Financial Crises in

Successful Emerging
Economies
Ricardo Ffrench-Davis,
Editor
UNITED NATIONS
BROOKINGS INSTITUTION PRESS
Financial Crises
in Successful
Emerging Economies
This page intentionally left blank
Financial Crises
in Successful
Emerging
Economies
Ricardo Ffrench-Davis
editor
UNITED NATIONS
Economic Commission for Latin America
and the Caribbean
BROOKINGS INSTITUTION PRESS
Washington, D.C.
anour niooxixcs
The Brookings Institution is a private nonprofit organization devoted to
research, education, and publication on important issues of domestic and foreign policy.
Its principal purpose is to bring knowledge to bear on current and emerging policy
problems. The Institution maintains a position of neutrality on issues of public
policy. Interpretations or conclusions in Brookings publications should be
understood to be solely those of the authors.
Copyright 2001
uxirio xarioxs icoxoxic coxxissiox ioi
iarix axiiica axo rui caiinniax
All rights reserved. No part of this publication may be reproduced or transmitted in any
means without written permission from the Brookings Institution Press.
Financial Crises in Successful Emerging Economies may be ordered from:
niooxixcs ixsriruriox iiiss, 1775 Massachusetts Avenue, N.W.,
Washington, D.C. 20036. Telephone: 800/275-1447 or 202/797-6258.
Fax: 202/797-6004. Internet: www.brookings.edu.
Library of Congress Cataloging-in-Publication data
Financial crises in successful emerging economies / Ricardo
Ffrench-Davis, editor.
p. cm.
Includes bibliographical references and index.
ISBN 0-8157-0211-6 (pbk. : alk. paper)
1. Latin AmericaEconomic conditions. 2. AsiaEconomic
conditions. 3. Financial crisesLatin America. 4. Financial crisesAsia.
I. FfrenchDavis, Ricardo. II. United Nations. Economic Commission for
Latin America and the Caribbean.
HC123 .F56 2001 2001003366
332.095dc21
9 8 7 6 5 4 3 2 1
The paper used in this publication meets minimum requirements of the
American National Standard for Information SciencesPermanence of Paper for
Printed Library Materials: ANSI Z39.48-1992.
Typeset in Adobe Garamond
Composition by Northeastern Graphic Services, Hackensack, New Jersey
Printed by R. R. Donnelley and Sons, Harrisonburg, Virginia
Preface vii
1 The Globalization of Financial Volatility: Challenges
for Emerging Economies 1
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio
2 Korea and Taiwan in the Financial Crisis 38
xaxuii i. acosix
3 Three Varieties of Capital Surge Management in Chile 65
iicaioo iiiixcu-oavis axo uiiiniiro raiia
4 From the Capital Surge to the Financial Crisis and
Beyond: The Mexican Economy in the 1990s 107
;aixi ios
5 An International Financial Architecture for
Crisis Prevention 141
sriiuax\ ciiiiiru-;oxis
Contributors 167
Index 169
Contents
v
This page intentionally left blank
T
his book is the result of a project developed by the
United Nations Economic Commission for Latin
America and the Caribbean (ECLAC), with support from the Ford Foun-
dation. The text encompasses five articles analyzing emerging economies
that were generally rated as successful by international financial institu-
tions and the financial press during episodes characterized by a broad sup-
ply of external funds. We include the cases of Chile, Korea, and Mexico in
the critical years of the 1990s and Chile in the deep crisis of the 1970s. All
of these economies were praised for their efficient public policies. They
all experienced episodes of an abundant supply of financial capital, and
they all suffered macroeconomic disequilibria as a result. We contrast these
cases with the positive experiences of Chile during the Tequila crisis and
of Taiwan during the Asian crisis.
Three of the articles are country studies, undertaken from a compara-
tive perspective. The paper by Manuel Agosin, professor at the University
of Chile, draws parallels between Korea and Taiwan. These two countries
achieved a similar performance from the mid-1960s through the early
1990s, but their paths then diverged. The study analyzes the national poli-
cies adopted in each case and the underlying motives.
The article by Ricardo Ffrench-Davis and Heriberto Tapia, both econ-
omists at ECLAC, compares three positive financial shocks experienced in
Chile: the liberalization of the capital account in the 1970s, which exploded
in a massive crisis in 1982; a substantial policy shift in 199194 in the di-
rection of a prudential macroeconomic management of the capital account,
Preface
vii
which kept Chile immune to the tequila crisis in 1995; and the capital surge
of 199597, which culminated in a rather severe adjustment in 1999.
The third study is by Dr. Jaime Ros, Mexican economist and profes-
sor at Notre Dame University, who addresses the contrasting experiences
of Mexico in 199194 and 199697. The paper examines the different do-
mestic and external variables that explain the marked differences in the
two episodes, and it evaluates the depth of the economic and social effects.
The fourth article, by Dr. Stephany Griffith-Jones of the University of
Sussex, analyzes the current architecture of the international financial sys-
tem and its incapacity for preventing crises or moderating the disequilibria
that generally lead to crises. The article analyzes several recent proposals, in-
cluding those of the author herself.
Finally, the paper by Jos Antonio Ocampo, Executive Secretary of
ECLAC, and Ricardo Ffrench-Davis, which opens the book, examines
why countries that were considered successful before the explosion of a cri-
sis incurred a level of macroeconomic disequilibria that made them vul-
nerable to a financial run. We start by considering the nature of supply,
focusing on investors who specialize in short-term, highly liquid opera-
tions. We then trace the evolution of the prices of financial assets, foreign
exchange, and stock markets in the receiving countries, and we identify
links with paths that culminate in unsustainable macroeconomic disequi-
libria. On the basis of this analysis, we expose five misconceptions that are
commonly held among proponents of full liberalization of the capital ac-
count.
Heriberto Tapia provided highly professional support in preparing the
final manuscript, verifying the technical content, and ensuring agreement
between the Spanish and English versions. Lenka Arriagada was excep-
tionally efficient in assisting with the presentation of the final manuscript.
We thank ECLAC for providing a stimulating environment for
policy-oriented research and the opportunity for independent analysis on
a most relevant issue today. Our deepest thanks also go to the Ford Foun-
dation for its support. Naturally, all the opinions presented here are the re-
sponsibility of the respective authors.

