resources poli
cy; normalization; role
________________________________________________________________ of private sector.
P e n s io n s y s t e m r e f o r m s a n d s a v in g s in L a t in A m e r ic a n a n d ... 213
ECLAC findings based on case studies of pension system evaluation

and reform alternatives show that some countries already have incorpora
ted capitalization schemes in the design of their pension-system reforms
or are studying the idea. These schemes may assume two forms: a compul
sory private scheme administered by private corporations created separate
ly from the public scheme, but regulated and supervised by the State, as
is the current system in Chile (Iglesias and Acuña, 1991)9; or private pension
schemes as a supplement to the public scheme. A mixture of these two
alternatives seems to be the basis for recent reform project proposals in
Colombia and Argentina (Ayala, 1992 and Schulthess, 1992)10. Several other
countries have historically encouraged voluntary private saving schemes to
supplement the public scheme (e.g., Brazil, Camargo, 1991)11.
The ECLAC findings, however, also recognize limitations arising from
characteristics of the regional economies. Government authorities are faced
with very strong political pressures as the result of a large share of fami
lies living below the poverty line and whose income is thus limited when
it comes to financing their basic consumption needs. The fact that these
segments work in the informal sector, where enforcement of social legis
lation is weak, is a major reason for their sparse sodal-security coverage.
Under such circumstances government authorities have difficulties in
accumulating reserve funds coming from contributions by the labour force
and contributors to the system. They are under pressure to redistribute
such funds among the poor and politically powerful groups eligible for
social-security defined benefits.
9 For details see Augusto Iglesias and Rodrigo Acuña, 1991. Sistema de Pensiones en Amé
rica Latina. Chile: Experiencia con un Régimen de Capitalización 1981-1991. CEPAL. Pro
yecto Políticas Financieras para el Desarrollo. Santiago, Octubre 1991.
10 For details see Ayala, Ulpiano 1992 Un Sistema Pensionai de Capitalización Individual pa
ra Colombia. Documento para el Proyecto Regional CEPAL/PNUD "Políticas Financieras
para el Desarrollo". CEPAL, Santiago, Chile Noviembre de 1992. And also Schulthess, Wal
ter, 1992 "Sistema Nacional de Previsión Social de la Argentina y Propuesta de Reforma"
Documento para el Seminario Regulación y Supervisión de la Banca" y "Reforma al Sistema
de Pensiones y Ahorro Institucional" en América Latina. CEPAL, Santiago de Chile,
Noviembre 1992.
11 Camargo, J. M. 1991, Pensión Funds in Brazil. Paper prepared for the Seminar on "Pension
Systems in Latin America: Diagnosis and Reform Alternatives" ECLAC Joint ECLAC/UNDP
Regional Project" Financial Policies for Development. Santiago de Chile.
214 A n d ras W . U th o ff
In order to avoid such problems, the question of ownership rights

relative to the accumulated fund is critical in the design of a pension system.
Reforms promoting the accumulation of a fund and avoidance of political
pressures on its allocation also endeavour to replace the public intergene-
rational transfer of contributions (from those in the labour force to those
who are retired) with some sort of life cycle private (common or individual)
savings and investment account of contributions for pension benefit pur
poses. There is undoubtedly a trade-off between the objectives for which
the pension system can be used. Public funds can be used discretionally
by the State for redistributive purposes in the first case, whereas such goals
must be pursued through the allocation of resources in the capital market
an d /or through a subsidiary role of the State in the second case.
Social security in Latin America in the context of economic adjustment
and restructuring must be carefully studied. Future perspectives and trends
need to be discussed within various contexts: (a) the redefinition of the
role of principal actors - mainly the private sector in providing health care
and pensions; (b) rethinking of ways to extend social protection in both
population and contingency coverage; and (c) the appraisal of financing
and cost efficiency alternatives (McGreevey, 1990, Mesa-Lago, 1991)12. Any
such discussion must bear in mind that in Latin America the private sector
has been traditionally isolated from these areas of public concern; in several
countries there is still very low population and contingency coverage by
current social security schemes, and the sparse capital markets are in their
initial stages of development.
Because pension systems are an individual policy instrument, they
cannot simultaneously contribute to the provision of reasonable pension be
nefits in relation to previous contributions, the redistribution of income by
increasing population and social-security contingency coverage, and the im
provement of domestic savings. The Latin American and Caribbean exper
ience shows that the answer to this policy dilemma is to keep the pension
12 For further details see, William Me. Greevey. Social Security in Latin America. Issues and
Options for the World Bank World Bank Discussion Papers 110. The World Bank,
Washington D.C. November 1990 and Carmelo Mesa-Lago 1991. "Social Security and Eco
nomic Adjustment-Restructuring in Latin America and the Caribbean. A Study for the
International Labour Office. University of Pittsburgh, December 1991.
P e n s io n s y s t e m r e f o r m s a n d s a v in g s in -L a t in A m e r ic a n and. 215
system financially and actuarially sound and separated from other govern
ment redistributive programmes. The conditions to achieve this goal are
given in the next section.
II. POPULATION, EMPLOYMENT AND RETURNS

TO PENSION SYSTEM FUNDS
The simulation in Table 3 is based on demographic and national account

statistics. It reports on the contribution rate in the hypothetical case of a
pension system designed to pay those over 65 a universal pension benefit
equal to per capita GDP. D ie given parameters are the population share
of those over 65 and the formal wage sector earnings as a percentage of
Table 3
EARNINGS PROFILES AND SIMULATIONS. SIMULATIONS OF CONDITIONS
TO ACHIEVE UNIVERSAL PENSION EQUAL TO PER CAPITA GDP
Parameters Simulations
Wage
GDP Share of earnings Contrib. Years needed
per capita population % Rate to achieve si
(approx. 1990) 65+ GDP milar pension
(US$ of 1980) (%) (%) (%)
Latin America 1,872 4.7 - - -
Colombia 1,379 4.1 37.9(88) 10.8 37

Chile 2,526 6.0 33.0(85) 18.2 30
Panama 1,461 4.8 50.6(89) 9.5 40
Uruguay 2,254 11.6 33.4 (88) 34.7 20
Brazil 2,020 4.7 34.8 (80) 13.5 34
Costa Rica 1,460 4.3 48.3 (89) 8.9 41
Ecuador 1,363 3.8 14.1 (89) 27.0 24
Mexico 2,280 3.9 25.9 (88) 15.1 33
Peru 896 3.8 24.4(89) 15.6 31
Venezuela 2,736 3.7 33.5 (89) 11.0 37
Bolivia 601 3.2 24.1 (86) 13.3 32
Guatemala 797 3.2 28.4(89) 11.3 37
Honduras 608 3.3 43.3 (89) 7.6 42
Paraguay 1,296 3.6 27.4(89) 13.1 34
Source: Estim ates by the author on basic of demographic and national account statistics.
216 A ndras W . U th o ff
GDP. By definition, the population share of those over 65 shown in column

(2) corresponds to the percentage of GDP needed to achieve the defined
pension benefit goal and is directly proportional to the ageing of the popu
lation. The maximum requirement is in the case of Uruguay where such
a defined pension benefit goal would necessitate the allocation of a 11.6
per cent of GDP for such purposes. The lowest requirements correspond
to those countries which are still well behind in their demographic transi
tion (Bolivia, Guatemala, Honduras, Paraguay).
A common feature of Latin American and Caribbean pension systems
is that their population coverage is limited to formal sector workers. Formal
sector wage-eamings as a share of total GDP is given in column 3. Column
4 shows the hypothetical average contribution rate were the entire cost of
providing the universal pension (equal to per capita GDP) to be financed
by contributors from the formal-sector workers system. This contribution
rate would range from a minimum of 7.6 per cent in Honduras (where
those aged 65 and above are only a small fraction of the population and
the wage-earning share of GDP is large), to a maximum of 34.7 per cent
in Uruguay (where there is a high percentage of elderly and, according
to Latin American standards, the wage-earning share of total GDP is at
an intermediate level).
If the same pension were to be offered to individual workers in a
capitalization scheme, with a yearly 5 per cent capitalization rate in real
terms (estimated as that needed for reasonable pension benefits in Table
1) column 5 indicates the number of years an individual worker would
have to contribute to the pension plan prior to the age of retirement, had
he earned a wage equivalent to per capita GDP during that period. There
is no doubt that the larger the contribution rate, the less convenient it
is for an individual to participate in a pay-as-you-go system and the more
convenient to participate in a capitalization scheme.
In spite of its simplicity and perhaps questionable assumptions, the
above exercise calls our attention to the parameters that determine the returns
derived from contributions to one scheme or another. In the pay-as-you-
go system, the returns on a given contribution rate can be approximated
by the sum of the rate of growth of contributors (those in the labour for
ce, employed and contributing to the system -most likely the rate of growth
of formal sector employment) plus the rate of growth of average net ear
P e n s io n sy st e m r e f o r m s a n d s a v in g s in L a t in A m e r ic a n and. 217
nings in real terms, less the rate of increase in life expectancy at the age
of retirement (e.g., 65 for males and 60 for females)13. That is why actua
rial studies of such systems often recommend, as some of the suggestions
in Table 2 do, either increasing contribution rates, reducing pension bene
fits (both in terms of coverage and the level of the benefit), increasing
the age at retirement, or a combination of all three.
These parameters are particularly relevant to Latin America and the
Caribbean. Demographic transition varies both between and within countries
according to the socio-economic level, affecting basic demographic para
meters which in turn influence pension system performance. On the average,
in 1950 the total population of the region (some 89 million) grew at a
rate above that of the 15 to 64 age group (the potential labour force parti
cipants). Statistics showed 78 persons at passive ages (below 15 and above
65 years old) per 100 persons at active ages. These elderly (above 65) repre
sented only 8 per cent of the total passive population, or one elderly per
son per 17 persons at active ages. From 1960 onwards, the total fertility
rate in Latin America declined sharply and the trends in the above indica
tors changed dramatically. In 1990, the total population was estimated at
260 million, more than triple the 1950 figures. The dependency ratio fell
to 69 persons at passive ages per 100 persons at active ages (after rising
to a maximum of 87 in 1970). One out of eight passive persons is elderly
and there are less than 13 persons at working ages per each person over
65 (CELADE, 1990)14.
These characteristics are far less important in terms of ageing than
has been the case in Western Europe. In that region, those aged 60 and
above currently represent 20 per cent of the total population and two-thirds
of the total passive population (Chesnais, 1989)15. However, as a distinctive
feature of Latin America, ageing is taking place together with relatively
high population growth rates in all age groups: those in need of education,
13 Larrain, Felipe and Gert Wagner 1982 Previsión Social, Algunas Consideraciones. Instituto
de Economía, Universidad Católica de Chile, Documento de Trabajo N2 84.
14 For further details, see CELADE, 1990 America Latina. Proyecciones de Población. Boletín
Demográfico. Año XXII, Ns 45, Santiago de Chile.
15 See Chesnais, J. C. 1989 L'inversion de la pyramide des ages en Europe: perspectives et
problems. "Congres International de la Population, N2 3. New Delhi.
those in need of jobs, and those in need of pension benefits. This brings
us to the important issue of the allocation of investment resources for growth
with equity and environmental sustainability. The investment in human ca
pital throughout the different stages of a person's life becomes a crucial
element which also strongly encourages political pressures on the alloca
tion of pension system reserve funds.
Another important feature of the region is the extremely heteroge
neous nature of the relationship between potential contributors and bene
ficiaries within each system. This is the result of a highly segmented labour
market. The poorer segments do not participate in formal pension fund
systems but instead are organized through intrafamiliar solidarity schemes,
operating via extended families in which those in the labour force contri
bute with their income to satisfy the needs of all the young and the aged
and those of working age, although well below the standards of any rea
sonable poverty line. Elsewhere, workers employed in the modem sectors
contribute to compulsory public pay-as-you-go schemes, whereby those in
the labour force finance the pension benefits of the retired with the expec
tation that future generations will in turn pay for them. At present, some
formal-sector workers have access to private capitalization schemes, in which
each generation contributes savings and investment needed to supplement
its pension benefits. The coexistence of all these formal and informal pen
sion systems is the result of the structure of the labour market in Latin
America. It is hard to believe that population coverage of formal systems
will increase without radical changes in labour market structures an d/or
the creation of very particular and innovative incentives to join the formal
compulsory institutional schemes. None of these events is likely to take
place in the near future.
In fact, between 1950 and 1980 Latin America's share of m odem or
formal employment, to which an institution pension-fund scheme is more
suitable, increased by only 5 percentage points (from 52 to 57 per cent)
of the total labour force. The rest of the employed pertained to traditional
agricultural activities (self-employed, peasants and non-remunerated family
workers) or to informal non-agricultural activities an d /or domestic service,
where control of compliance with labour legislation and the collection of
compulsory taxes for social security purposes is more difficult. This posi
tive trend in formal/modern employment creation was halted during the
eighties as a consequence of the debt crisis and overall recession. Among

non-agricultural activities, formal employment lost ground against informal
employment and occasionally unemployment (PREALC, 1982 and 1990)16.
Recent studies have shown that the low level of formal employment is
a major barrier to improving social security population coverage in general,
and that of pension systems in particular (CEPAL, 1985; ECLAC, 1989; Mesa
Lago, 1990)17.
This brings us back to the relevant issue of understanding the com
plex interrelations between demographic, employment, financial and
developmental variables. The core element is the provision of the necessary
savings to invest for redistribution and growth through a combination of
human and physical capital investment projects. Although investment in
human and physical capital is an important mechanism for the redistri
bution of income, it can also contribute widely to perpetuating inequality.
To the extend that an individual's access to capital markets determines his
opportunity for human and physical capital investment and his chances
for upward mobility as part of the work force, then capital-market segmen
tation will constitute a mechanism for the perpetuation of inequality.
III. DEVELOPMENT OF CAPITAL MARKETS AND

RETURNS ON PENSION SYSTEM FUNDS
Development of capital markets becomes an important issue in the dis

cussion of alternative designs for pension systems. If the objective of pro
viding good pension benefits is important, then financial markets must
warrant returns sufficiently to make it profitable to have invested in them
by the time the contributor is eligible for pension benefits. In accordance
with earlier estimates, Table 4 shows that for both males and females with
16 For further details on these trends see PREALC, 1982 Mercado de Trabajo en cifras 1959-
1980 PREALC-OIT Santiago de Chile, and also PREALC, 1990 Empleo y equidad, el desafío
de los 90 PREALC-OIT Santiago de Chile.
17 See CEPAL, 1985. El desafío de la seguridad social en América Latina Estudios e informes
de la CEPAL, Santiago de Chile. ECLAC, 1989 The Elderly in Latin America: A strategic
Sector for Social Policy in the 1990s LC/R 833 December 1989, and Mesa-Lago 1990 La se
guridad social y el sector informal Investigaciones sobre empleo 32. PREALC, Santiago.
similar age-eamings profiles, unless portfolio investment real returns exceed

5 per cent on the average during one's entire working life, neither men
nor women will obtain pension benefits equal to their last year's real ear
nings in the work force.
Table 4
ESTIMATED OLD-AGE PENSION BENEFITS:
ALTERNATIVE HYPOTHESES
(As percentage of net earnings from last 10 working years,)
Initial net Real benefits (5%) Real benefits (6%)

earnings Male Female Male Female
25,000 101.7 63.9 143.6 102.5
40,000 104.8 65.9 148.1 105.7
100,000 108.0 67.9 152.6 109.0
200,000 109.0 68.5 154.1 110.1
360,000 109.5 68.8 154.6 110.6
Based on the following assumptions:
(1) Fixed commisson o f 250 real monetary units.
(2) Net earning grow 2% per annum until contributor is 50 years old, and stabi
lizes thereafter.
(3) W orking period starts at age 22.
(4) Males retire at 65 and by then have a life expectancy of 16.
(5) Females retire at 60 and by then have a life expectancy o f 21.
(6) Percentage is estimated over the average net earnings of th e last 10 years.
For different initial net earnings brackets, Table 4 computes the pension
benefit an individual worker starting his/her working life at the age of
22 would receive had his/her net life earnings profile increased by 2 per
cent per annum in real terms up to age 50 and remained stable thereafter.
The difference by gender is due to the earlier retirement of women (at
age 60) as compared to men (at age 65). This means that the former con
tribute during a shorter period and need pension benefits for a longer pe
riod (due to their greater life expectancy at the age of retirement). This
explains why the relation between the monthly pension benefit for w o
men, compared to the average last ten years net earnings, is significantly
below that for men. But the table also shows that fixed commissionsimple
mented to bear the cost of administrating individual accountsareregressi
ve, in the sense that they lead to lower returns for those in the lower
net earnings brackets.
P e n s io n sy s t e m r e f o r m s a n d s a v in g s in L a t in A m e r ic a n and. 221
Capitalization of contributions has three basic inputs for increasing

an individual's pension fund every year: the contribution itself, the interest
paid on that contribution during that year, and the returns on the portfolio
in which the fund is invested. Under stable real rates of return on portfolio
investment over a period of time, in the earlier stages the largest addition
to the fund comes from contributions. From the very start, however, both
the fund and the returns from its investment grow exponentially. After
approximately 15 years at a 5 per cent rate of return on portfolio invest
ment, these returns constitute a larger input to the individual workers' fund
than do the contributions themselves. This is the advantage of a capitali
zation scheme in contrast to a pay-as-you-go system governed by reaso
nable parameters (positive growth rates in formal employment and in real
wages, and slight increases in life expectancy at age of retirement). This
also necessitates an awareness of the impact that pension funds may have
on an expanding financial market.
If reasonable returns from portfolio investments in domestic capital
markets are to be warranted, a twofold set of conditions must be met,
one dealing with the macroeconomic and the other the financial environ
ment.
In fact, and due to the interperiodical character of decisions involved
in institutional savings and resource allocation for future growth, autho
rities should take care to obtain a policy mix intended to maintain low
and controlled inflation rates, positive in real terms but moderate interest
rates and a credible exchange rate policy. In Latin America such policy
mixes should satisfy at least three conditions: (a) they should be aimed
at strengthening a financially balanced public sector; (b) they should be
aimed at reforming the banking system; and (c) they should be aimed at
developing the capital and insurance markets.
1. Public sector reforms
Three changes have been pointed out as the most important in order to
balance public-sector accounts: the financing of social-security system defi
cits, the autonomy and solvency of the Central Bank, and the elimination
and reduction of all quasi-fiscal deficits in the financial system.
In relation to the topic of this paper, the first change is straightforward.
Mechanisms need to be created to finance past deficits while avoiding fu

