GO

Financial crises are caused

by cross-border investment
flows, not misbehavior,
says economist Robert Aliber

FLOW
WITH

THE

By Nathan Jaye, CFA

Financial crises are widely believed to be caused by greed, Whats the main storyline of your research?
corruption, or lack of regulation. But what if the cause is Weve seen four waves of banking crises in
the past 30 years, all very similar. The first
simply the variability of cross-border investment inflows? wave was in Mexico, Brazil, Argentina, and 10
Thats the model developed by Robert Aliber, professor other developing countries in the early 1980s.
emeritus of international economics and finance at the Japan and several of the Nordic countries were
University of Chicago Booth School of Business. Aliber, involved in the second wave in the early 1990s.
The third wave was the Asian financial crisis
editor and co-author with Charles P. Kindleberger of the in July 1997, and the fourth is what I call the
1978 classic Manias, Panics, and Crashes: A History of Anglo-Saxon real estate crisis, which became
Financial Crises, predicted the Icelandic banking crisis apparent in September 2008.
18 months before it happened. In an interview with CFA Each country that experienced a banking
crisis previously had an economic boom and
Institute Magazine, Aliber offers a different view on the an increase in cross-border investment inflows,
cause of financial crises, discusses why banking crises which led to increases in the prices of its secu-
almost always coincide with currency crises, and explains rities and to an increase in the price of its cur-
why cross-border investment flows should be moderated. rencyunless the increase was forestalled by
central bank intervention.
These cross-border investment inflows are
too rapid to be sustained. Eventually, one or
several of the lenders recognize that borrow-
ers indebtedness is increasing too rapidly or
that the borrowers indebtedness is too large
relative to their incomes. When that happens,
lenders become more cautious. Borrowers will
not have enough cash to pay the interest. They
become distress sellers of real estate and secu-
rities. The prices of real estate and securities
decline, loan losses surge, and the country expe-
riences a banking crisis. Its also a currency crisis
in that many of the borrowers default on their
liabilities denominated in the foreign currency.

28 CFA Institute Magazine Nov/Dec 2014

How did you become aware of this pattern? calculated that the spreads between the interest rates they
Gradually, over more than 10 years. I was visiting a class- could earn when they bought mortgages and the interest
room at London Business School in April 2006. A young rates they had to pay when they sold their own IOUs were
man from Iceland was therehe spoke about his coun- sufficiently large so that the loans would be very profitable.
trys economy and the soaring stock pricesand I had a In my model, these firms are channels for the distribution
bizarre experience. I could complete all of his sentences, of credit. They compete fiercely for market share, but they
even though I knew nothing about Iceland. He had a set of do not determine the supply of credit.
facts, and I had a model. The two fit together quite nicely. The Washington policy establishment interpretation
I went to Iceland in June 2007 and spoke with 10 or 12 cannot explain why the US banking crisis occurred in 2008
economists at the central bank, in the private banks, at the rather than in 1988 or in 1998. The character of Lehman
University of Iceland. I was convinced there was a massive et al. did not change between 1988 and 2005; instead, the
asset price problem. A few months later, I wrote a paper change was in the credit market conditions. The Washing-
on Iceland, and I gave a lecture in Reykjavik in May 2008. ton policy establishment is unwilling to connect the dots
I said they were sitting on a volcano of credit and it would that link their own policies with the subsequent banking
soon implode. The price of the currency would fall, and the crisis. They want to ignore the relationship between the
price of securities would fall sharply. Four months later, in monetary policy in 2003 and 2004, the surge in property
September 2008, Iceland imploded. prices, and the subsequent bust. And they ignore the sim-
ilarity of events in the United States with those in many
How does your model explain the 2008 crisis other countries.
in the United States? The monetary instability of the
Beginning in 2003, cross-border investment flows to the last 30 years is unprecedented. The
United States increased sharply. Foreigners were buying title of Chairman Greenspans book
more US dollar-denominated securities; the Chinese trade is The Age of Turbulence. The book
surplus had surged. Chinas reserve managers bought sev- is in its second edition, but Green-
eral hundred billion dollars of IOUs in Fannie Mae and Fred- span still has not been able to iden-
die Mac, which enabled these institutions to buy more mort- tify the source of turbulence.
gages and more mortgage-backed securities.
At the same time, Fed Chairman [Alan] Greenspan fol- What are the sources
lowed an extremely expansive monetary policy. Thus, the of credit flows?
rapid increase in the domestic supply of credit complemented If we look at the 1982 crises, the
the increase in the supply of credit to American borrowers credit inflows came from the major
from the investment inflows. The United States had a mas- international banks that bought the Robert Aliber
sive housing boom. If Chairman Greenspan had been less US dollardenominated loans of
expansive, the United States still would have had a housing the governments and government-owned firms in Mexico,
boom and a subsequent crisis, but it would have occurred Brazil, and Argentina. The investment inflows that pre-
later and might not have been as severe. ceded the banking crises in Norway, Sweden, and Finland
Iceland, Ireland, Britain, and Spain had banking crises involved the sale of IOUs of the Nordic banks to banks in
at the same time as the United States. Every banking crisis London and other foreign centers; the Nordic banks repa-
is preceded by an excess supply of credit; the crisis occurs triated the money that they lent to commercial borrowers
when credit market conditions suddenly tighten. and real estate developers, who acquired the currency risk.