viii
Financial Crises
in Successful
Emerging Economies
This page intentionally left blank
1
:
O
ne of the outstanding features of modern financial
crises is that they occurred in emerging economies
that were generally viewed as very successful until the crises exploded.
Moreover, recent crises have been radically different from those typical
from the 1940s to the 1970s, which, in Latin America, displayed three
major features that have been absent or relatively less important in recent
experiences. First, past crises involved large fiscal deficits that were fi-
nanced with external loans or, in the absence of such financing, by central
banks. Second, domestic financial systems were repressed, a fact that was
generally accompanied by private sector access to rediscount or bank loans
at negative real interest rates. Finally, balance of payments crises were fre-
quently associated with a sharp worsening in the terms of trade or explicit
domestic policy decisions to overvalue exchange rates.
Over the past quarter century, a new variety of crises has gradually de-
veloped in Asia and Latin America. Four features differentiate them from
the previous type. First, the international capital market has been the major
source of shocks, whether positive or negative. Second, flows have largely
originated from and been received by the private sector. Fiscal deficits, in
contrast, have played a secondary role, and in most cases public finances
have been sound. Third, these financial crises have mostly hit emerging
iicaioo iiiixcu-oavis
;osi axroxio ocaxio*
The Globalization of
Financial Volatility:
Challenges for Emerging
Economies
*We appreciate the assistance of Angela Parra and Heriberto Tapia.
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio :
economies that were considered to be highly credible and successful. In
fact, the bulk of private flows has been concentrated on a small number of
relatively affluent and well-organized developing nations. Fourth, these
flows have been characterized by a lack of regulation, on both the supply
and demand sides. Domestic financial systems have often been liberalized
without the parallel development of a significant degree of domestic pru-
dential regulation and supervision.
In practice, the differentiation between old- and new-style crises is
naturally somewhat less clear-cut than the above description would sug-
gest. An early example of the new variety was the Chilean experience of
the 1970s and early 1980s, but the old type of crisis was still prevalent in
the rest of Latin America during that period, with other Southern Cone
countries in an intermediate position. In the 1990s, the new type gener-
ally predominated in both Latin America and Asia, but there were some
mixed episodes in which features of both new and old crises were inter-
mingled, with budget deficits and terms-of-trade fluctuations.
The evolution of private capital flows over the past three decades is well
known. During the 1970s, a large supply of funds was made available to
many developing nations. The 1980s brought a severe and widespread
shortage of financing, particularly for the Latin American countries. In
contrast, the Asian countries that were affected by the disturbances in the
international financial markets adjusted rapidly (with the exception of the
Philippines) and were able to leave the contagion effects behind. External
financing returned to Latin America in the 1990s, but it was volatile. The
resurgence of capital flows between 1991 and 1994 was followed by a sharp
scarcity, especially in Mexico and Argentina, with a rather generalized port-
folio outflow in late 1994 and early 1995. The so-called tequila crisis gave
way to renewed access in 199697, but in 19982000 external financing
was again in short supply as a result of contagion from the crisis detonated
in Asia in 1997. Worsening terms of trade aggravated the recession. On all
those occasions, changes in external financing were supply-led.
1
They had
a strong impact on the national economies on both sides of the cycle, with
contagion first of overoptimism and then of overpessimism.
Through 1996, the successful emerging economies of Asia appeared
to be immune to the instability associated with capital surges, as illustrated
by their performance during the tequila crisis. In reality, part of the out-
flows from Latin American countries were reallocated to Asia during that
1. Evidence shows that these changes have originated, to a large extent, in the sources of supply. See
Calvo (1998); Culpeper (1995); Griffith-Jones (1998); Larran (2000).
cionaiizariox oi iixaxciai voiariiir\ ,
episode. The subsequent events show that immunity was no longer a fea-
ture of the East Asian economies, and the two regions now faced common
destabilizing external forces.
The following section outlines the three capital surges experienced by
Latin American countries since the 1970s. Subsequently, the paper reviews
the main macroeconomic effects generated by capital surges and their pol-