ture ones. Transition to any alternative system different from the one in
operation should consider its public finance costs. The most common way
out has been by issuing government compensation bonds for those alrea
dy participating in the old system and not yet retired, and drawing from
general public tax revenues for financing those presently retired. The public
sector must make provisions for these payments in their current budge
tary exercises.
2. Banking regulation and supervision
A second set of conditions deals with reforms to the banking system (Held
and Szalachmann, 1991 and 1992, ECLAC, 1992)18. These can be presented
in two stages. First, macroeconomic adjustment needs to be complemented
by intermediation through the financial system in order to allow an efficient
allocation of resources. A precondition is the financial solvency of banks
and firms. Otherwise, financial liberalization and international competition
through gradual opening of the capital account may prove inimical to heal
thy development of the domestic financial sector. The steps taken to libera
lize the capital market in order to achieve monetary autonomy, free ca
pital flows and exchange rate management must start with a reasonably
solvent domestic banking system.
Regulatory structures through which the State may establish bank
and financial-institution "rules of the game" for deposits and loans is a
second stage. These include prudential regulation, to control banking risks,
financial insolvency and illiquidity; financial regulation, oriented towards
interest and exchange rates, currency denomination, instrument maturity
18 For further details on the regulation and supervision of banks and financial institutions, see
Held and Szalachman 1991, Regulación y Supervision de la Banca. Experiencias de América l a
tina y el Caribe. Argentina, Costa Rica, Chile, República Dominicana. CEPAL, PNUD. Proyecto
Regional Políticas Financieras para el Desarrollo, Santiago de Chile; also Held and Szalachman,
1992, Regulación y Supervisión de la Banca. Experiencias de América Latina y el Caribe. Volumen 2.
Guatemala, México, Paraguay, Perú, Venezuela. CEPAL, PNUD. Proyecto Regional Políticas
Financieras para el Desarrollo, Santiago de Chile, and ECLAC, 1992, "Regulation and Super
vision of Banks and Financial Institutions. Case Studies in Latin América and the Caribbean. Serie
Financiamiento del Desarrollo N2 8. Santiago dc Chile.
P e n s io n s y st e m r e f o r m s a n d s a v in g s in L a t in A m e r ic a n and. 223
periods, and other conditions related to the price and value of both do
mestic and foreign assets and liabilities; and finally, organizational regu
lation, to seek bank and financial institutional operational efficiency
through economies of scale, the integration of economic activities and the
promotion of competition among financial institutions.
Supervision of banks and financial institutions entails verification of
compliance with the prudential, financial and organizational rules and
regulations which govern their activities as financial intermediaries. (See
the chapter by Held in this volume.)
3. Security and insurance market regulation and supervision
Last but not least are the reforms needed to promote the development
of securities and insurance markets. Competition in these markets also re
quires three regulations: regulations for entry into the particular market;
prudential regulation for those operating in the market; and motivational
regulation for expanding the financial market.
Entrants into these markets must demonstrate financial solvency by
fulfilling legal and financial requirements as an intermediary, and provide
information about all financial instruments and conditions that will keep
the market transparent. This includes registration of instruments to be issued
and intermediated in the stock market, with the commitment to provide
public and periodical information on their financial status, and secondly,
registration of intermediaries in the stock market that will accept regula
tory norms and supervision of their activities by independent institutions
in order to maintain a minimum level of professional conduct, thus wa
rranting a competitive and equitative intermediation of securities.
Prudential regulation of these markets is mainly concerned with three
factors: market entry conditions -minimum capital and property require
ments; an investment insurance guarantee scheme; and portfolio quality
control standards and risk limitations. Two separate institutional develop
ments in this area are the Deposit or Custody of Securities and the Risk
Classification Commission. The first keeps track of all deposits, liquidations
and other transactions, in accounts which also perform multilateral elec
tronic balancing. The classification of risks is a service provided by spe
cialized agencies, in order to carry out ongoing evaluation of firms offering
instruments and to determine their financial capacity. The final objective

is to allow the public to take informed investment decisions. These agencies
must be registered and must follow very strict procedures and methods
in performing their duties.
Motivational regulation refers to the incentives needed to expand the
market in order to attract new businesses, intermediaries, and investors,
along with the emission of new stocks and securities. The purpose is to
improve the role of these markets in channeling savings into productive
investments. These incentives can assist in providing non-bank financing,
especially that which comes from institutional savings.
IV. THE CHILEAN EXPERIENCE WITH AN INDIVIDUAL

CAPITALIZATION SCHEME
The reform in Chile has drawn a great deal of attention due to its radical
transition from a pay-as-you-go public system to an individual capitalization
scheme, and also due to its relative success during its first ten years. This
experience has been described in ECLAC's case study (Iglesias and Acuña,
1991) and elsewhere (Arrau, 1992; and Montt 1992)19.
1. Main characteristics o f the new system
The system is based on mandatory contributions of 10 per cent of base

salary paid by the employee to a personal retirement account. This fund
continues to build until the age of 65, in the case of males, and 60 for
females. Funds belong to the worker and there are no defined benefits.
The latter will depend on past contributions and their corresponding ca
19 An exhaustive description of the reform is given by Iglesias, Augusto and R. Acuña. 1991
Sistema de Pensiones en América Latina. Chile: Experiencia con un Régimen de Capitalización 1981-
1991. Proyecto Políticas Financieras para el Desarrollo, CEPAL/PNUD Santiago de Chile.
Summary reviews of the reform are given by Arrau, Patricio. 1992 El Nuevo Régimen previsio-
n d Chileno Seminario Internacional de Reformas al Régimen Pensionai, Bogota, Colombia,
Mayo 20-21 de 1992; and Montt, Diego, in The Chilean Private Pension Fund System Superinten
dencia de Administradoras de Fondos de Pensión-Chile presented at Workshop on "Emerging
Markets, Current Issues", Bolsa de Comercio de Buenos Aires, 23,24 November 1992.
pitalization. Management of these accounts for capitalization purposes is

handled by private pension fund managers, thus being firms incorporated
into the private pension fund system and which compete for the funds
on the basis of quality of service and commissions. The latter are set at
about 3.25 per cent of reported salary and covers account administration,
overhead, marketing, a group insurance policy of approximately 1 per cent
of that salary for invalidity and survivor's benefits, and profits, if any.
The Pension fund managers have but one duty: to manage pension
funds. The State is responsible for regulating their firms and supervising
their market. The State also sets conditions to guarantee adequate returns
according to average market conditions and a minimum pension for those
participants unable to accumulate sufficient funds.
The reform sets benefits depending on past contributions and capi
talization returns. Employees withdraw the fund upon retirement and pur
chase a life-insurance policy that guarantees a minimum family retirement
revenue. Any surplus above the present value of the legal minimum may
be withdrawn from individual accounts under a specially-managed retire
ment program. In addition, the contribution to a group insurance policy
provides for resources in the event of other contingencies related to pension
benefits: invalidity, survivor's coverage, and severance allowances.
Separation of social security programmes is also warranted by the
reform. Sickness, maternity, work injury, unemployment benefits and fami
ly allowances are carried under separate national security or social insu
rance schemes that are normally government-sponsored but with varying
degrees of private-sector involvement.
2. Prudential regulation o f pension fund investments
The system operates under very strict regulations which can be summari
zed as follows:
1 - Pension funds can only be invested in financial instruments authorized

by law.
2 - Pension funds are subject to very strict regulations for portfolio diversifi
cation by financial instrumens and by issuers.
3 - All financial instruments subject to investment by pension funds must be
evaluated and approved by the national risk commission.
4.- The value of all pension-fund instruments is adjusted daily in accordance

with prices established by the supervisory authority: (85 per cent of portfo
lios are valued at current stock exchange prices).
5 - The management of pension funds is subject to a minimum level beneath
which profitability may not drop.
6.- Custody of pension fund investment securities: at least 90 per cent of
financial instruments must be in custody at the Central Bank.
Table 5 indicates the limits of pension fund investments, according

to different instruments. These limits are set by law, and pension fund
managers must obey very strict norms for diversification of their portfolio
according to both investment instruments and issuers. The selected instru
ments are evaluated by the Risk Classification Commission, and must be
intermediated in well-established stock and secondary markets. The value
of the fund is adjusted daily according to market prices reported by the
Superintendency of pension fund managers, and periodical reports must
be given to contributors. Eighty-five per cent of the portfolio is evaluated
at current stockmarket prices, but the overall value of the fund also varies
in relation to changes in the market interest rate when estimating its cu
rrent present worth.
In order to eliminate solvency risks due to bad investment mana
gement, pension fund managers are subject to a requirement concerning
minimum profitability. To bear relation with market conditions such mini
mum is set at a level equal to the lower bound between the average profi
tability of all PFMs less 2 percentage points and half the average profita
bility of all PFMs. In the event of a cease of activities PFMs must make
use of their reserve fund and legal capital requirements.
3. Pension system results and financial market developments
The effects of the pension system reform will now be analyzed in terms
of four important results: population coverage; level of pension benefits;
financial market developments; and national savings and investment.
In general, these types of evaluations are very hard to perform, largely
due to the difficulty in identifying the standard of comparison. If, as is
often done, the results of the new systems are compared to those of the
old system, this is said to be unfair since it is precisely due to its poor
P e n s io n sy s t e m r e f o r m s a n d s a v in g s in L a t in A m e r ic a n and. 227
results that the latter is being replaced. If, quite to the contrary, the system
is assessed on absolute terms, then supporters claim unfairness because it
ignores the starting conditions. On the whole, as will be made clear, no
definite evaluations have been made, and further research is under way
on these matters.
Table 5
PRUDENTIAL REGULATION OF PENSION FUND INVESTMENTS:
CHILEAN LIMITS SET BY LAW ACCORDING TO INSTRUMENTS
Instrument Limits (% of Fund)
1. Government instruments 45
(Treasury and Central Bank)
2. Mortgages 80
3. Long-term savings, other securities
from approved financial institutions 50
Short-term (less than one year) 30
4. Public and private sector bonds 50
5. Shares from other pension funds 20
6. Corporate shares
a. non-concentrated property 30
b. real state corporations 10
c. concentrated property 10
7. Shares of investment funds 10
8. Commerce 10
9. Finandal instruments from abroad 10(a)
(a) From O ctober 1996 onwards
Source: Iglesias y Acuña 1991, op. cit.
a) Pension system population coverage
Social security population coverage in Latin America in general, and that

of pension systems in particular, has historically been very low. As discussed
earlier, this is due to the structure of labour markets, with large segments
of the labour force outside the formal sectors of the economy where so
cial security legislation is easily enforced. The Chilean case has often been
quoted as relatively successful in terms of population coverage (Messa Lago,
op cit.). With the reform, lower social security taxes and increases in net
earnings were considered important incentives to induce independent wor
kers to contribute voluntarily to the system. Nevertheless, workers in infor-
228 A n d ras W . U th o ff
mal sectors who had not contributed to the old system were ineligible for
the main incentive, that being a compensation bonus.
As given in Table 6, population coverage measured by share of the
labour force shows that whereas affiliates have come to represent 86 per
cent of the labour force, those actually contributing to the system have
never increased above 55 per cent. The causes of this poor result should
be eliminated in order for coverage to increase. As mentioned before, this
responds to the labour market structure in Chile, still with a high share
of independent informal workers, together with a large incidence of p o
verty (Uthoff and Pollack 1990)20. Other authors have also mentioned that
the situation results from loss of jobs; lack of incentive for independent
workers to contribute; delays in contributions by employers; people for
merly in the labour force w ho are retired and do not qualify; and partici
pants in the old system (estimated below 10 per cent of the labour force)
(Arrau, op. cit.).
Tableó
POPULATION COVERAGE UNDER THE NEW PENSION SYSTEM
(as a porcentage of total labour force)
Affiliates Contributors In old system
1981 39.0 - -
1982 39.3 24.8 -
1983 43.0 28.0 -
1984 49.6 29.3 -
1985 56.8 32.9 11.3

1986 60.7 35.0 10.4
1987 66.4 38.5 10.0
1988 69.9 38.9 9.3
1989 74.4 41.0 8.8
1990 79.1 41.5 8.5
1991 85.6 51.8 -
Source: Iglesias and Acuña 1990, Chile: Experiencia con un Régimen de
Capitalización 1981-1991, ECLAC, Políticas Financieras para el Desa
rrollo, and updated with figures from Boletín Estadístico de AFP.
20 Uthoff, Andras and Molly Pollack. 1989. Poverty and the Labour Market: Greater Santiago, 1969-
85 in Rodgers Gerry, ed. Urban Poverty and the Labour Market. Access to Jobs and Incomes
in Asian and Latin American Cities. ILO, Geneva.
P e n s io n sy s t e m r e f o r m s a n d s a v in g s in L a t in A m e r ic a n a n d .. 229
b) Pension benefits, now and then
It has not been possible to make a definite and general comparison bet
ween the amount of pension benefits received in the old and new system.
The problem lies in the difficulties in estimating the pension benefit after
full maturity of the new system. Such an exercise involves an estimate
of the expected active life-cycle income (which in turn requires assumptions
concerning labour productivity increases according to human capital en
dowments to estimate longitudinal earning profiles). It also involves an es
timate of returns during the capitalization period and standarization proce
dures with respect to age and life expectancy at retirement.
Table 7
FINANCING AND CHARACTERISTICS OF PENSIONS
1. Financing for the following benefits

Old-age Pension Benefits Disability and survivors pension benefits
A. Accumulated capital A. Accumulated capital
1. (+) Contributions (compulsory and vo 1. (+) Contributions (compulsory and vo
luntary) luntary)
2. (+) Investment returns 2. (+) Investment returns
3. (-) Commission on individual account 3. (-) Commission on individual account
B. Compensation bonus B. Compensation bonus
C. Transfers from voluntary savings account C. Transfers from voluntary savings account
(if worker so wishes) (if worker so wishes)
D. Government contribution if required to D. Additional contributions (funded with in
achieve minimum pension surance)
E. Government contribution if required to
achieve minimum pension
2. Number of pension benefits paid up to 1990

Old age % Contri- Disabi- % Contri Survi % Contri
butors lity butors vors butors
1989 19,953 0.57 14,388 0.42 35,094 1.01
1990 29,666 0.79 15,777 0.42 41,618 1.11
3. Average pensions (Thousands of pesos at December 1990; US$ 1 = $300)

Old age Disabi Widow Survivors Average
lity hood Child Other earnings
1989 37.4 65.2 28.5 9.2 15.8 81.2
1990 39.4 63.9 27.8 9.5 16.8 84.4
Part 1 of Table 7 shows that results from comparisons of pension

benefits provided by the two systems are highly sensitive to the transition
period. This, is due to the importance of the compensation bonus and the
way it was calculated relative to the overall result of the new system pen
sion benefit. In turn, such a comparison is also sensitive to the amount
of subsidies that pension benefits in the old system received as compared
to those in the new system.Finally, we must also bear in mind the beha
viour of the capital marketduring the capitalization period and its repli
cability in the future.
Parts 2 and 3 of Table 7 show that the system is still too new to
make valid comparisons. In part 2 of the table we can see that so far
less than 2 per cent of contributors have received benefits from the fund
so as provided by the system (old-age pension benefits, disability benefits
or survivors benefits). In turn, the income level of benefits that are being
provided at present are quite below the average net earnings reported by
contributors to the systems (Part 3 of the table).
The best one can do is recall the results in Tables 1, 3 and 4, to
a large extent attributing the success of the system to returns above 5 per
cent on the pension-fund investment portfolio. The role of the government
is to enforce prudential regulation to try to reach such a goal based on
market conditions.
c) Pension Funds and Financial market developments
Table 8 indicates the amount of the cumulated fund and its portfolio invest
ment composition. Only after 1985 were pension fund managers authori
zed to invest in the stock market. But it was not until 4 years later, in
1989, that these firms decided to go about it professionally. Between 1989
and 1992, the market value of common stock in pension fund portfolios
multiplied by 10, from US$ 350 to US$ 3.3 billion. A note of caution must
be sounded before extrapolating this behaviour. Again it is hard to isolate
the impact of the pension-fund purchasing power from that of at least three
other phenomena that can also help to explain the same figures: first, dur
ing the same period, capital flows from abroad increased significantly and
currency revalued in United States dollars by about 20 per cent; second,
Chile has witnessed seven years of strong economic growth that started
P e n s io n s y st e m r e f o r m s a n d sa v in g s in L a t in A m e r ic a n and. 231
in 1985, just when pension funds were authorized to invest in the stock
market, and has averaged above 6 per cent per year; and third, since 1985
the real average price index of shares (i.e., after allowing for inflation) has
recorded a 500 per cent increase, of which half occurred after 1989 (see
Table 9).
An important conclusion with respect to the relationship between pen-
sion-fund developments and financial market behavior is their contributions
to expanding the financial system. This is the result of a relatively young
system, which, together with the exponential accumulation of investment
income in the individual funds of those of contributing age, justifies a lar
ge share of resources to be intermediated in financial markets.
The last column in Table 8 contains one indicator of financial expan
sion in Chile. Pension funds as a percentage of GDP grew from less than
one per cent in 1981 to above 38 in 1992. Projections show that this share
will level out at above 100 per cent of GDP by the year 2020 and before
the system has completely matured.
The portfolio investment composition in Table 8 also shows extensive
diversification and changes as the result of very strict regulations which ha
ve subsequently been modified. Highly concentrated on bank deposits and
bonds (62 per cent) in the earlier stages, it is now more heavily weighted
toward Central Bank and Treasury bonds (37 per cent) and common stocks
(28 per cent).
Despite conservative regulations to avoid risks to institutional inves
tors, returns on these investments have averaged, during the first 12 years
of existence of the system, well above real interest rates in the financial
system for the same period (see Table 10). Nevertheless a note of caution
is also needed here. When analysing the results for 1992, one must obser
ve that returns in real terms have dropped to 4 per cent, which contrasts
with the excellent performance of Chile's economy during that same year,
considered to be the best in the last 30 years (CEPAL, 1992)21.
21 CEPAL, 1992. Panorama Económico de América Latina y el Caribe Santiago de Chile.