How does your model differ from the dominant What factors determine supply of credit?
interpretation of the 2008 crisis? Often, the increase in the cross-border investment inflows
The dominant interpretation (the Washington-policy-estab- is stimulated by a boom in the economy of the country
lishment consensus, including nearly everyone connected that experiences the investment inflow. At other times, the
with the Federal Reserve) is that the US banking crisis was increase is a response to a relaxation of regulations that pre-
the fault of lenders, such as Countrywide Financial, Lehman viously had limited cross-border investment flows, which
Brothers, Bear Stearns, Washington Mutual, and several was the Nordic experience.
hundred others, because they acquired too many risky In the 1990s, Mexico was being prepared for adherence
loans. The public officials have been successful in creat- to the North American Free Trade Agreement, and the lib-
ing the impression that the crisis would not have occurred eralization of economic regulations in Mexico was exten-
if the private lenders had behaved responsibly. sive. Moreover, macroeconomic initiatives to reduce infla-
But these crises are not caused by the misbehavior of tion (after several years when inflation was higher than
the private sector lenders. If the credit is there, it has to 100%) led to extraordinarily high real interest rates on
go someplace. Why does subprime become important? peso-denominated securities, which attracted money market
Because there arent enough prime borrowers. The only mutual funds. American, Japanese, and European firms
reason Countrywide and Washington Mutual went scroung- were investing in Mexico as a low-cost source of supply for
ing for borrowers was because the credit was there. They the American market.

Nov/Dec 2014 CFA Institute Magazine 29

Does the boom cause the credit inflow, or vice versa? to its GDP, both domestically and externally. It was predict-
Cross-border investment inflows contribute to the booms by able that when the lender stopped providing money in the
elevating securities prices in the countries experiencing the form of the loans to the borrowers, some of the borrowers
inflows; as household wealth increases, consumption spend- would default and the currency would collapse.
ing increases. Higher levels of spending lead to increases When the krona collapsed, many of the IOUs of the Ice-
in GDP growth rates and higher anticipated returns on the landic borrowers were denominated in foreign currencies.
securities available in these countries, which attracts even For example, the Icelandic banks were helpful in enabling
larger investment inflows. Some countries have had booms Icelandic households to borrow, to finance the purchase
without increases in investment inflows, but investment of cars and homes with loans denominated in the Japa-
inflows have almost always led to booms when currencies nese yen, the Swiss franc, the euro. When the price of Ice-
have not been anchored to parities. landic krona fell very sharply, the krona equivalent of the
liabilities denominated in the foreign currency increased in
How do you distinguish between structural proportion to the decline in the price of the krona and many
and monetary causes of shocks? of these borrowers were then bankrupt. In this way, the cur-
One of the major arguments advanced in support of a move rency crisis intensifies the banking crisis.
toward a floating exchange rate system in the 1950s and If the borrowers expenditures (exclusive of interest pay-
1960s was that economies would be better able to adjust ments) are larger than revenues, then it must follow that the
to structural shocks. Structural shocks are oil price shocks, increase of indebtedness is larger than interest payments.
bad harvest shocks, new discoveries of North Sea gas, etc. Thats an explosive relationship; it cannot continue. This
Monetary shocks are largely (not exclusively) changes logic can be applied to a family, to a firm, and to a govern-
in investor demand for securities dominated in the foreign ment. Indebtedness cannot grow more rapidly than interest
currency, which immediately leads to a change in the price payments. It may for one, two, or several years, but indebt-
of the countrys currency. The shocks in my narrative are edness cannot grow more rapidly than interest payments
virtually all monetary shocks. for an extended period of time.