icy implications. We then compare the specific experiences of the emerg-
ing economies covered by this research project: Korea and Mexico in the
1990s and Chile in the 1970s were all regarded as highly successful until
the sudden outbreak of severe crises. The study contrasts these experiences
with Chile in the early 1990s and Taiwan throughout the last decade,
which provide examples of economies that deployed a set of prudential
macroeconomic policies and thereby avoided domestic disequilibria and
mitigated contagion. The paper then addresses five misconceptions that
are currently in fashion. The final section summarizes some robust lessons
for domestic policies and reform of the international financial architecture.
Three Financial Capital Surges to Emerging Economies
since the 1970s
Purely financial factors have been changing in the world at a much
faster pace than international trade and the globalization of production.
During the 1970s and 1980s, many countries began to liberalize their fi-
nancial sectors and to relax or eliminate foreign exchange regulations.
2
This contributed to a boom in international flows, which was facilitated
by the revolutionary innovations in data management and telecommuni-
cations technology and the emergence of increasingly sophisticated finan-
cial techniques. The financial booms generally occurred within a
framework of lax or nonexistent regulations and supervision, and most ex-
isting regulations were in fact procyclical.
3
Net capital inflows to Latin America averaged nearly 5 percent of gross
domestic product (GDP) in 197781, 199194, and 199697. Exchange
rates appreciated in all three periods, which naturally led to a rapid in-
crease of imports relative to exports, with the corresponding current ac-
count deficit being financed (indeed, overfinanced) by a sharp rise in the
2. Daz-Alejandro (1985); Devlin (1989).
3. Ocampo (1999, 2001); Griffith-Jones (in this volume); United Nations (1999); Turner (2000).
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio
stock of external liabilities.
4
All these macroeconomic variables experi-
enced some overshooting.
5
Adjustment was frequently anchored to one
dominant balance, which generated imbalances in other macroeconomic
variables, as in a falling inflation rate associated with real exchange rate ap-
preciation and climbing external deficits. Such adjustment was obviously
facilitated by access to external financing, which most probably would
have been absent in a dry foreign supply.
The increased supply of external funding in those three episodes gener-
ated a greater demand for such financing. Recipient countries that adopted
procyclical or passive domestic policies experienced real exchange revalua-
tion and large current account deficits. Because these were heavily financed
by volatile flows of mostly short-term, liquid capital, the economies tended
to become increasingly vulnerable to changes in the mood of external cred-
itors; the outstanding case was Mexico in 199194 (see Ros, in this volume).
Creditors with financial assets placed in the region became more sensitive to
bad news as their exposure increased. The sensitivity rose steeply with the
size of net short-term liabilities.
6
The dramatic increase of international financial flows was more diver-
sified in the 1990s than in the 1970s. The situation is potentially more un-
stable, however, inasmuch as the trend has shifted from long-term bank
credit, which was the predominant source of financing in the 1970s, to
portfolio flows; medium- and short-term bank financing; time deposits;
and foreign direct investment (FDI; including acquisitions and other than
greenfield investment). A very high share of the newer supply of financing is
short term or liquid or both. The region thus saw a paradoxical diversifica-
tion toward volatility in the 1990s. The relative improvement after the te-
quila crisis, with a rising share of FDI, still included a significant proportion
of volatile flows.
7
The foundations of a broad liquid market for portfolio in-
vestment were laid down in the late 1980s through the Brady bonds and de-
veloped vigorously in the 1990s, with Latin America as a major destination
for both bond and stock financing. This market offered the expectation of
high rates of return during the upswings of the two cycles in the 1990s.
4. ECLAC (1995, 1998, 2000a).
5. When significant macroeconomic disequilibria persists despite repeated statements on the need
to maintain equilibria, it reveals an inadequate understanding of how to achieve sustainable equilib-
ria that are consistent with development. See Ffrench-Davis (2000, ch. 6).
6. Rodrik and Velasco (1999).
7. The positive link between FDI and productive investment is well documented (see Ffrench-Davis
and Reisen, 1998, chap. 1). The link was weakened, however, by the fact that about 40 percent of FDI
inflows in 199799 corresponded to acquisitions of Latin American firms rather than the creation of
new capacity (ECLAC, 2000b, chap. 1).