Table 8
PORTFOLIO COMPOSITION OF PENSION FUNDS IN % (CHILE)
Central Bank Bank Firm Funds Funds

Year and Treasury deposits Mortgage bonds and Common millions USS %
bonds and bonds Securities debentures stock equivalent GDP
1981 28 62 9 1 0 219 0.9

1983 44 3 51 2 0 1,223 6.4
1985 43 23 36 2 0 2,228 10.9
1987 42 28 21 3 6 3,570 15.5
1988 35 30 21 6 8 4370 16.5
1989 42 21 18 9 10 5388 19.7
1990 44 17 16 11 11 7,136 26.5
1991 38 12 13 13 24 10,078 34.4
1992 (Jun) 37 10 13 12 28 11,922 38.1
Projection
2000 49-54
2010 77-87
2020 88-109
Source: Iglesias y Acuña, 1991. op. cit., and updated with data from the Superintendency reports.
Table 9
STOCK MARKET AND PENSION FUND PORTFOLIO
Common Stock Pension Funds

Volume Market
Year Price traded value (US$ Common stock
index index millions) (US$ millions)
1981 105 592 5,000 -
1983 70 369 2,783 -
1985 100 100 2,919 -
1987 313 908 6,852 223

1988 425 1,094 8,438 354
1989 581 1372 11,375 545
1990 750 1,155 14,564 806
1991 1,682 2,651 27,706 2,406
1992 (Jun) 2,122 3,117 34,444 3347
Source: Montt, Diego 1992. The Chilean Private Pension Fund System. Superintendencia
de Administradoras de Fondos de Pensiones. W orkshop on "Emerging Markets,
Current Issues", Bolsa de Comercio de Buenos Aires, 23-24 November.
P e n s io n s y s t e m r e f o r m s a n d sa v in g s in L a t in A m e r ic a n and. 233
Table 10
PENSION FUNDS: CHILE
Real Rate of Return on Fund Investments (%)
Year Pension Fund (a) Financial system (b)

System Interest rates
1981 27.5 13.2
1983 21.2 7.8
1985 13.4 8.2
1987 5.4 4.3
1988 6.5 4.6
1989 6.9 6.8
1990 15.6 9.4
1992 4.0 (E) q 5.8 (E)
Averages
1981-91 15.4 8.4
1987-92 11.2 6.4
(a) Average of Pension Funds
(b) Real rates of interest o f 90-360 days deposits.
Source: Iglesias y Acuña, 1990. op. cit. and updates by the author.
In summary, there has been a joint and significant increase in the

volume and value of stock-market operations and pension-fund stock invest
ments. The 300 per cent increase in pension-fund holdings matches the
pattern of transactions, volume and prices on the Santiago Stock Exchange.
However, there is room for believing in significant repercussions in the
real economy. For example, sixty per cent of all mortgage securities are
financed from pension funds, the same as with firm bonds and debentures.
Nevertheless, due to the fact that this period has been accompanied by
high and stable GDP growth rates, together with important incentives for
arbitrage of international interest rates causing large capital inflows and
currency appreciation, we are unable to draw any direct causal effect bet
ween pension funds and stock-market developments. Further research is
also needed here. Relative to financial expansion, pension funds represent
38 per cent of total GDP and are distributed among different instruments
according to very strict regulations. It is worth mentioning that high re
turns from capital markets will benefit workers to the extent that during
their working lifetime these returns exceed 5 to 6 per cent on the average
in real terms. During the last 12 years, returns have exceeded 15 per cent.
This result, however, is not sustainable in the long run, as would seem from
the fact that for 1992 real returns on such investments fell to four per cent.
d) Pension Funds, savings and investment
The effect on savings and investment is very hard to decipher. As expected

from figures on labour earnings as a percentage of GDP (Table 2), and
a contributory tax of 10 per cent, sodal-security savings increased to above
3 per cent of GDP during the last three years for which we have records.
At the same time, total savings also increased to around 20 per cent of
GDP (22 per cent is estimated for 1992). There is no dear relationship
between these two trends. Given the significant increase in sodal-security
savings, national savings would have increased were no other forms of
savings to have changed during the same period. Strong economic growth
during this period also interacted with the increase in savings and capital
formation. There are no specific studies regarding these events.
Figures in Tables 11 and 12 show the most recent official estimates
on savings. There is a twofold trend on the issue of the relationship between
sodal-security and other forms of savings. The first is that, together with
private sodal-security savings, other forms of private savings have also in-
Table 11
SAVINGS AND INVESTMENT IN CHILE
(% of GDP)
Private Savings Public National Foreign Total savings

Pension Other private Total savings savings savings = investment
Year savings savings (1) (2) (3) = (1) + (2) (4) (5) = (3)+ (4)
1980 0 2.8 2.8 1 1 .0 13.8 7.1 21.0

1981 0.9 -1.8 -0.8 8.6 7.8 14.3 22.0
1983 1.7 8.2 9.9 4 .5 5.4 5.7 1 1 .0
CO
1985 1.8 8.6 10.3 9.5 7.8 17.3

1987 2.1 6.4 8.5 4.1 12.6 4.3 16.9
1988 2.7 6.0 8.7 7.6 16.3 0.8 17.0
1989 3.1 6.3 9.4 7.9 17.3 3.0 20.3
1990 3.3 8.5 11.8 5.5 17.3 3.0 20.2
1991 3.3 10.9 14.3 4.9 19.2 -0.3 18.8
Source: National accounts and Ministry of Finance
P e n sio n sy st e m r e f o r m s a n d sa v in g s in L a t in A m e r ic a n a n d . 235
creased quite significantly. The second is that despite the unusually high
public-sector savings figures for 1980-81, what really happened is that pu-
blic-sector savings capacity was affected, reverting to its more traditional
figures. This is^xiue l o large increases in social-security deficits resulting
from the transition from one^system to another. This deficit is estimated
to have reached close to 5 per cent of GDP during 1989-91.
Table 12
SOCIAL-SECURITY DEFICIT ASSOCIATED WITH REFORM
(% GDP)
Year Recognition Operational Total

Bonus Deficit Deficit
1981 0.01 1.19 1.20
1983 0.17 3.53 3.70
1985 0.24 3.36 3.60
1987 0.38 3.42 3.80
1988 0.36 3.04 3.40
1989 0.53 4.17 4.70
1990 0.63 4.13 4.76
1991 0.71 4.08 4.79
Projections
1995 0.71 3.84 4.55
2000 0.95 3.18 4.13
2005 1.11 2.28 3.39
2010 0.94 1.47 2.41
Source: P. Arrau, op. cit.
In summary, specialists have reached no final conclusions concerning

the effect of social-security savings over national savings. This is because
such an effect can not be easily isolated from others resulting from changes
in income distribution, economic policy management and the composition
of savings. Further research is needed on this issue.
4. Other problems pending
Aside from the above issues related to the financial market, there still are
some challenges linked to its current structure and functioning. There are
also problems of a different nature.
a) Challenges related to financial markets
In relation to financial markets, we can identify a twofold challenge. First,

that of the vulnerability of an individual worker's pension benefits to fi
nancial market shocks immediately before retirement. "Positive" shocks can
substantially increase personal pension-fund benefits whereas "negative"
shocks can substantially reduce them. Steps should thus be taken in order
to make funds less vulnerable to sudden shocks in financial markets. The
idea of a stabilization fund to render average returns compatible with nor
mal financial market conditions during a worker's lifetime should be given
some thought.
The second challenge concerns the potential control over firms and
economic activity that pension-fund management firms can achieve based
on financial capital markets conditions and operations. As things now stand,
although these firms are not owners of the fund, they can gain considerable
economic power by placing some of their members on corporate boards
of directors. New regulations must be developed to avoid control over firms
and economic sectors by economic groups linked to these pension-fund
management firms. Estimates show that out of 15 firms the 3 largest account
for about 70 per cent of all contributors, and the 5 largest for about 85
per cent of all contributors. Thus concentration of pension-fund management
firms in the market may very well be a potential source for control over
the economy; further regulations must be created to avoid this situation.
In addition, and with a positive outlook for capital-market developments,
mechanisms should be created to have fund management improve the link
between financial savings and capital formation, especially that which is
needed to improve redistribution and growth.
b) Other challenges
Concerning challenges indirectly linked to the structure and functioning of

financial markets there is one of great importance: the market conditions
for pension-fund managers.
Market conditions for pension-fund management firms are also in need
of further regulations to correct various factors affecting the system's cost
efficiency and redistribution goal: concentration of contributors in the hands
of a small group of firms is common and responds to important econo

mies of scale; marketing costs are very high as the result of product com
petition and of the high returns on commissions; there is very poor trans
parency and participants lack all the necessary information for rational
decision-making with regard to market opportunities. Lastly, due to the
existance of fixed commissions, the system operates regressively in the allo
cation of returns based on the net-eamings brackets of participants.
V. CONCLUSIONS
Pension systems operate in an interrelated manner with other important

sectors of the economy: the labour market, the public sector, the financial
market, the securities and insurance markets, and so forth. Reforms must
be sought via an integrated approach. Both new and old pension-system
designs can be perfectly operational under ideal conditions. Regulation and
supervision mechanisms must be developed to meet them. The later in
volve not only those needed for the structure and functioning of the pen
sion system itself but also those concerning other sectors and markets with
which the system interacts.
The Latin American and Caribbean experience shows that regardless
of their original design, most systems have developed into pay-as-you-go
schemes. This has been a consequence of inadequate returns on portfolio
investment due to inflation and its concentration on government non-in-
dexed paper. Pay-as-you-go systems are facing, or will face, actuarial im
balances as the result of demographic changes, poor performance of la
bour markets and bad administration in terms of both the provision of
pensionable benefits (which bear no relation to the individual's contribu
tion) and an excess of overhead expenditures. The non-credibility of current
systems have further deteriorated their financial situation by causing contri
butors to avoid their obligations and beneficiaries to claim benefits unre
lated to past contributions. The impact of pension-system deficits on public-
sector finance and macroeconomic imbalances close a sort of vicious circle
by creating a negative environment for adequate investment decisions.
Reforms are being sought with expectations of breaking such a vi
cious circle. The key is to develop capital markets where compulsory sa-
238 A n d ra s W . U th o ff
vings can contribute to three objectives: accumulating funds for better

pension benefits, contributing to capital formation for sustainable growth,
and eliminating capital market segmentation in the access to investment
funds. Other redistributive goals must be sought through general taxes on
other sources of income.
The most radical reform has been undertaken in Chile, where a pay-
as-you-go system has been replaced by an individual capitalization scheme.
The fundamental change is from public management of tax collecting for
pensionable purposes to private pension-fund management by firms who
compete for individual accounts. Another change is from a pay-as-you-go
scheme, where the rate of return depends on employment and real-wage
growth and the rate of change in life expectancy at age of retirement, to
a capitalization scheme, where the rate of return depends on the portfolio
investment composition of the fund and the financial market results du
ring the working lifetime as well as on the life expectancy at age of re
tirement.
The effect on savings of such a reform is not easy to discern. On
the one hand, the transition implies sizeable obligations from the public
sector with respect to those already retired within the old system (opera
tional deficit) as well as those active and transferred to the new system
(recognition bonus). Estimates of such a burden show that if will increase
to as much as 4.8 per cent of GDP by 1991 and then gradually decrea
se to figures of 2.4 per cent of GDP by the year 2010. This undoubtedly
affects public-sector savings capacity. With regard to the private sector, little
research has been done with respect to the switch from compulsory insti
tutional savings to other voluntary forms of savings. Social-security savings
have increased up to 3.3 per cent of GDP from 1980 to 1991, but other
forms of savings have also increased considerably in that same period. The
definite causal rela-tionship between changes in saving rates and all events
(e.g., changes in the redistribution of income, balance of payments dere
gulation, financial liberalization, labour market reforms) occurring simulta
neously with pension-system reforms makes it impossible to draw a defi
nite conclusion.
Definite conclusions, however, can be drawn with respect to financial
expansion. The fact that the new system is based on accumulation (capitali
zation) of pensionable funds, together with the relatively young age of
P e n s io n s y st e m r e f o r m s a n d s a v in g s in L a t in A m e r ic a n and. 239
participants, explains the rapid growth of the total fund to be interme

diated in the financial sector as a share of GDP. This share increased from
zero to 34 per cent in 12 years and is expected to exceed 100 per cent
by the year 2020. The speed at which the financial market expands as a
consequence of pension-funds accumulation, and the responsibility of the
State to justify rates of return that will bring about significant improve
ments in pension benefits (for which purpose real rates of return should
be above 5 per cent) puts a lot of pressure on financial market perfor
mance. Macroeconomic stability, together with regulation and supervision
of both pension-fund management and the financial markets, become im
portant components of such reform. But they should be limited in order
to allow markets to play their proper role in allocating resources efficient-
!y-
Despite some impressive advances, other challenges still remain un
solved: the vulnerability of an individual worker's pension benefits to fi
nancial fluctuations at the time of retirement; the risks that economic power
will become concentrated in the hands of one or two pension-fund Mana
gement firms; and the difficulties of increasing population coverage by
system despite large and important individual incentives. Finally, there is
the industrial organization of the pension fund managers' market in order
to reduce its concentration on two or three firms, reduce marketing costs,
improve the transparency of its operations, and avoid fixed commissions
so as to improve the distribution of returns among participants at different
income levels.
BANK REGULATION, LIBERALIZATION AND
FINANCIAL INSTABILITY IN LATIN AMERICAN
AND CARIBBEAN COUNTRIES*
Günther Held Y.**
* This paper draws heavily on country case studies undertaken for the joint ECLAC/UNDP
regional project on Financial Policies for Development.
** I would like to thank Luis F. Jiménez for his very helpful comments. However, responsibi
lity for this paper rests solely with the author.
CONTENTS
INTRODUCTION .......................................................................................................................... 245
I. BANK REGULATION AND SUPERVISION ................................................................ 247

1. Financial regulation..................................................................................................... 247
2. Organizational regulation........................................................................................... 249
3. Prudential regulation................................................................................................... 250
a) State guarantee on deposits .................................................................................. 251
b) Deposit insurance.................................................................................................... 252
c) Norms controlling solvency.................................................................................. 253
4. Bank supervision ......................................................................................................... 255
5. Arrangements for the regulation of bank solvency ............................................... 256
a) Free or non regulated banking (as regards solvency)........................................ 257
b) Public sector controlled banking .......................................................................... 257
c) Bicontrolled ban k in g .............................................................................................. 257
d) Uncontrolled banking ............................................................................................ 258
II. BANK REGULATION IN THE CHILEAN FINANCIAL

LIBERALIZATION EXPERIENCE .................................................................................. 259
1. Measures and financial consequences of bank reform ............................................ 260
a) Unfavorable macroeconomic conditions ............................................................ 263
b) Prudential regulation failures .............................................................................. 265
c) Loan-portfolio growth and financial crises
as an outcome of "uncontrolled" banking............................................................ 267
2. The bank reform of 1986 .............................................................................................. 268
a) Transparency vis-a-vis credit risk and effective capital.................................... 268
b) Assumptions regarding financial stability
and self-regulating solvency measures................................................................. 270
c) Control of deposit interest ra tes............................................................................ 271
3. Bank reform, bank performance and financial liberalization ................................ 271
III. OVERVIEW OF RECENT FINANCIAL INSTABILITIES

IN LATIN AMERICAN AND CARIBBEAN COUNTRIES.......................................... 273
1. Sample of countries exhibiting financial instability ............................................... 273
2. Defective bank regulation and supervision
as a factor of financial instability................................................................................ 284
IV. CONCLUSIONS.................................................................................................................. 286
ANNEX ............................................................................................................................................. 289

B ank r e g u l a t io n , l ib e r a l iz a t io n a n d f in a n c ia l in s t a b il it y in . 245
INTRODUCTION
Since the mid-1970s financial liberalization has been part of structural re
form in Latin American and Caribbean countries. Liberalization of bank
credit and of interest rates and related measures on funding and on the
structure of the banking system have characterized policy measures. Only
recently have security markets and stock exchanges in some of the more
developed countries of the region become important to financial develop
ment.
Solvency and efficiency are closely interrelated criteria of bank per
formance. A solvent bank is able to pay its deposits and other liabilities
without drawing on its capital or own funds1. Bank efficiency can be apprai
sed at two levels: operational efficiency in carrying out financial and servi
ces transactions at minimum cost, that is, at a minimum "spread" between
lending and deposit rates of interest; and allocational efficiency when mak
ing loans and granting credit to socially profitable projects. Allocational
efficiency includes operational or cost efficiency as well as the capacity of
banks to attract deposits and make loans at competitive interest rates2.
This paper seeks to highlight the factors which determined the out
come of financial liberalization experiences in Latin American and Caribbean
1 Solvency is taken up when dealing with prudential regulation.