Are banking crises always associated Do you use an actual model?

with currency crises? I count the cranes [for construction] in the urban landscape
When I started looking at this, what struck me was that and whether they are moving or stationary. I look at the rela-
banking crises and currency crises were twinned. There tion between rental rates of return and mortgage interest
was a very strong overlap. I began to wonder, what was the rates. If rental rates are less than mortgage interest rates,
relationship between them? Did one cause the other? Did a property prices are in bubble territory. Moreover, I look at
banking crisis cause the currency crisis? the pattern of cash flows and the relationship between the
Ninety percent of banking crises have been associated increase in the indebtedness of the borrowers and their
with a currency crisis, and every currency crisis has been incomes and also the increase in the borrowers indebted-
associated with a banking crisis. But I now realize these ness and the interest rates on their indebtedness. If some
are not different crises. These are different manifestations of the borrowers have primary deficits and are dependent
in different markets of reductions in investment inflows. on the increase in their indebtedness for some of the cash
to pay interest on their loans, then I know that the borrow-
Are you saying a banking crisis is predictable? ers are on a non-sustainable trajectory and that they will
Yes. Banking crises are predictablewith uncertain dates. be obliged to reduce their consumption when the lenders
Are earthquakes predictable? If you live along the San become more cautious.
Andreas Fault or the Hayward Fault, earthquakes are pre- One question is when the lenders will realize that the
dictable, but you cant really predict the exact date when borrowers are on a non-sustainable trajectory. Another ques-
one will occur. But Ive predicted some of these crises, tion is whether the borrowers can adjust to the decline in
including those in Iceland, Mexico in 1994, Thailand and their ability to increase their indebtedness and adhere to
Malaysia in 1997, and Argentina in 2001. their debt-servicing commitments.

Is there a tipping point when a banking crisis What are the implications of your research?
is inevitable? One set of implications is for investors: when to buy for-
I fly small airplanes, and in one of my Walter Mitty moments, eign stocks and bonds and when to sell domestic stocks and
I imagine that Im at Roosevelt Field in 1927 advising Charles bonds. Investment practitioners should follow the money
Lindbergh. I say, Charlie, when you cross the 19th merid- and momentum strategies; increases in prices of securities
ian, you cant turn back. The winds will be against you, are correlated with increases in the prices of currencies.
and you wont have enough fuel. Thats the concept of the And they should always ask, How long can the borrowers
point of no return, which can be modified to the date of no continue to have a primary deficit?
return. What is the date after which a crisis is inevitable? The second set of implications is for domestic financial
In the Icelandic case, Iceland had a massive capital account regulations: whether bank capital requirements should be
surplus after 2005. It had a very high level of debt relative high or low, whether banks need living wills, and whether