cionaiizariox oi iixaxciai voiariiir\ ,
In 1991 the stock of assets invested in Latin America by the new in-
vestors that had discovered the emerging markets was evidently below their
desired stock level, but by 1994 it had become considerably larger. Net
capital inflows were used to finance rising current account deficits, and ex-
ternal liabilities accumulated through time. This was sometimes accompa-
nied by significant mismatches in the maturity structure of the balance
sheets of domestic financial intermediaries, when short-term external
funds were used to finance longer-term domestic credits. This issue was
particularly severe in the dollarized segment of the domestic financial sys-
tem and in those cases in which external interbank credit lines were used
as a major source of domestic financing. Consequently, the region moved
into a vulnerability zone, with the economy becoming increasingly sensi-
tive to adverse political or economic news and hostage to the whims and
fancies of a few country analysts in London, Frankfurt and New York.
8
This situation was likely to put the economy at the mercy of the capital
markets occasionally whimsical moods.
9
The longer and deeper the econ-
omys incursion into that zone, the more severe was the financierist trap in
which authorities could get caught, and the lower the probability of leav-
ing it without undergoing a crisis.
10
Mexico and Argentina were particularly vulnerable in 1994, while
Chile had deliberately avoided venturing into the vulnerability zone.
Meanwhile, East and Southeast Asian countries were just starting to take
that risk in the first half of the 1990s, and the resulting mismatches in the
maturity structure of the balance sheets of domestic financial intermedi-
aries proved to be even more severe than a worsening net debt position.
By the next cycle, several economies in both Asia and Latin America had
penetrated deep into the vulnerability zone. Both regions suffered severe
crises when the mood of the external market changed.
Worsening of Macroeconomic Fundamentals Led by Inflows
The economic activity of Latin American countries exhibited signifi-
cant vulnerability to changes in international financial markets over the
last three decades, which worked as an intensely procyclical factor for the
8. Rodrik (1998).
9. Calvo (1998).
10. The financierist trap refers to a macroeconomic policy approach that leads to an extreme pre-
dominance of or dependency on agents specializing in microfinance, positioned in the short-term and
liquid segments of the market.
emerging economies. This vulnerability was associated with the volatility
of international markets since the 1970s, as well as with the procyclical
macroeconomic policies adopted by recipient countries. Several Asian
emerging economies followed suit in the 1990s.
Annual GDP growth rose in Latin American countries from 1.3 per-
cent in the 1980s to 4.1 percent between 1991 and 1994 and 4.5 percent
in 1996 and 1997. Recessive adjustments took place in 1995 and 199899
(see table 1-1). Overall, GDP rose by a mere 3.3 percent in the decade
(19912000). Given that GDP was highly unstable, however, the precise
figure depends on the period chosen. For 199099 the growth rate was 2.8
percent, because the period starts and ends with a recession. From peak to
peak (between 1989 and 2000), average growth was 2.9 percent.
A growth of productive capacity of around 3 percent (1.3 percent per
capita) is remarkably low compared with the expectations generated by the
structural reforms. Comparison with the previous golden age is striking.
During the three decades from 1950 and 1980, GDP growth averaged 5.5
percent a year (2.7 percent per capita), with rather high domestic invest-
ment ratios sustaining these vigorous rates. In the 1980s, gross domestic
investment dropped sharply, by 7 percentage points of GDP. The recovery
in the 1990s was weak (see figure 1-1). In fact, investment grew much less
during this decade than did capital inflows; a significant proportion of ex-
ternal flows thus financed increased consumption, crowding out domestic
savings.
11
Recovery from Recession
The domestic conjuncture has crucial implications for the link be-
tween capital flows and economic activity. When there is a binding exter-
nal constraint, any inflow will contribute to relaxing it, thus facilitating a
recovery of economic activity. Binding external constraints predominated
during several episodes in many Latin American countries, and they were
particularly widespread from the early 1980s up to 1990, in 1995 and in
19982000.
In the early 1990s, renewed capital inflows thus contributed to a recov-
ery of economic activity, and they facilitated the adoption of successful
anti-inflationary adjustments. Argentina and Peru, for example, both fea-
tured huge underutilization of capacity and hyperinflation; the disappear-
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio o
11. See Ffrench-Davis (2000, chap. 1 and 5); Uthoff and Titelman (1998).