2 A bank which is perceived as solvent by depositors can attract funds and lend at lower
interest rates than otherwise, while attracting high-quality borrowers; a cost-efficient or low-
spread bank is also able to lend at lower interest rates than otherwise and attract low-risk
borrowers.
246 G ün th er H eld Y .
countries in the 1970s and 1980s. Emphasis is given to country cases where
these experiences led to solvency problems or severe financial crises.
The performance of banks as regards solvency and efficiency depends
mainly on the following:
a) Macroeconomic factors and policies determining output variability
growth and relative price changes. These factors affect the level and
variability of profit and income, i.e., the "first" loan repayment sour
ces of firms and households; they thus also affect risks of loss in
bank loan and asset portfolios.
b) Regulations and norms which set the "rules of the game” of banking
and the supervision of their compliance.
c) Internal bank management of loans, credit and financial investments
in terms of "safe and sound" portfolio practices.
Macroeconomic conditions and policies can have a major effect on
bank solvency and efficiency. Here, only a brief outline of such conditions
and policies will be provided as part of the environment in which finan
cial liberalization experiences took place. Bank management also ranks high
in determining bank performance. However, it will not be taken up except
for noting that loan, credit and financial-investment practices depend signi
ficantly on bank regulation and supervision.
The main issues of this paper refer to bank regulation and supervi
sion and their relations to liberalization and financial instability. The paper
consists of three sections. The first provides an account of the different
types of regulations which shape the framework of banking activities. It
deals with financial liberalization, prudential regulation and alternative
arrangements for controlling bank solvency. The second section looks at
the financial liberalization of the Chilean banking system. This experience
is of particular interest since it has gone full circle -from swift liberaliza
tion in the 1970s, to a heavy financial crisis at the beginning of the 1980s,
a far-reaching bank reform in the mid-1980s, and since then, to renewed
financial liberalization and proposed additional reforms to the banking
structure. The third section contains an overview of a large sample of fi
nancial liberalization experiences in Latin American and Caribbean countries
during the 1970s and 1980s which resulted in solvency problems and fi
nancial instability. This section seeks to highlight common factors behind
this outcome. The paper ends with a summary of conclusions.
I. BANK REGULATION AND SUPERVISION
Banking activities are usually subject to considerable regulations or norms

governing authorized financial transactions and services, financial flows and
stocks, the legal and accounting framework of banks, and other aspects.
Based on Marshall, three types of regulations will be distinguished: finan
cial, organizational, and prudential3.
1. Financial regulation
Norms seeking to control the flow and intermediation of funds are mainly
geared to price stability and efficient allocation of resources. The Central
Bank (or other financial regulator) may set norms on interest rates and
the exchange rate, taking into account macroeconomic conditions and the
competitiveness of domestic financial instruments (the latter should be able
to attract and keep funds denominated in domestic currency). It may also
regulate bank reserve requirements and foreign-capital flows to control the
growth of the supply of money, the channelling of credit in certain directions
(deemed socially profitable), and other aspects related to the pricing and
availability of financial resources.
Financial regulations may set interest rates a n d /or the exchange rate
too low, or below equilibrium levels. They may impose heavy reserve
requirements, low-yield public-sector financial investments and other "quasi
fiscal" burdens on banks, and put other restrictions on raising and allocat
ing funds, all of which reduce the real rate of growth and the real size
of funds in relation to GDP or other non-finandal magnitudes. According
to the financial liberalization approach, in this case, regulations cause fi
nancial "repression" and lead to a backward or "shallow" financial system,
represented typically by a few oligopolistic banks and a fragmented finan
cial system. Furthermore, in real terms, regulations on interest rates and
exchange rates would discourage savings and hence capital formation, ex
ports and the production of tradeables4.
3 E. Marshall, "El Banco Central como Regulador y Supervisor del Sistema Financiero", Reu
nión de Gobernadores de Bancos Centrales, Santiago, October 1991.
4 E. Shaw, Financial Deepening in Economic Development, (Oxford University Press, 1973).
248 G ü n th er H eld Y .
The financial liberalization approach interprets financial repression as

typical of inflationary economies with weak public finances and heavy
government "interventionism". On the basis of a sufficiently stabilized eco
nomy and strengthened public finances5, this approach advocates a policy
of financial deregulation or liberalization of interest rates and exchange ra
tes, as well as the abolition of directed credit allocation and other norms
hampering market forces (as part of a wider shift toward market oriented
policies)6.
High real interest rates and competitive rates of exchange are expec
ted to lead to financial "deepening", as the ratio of financial assets to GDP
increases. To sustain this process, financial and organizational regulations
have to lay the legal and institutional framework for new financial ins
truments and institutions able to pool and diversify funds on the savings
and investment sides of capital accumulation. Accordingly, the real effects
of financial liberalization are expected to show up as an increase in the
ratio of private savings (and hence of investment) and exports to GDP,
thus "reviving" growth7. As the financial and real effects of financial libera
lization become visible, market-determined interest rates and the rate of
exchange are expected to come down gradually and fall in line with the
reference standards for these macro-prices in an orderly managed economy:
positive but moderate real interest rates, and credible and sustainable ex
change rates (as approximations to their equilibrium prices in the medium
to long term).
5 A low and controlled rate of inflation is an important condition of bank performance and
financial liberalization. High and variable inflation rates change relative prices and profits in
ways which are difficult to predict. They foster higher credit risks, higher interest rates and
a shortening of the period of financial transactions.
A public sector with a financial surplus is also important to bank performance and financial
liberalization. A frequent policy to provide resources to an inadequately funded public
sector involves imposing high reserve requirements and other "quasifiscal" charges on banks.
The latter increase the spread through higher lending or active rates of interest and lower
deposit or passive interest rates. Strong public finances therefore mean both lower interest
rates and a greater availability of funds to the banking system.
6 R. Me Kinnon, Money and Capital in Economic Development, Washington D.C., (The Brookings
Institution, 1973).
7 E. Shaw, op. cit., pp 3-14.
B ank r e g u l a t io n , u b e r a u z a t io n a n d f in a n c ia l in s t a b il it y i n ... 249
2. Organizational regulation
Norms concerning the "industrial" structure of banking are aimed at cost

or operational efficiency of banks8 and fortifying the banking system as
a whole relative to financial stability.
Cost efficiency calls for norms promoting competition through entry
barriers which allow a sufficient number of banks in order to prevent mo
nopolistic practices, as well as banks of a sufficient size in order to take
advantage of economies of scale. Cost efficiency also depends on norms
defining the range of financial services, beyond granting loans and credit,
which are open to banks. These norms aim at broader defined banking
activities while taking advantage of scope economies in complementary fi
nancial services.
The large-scale capacity of banks to handle information and the in
creasing range o f financial services in conjunction with financial develop
ment have fostered the expansion of specialized short-term commercial
banking into "multibanking" and even into "universal” banking services9.
This expansion exposes traditional commercial banks to the risks of a wi
dening range of activities as well as to conflicts of interest stemming from
a "principal agent" problem in so far as better results of the latter may
affect the results of the former10. Arguments on behalf o f prudential regu
lation set out below emphasize that commercial banks are financially fra
gile. Therefore, when they move into multibanking or into universal bank
ing, norms regarding the structure of banking have to seek ways to "shield"
the capital and results of core banking activities from the performance of
parent firms offering other financial services11.
8 Cost efficiency translates into low spreads or margins between lending or active interest
rates and deposit or passive interest rates.
9 Universal banks are able to offer a full range of financial services.
10 Consider for instance a commercial bank which sets up a wholly owned mutual fund. If the
bank issues bonds which are bought up by the mutual fund at above market interest rates,
the profits of the bank will be lower and the profits of the mutual fund will be higher than
otherwise.
11 F. Morandé and J. M. Sánchez, "La Expansión del Giro Bancario Tradicional a los Negocios
No Bancarios", in La Banca Ante Nuevo Negocios y Mercados, (Santiago, Asociación de Bancos
e Instituciones Financieras de Chile, December 1992).
250 G ü n t h e r H e l d Y.
3. Prudential regulation
For the most part, prudential norms aim at preserving the solvency of
banks. The latter are financially fragile for two main reasons. First, a dis
tinctive feature of banks as firms is that payment of deposit or passive
interest rates and the stock value of deposits and liabilities are contracted
as fixed with the public, although lending or active interest rates and the
recovery of loans and financial investments (at currency denominations
which may partly differ from those pertaining to deposit and liabilities)
are subject to risk of loss, and are thus variable in nature. Second, the
financial structure of banks is highly leveraged12. Thus, the loss of a small
proportion of assets can wipe out all capital13 unless other of the bank's
own funds are at hand.
The (accounting) capital of banks represents an effective capital or
patrimony, provided all expected losses in assets (stemming from risk tak
ing) have been properly measured and fully provisioned (out of profits).
In this case, capital is an indicator of solvency and a measure of the abili
ty to withstand unexpected losses. As long as effective capital is positive,
a bank is able to pay all its deposits and liabilities out of its loans and
financial investments (without having to draw on its own funds)14. De Juan
has remarked that in the case of banks, insolvency invariably comes be
fore illiquidity (while in the case of firms located in the real sector, illi
quidity usually precedes insolvency)15. Thus non-payment of deposits and
other liabilities by a bank is a clear sign that it is fully insolvent.
12 The ratio of deposits and liabilities, or else their counterpart of loans and finandal invest
ments, usually exceeds capital by ten times or more.
13 If loans and finandal investments are equal to deposits and liabilities, this loss is equal to
the inverse of the leverage ratio.
14 Capital is the difference between correctly-valued assets and liabilities. If capital is positive,
assets exceed liabilities.
15 Banks are able to raise funds from the public and to sell finandal instruments out of thrir
portfolio to make cash. This allows them to meet their obligations though they may face
acute insolvency. See Aristóbulo de Juan, "From Good Bankers to Bad Bankers: "Ineffective
Supervision and Management Deterioration as Major Elements in Banking Crises", World
Bank, Washington, Country Economic Department, 1987, mimeo.
B ank r e g u l a t io n , l ib e r a l iz a t io n a n d f in a n c ia l in st a b il it y in .. 251
Non-payment of deposits and other obligations by some banks, parti

cularly the larger ones, runs the risk of causing major external negative
repercussions of a macroeconomic nature. It may lead to a "run" on banks
and a massive deposit withdrawal, or a contraction of liquidity and credit
in domestic financial markets as the public loses confidence in the banking
system. In foreign markets, it can undermine the credit-worthiness of the
country as a whole, and not only of insolvent banks. Similarly, a banking
system which exhibits a strong solvency position produces positive macro-
economic external repercussions. It boosts public confidence in the payment
system facilitating economic transactions; it diminishes the risks faced by
agents and firms located in the real sector enhancing financial transac
tions; it mobilizes domestic financial savings which may otherwise remain
idle, and allows access to foreign financial resources and the financing of
projects which may not be undertaken otherwise. The above positive and
negative external considerations are strong arguments in favour of con
trolling bank solvency.
Bank solvency problems can be dealt with via two principal mea
sures. The first deals with causes of insolvency which can be prevented
through "prudential" regulations and norms. The second refers mainly to
measures avoiding the risk of bank runs or large-scale deposit withdrawals
(which may result from non-deposit payment by some insolvent banks).
The latter measures will be considered first.
a) State guarantee on deposits
The State can provide a guarantee on deposits to limit the risk of bank
runs. However, insolvency risks increase due to the fact that a State gua
rantee "distorts" the retum-risk relation by diminishing or eliminating the
risk of losses to depositors and by stimulating banks to take on larger
risks.
The return-risk relation calls for depositors and savers to make their
decisions considering both the deposit interest rates offered by banks and
the repayment prospects of their funds (as determined by the riskiness of
banks' asset portfolios). However, as State guarantee coverage expands, de
posit interest rates offered by banks have an increasingly greater influence
on savers' and depositors' decisions as to which bank to choose. Evidently,
when there is a full State guarantee on deposits there is no asset risk for
depositors; the only variable guiding depositors and savers is the level of
the interest rate offered by different banks. The State guarantee on de
posits also distorts the retum-risk relation in bank portfolio management.
A broad State guarantee allows banks to assume greater risks when mak
ing loans and financial investments without being exposed to deposit
withdrawals themselves. Banks are therefore tempted to relax their credit
and investment allocation processes. They are also able to enter more risky
market segments without having to increase their deposit interest rates. Thus
the larger the guarantee, the greater the probability that the Central Bank,
as "lender of last resort", will have to cover the deposits of insolvent banks
and assume bank losses.
A very important characteristic of the State guarantee on deposits is
that it can be explicit or implicit. In the first case, there are written norms
which set out the terms of the guarantee. In the second case, the public
clearly understands that the Central Bank (or other public sector institu
tion) will stand behind the deposits of insolvent banks. While the explicit
State guarantee may be partial (leaving an element of risk in savings and
deposit decisions), the implicit guarantee is distinguished by the fact that
depositors and savers behave as if their funds are fully protected (that
is, their decisions will depend only on the level of deposit interest rates
offered by banks)16.
b) Deposit insurance
Deposit insurance has an important conceptual advantage in comparison

to the State guarantee on deposits. Insurance premiums payable by banks,
and eventually by depositors themselves, could be differentiated accord
16 Various factors can shape one's perception of an implicit State guarantee on deposits. The
following are all important; lack of bank liquidations in the past due to the rescue of banks
with solvency problems; the existence of public banks and large private banks in relation to
the size of the domestic credit market, and foreign banks, which depositors believe will not
be allowed to go under; introduction of a partial guarantee on certain deposits; and lack of
transparency of bank solvency. In the latter case, economic agents are unable to distinguish
among banks according to their risks. The State thus assumes somewhat of a moral duty to
protect the savings and deposits of banks which become insolvent.
B ank r e g u l a t io n , l ib e r a l iz a t io n a n d f in a n c ia l in s t a b il it y i n . 253
ing to the risk of banks losing their assets. However it is very difficult
to make practical use of this advantage. So far, measurement of the va
rious risks faced by banks does not rely on generally accepted procedures,
nor are accepted methods at hand to sum up these risks in an overall
index. For these reasons, insurance funds are generally built up on the
basis of uniform premiums charged on all banks, to cover listed deposits
up to certain amounts in cases of insolvency. Hus type of insurance has
similar effects as a partial guarantee on deposits17.
c) Norms controlling solvency
These norms seek to limit risks and control bank solvency mainly by re
lying on "self regulation" by market participants. They aim to encourage
bank directors and bank managers (in representation of shareholders) to
adopt "safe and sound" loan and financial investment portfolio policies and
decisions (taking on risks compatible with the high leverage of banks). At
the same time, they seek to encourage depositors to exert market disci
pline on banks by taking into account both deposit (or passive) interest
rates and the risk of loss of their funds at different banks when making
their decisions.
Table 1 shows the objectives and content of a typical set of pruden
tial norms intended for self-regulation of bank solvency. These norms
address the main factors acting in favour of the financial stability of banks
taking into account the macroeconomic external repercussions related to
bank solvency and insolvency.
First, strict norms for entering the banking system. These norms re
quire the "screening" of entrants and a barrier of minimum capital18.
Second, norms maintaining a solid capital base, net of expected risk
of loss, for established banks. Accordingly, solvency norms aim at:
• Limiting the risks which banks are able to take on when making
17 Depositors and savers tend to overvalue the deposit rate of interest banks in their decisions;
banks can allocate funds to more risky market segments without fearing a withdrawal of
deposits.
18 Entry conditions for solvency reasons may be more demanding than the barriers on account
of organizational regulations promoting competition.
254 G u n t h e r H e l d Y.
Table 1
PRUDENTIAL REGULATION OF BANK SOLVENCY
Objective of controls Content of Controls

Strict entry - Fairly high minimum entry capital
conditions ■ Standards of main shareholders, directors, and the head manager
Risk-taking •Ample loan and financial investment portfolio diversification (to

compatibl control risks arising from fund concentration in economic agents,
with high in economic sectors and in financial instrument)
leverage ratios •Written norms on loans and collaterals (to control individual cre
dit risks, particularly, of "related" loans (a)
■ Limits on active/passive funds with maturity mismatching accor
ding to their time profile (to control interest rate risks)
•Limits on active/passive funds with currency mismatching accor
ding to their currency denomination (to control exchange rate
risks)
■ Limits on fixed or immobilized assets and requirements of liquid
financial instruments (to control liquidity risks)
■ Control of other risks affecting banks (b)
Full recognition Accurate measurement of risks or expected losses in asset portfo

of measured risks lio
or expected losses Full risk provisioning or built up of reserves (out of profit)
in asset portfolio Suspension of interest accruals on very risky loans
Solid capital Speedy replacement of capital losses

base Minimum capital requirements according to the risks of different
assets
Transparency of Periodic information to depositors and to the public on the sol

bank solvency to vency position of banks (including measured risks in their assets,
depositors and amounts of provisions or reserves to cover expected losses, and
to the public effective capitals)
Orderly exit from the Orderly liquidation of insolvent banks

banking system Clear rules on preferential payment of certain deposits and obliga
tions (such as checking account, small-scale savings, and Central-
Bank credits)
(a) Loans "related" to shareholders, director, and bank managers

(b) Mainly refers to off-balance sheet transactions.
loans and financial investments, particularly credit risks (or the risk
of loan default), foreign-exchange risks (derived from lending or
investing in a currency denomination which differs from the one
pertaining to deposits and liabilities), and interest-rate risks (arising
from lending or investment at fixed interest rates at longer terms
than those for deposit and liabilities).
• An accurate measurement and full provisioning of all risks of loss
in the asset portfolio of banks, and the suspension of interest
accruals on high-risk loans.
• Building up the capital base of banks according to the risk profi
le of their loan and financial asset portfolio19.
The effectiveness of these norms in preserving the solvency of banks
depends significantly on the timely and periodic furnishing of reliable
information to depositors and the public concerning risk and capital in
dicators of banks, as well as how depositors and bank shareholders per
ceive their funds as being subject to losses depending on the riskiness of
bank assets.
Third, norms dealing systematically with financial instabilities of dif
ferent magnitudes, including timely recapitalization of banks with emer
ging solvency problems and orderly liquidation of insolvent banks. In the
latter case, norms must clearly indicate the precedence of payments of de
posits and liabilities out of the liquidation of assets, in favour of small-
scale savers and of the Central Bank as "lender of last resort".
4. Bank supervision
Bank supervision involves the activities of a public sector financial authori

ty (usually a Bank Superintendency or possibly the Central Bank) geared
to ensuring compliance with financial, organizational, and prudential re
gulations.
The efficacy of supervision in seeing that banks effectively comply
with the prevailing norms depends mainly on the following factors:
1. A strong supervisory authority endowed with legal and financial
19 Like the capital requirements agreed in the Basle Accord.