30 CFA Institute Magazine Nov/Dec 2014

large banks are too big to fail. The current stance of bank WHY DO WE HAVE SO MANY
regulation increases the costs that banks (and hence both
borrowers and depositors) incur. BANKING CRISES? THE
The third set of implications is for central banks and INTERNATIONAL MONETARY
the management of monetary policy. Many central banks,
including the Federal Reserve, operate as if they were in FUND (IMF) PRESIDES
a windowless silo. They ignore the impacts of changes in OVER A DYSFUNCTIONAL
cross-border investment inflows on the prices of securities,
household wealth, and consumption spending. FINANCIAL ARRANGEMENT.
The fourth set of implications involves the design of the
international financial arrangement for resolving imbal-
ances in payments among countries. Why do we have so
many banking crises? The International Monetary Fund to my account at LaSalle Bank in Chicago. I had to pay a
(IMF) presides over a dysfunctional financial arrangement; $75 fee. I sold my 1998 Cessna and went to the factory and
the IMF is like the three monkeysdeaf, dumb, and blind. bought a new Cessna.
Every banking crisis has resulted from highly variable cross- MBNA essentially was going to make me a zero inter-
border investment flows, and investment flows are much est rate loan because the supply of credit available through
more accelerated when currencies are floating than when MBNA was extremely large and very cheap. And MBNA had
they are pegged. When money begins to flow into a country, calculated for every 100 people who take advantage of this
whether its to Iceland or to the United States, the coun- offer, 95 would be on the hook at the end of the first year
try has a boom, rates of return increase, and the inflow of and would then pay an annual interest rate of 15% or 20%.
money brings even more money. This anecdote demonstrates that the supply of cheap credit
was super-abundant.
Should credit flows be moderated or regulated?
At the minimum, we need to do something to moder- Why does credit funnel into consumption whereas in
ate cross-border investment flows. We could go back to a the past it financed infrastructure?
modified Bretton Woods arrangement with a much larger I dont have a good answer. But if there is a large flow of
range of movement of currencies around a central parity. money, its going to go into the housing market and to con-
We could rely on exchange controlsor on some combi- sumers; housing loans offer the lenders the security of
nation of the two. collateral. One difference is that in the 19th century, the
cross-border investment flows were long term. More recently,
What are the chances of this happening? these investment inflows are short term or term loans with
Every now and then, the IMF makes a noise as if it will interest rates that change when a base interest rate changes.
allow some sort of margin control or prudential require- In the 19th century, infrastructure investments were financed
mentthats the term they use; it sounds less offensive than in the private sector, but now, theyre in the public sector
controlsome form of ad hoc control. But I want to min- for a variety of reasons. The public sector is reluctant to
imize the ad hoc controls. I want to return to a system. We increase borrowing to finance infrastructure investments.
dont have a system now. We now live in a world in which
anything is feasible and many of the measures adopted How does your model contradict monetarist
by foreign countries have had very high cost to the United economic theories?
States, to the US export industry. Asian countries in partic- The constitution for the international financial arrange-
ular have kept the prices of their currencies extraordinarily ment that we have today is founded on a set of articles by
low. Thats cost millions of manufacturing jobs in the United my colleagues at Chicago (Milton Friedman and Harry John-
States. Thats because we do not have a rule-based system. son) and by other scholars, such as Fritz Machlup at Prince-
ton and Gottfried Haberler at Harvard. Their description of
Havent we always had changes in credit flows? the adjustment process when currencies are floating is the
In the 19th century, cross-border investments often financed counterpart of the rules of the game of the gold standard.
large infrastructure investments. The United States bene- The proponents said that if currencies were allowed to
fited greatly from the investment inflows from Britain and float, the changes in the prices of currencies would be grad-
a few other countries that financed much of the US rail- ual and that the deviations of the market prices of curren-
road system. In the last few decades, investment inflows cies from long-run equilibrium prices would be much smaller
have stimulated consumption booms and real estate booms. than when currencies were pegged. They told us that there
Ill give you a personal anecdote. In August 2004, I would be fewer currency crises and that the demand for
received an announcement from my credit card company international reserves would be smaller. They claimed that
MBNA, and the announcement was that I had been preap- countries would be better insulated from foreign shocks.
proved for a balance transfer for one year at zero interest Every one of their claims is now challenged by the data.
rate. When I called MBNA, I asked, What is my credit line? Nathan Jaye, CFA, is a member of CFA Society San Francisco.
They said US$100,000. I said, Fine, please wire $100,000

Nov/Dec 2014 CFA Institute Magazine 31