T
a
b
l
e

1
-
1
.
L
a
t
i
n

A
m
e
r
i
c
a

a
n
d

E
a
s
t

A
s
i
a
:

G
r
o
s
s

D
o
m
e
s
t
i
c

P
r
o
d
u
c
t
,

1
9
7
1

2
0
0
0
A
n
n
u
a
l

g
r
o
w
t
h

r
a
t
e

(
p
e
r
c
e
n
t
)
R
e
g
i
o
n

a
n
d

c
o
u
n
t
r
y
1
9
7
1

8
0
1
9
8
1

8
9
1
9
9
0
1
9
9
1

9
4
1
9
9
5
1
9
9
6

9
7
1
9
9
8

9
9
1
9
9
1

2
0
0
0
L
a
t
i
n

A
m
e
r
i
c
a

(
1
9

c
o
u
n
t
r
i
e
s
)
a
5
.
6
1
.
3
-
0
.
6
4
.
1
1
.
1
4
.
5
1
.
2
3
.
3
A
r
g
e
n
t
i
n
a
2
.
8
-
0
.
7
-
2
.
0
8
.
0
-
2
.
9
6
.
7
0
.
4
4
.
2
B
r
a
z
i
l
8
.
6
2
.
3
-
4
.
6
2
.
8
4
.
2
3
.
0
0
.
5
2
.
7
C
h
i
l
e