256 G ü n th er H e ld Y .
autonomy, capable of basing its activities on technical criteria and

of fending off political interests and pressure groups.
The supervisory authority should be sufficiently endowed with
highly qualified personnel in legal, accounting, financial, and in
formation processing matters, in order to provide technical sup
port for banking norms and their fulfillment. The cornerstone of
solvency supervision is a periodic assessment of the risks of loan
losses in the finandal-asset portfolios of banks, on the basis of a
reliable and uniform procedure.
2. A clear division of tasks among supervisory authorities of the fi
nancial system, once banks expand towards multibanking or uni
versal banking, in order to cover all financial transactions and ser
vices and to avoid the overlapping of supervision20. If a deposit
insurance fund is in operation, it is also important to separate its
tasks from those of the supervisory authority of the banking sys
tem in cases of bank insolvency.
3. Clearly-specified banking norms and well-defined and calibrated
sanctions to discourage non-compliance.
5. Arrangements for the regulation o f bank solvency
The regime pertaining to State guarantee or insurance on deposits (and

other bank liabilities) together with the set of "prudential" controls or
norms with regard to solvency, are the two main components of a sys
tem of bank solvency regulation21. According to Feller, the combination of
these two components yields the four institutional arrangements on regu
lation of bank solvency shown in Table 2.
20 This is easier to achieve by organizing supervision according to functions or financial ser

vices, i.e. credit, loans, insurance and others, instead of financial institutions, i.e. banks,
insurance companies, and so on.
21 This section admits that if deposit insurance is available, premiums, are uniform, and thus
insurance is similar to a partial guarantee on deposits.
B ank r e g u l a t io n , u b e r a l iz a t io n a n d f in a n c ia l in s t a b il it y in .. 257
a) Free or non-regulated banking (as regards solvency)22
In a system of free or non-regulated banking, there is no State guarantee

on deposits nor rules controlling bank solvency. Depositors and savers
assume substantial risks when placing their funds. Accordingly, they are
expected to exercise great precaution when choosing a bank. Macroecono
mic externalities of banking activities in developed or formal financial
markets appear to have turned free banking into a theoretical alternative.
b) Public-sector controlled banking
A large, explicit State guarantee on deposits, or an implicit State guaran

tee, induces savers and depositors to fully ignore the insolvency risks of
banks. They are motivated only by the highest deposit interest rates
offered by banks for their funds. Thus, maintaining bank solvency de
pends entirely on prudential regulation controls (and on its supervision).
This task calls for a rigorous definition of the set of norms mentioned
in table 1.
c) Bicontrolled banking
The absence of an explicit or implicit State guarantee, or of deposit in

surance, encourages depositors and savers, especially those handling large
amounts, to assume an active market-disciplining role by assessing banks
according to their risks and deposit interest rates23. To perform this func
tion, depositors and savers must be informed about bank solvency indi
cators. At the same time, prudential regulation norms limit bank risks and
establish risk provisioning and capital requirements. Since both depositors
and a public-sector financial authority assume solvency control functions,
this arrangement is tantamount to "bicontrolled" banking24.
22 Free banking differs from non-regulated banking as regards solvency, in that the former is
not subject to financial and organizational norms nor to prudential norms.
23 Banks carrying larger risks in their asset portfolios will have to pay higher deposit rates of
interest.
24 In this arrangement prudential norms should be gauged taking into account the market dis-
Table 2
INSTITUTIONAL ARRANGEMENTS FOR BANK SOLVENCY REGULATION (a)
Solvency controls
Deposit Without solvency With prudential
Guarantee controls regulation controls
Without deposit Free or non Bicontrolled banking

guarantee regulated Banking Banks subject both to the
Banks subject to the control of depositors and
control of depositors of a public-sector financial
authority
With deposit Uncontrolled Banking Public-sector controlled

guarantee (explicit or Banks not subject to banking
implicit) controls of depositors Banks subject to the control
nor of a public-sector of a public-sector financial
financial authority authority
(a) Adapted from A. Feller, "Supervisión, Regulación y Riesgos Bancarios", Superintendencia de Bancos
e Instituciones Financieras, Información Financiera (Santiago, Chile, May 1985), pp. ix-xi.
d) Uncontrolled banking
A total explicit or implicit State guarantee on deposits leads depositors and

savers to take into account only interest rates when making their deci
sions, fully ignoring bank solvency risks. At the same time, there is no
prudential bank regulation. Banking norms refer mainly to accounting
procedures and financial norms concerning reserve requirements, directed
credit allocations, and financial investment in public-sector instruments. In
this arrangement neither depositors nor a public-sector financial authori
ty are concerned with bank solvency risks, thus representing a situation
of "uncontrolled" banking as regards solvency.
Deregulating or liberalizing credit in this context can easily lead to
dpline role of depositors and savers. Too rigorous norms can lead banks to adopt highly
conservative loan and financial policies, taking on less risks than are socially desirable.
Loans for small -and medium- scale firms, especially agents wishing to start an economic
activity, may get little credit since prudential norms may consider that they are too risky.
Thus bank solvency indicators may be very good, although the establishment may not
finance socially profitable investment projects.
B ank r e g u l a t io n , l ib e r a l iz a t io n a n d f in a n c ia l in st a b il it y in . 259
excessive high risk credit due to the lack of "filters" or restrictions on de
mand for bank funds. At the level of firms and banks, this translates into
a significant number of high-risk projects, uncollectable loans and large ex
pected bank losses. The counterpart at the macroeconomic level is excessi
ve aggregate expenditure, high real interest rates (and inflationary pressu
re). This process can be accumulative as very high real interest rates drag
down solvent bank debtors25.
II. BANK REGULATION IN THE CHILEAN FINANCIAL

LIBERALIZATION EXPERIENCE
The Chilean financial liberalization of the mid-1970s resulted from an eva

luation of economic policies which had "repressed" bank credit, savings and
exports owing to import substitution and interventionism in previous de
cades26. These reforms also envisaged opening the economy up to foreign
trade and capital, and shifting resources and decision-making to the priva
te sector. Since the economy was faced with triple-digit inflation between
1974 and 1975, stabilization measures were implemented simultaneously
with structural reforms27.
The banking system which prevailed in 1974-75 was made up of spe
cialized banks, mainly commercial banks, limited to short-term transactions.
25 As Feller points out, there comes a point when the government perceives that the situation
is unsustainable. But at this stage, the restoration of solvency of both banks and debtors
invariably means large financial losses for the State. (A. Feller, 'Supervisión, Regulación y
Riesgo Bancario”, Información Financiera, Santiago, Superintendencia de Bancos e Institucio
nes Financieras, May 1985).
26 The setting of (nominal) interest rates and the rate of exchange in inflationary conditions
frequently led to negative real rates of interest and, most of the time, to a substantially
overvalued domestic currency. At the outset of liberalization policies in 1975-76, bank loans
averaged only 9.3 per cent of GDP, national savings 11.2 per cent of GDP (though 14 per
cent represented historical levels), and exports of goods and services 25.2 per cent of GDP
(though 20 per cent were closer to historical levels).
27 For short but forceful statement regarding the evaluation and changes in the economic
policy régime see S. de Castro, "Exposición sobre el Estado de la Hacienda Pública”, Boletín
Mensual, Banco Central de Chile, January 1978, and P. Barahona and M. Gómez, "Esfuerzos
de Estabilización en Chile", Boletín Mensual, Banco Central de Chile, October 1975.
260 G ü n t h e r H e ld Y.
Virtually all banks had also been taken over by the State in 1970-73. The
banking reform thus had two main components. First, financial liberaliza
tion of credit and interest rates and related measures concerning reserve
requirements, a broader funding base and access to foreign capital. Second,
organizational measures aimed at a private and competitive banking sys
tem and the first steps toward multibanking by authorizing banks to un
dertake longer term transactions and mortgage operations.
1. Measures and financial consequences of bank reform
Bank privatization proceeded swiftly after September 1975. Virtually all

banks had been sold to the private sector by the end of 1978, with the
exception of the State Bank. Banks were privatized at the same time as
a large number of firms in the real sector, which had also been taken
over by the State in 1970-73. This procedure soon led to economic "groups",
as bank owners used their banks' credit to purchase firms offered for sale.
Interlocking ownership and management of banks and firms in the real
sector would enhance the meaning of "related" loans in credit liberaliza
tion.
Administrative and legal barriers to the banking system were lifted
with the aim of fostering competition. Foreign banks could enter the do
mestic market at the end of 1974 while foreign investment in banking be
came possible in January 1977. At the end of 1974 the Superintendency
of Banks regulated the activities of financial establishments28. Between 1974
and 1981, the number of banks increased from twenty-one to forty-five,
foreign banks went from one to eighteen, while seventeen financial esta
blishments were in operation at the end of 1981.
Quantitative credit allocations were lifted at the end of 1973. In June
1977, the authority of the Central Bank to establish credit guidelines was
severely limited. After partial deregulation of interest rates in 1974-75, they
could be freely negotiated starting in December 1975. Taking account of
ongoing inflation, all debt transactions due in more than ninety days
28 Financial establishments are unable to offer checking accounts and to finance foreign trade
transactions. They devote themselves mainly to financing household purchases of durable
consumer goods.
B ank r e g u l a t io n , u b e r a u z a t io n a n d f in a n c ia l in s t a b il it y in .. 261
could be indexed to the CPI after July 1976. Reserve requirements for sight
and time deposits were reduced starting in 1975 and 1976, and levelled
out at 10 per cent and 4 per cent respectively in 1980. Restrictions on the
acquirement of foreign debt by banks were gradually removed after De
cember 1977 and virtually abolished in April 1980. However, controls were
kept on outflows of capital. These were instrumental in maintaining foreign
funds in the banking system and in increasing the financing role of the
latter. In 1981 all loan and finandal-asset portfolio growth of the banking
system was due to foreign debt financing.
The financial liberalization measures were soon followed by abnor
mally high rates of growth of the banking system's loan (and finandal-
asset) portfolio and of interest-rate levels, as well as by extremely large
interest spreads29. Column 1, Table 3 shows these figures for 1978-80, four
years after financial liberalization got under way in order to dampen the
effects of inherited financial "repression" on the behaviour of banking va
riables. Loan (and finandal-asset) portfolios grew on the average at almost
40 per cent a year in real terms (equivalent to five times the GDP growth
rate), while average real rates of interest on the outstanding stock of loans
(and finandal assets) stood at around 20 per cent a year (equivalent to
two and a half times the GDP growth rate). Gross spreads were almost
10 per cent a year. They covered provisions and write offs of risk-prone
loans of 1.7 per cent a year (indicating an apparently sound portfolio growth),
and hefty administrative and other current expenses of 5.2 per cent a year
(suggesting operational ineffidendes). Gross spreads also provided for a
substantial profit rate of 22 per cent on capital (achieved with an average
asset ratio of nine). The above indicators reflected a booming banking sys
tem which looked profitable and safe.
Yet a solvency crisis of unexpected magnitude erupted in 1981, lead
ing to large-scale intervention of the banking system. Between 1981 and
1984 the Superintendency of Banks and Financial Institutions intervened
in fourteen private domestic banks, induding the two largest private banks
in 1983, and eight finandal establishments, out of twenty-six domestic pri-
29 On the behaviour of interest rates see R. Zahler "Las tasas de interés en Chile: 1975-82", El De
sarrollo Financiero de América Latina y el Caribe 1985, (Caracas, Instituto Interamericano de
Mercados de Capital, 1985).
Table 3
CHILE
Bank Behaviour at the Peak of Financial Liberalization (1978-80) and after
the Recovery of Solvency and Prudential Regulation Reform (1986-91)
(in %, with the exception of the asset leverage ratio)
1978-80 1986-88 1991-92

Real rate of growth per year (%)
a) Loan and financial-asset portfolio 38.7 -3.1 6.5
b) Deposits and liabilities 39.8 4 .9 7.5
Gross Domestic Product (GDP) 8.1 6.2 8.2
Real average rates of interest per year (%)
a) Loan and financial-asset portfolio 19.6 8.2 8.7
b) Deposits and liabilities 13.0 5.9 6.1
Yearly spread as % of assets (a)
a) Interest rate spread 6.9 2.6 3.0
b) Gross spread 9.8 5.0 6.1
Yearly expenditure as % of assets (a)
a) Provisions and loan write offs 1.7 1.0 1.0
b) Administration expenses 5.2 2.2 3.5
Results before tax as % (a)
a) Assets (a) 2.4 1.5 1.5
b) Capital and reserves 21.7 21.3 19.2
c) Asset leverage ratio = (b/a) 9.0 14.2 12.8
(a) Total bank assets.
S ou rce Información Financiera, Superintendencia de Bancos e Instituciones Financieras
Santiago, December issues of 1978 to 1991
vate banks and seventeen finandal establishments which were in operation

in 1981. Banks and financial establishments which had grown at specta
cular rates before the crisis virtually all underwent intervention and liqui
dation. Eight banks and all finandal establishments affected by interven
tion were dosed down.
The generalized financial crisis among domestic banking institutions
and debtors called for large-scale recovery measures. Bad loans which had
been sold to Central Banks by the end of 1985 were equivalent to 28 per
cent of the outstanding loan portfolio, 18 per cent of GDP, and on the
average over three times the capital of banks and of finandal establish
ments which made use of this alternative. But if one takes into account
bad loans made by dosed banks and finandal establishments, and interest
B ank r e g u l a t io n , l ib e r a l iz a t io n a n d f in a n c ia l in s t a b il it y i n .. 263
rate and exchange rate subsidies involved in large-scale debt reprogram

ming, the estmated financial losses of the Central Bank were around 40
per cent of GDP of 198930.
In seeking to explain these extraordinary losses under financial libera
lization, two main factors stand out: unfavourable macroeconomic condi
tions and the widespread failure of prudential regulation and supervision.
a) Unfavourable macroeconomic conditions
The simultaneous implementation of structural reforms and stabilization

policies (and occasionally inconsistent economic policies, economic shocks,
overly ambitious growth expectations, and other events), frequently led to
sharp changes in relative prices and profitabilities among firms and econo
mic sectors, with subsequently erroneous profit and risk signals31. Excessi
ve bank lending, along with high credit risks and interest rates, are thus
interpreted as "transmission mechanisms" of macroeconomic factors.
The following macroeconomic policies led to large relative price and
profit shifts between economic activities, and consequently to substantial
credit risks and solvency problems among banks and debtors:
• In 1974-77, a foreign-trade reform which lowered nominal tariffs
from an average of 105 per cent in 1974 to 10 per cent in 1977,
amid macroeconomic adjustment which reduced GDP by 13 per
cent in 1975 and a stabilization policy which brought inflation
down from the triple-digit level in 1974-76 to around 60 per cent
in 1977.
• In 1978-82, the effects of the adoption in February 1978 of the no
minal exchange rate as a stabilizing factor in the form of fixed,
preannounced and diminishing rates of devaluation, in the presence
o f high domestic interest rates and wage indexation clauses. This
policy brought inflation down, but a rate comparable to that of
industrialized countries was not achieved until early 1982. In the
30 N. Eyzaguirre and O. Larrañaga, Macroeconomia de las operaciones cuasífiscales en Chile, (Santia

go, SECLAC/UNDP, August 1990).
31 S. Edwards, Stabilization with Liberalization: An Evolution of Ten Years of Chile's Experiment
with Free Market Policies 1973-83, (EDCC, University of Chicago Press, January 1985).
meantime, the virtual elimination of foreign-exchange risk had pro

duced a wide gap between real lending rates in domestic currency
and in foreign currency (of ex-post 13 per cent in 1980 and 40
per cent in 1981), and stimulated large foreign-debt-financed capital
inflows. The latter led to yearly domestic expenditure increases of
over 10 per cent in real terms between 1978-81, and in conjunction
with wage indexation, to an accumulated drop in the real exchan
ge rate of 33 per cent by mid-1982. The profitability and solvency
of export and import competing sectors had gone aground, while
real estate and other non-tradeables were getting a larger share of
bank credit32.
• In 1982-1983, the effects of a drastic balance of payments adjust
ment, on account of the abrupt interruption of foreign capital flows
following the Mexican debt moratorium of March 1982. GDP fell
15 per cent, and after the exchange rate started to float in June
1992, its real value increased around 70 per cent by the end of
1983. These adjustments caused solvency problems for debtors lo
cated in non-tradeable activities and in other sectors.
The above macroeconomic conditions are certainly important in ex
plaining the large-scale financial crisis of the early 1980s. However, for the
following reasons they are also quite insufficient. The banking-system loan
portfolio had grown by around 38 per cent in 1980 (as compared to 9
per cent in 1975-76), suggesting that credit "repression" was no longer im
portant. Yet, the loan portfolio was growing at 46 per cent in real terms
in that year and real lending rates of interest still stood above 18 per cent,
both quite out of line with GDP growth or other macroeconomic activity
indicators. Then, the financial crisis started in 1981, after five years of strong
growth, while the economy was growing at 5.5 per cent a year, and befo
re the severe macroeconomic adjustment of 1982.
32 J.A. Fontaine, "Crecimiento, recesión y mercado", Estudios Públicos (Santiago), Ne 11,1983; V.