2
.
5
3
.
0
3
.
3
7
.
5
9
.
1
6
.
9
1
.
5
6
.
1
C
o
l
o
m
b
i
a
5
.
4
3
.
7
4
.
3
4
.
3
5
.
2
2
.
8
-
2
.
1
2
.
6
M
e
x
i
c
o
6
.
7
1
.
5
5
.
1
3
.
5
-
6
.
1
6
.
1
4
.
3
3
.
5
P
e
r
u
3
.
9
-
0
.
7
-
6
.
0
5
.
1
8
.
6
5
.
4
1
.
0
4
.
5
V
e
n
e
z
u
e
l
a
1
.
8
-
1
.
5
5
.
5
3
.
2
-
1
.
9
3
.
4
-
3
.
6
2
.
1
1
9
7
1

8
0
1
9
8
1

9
0
1
9
9
1

9
2
1
9
9
3

9
6
1
9
9
7
1
9
9
8
1
9
9
9
1
9
9
1

2
0
0
0
E
a
s
t

A
s
i
a

(
6

c
o
u
n
t
r
i
e
s
)
b
8
.
1
7
.
0
7
.
3
7
.
3
4
.
8
-
4
.
2
6
.
5
5
.
6
I
n
d
o
n
e
s
i
a
7
.
7
5
.
5
8
.
1
7
.
7
4
.
7
-
1
3
.
2
0
.
2
4
.
2
K
o
r
e
a
9
.
0
8
.
8
7
.
3
7
.
3
5
.
0
-
6
.
7
1
0
.
7
6
.
1
M
a
l
a
y
s
i
a
7
.
8
5
.
2
9
.
2
9
.
7
7
.
5
-
7
.
5
5
.
4
7
.
1
P
h
i
l
i
p
p
i
n
e
s
5
.
9
1
.
7
-
0
.
1

4
.
2
5
.
2
-
0
.
6
3
.
3
2
.
8
T
a
i
w
a
n
9
.
3
8
.
5
7
.
5
6
.
7
6
.
7
4
.
6
5
.
7
6
.
5
T
h
a
i
l
a
n
d
7
.
9
7
.
9
8
.
3
8
.
0
-
1
.
7
-
1
0
.
2
4
.
2
4
.
5
S
o
u
r
c
e
:

F
o
r

L
a
t
i
n

A
m
e
r
i
c
a
,

t
h
e

U
n
i
t
e
d

N
a
t
i
o
n
s

E
c
o
n
o
m
i
c

C
o
m
m
i
s
s
i
o
n

f
o
r

L
a
t
i
n

A
m
e
r
i
c
a

a
n
d

t
h
e

C
a
r
i
b
b
e
a
n

(
E
C
L
A
C
)
.

F
o
r

E
a
s
t

A
s
i
a
,

t
h
e

I
n
t
e
r
n
a
t
i
o
n
a
l
M
o
n
e
t
a
r
y

F
u
n
d

(
I
M
F
)
,

I
n
t
e
r
n
a
t
i
o
n
a
l

F
i
n
a
n
c
i
a
l

S
t
a
t
i
s
t
i
c
s
,

N
o
v
e
m
b
e
r

2
0
0
0
;

A
s
i
a
n

D
e
v
e
l
o
p
m
e
n
t

B
a
n
k
;

J
.
P
.

M
o
r
g
a
n
.
a
.