Cortx), J. de Meló and J. Tybout, "What went wrong with the recent reforms in the Southern
Cone", (Washington D.C.: World Bank, 1985, Report DRD128).
b) Prudential regulation failures
Vigorous prudential regulation and supervision were particularly required

in a rapidly liberalized banking system with insufficient loan management
experience, and which started extending credit to firms very short of funds
following the economic and political turbulence of 1970-73. However, the
only noticeable measure in favor of bank solvency which was taken dur
ing financial liberalization was the increase, on various occasions between
1974 and 1980, of minimum capital requirements for banks and financial
establishments33.
By 1977, the distinctive features of "uncontrolled" banking as regards
solvency (discussed in section I) can clearly be seen: lack of prudential
norms controlling solvency and a full State guarantee on deposits. Under
these conditions, banks were not merely recipients of high-risk loans due
to macroeconomic imbalances and distorted macroeconomic prices. They
were able to put in place a high-growth high-risk credit "mechanism"
which generated solvency problems as soon as credit and interest rates
had been liberalized. Thus excessive loan-portfolio growth reflected both
defective norms controlling the demand for bank credit, as well as an inabi
lity to regulate the supply of funds due to the State guarantee on deposits.
Public-sector financial authorities exerted no prudential regulation and
supervision, nor did depositors play a role in disciplining the market.
i. Defective nom s controlling solvency
Essential aspects of the credit process were at fault: guarantees or colla

terals as a second source of loan repayment, limits on individual credit
to avoid risk concentration, and the handling of high-risk overdue loans.
Very weak collaterals are evident in "goods received in payment" for
bad loans. Of the loans which banks and financial establishments sold to
the Central Bank in 1981-83 as part of measures to rescue troubled insti-
33 The minimum capital of banks went up from US$ 2.5 million to around US$ 10 million; that
of financial societies from US$ 0,4 million to around US$ 5 million. The financial crisis of the
early 1980s demonstrated the pitfalls of attempting to face solvency problems by merely
strengthening capital bases. The latter proved totally insufficient to cover accumulated losses.
tutions, only 4 per cent were backed up by such goods at the end of
1985. Limits on individual credit were amply exceeded in the case of "re
lated" loans by lending to fictitious firms and other practices. It is estima
ted that up to 50 per cent of bad loans which banks undergoing inter
vention in 1981-83 sold to the Central Bank were "related" (and had vir
tually no collateral). Indirect evidence regarding interest capitalization and
the rolling over of overdue loans is provided by the fact that while loans
were growing at abnormal rates, interest-income accruals by banks were
very high; after they had sold their large stock of bad loans to the Central
Bank, both the growth rate of their loan portfolio and the lending rates
dropped sharply (see second and third column of Table 3).
Not until 1980 did the Superintendency of banks introduce a method
to classify loans according to risk categories. The public was evidently not
informed of expected losses; however, these losses led to intervention in
the banking system starting in 1981, in order to gauge the magnitude of
the recovery measures. Therefore, neither the Superintendency nor depo
sitors and shareholders had access to reliable information on portfolio qua
lity -and consequently on effective results and capital figures of bank and
financial establishments- in the course of financial liberalization. The high
profitability shown in column 1 of Table 3 was fictitious as it followed
mainly from accounting as income interest accruals on high-risk loans and
from insufficient provisioning of these risks.
ii. State guarantee on deposits
Lack of public information on portfolio risks and veritable solvency indi

cators of banks and financial establishments opened the door to an implicit
State guarantee on deposits (and other liabilities). An unmistakable step
in this direction had already been taken in 1975-76 when an emerging fi
nancial crisis led to intervention in and recovery of a medium-sized com
mercial bank, and to the introduction in January 1977 of a State guarantee
on deposits denominated in domestic currency equivalent to around US$
3,000 (at the current rate of exchange). In January 1991, this guarantee was
complemented by a voluntary deposit insurance covering additional depo
sits in domestic currency up to the equivalent of around US$ 3,500. De
positors completely overlooked this option, in the belief that the State was
fully protecting their funds.
B ank r e g u l a t io n , u b e r a u z a t io n a n d f in a n c ia l in s t a b il it y in . 267
c) Loan-portfolio growth and financial crises as an outcome of

"uncontrolled" banking
The mechanism of abnormal loan-portfolio growth contained large flows

of "related" and other non-collateralized credit, the capitalization of high
accrued lending rates of interest and the rolling over of loans subject to
risk. The process leading to a credit "bubble" can be described by dividing
the real rate of growth of bank loan portfolios into two components: the
rate of growth of normal-risk or "legitimate" credit (1), and the rate of growth
of high-risk or "false" credit (f), defining both rates in the outstanding port
folio and when granting or renewing loans. False credit may in turn be
split up into the rate of growth of new "false" credit (fl) and capitalized
interests stemming from the rolling over of high-risk credit (f2). If "a" is
the proportion of interest accruals which is paid and "r" is the real interest
rate on the loan portfolio, f2 = (1 - a)r stands for capitalized high-risk
interest accruals. The real rate of growth of the loan portfolio (p) can thus
be written:
p = 1 + f = 1 + fj + (1 - a)r (1)
The annex contains a rough simulation with a breakdown of the 39

per cent yearly growth rate of the loan and financial asset portfolio of
the banking system in 1978-80. "Legitimate" credit would have expanded
at around 22 per cent in real terms a year, or at three times the GDP
growth rate, while "false" or very high-risk credit would have grown at
17 per cent in real terms a year, or at twice the speed of GDP. Of the
latter, about 10 per cent may stand for the capitalization of interest rates
and 7 per cent for the granting of new "false" credit.
The implicit State guarantee on deposits and liabilities stands out
when one seeks to explain the behaviour of debtors and bankers as regards
agreeing on "false" credits, and of depositors and lenders to the banks as
regards providing finance. Once the former perceived that insolvency pro
blems were becoming widespread, collusion on rolling over bad loans,
capitalizing interest, and extending related credit became profitable in the
belief that the State would provide a large-scale bail-out34. Once the latter
perceived that they could count on a State guarantee, depositors and even
foreign banks were willing to provide funds.
2. The bank reform o f 1986
The bank reform of November 1986 aimed at an institutional arrangement

of "bicontrolled" banking as regards solvency regulation (as set out in Ta
ble 2 above), and at multibanking vis-à-vis the structure of the banking
system.
The reform introduced all the prudential solvency norms mentioned
in Table 1. At the same time, it withdrew the State guarantee on deposits,
excluding only sight deposits which earned no interest and small-scale sa
vings35, with a view to incentivating the market-disciplining role of depo
sitors. Multibanking was extended by authorizing complementary finan
cial services to grant loans and credit (leasing, administration of mutual
funds, securities dealings and investment bank activities). The provision of
these services require affiliated but independent establishments with their
own capital, and are subject to limits as regards bank lending, so as to
fully shield the banks' own capital and results.
The prudential regulation framework contained novel features which
provided transparency concerning risk and capital among banking institu
tions, set up self-regulating measures to recover solvency, and relied on
the level of deposit rates of interest as an indicator of solvency problems.
a) Transparency vis-à-vis credit risk and effective capital
Table 4 defines credit risk categories for the loan portfolio and an example
of credit risk assessment of two banks starting in May 1987. The lower
part of this table shows the "effective" capital position of the same two
banks also starting May 1987, when the above information was made pu
blic for the first time by the Superintendency of Banks. The capital p o
sition is shown by means of a coefficient where the numerator is "effec
tive" capital, the denominator accounting capital, and with the following
variables:
34 N. Eyzaguirre, "Crisis Financiera, Reforma y Estabilización: La experiencia Chilena", Banco

Central de Chile, Santiago, 1992, mimeo.
35 The State guarantee covers time deposits by natural persons equivalent to around US$ 2,400
(at the rate of exchange at the beginning of 1993), with a limit of 90 per cent of this amount.
C = Accounting capital and reserves (excluding provisions for assets

subject to risk)
Prov = Provisions on risks of loss of loans and financial assets
R = Current result (profit or loss)
L = Expected risk of loss in loan andfinancial-assetportfolio
The effective capital coefficient (ecc) is36:
ecc = ((C + Prov + R - U /Q 1 0 0 (2)
Table 4
CHILEAN PRUDENTIAL REGULATION BANK REFORM OF 1986 TRANSPARENCY
OF CREDIT RISK AND OF EFFECTIVE CAPITAL OF BANKING INSTITUTIONS
1. Credit risk categories Category Expected losses in %

Normal risk A 0
Potential risk higher than normal B 1
With expected losses B- 20
With substantial expected losses C 60
Non-recoverable D 90
2. Expected losses as percentage of loan portfolio

Bank May 1987 May 1990 February 1993
American Express 3.51 3.36 1.98
Concepción 3.51 3.58 3.24
Average for banking system 6.58 4.18 2.43
3. Effective capital coefficient, ecc = (Effective capital/Accounting capital) 100

Bank May 1987 May 1990 February 1993
American Express 105.7 106.9 102.6
Concepción 108.8 129.5 116.3
Average for banking system 107.0 116.9 106.4
Source: Superintendencia d e Bancos e Instituciones Financieras, Información Financiera, Santiago, diffe
rent issues.
36 Thus, if ecc > 100 or if ecc < 100, effective capital exceeds or falls short of accounting capital
by the percentage difference to hundred.
270 G ünther H eld Y.
b) Assumptions regarding financial stability and self-regulating

solvency measures
Assumed financial problems are defined on the basis of the "percentage

of capital at risk" (per), which is the complement of the effective capital
coefficient when the latter is less than 100 per cent (that is, when effective
capital falls below accounting capital):
per = (100 - ecc)%, ecc < 100 (3)
If the percentage of capital at risk is approximately between 20 per

cent and 40 per cent, that is, when expected asset losses (not covered by
provisions) threaten 20 per cent to 40 cent of the accounting capital37, pru
dential regulation presumes "financial instability". To overcome this situa
tion, shareholders are required to supply fresh capital on short notice38.
If on the other hand, the percentage of capital risk reaches 60 per cent
or more, that is, if expected asset losses (not covered by provisions) threa
ten that percentage of accounting capital, or more, prudential regulation
presumes "severe solvency problems". In this case, in addition to fresh ca
pital supplied by shareholders, a committee of bank creditors can trans
form deposits into capital and take other measures to reduce deposits and
liabilities in order to re-establish the solvency of the bank or financial so
ciety39.
37 The precise percentage of capital at risk depends on the leverage ratio of the bank of
financial establishment. The higher the leverage ratio, the lower the admitted percentage of
capital at risk.
38 If this capital is not forthcoming or deemed insufficient by the Superintendency, the bank or
financial establishment in question is not permitted to expand its loan and financial-asset
portfolio, with the exception of the purchase of financial instruments issued by the Central
Bank.
39 Full insolvency is not defined, but banks and financial establishments assumed to incur
"severe solvency problems" run the risk of being liquidated if no corrective measures are
forthcoming. If the Superin tendency of Banks decides to liquidate, sight deposits and small
savings are protected by the State guarantee, while the Central Bank enjoys priority in
recovering both its loans to the affected financial establishment and its disbursements to
cover protected deposits.
B ank r é g u l a t io n , l ib e r a l iz a t io n a n d f in a n c ia l in s t a b il it y i n .. 271
c) Control of deposit interest rates
The link between high growth of non-performing loans and pressure on

deposit interest rates (as the former do not provide a cash flow and hen
ce require the raising of funds) led to a further assumption of financial
instability. Banks or financial establishments which paid interest rates ex
ceeding the average rate paid by the same kind of financial institution in
three or more months of a same year by a fifth or more are also presumed
unstable. Consequently, these banks and financial establishments are not
allowed to increase their loan and financial-asset portfolio (with the excep
tion of financial instruments issued by the Central Bank), until their share
holders supply fresh capital deemed adequate by the Superintendency of
Banks.
3. Bank reform, bank performance and financial liberalization
The vigorous prudential regulation reform implemented at the end of 1986,

the related measures to restore the solvency of banks and debtors between
1982-86, and the orderly macroeconomic conditions prevailing since 1985-
86 led to a structural change in the performance of banks and to a new
environment for financial liberalization.
The change in bank performance took place while fully maintaining
credit and interest-rate liberalization40. All banking institutions have remai
ned solvent since 1986. At the same time, as the second and third columns
of Table 3 show, after initially falling, the rate of growth of the banking
system's loan and financial-asset portfolio has recently kept in line with
the rate of growth of GDP. Lending (or active) and deposit (or passive)
interest rates dropped by half, although the latter have remained attractive
in real terms; spreads also dropped by half on account of similar reductions
in portfolio expenditures, approaching international standards.
National savings increased from 10 per cent of GDP in 1985-86 to 22
per cent in 1990-92, an increase made up mainly of private savings (of
40 The Central Bank "suggested" short-term deposit interest rates between 1982-86, while the
financial crisis was overcome.
which around 3 per cent is due to the pension-fund reform of 1981); exports
of goods and services reached a peak of 35 per cent GDP in 1990-92, backed
by record high real export rates of exchange in 1988-91; inflation has been
kept at nearly 20 per cent or less a year (in a highly indexed economy)41
and GDP has been expanding at above 6 per cent a year, both since 1986.
The "revival" of growth and the behaviour of financial and real variables,
quite akin to the results predicted by the financial liberalization approach,
merit close attention under this policy and relative to the interrelationship
among these variables42.
In the meantime, the highly successful bank reform of 1986 created
new issues. It only partially achieved the objective of establishing a regime
of "bicontrolled" banking as regards solvency. Although pension funds,
insurance companies and other institutional investors follow closely the sol
vency indicators of banking institutions shown in Table 4, there is evidence
that a large proportion of depositors still behave as if their funds were
fully protected by the State43. What was achieved may be closer to public-
sector-controlled banking. This has led to proposals which enlarge the role
of the private sector in evaluating the risk of the loan and financial-asset
portfolio pertaining to banking institutions, and which render the risk of
deposit capitalization and of deposit losses in cases of insolvency feasi
ble44. Rapid development of the capital market since the mid-1980s brought
about by the pension fund reform of 1981 has also led the banking sys
41 The CPI measured inflation was 27 per cent in 1990 but approached one digit rates at the
end of 1992.
42 While the real rate of exchange appears as essential in explaining the expansion of exports
and tradeables; it can be argued that a further fall in interest rates requires more national
savings (instead of attempting to rely on higher rates to incentívate household savings).
43 The following factors may have contributed to this result: the explicit guarantee on sight
deposits and on small scale deposits, the preponderant role of the Superintendency of Banks
in measuring risks and in controlling solvency, and the requisite that banks and financial
societies have to repurchase the bad loans which they sold to the Central Bank and are thus
hardly expected to break down.
44 Since Institutional investors keep themselves informed on the solvency position of banks
and financial societies, they will withdraw their deposits before the "capital at risk" of a
bank reaches the critical ratio of 60 per cent, that is, when deposit capitalization under a
creditors committee is to take place.
tem to seek further expansion of its activities including insurance, pension-

fund management, and a wider array of international financial transac
tions45.
III. OVERVIEW OF RECENT FINANCIAL INSTABILITIES IN

LATIN AMERICAN AND CARIBBEAN COUNTRIES
1. Sample o f countries exhibiting financial instability
Table 5 provides an overview of countries in the region where banks and

other credit institutions have faced solvency problems in the last two de
cades.
Argentina, Chile and Uruguay make up the well-studied "Southern
Cone" experiences of the 1970s dealing with stabilization, economic reform
and financial liberalization46. On account of the entrenched import-substi-
tution industry in Argentina and Uruguay, financial liberalization took place
before commercial reform. Between 1974 and 1979 in both countries a num
ber of measures liberalized interest rates and bank credit, the exchange
rate and foreign-exchange operations, as well as reducing reserve require
ments and other quasi-fiscal charges on banks. In 1974 both countries libe
ralized private inflows and outflows of capital.
Argentina and Uruguay followed Chile's lead in carrying out financial
liberalization amidst a very unstable macroeconomic environment and se
vere shortcomings in prudential regulation an d/or supervision.
Yearly inflation rates of 170 per cent in Argentina and 50 per cent
in Uruguay in 1977-78 led both countries to adopt the nominal rate of
exchange as an anchor of stabilization at the end of 1978. Annualized in
flation rates still stood at 70 per cent in Argentina and 25 per cent in
Uruguay at the beginning of 1981, while the real rate of exchange had
dropped to half in both countries. Contributing to this situation were pre-
45 La Banca ante Nuevos Negocios y Mercados, Santiago, Asociación de Bancos e Instituciones

Financieras de Chile, 1992.
46 ECLAC, Estabilización y Liberación Económica en el Cono Sur, Santiago, Estudios e Informes
CEPAL, 1984, Na 38.
Table5
BANK SOLVENCY PROBLEMS IN LATIN AMERICAN AND CARIBBEAN COUNTRIES
Key Prudential re Out Come