N
a
t
i
o
n
a
l

a
c
c
o
u
n
t
s

e
x
p
r
e
s
s
e
d

i
n

U
.
S
.

d
o
l
l
a
r
s

a
t

1
9
8
0

p
r
i
c
e
s

f
o
r

1
9
7
1

8
0
,

a
t

1
9
9
0

p
r
i
c
e
s

f
o
r

1
9
8
1

8
9
,

a
n
d

a
t

1
9
9
5

p
r
i
c
e
s

f
o
r

1
9
9
0

2
0
0
0
.
b
.

I
n

e
a
c
h

p
e
r
i
o
d
,

e
a
c
h

c
o
u
n
t
r
y

s

G
D
P

w
a
s

w
e
i
g
h
t
e
d

b
y

i
t
s

a
v
e
r
a
g
e

s
h
a
r
e

i
n

t
h
e

r
e
g
i
o
n
a
l

o
u
t
p
u
t
,

e
x
p
r
e
s
s
e
d

i
n

c
u
r
r
e
n
t

d
o
l
l
a
r
s
.
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio
Figure 1-1. Latin America: Gross Fixed Investment, 19772000
Source: Authors calculations, based on ECLAC figures, scaled to constant 1995 prices.
26
24
22
20
18
0
Percent of GDP
1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 19992000
25.9
23.7
19.2
19.8
ance of the binding external constraint and the reintroduction of macroec-
onomic discipline to combat hyperinflation were strongly complementary.
The monetary effects of reserve accumulation and the wealth effects of ex-
change rate appreciation tended to push up aggregate demand, which facil-
itated the recovery of economic activity. At the other end of the spectrum,
capacity underutilization was not significant in Chile and Mexico. Conse-
quently, the positive link between capital flows and GDP growth was not
automatic, but rather was contingent on the capacity to transform addi-
tional external financing into increased productive investment.
On average, Latin American GDP rose faster in 199194 than the ex-
pansion of the production frontier thanks to increased capacity utilization.
An estimated one-third of the 4.1 percent annual GDP growth in
199194 was based on this factor. In 1995, the binding external constraint
again became a crucial variable, with GDP growth lagging behind capac-
ity growth. Renewed capital inflows in the following years contributed to
a recovery of economic activity, based to some extent on the excess capac-
ity generated in 1995. However, the return of a binding external constraint
in 199899, particularly in South America, led to a new recession.
One implication of this analysis is that any serious research should
control for the huge swings in the rate of capacity utilization when meas-
uring productivity and the performance of policies and reforms. In the
cionaiizariox oi iixaxciai voiariiir\ ,
presence of excess capacity, recovery naturally yields high private and so-
cial returns, but they are built on preexisting disequilibria, that is, on for-
gone profits, wages, taxes, and employment that exist whenever the
economy is operating below its productive frontier or economically poten-
tial GDP. Whether economic recovery opens the way to more sustained
growth depends crucially on two dimensions. First, the speed at which ca-
pacity is expandedthrough physical investment, investment in people,
and productivity gainsdetermines future potential growth. Second, the
sustainability of the macroeconomic environment that develops during
the recoverynamely, exchange and interest rates, current account defi-
cit, domestic financial vulnerability, fiscal accounts, and asset pricesde-
termines whether growth in aggregate demand can be sustained or
whether it will be subject to corrections associated with imbalances accu-
mulated during recovery.
Overshooting in Emerging Asia and Latin America
The increased availability of financing in the 1990s removed the bind-
ing external constraint that had been responsible for the decade-long re-
cession in Latin America. The bases for growth were not laid down,
however, as investment did not increase rapidly and macroeconomic im-
balances built up. Effective output thus approached the production fron-
tier, while exchange rate appreciation led to overvaluation.
12
Asset markets
also overshot, and a large stock of mostly liquid external liabilities accu-
mulated (see figure 1-2). The regions economies therefore became more
vulnerable to future negative external shocks. With some variation, this
story applies to both 199194 and 199597, reproducing the path toward
the crisis of 197681.
In 1995, the tequila crisis had negligible effects on the Asian region,
even in economies with large current account deficits, such as Malaysia
and Thailand. Many outstanding researchers and observers therefore as-
serted in 1996 that such deficits were not relevant if investment ratios and