Country Period Financial regulation macroeconomic gulation and Stability of fi Financial
conditions supervision nancial system prices
Argentina 1974-81 Credit and interest rate Fiscal imbalance and Severe shortco Large banking cri High real rates of
liberalization. high inflation. Stabiliza mings. ses. interest. Lagging
Early opening of capi tion based on the rate of rate of exchange.
tal account. exchange (1979-81).
Chile 1974-82 liberalization of credit Structural reforms. Stabi Severe shortco Generalized bank High real rates of
and interest rates. lization based on the rate mings. ing crises. interest. Lagging
of exchange (1978-1982). rate of exchange.
Uruguay 1974-82 Credit and interest rate Current account imba Severe shortco Large banking cri High real rates of
liberalization. lance. Stabilization based mings. ses. interest. Lagging
Early opening of capi on the rate of exchange rate of exchange.
tal account. (1979-1982).
Colombia 1979-82 Sectoral credit alloca Fairly stable macroeco Severe shortco Large banking cri Fragmented inte-
tions. nomic conditions. mings. ses. ' rest-rate structu
G ünther H eld Y.
re.
El Salvador 1980-89 Directed credit and set Large fiscal deficit. Infla Severe shortco Large banking cri Financial repres
ting of interest rates. tionary pressure. mings. ses. sion.
Table 5 (Cont.)
Bank
Out Come
regulation, liberalization
Key Prudential re
Country Period Financial regulation macroeconomic gulation and Stability of fi Financial
conditions supervision nancial system prices
Dominican 1982-90 Sectoral credit limits Fiscal imbalance. Grow Severe short Financial crises in Financial repres
Republic and setting of interest ing inflation. comings. non-regulated fi sion of regulated
rates. nancial sector. banks. High real
Hidden solvency rates of interest in
problems in regu non-regulated
and
lated banks!?). sector.
financial
Costa Rica 1983-87 Credit liberalization Stabilization and balance Important im Financial crises in High real rates of
and gradual deregula of payment adjustment. provements in non-regulated fi interest in non-re
instability
tion of interest rates. prudential regu nancial institu gulated financial
lation and su tions. sector.
pervision.
in ...
Bolivia 1985-90 Credit and interest-rate Stabilization and struc Severe Some insolvent Very high real
liberalization. tural reforms following shortcomings. banks. Weakened lending rates of
large fiscal deficits and solvency of bank interest.
hyperinflation. ing system.
Peru 1990-92 Credit and interest-rate Stabilization and struc Shortcomings in Solvency pro Slow decline of
liberalization. tural reforms, following provisions and blems in commer very high real
large fiscal deficits and recapitalization cial banking sec lending rates of
hyperinflation. norms. tor. Ínteres.
.
276 G u n t h e r H e l d Y.
vailing public-sector deficits -particularly in Argentina- tariff reductions ca

rried out at this stage of economic reform, and very large foreign-debt-
financed capital inflows. The latter resulted from the huge differential in
the real lending rates of interest in domestic currency and in foreign currency
(of expost 50 per cent a year in Argentina and 20 per cent a year in Uruguay
in 1979-80), and from an exchange-rate policy which had virtually elimi
nated foreign-exchange risk.47
Real domestic lending rates of interest also declined while fluctuating
drastically due to high and variable inflation rates. However, they started
to climb in Argentina in 1980 and in Uruguay in 1981, when it became
clear that current exchange rates were unsustainable (and also because
international interest rates increased to unusually high levels). Domestic
lending rates of interest rose to 40 per cent a year in real terms in the
second half of 1980 and the beginning of 1981 in Argentina, and in 1982
in Uruguay. Nevertheless between 1979 and 1983 capital flight equivalent
to more than 90 per cent of the foreign debt of the former and to more
than 60 per cent of the foreign debt of the latter took place.48
Banking regulations provided and implicit State guarantee on depo
sits in Uruguay and an explicit State guarantee in Argentina.49 In Uruguay
there were also no "minimum" rules controlling bank solvency as regards
entry conditions, credit limits, provisioning, and bank information for de
positors.50 In Argentina the banking law of 1977 set out very general norms
on solvency and liquidity, referring mainly to credit limits, constitution of
guarantees, restrictions on fixed assets, relations between active and passi
ve funds, and minimum capital requirements. However, it was not until
1985-90, after the financial crisis had surfaced, that rules on related credit
47 V. Corbo, J. de Melo and J. Tybout, "What Went Wrong with the Recent Reforms in the
Southern Cone", Washington D.C., World Bank, 1985, Report DRD128.
48 Ibid.
49 A full State guarantee on deposits was reintroduced in Argentina in 1973. The guarantee on
deposits of firms became negotiable with the Central Bank in 1977. However, all deposits in
banks which had undergone intervention up to the beginning of 1982 were paid. The im
plicit State guarantee on deposits in Uruguay stemmed mainly from full payment of de
posits during the financial crisis of 1971.
50 A. Banda, "Regulación Bancaria, Crisis Financiera y Políticas Consecuentes. El Caso
Uruguayo", in G. Held and R. Szalachman eds., A horroyAsignacióndeRecursosFinancieros,
(Buenos Aires, GEL, 1990).
B ank r e g u l a t io n , u b e r a u z a t io n a n d f in a n c ia l in st a b il it y i n . 277
were tightened. N ow loans had to be classified according to risk category,

and stricter rules on provisions were implemented. Norms prevailing in
the 1970s were considerably weakened by the virtual absence of bank
supervision. While each bank had the chance of being inspected once every
two years in 1976-77, this frequency dropped to once every six years in
1978-82.51 Thus in both countries, as in Chile, financial deregulation took
place under conditions of "uncontrolled" banking as regards solvency.
In Argentina private-sector bank credit grew in real terms at an avera
ge of 10 per cent a year in 1976-78 and 20 per cent a year in 1979-81,
with a stagnant economy. In Uruguay private-sector bank credit expanded
at average yearly rates of 20 per cent in real terms in 1976-81 while GDP
was growing at 3.5 per cent a year on the average. The sharp decline
in the real exchange rate as of 1979, and the subsequent increases in the
real rates of interest, which took place in both countries, soon eroded the
profitability and solvency of import competing and export sectors. Loan
rollovers and interest capitalization became an increasingly larger com po
nent of bank credit, as did speculative loans channelled at the time into
profitable real estate and other non-tradeable activities.
The financial crisis exploded in Argentina in March 1980 with inter
vention in three fast-growing commercial banks, including the largest de
posit holding bank. An impeding balance-of-payment crisis (fueled by
large-scale capital flight) and the abandonment of the nominal rate of ex
change as a stabilization device in the second quarter of 1981 exacerbated
the financial crisis. An immediate increase of 60 per cent in the real ex
change rate and an accumulated GDP drop of 11 per cent 1981-82 caused
severe solvency problems for debtors located in non-tradeable activities and
in the economy at large. In December 1988, the Central Bank of Argen
tina was still liquidating 206 financial institutions, of which 84 had under
gone intervention before the end of 1982.52
51 E. Salama, "Regulación y Supervision de la Banca e Instituciones Financieras en Argentina",

in G. Held and R. Szalachman eds., Regulación y Supervision de la Banca. Experiencias en Améri
ca Latina y el Caribe, (Santiago, ECLAC/UNDP, 1991).
52 L. A. Giorgo, "Crisis Financieras, Reestructuración Bancaria e Hiperinflación en Argentina",
(México, CEMLA, August 1991, working paper.
Bank solvency problems had also surfaced in Uruguay in 1980-81. They

developed into a financial crises in 1982, when the Latin American debt
crises, capital flight, and a large public-sector deficit forced a drastic balan
ce of payment adjustment. GDP dropped 11 per cent in 1982 and another
6 per cent in 1983-84. Banks attempted to protect themselves from the fo-
reign-exhange risk by denominating loans in dollars, only to find that cre
dit risk had become unmanageable once the nominal real rate of exchange
was abandoned as an anchor of stabilization in November 1982. Floating
led to a 100 per cent rise of that rate in a few days. In -1982-85, severe
solvency problems of banks and debtors led to large-scale domestic debt
refinancing and to the purchase of bad-debt portfolios of smaller banking
institutions by the Central Bank. A group of private banks, including the
three largest commercial banks, which had shown growing solvency pro
blems since the beginning of the 1980s, underwent intervention in 1985.53
The solvency problems faced by the banking systems of Colombia,
El Salvador and the Dominican Republic at the end of the periods men
tioned in Table 5, took place in different macroeconomic conditions and
financial policies. But in every case, they are mainly the result of short
comings in prudential regulation an d /or supervision.
Annual inflation in Colombia during the second half of the 1970s sta
bilized in the 20-25 per cent range, with average GDP growth rate of around
5 per cent a year. A conservative foreign debt policy also protected Co
lombia from the Latin American debt crisis in 1982. The latter contributed
to a recession, although GDP still grew 1 per cent in that year.
In the 1970s private-sector bank credit in Colombia expanded in keep
ing with GDP. However, the country experienced a "bubble" in 1980-81
when bank credit grew 41 per cent in real terms over the two years. The
large-scale banking intervention unitiated in 1982 demostrated that the cre
dit bubble consisted mainly of high-risk "related" loans to economic and
financial groups, including the use of bank credit to buy up common stock
of firms in the real sector. Unsafe loan practices had overstepped establi
shed credit norms and limits. At the same time, Supervisory authorities
had failed to verify the quality of bank loan portfolios. After the interven
53 A Banda, op. cit.

tion, there were eight "official" banks, administered by the public sector,
as compared to three in 1982, and nine private banks as compared to sixteen
in 1982. By 1989, the former accounted for 70 per cent of the assets of
the banking system.54
The commercial banking system of El Salvador had accumulated losses
equivalent to 3.4 times its capital at the end of 1989. Overdue and refinanced
loans representing 39 per cent and 26 per cent of the loan portfolio signalled
further outstanding losses. Three factors are particularly noteworthy when
seking to explain the financial crises amidst a stagnant economic activity
and die severe political turbulence of the 1980s: public-sector interference
in the credit process, financial repression and shortcomings in the regula
tion and supervision of banks.55
The nationalization of commercial banks in March 1980 and a large
Central Government deficit (equivalent to 15 per cent of CDP in the mid-
1980s) facilitated financial repression and made bank credit prone to politi
cal influence and sectoral pressure groups. The Central Bank set both de
posit and lending rates of interest, and banks had to meet a reserve requi
rement of 20 per cent on deposits while investing 10 por cent of their
asset portfolios in low-yield public-sector financial instruments. In the 1980s,
yearly inflation rates averaging 20 per cent led to negative real deposit
interest of 6 per cent a year on average. While deposits stagnated, bank
credit dropped by 35 per cent in real terms between 1980-81 and 1988-89.
Public-sector and Central Bank authorities took on an increasing role
in the management of banks and the allocation of credit through appoint
ment of personnel and credit directives. Under such conditions, the bulk
of bank regulation and supervision fell to financial norms and their com
pliance. Prudential regulation and supervision were sidestepped. Individual
credit limits were bypassed, credit risks inadequately assessed, provisions
on risk-prone loans delayed, bad loans rolled over and overdue loans not
collected. Since deposits enjoyed an implicit State guarantee after the na
tionalization of banks, financial repression in El Salvador took place in a
54 L. A. Zuleta, "Estructura del Proyecto de Crédito para Inversion en Colombia", in C. Massad

and G. Held eds., Sistema Financiero y Asignación de Recursos. Experiencias Latinoamericanas y
del Caribe, (Buenos Aires, GEL, 1990).
55 J. A. Belloso, "Regulación y Supervision de la Banca en El Salvador" (Santiago ECLAC/
UNDP, 1992), working paper.
climate of "uncontrolled" banking as regards solvency. Thus, once overdue

loans had to be recorded, their share of the loan portfolio increased from
14 per cent in 1988 to 39 per cent in 1989.56
Financial "repression" became important in the Dominican Republic
in the 1980s. The Central Bank maintained financial regulations that set
nominal interest rates, sectoral credit limits, and credit-linked reserve re
quirements, while inflation rates increased until reaching 100 per cent a
year in 1980. Frequent negative real deposit rates of interest in the bank
ing sector led to a decline in private-sector bank funding.57 Banks "own
credits" diminished around 7 per cent in real terms in the decade (while
GDP increased by around 20 per cent).
Financial repression plus a lax prudential regulation and supervision
fostered a flourishing free financial sector. The latter was made up o f fi
nancial establishments, financial groups, real-estate financiers, leasing firms,
credit-card firms and other agents. The number of these agents increased
from around 300 in 1979 to 650 in 1987, at which time their financial acti
vities involved a quarter of the credit and a third of the deposits in the
regulated banking sector.
In 1983 the Central Bank tightened limits on related credit and set
out provisioning norms taking into account overdue loans and guarantees.
In an attempt to counter the high credit risks which non-regulated finan
cial agents were assuming in 1987 the Central Bank instituted credit limits,
reserve and minimum capital requirements. However, even regulated bardes
have fallen far short of complying with credit limits and reserve require
ments. Nor do available procedures and the capacity of Supervisory au
thorities allow for evaluating credit risks and establishing the solvency po
sition of banks and financial institutions. The shortcomings of prudential
regulation and supervision at the end of the 1980s were singled out as
the main problem facing public-sector financial authorities in the current
drive to modernize the financial system of the Dominican Republic.58
58 J. A. Belloso, op. dt.

57 A. Veloz, "Sistema Finandero Dominicano y Asígnadón Selectiva del Crédito", en C. Massad
and G. Held eds., Sistema Financiero y Asignación de Recursos, (Buenos Aires, GEL, 1990).
58 H. Guilliani and J. Aristy, "Regulation y Supervisión del Sistema Finandero en la República
Dominicana" in G. Held and R. Szalachman eds., Regulación y Supervision de la Banca.
Experiencias de América latina y el Caribe, (Santiago, ECLAC/UNDP, 1991).
B ank r e g u l a t io n , l ib e r a l iz a t io n a n d f in a n c ia l in st a b il it y in . 281
A financial crisis exploded in 1989, when dozens of previously non-

regulated financial agents failed and when some banks in the regulated
sector also became insolvent. Yet interest rates (and the exchange rate)
were liberalized at the beginning of 1991, as part of a stabilization policy,
under conditions of undetermined bank solvency and very weak pruden
tial regulation and supervision. Real lending rates of interest among com
mercial banks immediately jumped to an equivalent of 50 per cent a year,
but then declined as inflation dropped from 100 per cent in 1980 to around
5 per cent a year in 1991-92. Central Bank interest rates pointed to real
annualized lending rates above 25 per cent a year at the end of 1991 and
above 20 per cent a year at the end of 1992.59 These rates have raised
doubts as to the solvency of certain banks and financial agents.
Costa Rica stands for noteworthy macroeconomic and financial po
licies in the eighties, but also for varying solvency results depending
upon prudential regulation and supervision. It adopted stabilization and
structural reforms in 1983, in the aftermath of a severe macroeconomic cri
sis at the beginning of the 1980s and its foreign-debt moratorium in 1981
(one year before that of Mexico). Financial reform led to an expansion of
private-sector banking, liberalization and devolution of credit decisions to
banks,60 and positive real rates of interest and their gradual deregulation.61
At the same time, prudential regulation and supervision were strengthened
on the basis of a uniform classification of the loan portfolio of banks ac
cording to risk categories, stricter rules on risk provisioning,62 and a set
59 Banco Central de la República Dominicana, B oletínMensuâ,Santo Domingo, October-Decem-

ber 1991 and 1992, Annex.
60 The Central Bank reintroduced some sectoral credit limits in favor of primary and export
sectors along the way.
61 Interest rates were deregulated in three stages. In the first two, the interest-rate structure
privoted around a 'basic" interest rate set by the Central Bank. In the first stage, interest-rate
dispersion was reduced and the interest-rate structure became flexible through movements
of the basic rate. In the second stage, commercial banks could set their deposit rates freely
within limits, while the interest spread (and thus their lending rates) were set in relation to
their deposit rates. In the last stage, commercial banks were free to negotiate their deposit
and lending rates of interest, with the exception of subsidized credit lines in favour of small-
scale producers. See F. de Paula, "Costa Rica: Intermediación Financiera y Asignación de
Recursos", in C. Massad and G. Held eds., Sistem aFinancieroyAsignacióndeRecursos,(Buenos
Aires, GEL, 1990).
62 Full-risk provisioning was a goal to be achieved in the forthcoming years.
282 G ü n t h e r H e ld Y.
of bank-performance indicators of which depositors and the public were

to be informed. Financial reform also opened the door to "free" financial
establishments which soon started to grow in number, operating at high
real rates of interest in the more risk-prone segments of the credit market.
The results of stabilization and adjustment policies were encouraging.
GDP grew on an average of 4.3 per cent a year in 1983-87 while annual
inflation dropped from 33 per cent in 1983 to an average of 15 per cent
in 1984-87. At the end of 1987 a tightening of monetary policy led to liqui
dity shortages and to the breakdown of all (non-regulated) financial esta
blishments. However, no supervised bank became insolvent in this finan
cial debacle.63
The financial liberalization experiences of Bolivia and Peru typify the
extent to which solvency problems are brought out into the open by
macroeconomic upheavals. However, macroeconomic conditions and poli
cies have also overshadowed the shortcomings of prudential regulation and
supervision.
In August 1985, the government of Bolivia adopted a radical stabi
lization policy and structural reforms aimed at an open market economy.
The reduction of the fiscal deficit from 25 per cent of GDP in 1984 to
2.5 per cent in 198664 was decisive in controlling hyperinflation. Inflation
fell from 8,200 per cent in 1985 to 11 per cent in 1986 and has since remained
in the 10-20 per cent range. Trade reform reduced tariff barriers from an
average of 40 per cent in 1985 to 10 per cent in 1990.65
As part of the new policy régime, prices, interest rates, the rate of
exchange66 and bank credit were liberalized in August 1985. Prevailing
uncertainty as to the behaviour of prices led first to an unprecedented real
interest rate jump and then to a decline in interest rates as inflation drop
63 R. Díaz "Regulación y Supervision de la Solvencia Bancaria en Costa Rica", in G. Held and

R. Szalachman eds. Regulación y Supervision de k Banca. Experiencks en América Latina y el
C arée, (Santiago, Chile, ECLAC/UNDP, 1991).
M A tax reform was introduced in 1986 to increase fiscal revenue.
65J.A. Morales et. a l, Bolivk: Ajuste Estructural. Equidad y Crecimiento (La Paz, Baremo, 1991),
Introducción.
66 The rate of exchange was liberalized through auctions of foreign exchange by the Central
Bank.
B ank r e g u l a t io n , u b e r a u z a t io n a n d f in a n c ia l in s t a b il it y i n . 283
ped. The real lending interest rates of banks went from around minus 80
per cent a year in 1984 (when interest rates were still set by the Central
Bank) to nearly 100 per cent in 1985, 70 per cent in 1986, 40 per cent
in 1987 and an average of 20 per cent in 1989-91. Though real lending
rates of interest have fallen, they have none the less remained at high
levels.
Solvency problems led to the closure of four banks in 1987. The very
high real interest rates and the changes in relative prices and sectoral
profitabilities brought about by rapid disinflation and the trade reform
loom large in this situation. Inadequate norms controlling credit risk, an
implicit State guarantee on deposits, and week supervision also played a
role. Since 1987, a number of measures have been taken to limit credit
concentration and related credit, increase provisions on risky credit, and
shore up supervision. Yet, overdue loans still stood at 14 per cent of the
loan portfolio of banks in 1989.67 The prevailing high real lending rates
of interest may thus partly reflect remaining high credit risks in bank
portfolios.
A handful of bank insolvencies in the first half of the 1980s alerted
public-sector financial authorities in Peru as to defective norms controlling
solvency and the risks stemming from an implicit State guarantee on de
posits. Accordingly, a number of measures considerably improved pruden
tial regulation in the second half of the 1980s. Credit classification to risk
categories and stricter limits on individual and related loans became offi
cial in 1989. The new banking law of 1991 expanded capital requirements,
introduced the sharing of information on debtors among banks, required
banks to provide the public with information regarding the quality of their
portfolios and capital positions, and empowered Supervisory authorities to
levy clearly defined sanctions. At the same time, the banking law of 1991
set up an explicit and compulsory deposit insurance fund. The fund gua
rantees bank deposits up to US$ 3,500 (at the rate of exchange prevail
ing in 1991) against a premium payable by banks. The latter are to be
differentiated according to the risk of their portfolios.68
67 G. Afcha de la Parra, "Desintermediarión y Liberalization Financiera en Bolivia: 1980-88", in

G. Held and R. Szalachman eds., A horroy AsignacióndeRecursosFinancieros, (Buenos Aires,
GEL, 1990).
68 G. González Arríela, "Regulación y Supervisión de la Solvencia del Sistema Financiero en el
The Peruvian economy moved towards hyperinflation and severe

macroeconomic imbalances in the second half of the 1980s. Inflation reached
2,800 per cent in 1989 while GDP fell 21 per cent in 1988-89. In August
1990 the government implemented a drastic stabilization policy and started
structural reforms directed towards an open market economy as Bolivia
had done in 1985. As part of the new policy regime, prices, interest rates,
the rate of exchange and private sector bank credit were quickly libera
lized. Inflation peaked at 7,600 per cent in 1990 while GDP dropped another
5 per cent. Since then, disinflation made less progress, the interest jump
has been higher, and the decline in interest rates has been slower as com
pared to the Bolivian experience. In the first halves of 1991, 92 and 93
the yearly inflation rates were 200 percent, 65 percent and 55 percent, whi
le the real annualized lending rates stood at around 170 per cent, 50 per
cent and 35 per cent.69
The above macroeconomic conditions were bound to create solvency
problems for banks and debtors. In March of 1993 overdue loans of do
mestic commercial banks were a 22 percent of their loan portfolios. H owe
ver provisions for risky loans stood at only 12 percent pointing also to
defective norms on provisioning and on a timely recapitalization o f banks,
in view of the severe erosion of their capital bases. A weakened supervision
may have contributed to this outcome on account of the loss of qualified
professional staff stemming from the drastic adjustment of public sector
expenditure.
2. Defective bank regulation and supervision

as a factor o f financial instability
Severe shortcomings in prudential regulation and supervision are a

common feature of all banking crises mentioned in Table 5. Faults regard
ing solvency controls on risks, provisions, public information on banks, and
supervision, together with a implicit or explicit State guarantee on depo
Perú", en G. Held and R. Szalachman eds., Regulación y Supervision de la Banca. Experiencias de

América Latina y el Caribe, (Santiago, ECLAC/UNDP, 1992).
69 On loans in domestic currency due in less than a year. See: Banco Central de Reservas del
Perú, 'Tasas de Interés Activas y Pasivas de la Banca Comercial" (Lima, May 1993).
sits, placed virtually all macroeconomic and financial policies in a context

of "uncontrolled" banking as regards solvency.
The banking crisis in Colombia at the beginning of the 1980s illus
trates how uncontrolled banking alone can create high credit risks and
unsustainable capital losses even under quite stable macroeconomic con
ditions.
The liberalization of bank credit, interest rates and other financial va
riables in Argentina, Chile and Uruguay in the 1970s were all carried out
under very unfavourable conditions for the stability and efficiency of the
banking system. Uncontrolled banking together with unstable macroeco
nomic conditions and policy inconsistencies led to greatly unbalanced
exchange and interest rates and to relative price and profit shifts among
economic activities. Although it is difficult to rank the importance of the
se factors in explaining the banking aises which exploded in the three
Southern Cone countries at the beginning of the 1980s, the severe short
comings in prudential regulation and supervision count for a great deal
in all of them. It is significant that in Argentina and Chile, and partly
in Uruguay, solvency problems erupted before balance-of-payments and
foreign-debt crises led to the need for major macroeconomic adjustments.
In the mid-1980s, Bolivia liberalized interest rates, bank credit, and
other financial variables in light of macroeconomic turbulence and uncon
trolled banking. Peru took the same steps in 1990, in conditions of soft
risk provisioning and recapitalization norms. The magnitude of the stabili
zation and adjustment effort, together with very high real rates of interest
in the commercial banking sector, appear to be the leading causes of the
solvency problems which have surfaced so far. However, high real lending
rates of interest also signal high credit risks in the loan portfolios of banks
and shortcomings in provisions and in the write-off of bad loans from their
asset portfolios.
Defective prudential regulation and supervision have also created sol
vency problems under conditions of financial repression, such as in El Sal
vador and the Dominican Republic in the 1980s. The same holds true in
the case of "free" financial institutions in Costa Rica and in the Domini
can Republic in the second half of the 1980s.
Of all the countries mentioned in Table 5, Costa Rica is the only one
where the liberalization of bank credit and the (gradual) liberalization of
interest rates in the 1980s did not cause financial problems in the regulated
banking system. It is also the only case where prudential regulation and
supervision had been substantialy improved.
IV. CONCLUSIONS
A. Prudential regulation and supervision have played an outstanding role

in determining the solvency (and efficiency) of banks and financial
institutions undergoing various financial experiences in Latin Ame
rican and Caribbean countries during the last two decades. Severe
shortcomings in prudential regulation a n d/or supervision, as exem
plified by the arrangement of uncontrolled banking, has caused or
fueled banking aises and financial instability during times o f finan
cial liberalization, financial repression and free financial activities.
B. The outcome of the liberalization of interest rates, a ed it and other

financial variables, as regards rendering banks solvent and efficient
while fostering an efficient allocation of resources, depends on the
fulfillment of a number of conditions. Among these, the following
stand out in Latin American and Caribbean countries:
1. A sufficiently stabilized m aaoeconom y based on a strong finan
cial position of the public sector and a real exchange rate and real
rates of interest in line with medium and longer term economic
conditions.
2. Prudential regulation and supervision capable of limiting and con
trolling a ed it risks, foreign-exchange and other risks which banks
and economic agents have to face in freer or more liberalized a e
dit, foreign exchange and other markets.70
3. An "industrial" banking organization whereby banks can engage in
multiple financial services with a view toward competitiveness;
taking advantage of economies stemming from complementarities
among financial services, the scale of operation and other factors;
70 It is also important to make sure that banks are solvent at the outset of finandal liberali
zation.
and achieving financial "deepening" as the volume of funds in

creases (through financial instruments extending the period of tran
sactions and spreading and transforming risks).
While the financial liberalization approach has stressed the first of
the above conditions, it has certainly overlooked the meaning of pru
dential regulation and supervision.
C. The liberalization of bank credit and of interest rates under conditions

of severe shortcomings in prudential regulation and/or supervision
and of a high macroeconomic instability can easily end up adding
risk to loan portfolios and creating unduly high real lending rates
o f interest, frequently in the form of a credit "bubble". These were
common features of financial liberalization in the Southern-Cone coun
tries in the 1970s. However, later experiences have also shown financial
liberalization not to be accompanied by prudential regulation, stron
ger bank supervision, and sufficiently stabilized macroeconomic envi
ronments. Moreover, the often prevailing arrangement of "uncontro
lled" banking as regards solvency has by itself been conducive to fi
nancial instability.
D. High real lending rates of interest have been a common feature of

most financial liberalization experiences in Latin American and Cari
bbean countries during the last two decades. While a number of fac
tors may contribute to these rates, one stands out in relation to insol
vency: extremely high real "accrued" lending rates which are not
paid by debtors but are instead rolled over, thus giving rise to a
very rapid increase in the volume of credit and adding to its riski
ness.
E. Many of the countries mentioned in Table 5 have in the meantime

taken measures to strengthen prudential regulation and supervision
of their banking systems and to expand the scope of their activi
ties to include multibanking. However, prudential regulation measu
res have generally not included freely informing depositors and the
public with regard to credit and asset risk and the effective capital
positions of banks (as Chile did in 1986). They have thus opted for
288 G ü n th er H e ld Y .
banking controlled by the public sector (see Table 2), though some
countries have explicitly limited the guarantee or insurance on de
posits.
F. Unstable macroeconomic conditions and defective prudential regula

tion and supervision do not provide a testing ground for financial
liberalization in most of the financial liberalization experiences of La
tin American countries during the last two decades. However, im
proved prudential regulation and supervision, along 'with more sta
ble macroeconomic conditions in a number of countries of the region,
may do so in the 1990s. Among the countries mentioned in Table
5, Chile has advanced the most as regards financial liberalization of
credit and capital markets emphasizing prudential regulation and
supervision. The financial and real sector results which have been
achieved since the mid-1980s deserve close attention.
ANNEX
Decomposition of credit growth in 1978-80 in

the financial liberalization experience of Chile
An important argument can be derived from the performance of The Sta

te Bank when simulating a decomposition of the 39 per cent yearly
growth of the loan and financial asset portfolio of the Chilean banking
system in 1978-80. The share of the State Bank in the banking-system loan
portfolio stood at a substancial 20 per cent in 1978, its loans grew at a
yearly rate of 22 per cent in real terms in 1978-80, and it did not have
to sell bad loans to the Central Bank.3 Its rate of growth can therefore
be taken as representative of the growth rate of normal risk or "legiti
mate" credit in the banking system. Under this assumption, the latter
would have expanded at around three times the GDP growth rate, while
"false" or very high risk aedit would have grown twice as fast as GDP,
or at 17 per cent in real terms annually.
The rate of growth or "false" a edit can be broken down making use
of the fact that the real lending rates of interest shown in the second and
third column of table 3 fell to around a half, once the accumulated stock
of bad loans had been sold to the Central Bank and risk-prone loans
reprogrammed. Since solvency norms as of 1987 required ¿ a t interest
accruals also meant cash income, nearly half of accrued interest on loans
in 1970-80 may have been capitalized, that is, (1 - a)r = 0.5 x 19.6 = 10%.
The remainder of high-risk loan growth would stand for the yearly
growth of new "false" credit (f1 = 7%). Thus:
p = 1 + fj + (1 - a)r = 22% + 7% + 10% = 39% (la)
* The State Bank was not authorized to sell bad loans, but it did not require Central Bank
assistance to overcome its loan-portfolio probles.
The corresponding growth rate of deposits and other liabilities which

finances the growth of the loan and financial asset portfolio isb:
d = 1 /e (i + 1 + fj + g - a.r) (2)
where "a.r" stands for the proportion of interest accruals on loans paid
in cash and where non previously mentioned symbols mean:
e = average ratio of reserve requirements on deposit and other lia

bilities.
i = Average real rate of interest on deposits and liabilities,
g = Administration and other current cash expenses as a percentage
of the loan (and financial assets) portfolio.
Filling in the variables o f this equation with 1978-80 data on the bank
ing systems yields a rate of growth similar to that of the loan andfinan
cial asset portfolio of the banking system:
d = 1/0.98 (13% + 22% + 7% + 6% - 0.5 x 19.6%) = 39% (2a)
b A simple bank model is assumed where loans and financial assets are equal to deposits and
liabilities. See G. Held and R. Szalachman, Regulación y Supervisión de la Banca en la Expérien-
cia de Liberalización Financiera en Chile (1974-1988), (Santiago, ECLAC, 1989), Annex B.
NOTE ON THE AUTHORS
Alice H. Amsden Leo Model Professor of Economics, New School for

Social Resarch, New York.
Yilmaz Akyüz Chief, Macroeconomic Section, Global Interdependen

ce Division, UNCTAD, Geneva.
Donald Hanna Senior Economist, World Bank, Resident Staff, Jakarta,

Indonesia.
Günther Held Coordinator, Regional Joint Project ECLAC/UNDP

"Financial Policies for Development", ECLAC, Santia
go, Chile.
Machiko Nissanke Senior Researcher, Overseas Development Institute,

London; School of Oriental and African Studies, Uni
versity of London.
Andras Uthoff Regional Adviser on Monetary and Financial Policy,

ECLAC, Santiago, Chile.

S332A315FR en

Uploaded by

Copyright:

Available Formats

S332A315FR en

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

S332A315FR en

Uploaded by

Copyright:

Available Formats

FINANCE

United Nations University/W orld Institute fo r Development

Collection: Social and Political Studies

I.S.B.N.: 956 - 7363 - 02 - 1

© S.R.V. Impresos S.A., Tocomal 2052, Santiago - Chile.

Compositor: Patricio Velasco G.

Santiago, September 1993

FINANCIAL LIBERALIZATION: THE KEY ISSUES....................................................... 19 \/

EAST ASIAN FINANCIAL MARKETS: WHY SO MUCH (AND

SAVINGS AND FINANCIAL POLICY ISSUES IN SUB-SAHARAN A FR IC A ................... 103

THE INDONESIAN EXPERIENCE WITH FINANCIAL-SECTOR REFORM ..................... 149

j BANK REGULATION, LIBERALIZATION AND FINANCIAL

NOTE ON THE AUTHORS ........................................................................................... 291

This book is a product of a joint work programme on "Financing for Deve­

the fast-growing ones they have tended to be allocated according to the

Santiago and Geneva, August 1993.

I. INTEREST RATES AND SAVINGS ..................................................................... 24

II. FINANCIAL LIBERALIZATION AND DEEPENING............................................ 27

III. ALLOCATIVE EFFICIENCY ................................................................................ 30

IV. PRODUCTIVE EFFICIENCY AND COST OF FINANCE ....................................... 35

V. REGULATION OF FINANCE AND FINANCIAL STABILITY .............................. 37

VI. OPTIONS IN FINANCIAL ORGANIZATIONS.................................................... 42

VII. EXTERNAL LIBERALIZATION AND FINANCIAL OPENNESS........................... 49

In recent years financial policies in both industrial and developing countries

borrowers and lenders found themselves overcommitted: debtors tried to

I. INTEREST RATES AND SAVINGS

degrees and under different circumstances.

The historical experience of major industrial countries holds some use­

II. FINANCIAL LIBERALIZATION AND DEEPENING

It is generally agreed that financial liberalization raises financial activity

III. ALLOCATIVE EFFICIENCY

Financial markets and institutions can be said to be allocatively efficient

provides a rationale for devoting resources to financial activity.

Governments in many countries have therefore acted to influence the allo­

degenerating into inflationary finance, to resist excessive credit expansion

As noted above, the main impulse to financial liberalization in developing

tion requires equalization of marginal cost of borrowing and marginal rate

IV. PRODUCTIVE EFFICIENCY AND COST OF FINANCE

The traditional concept of productive efficiency refers to microeconomic

1. Risk, uncertainty and interest rates

The Keynesian notions of lender's and borrower's risks provide an appro­

The second component of the cost of finance, namely the intermediation

ced by operating expenses, legal reserve and liquidity requirements, and

V. REGULATION OF FINANCE AND FINANCIAL STABILITY

The preceding discussion has shown that stability is an essential attribute

The theory of finance suggests that because information is imperfect and

however, regulations restricting excessive risk-taking and/or covering such

In short, competition among financial institutions can easily result in esca­

VI. OPTIONS IN FINANCIAL ORGANIZATIONS

1. Bank-oriented and market-oriented finance

It is possible to distinguish between two broad types of financial arran­

2. Efficiency o f alternative systems

In recent years many developing countries have been seeking to institute

tions tends to lower the rate of return required by investors to underta­

3. Requirements for an efficient bank-oriented system

These are particularly important considerations to be taken into account

oriented system. In particular, firms should not be allowed to own and

4. Control and regulation o f stock-markets

This book is a product of a joint work programme on "Financing for Deve

The historical experience of major industrial countries holds some use

Governments in many countries have therefore acted to influence the allo

The Keynesian notions of lender's and borrower's risks provide an appro

In short, competition among financial institutions can easily result in esca

It is possible to distinguish between two broad types of financial arran

tions tends to lower the rate of return required by investors to underta

Recent years have witnessed the increased integration of developing coun

Widespread liberalization has occurred on all three fronts. Inward transac

an ex post attempt to explain why the Southern Cone liberalization expe

The focus of financial policies in developing countries should be industria

of Central Banks of the American Continent, Barbados, 16-20 Novem

The basic Heckscher-Ohlin trade model assumes identical production func

administrative frameworks generally. Over time, more direct microecono

In comparing the behaviour of more and less economically- successful la

The financial system has incorporated an intangible set of dependen

the United States Securities and Exchange Commission to investigate a fi

ing to US $60 million, was established in May 1984. It aimed at long

In the 1980s, credit continued to be allocated preferentially, but real de

Without the competitive asset of new products and processes, late-indus

Ratio of m anufacturing N et value of m anu