economic growth were high. Several Asian countries had successfully reg-
ulated capital inflows and foreign exchange markets for long periods.
13
Ec-
onomic growth was actually sustained and extremely high. From 1970 to
12. Several Latin American countries implemented sharp import liberalization at the same time that
the exchange rate was appreciating. See Ffrench-Davis (2000, chap. 3) and ECLAC (1998, chap. 5;
for an English version, see ECLAC 1995). The average import tariff was cut from 45 percent in the
mid-1980s to 13 percent in the mid-1990s; nontariff restrictions were also reduced significantly.
13. On Malaysia, Indonesia, and Thailand, see Sachs, Tornell and Velasco (1996); on Korea and
Taiwan, see Agosin (in this volume).
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio :c
Figure 1-2. Latin America and East Asia: Current Account Balance,
19902000
a
Source: ECLAC; IMF.
a. East Asia includes Indonesia, Korea, Malaysia, the Philippines, and Thailand. Latin America includes nineteen
countries.
10
8
6
4
2
0
2
4
Percent of GDP
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Surplus
Deficit
East Asia
Latin America
1996, yearly GDP growth averaged 8 percent in Korea, Indonesia, Malay-
sia, Taiwan, and Thailand (see table 1-1 above). The investment ratio fre-
quently exceeded 33 percent, with domestic savings ratios at a similar level.
Inflation was low (in the range of 5 percent a year), and fiscal budgets were
generally balanced or in surplus (with the exception of Taiwan). Mean-
while, average GDP growth in Latin America was 2 percent and the in-
vestment ratio fluctuated around 20 percent.
What is the explanation for the sudden worsening in Asia? First, export
performance in several Asian economies weakened. Export sectors that had
been experiencing notably dynamic demand suddenly faced tightening
markets, either as a result of a temporary excess supply or because specific
markets were reaching maturity.
14
The long Japanese crisis contributed to
the intensity of these problems.
Second, the global drive toward financial liberalization had also per-
meated several Asian economies in the 1990s.
15
China, India, and Taiwan
were three notable exceptions. Capital inflows and current account deficits
increased substantially in Korea and Thailand from 1993 on. External im-
14. Radelet and Sachs (1998).
15. Agosin (in this volume); Ariff and Ean (2000); Akyz (1998); Furman and Stiglitz (1998); Jomo
(1998); Wang (2000).
cionaiizariox oi iixaxciai voiariiir\ ::
balances were not associated with public deficits and did not imply losses of
international reserves: in fact, in Korea, Indonesia, Malaysia, and Thailand,
international reserves, fed by capital inflows, accumulated consistently be-
tween 1992 and early 1997, more than doubling in the period. All the data
indicate that the cause of disequilibria was a rise in private expenditure led
by capital inflows, which allowed liquidity constraints to be relaxed. The
induced domestic lending boom was accompanied by bubbles in real estate
and stock market prices. In some cases, real exchange rate appreciation and
import liberalization led to cheaper imports, which further fed the import
boom.
Third, most inflows were short term or liquid, including a large pro-
portion of interbank lending.
16
Domestic balance sheets became quite vul-
nerable as a result of maturity and currency mismatches and the rapid rise
of firms leverage.
17
Weak prudential supervision of the financial system,
which had not been a real threat in the previously repressed domestic mar-
kets, became evident with financial liberalization and the lending boom.
In these Asian economies, vulnerability was thus associated with wors-
ening macroeconomic fundamentals led by a capital surge, which carried
over to an exchange rate appreciation (moderate compared to that in Latin
America), a boom in aggregate private demand (with a significant enlarge-
ment of the current account deficit by 5 percentage points of GDP in
Korea, 2 points in Indonesia, and 3 points in Thailand), and an increased
vulnerability of the balance sheets of domestic financial intermediaries.
The disequilibrium was recognized by financial markets only in 1997 and
resulted in a weighty bill in 1998. The policy failure was an error shared
with the rather similar financial reforms conducted in Chile in the 1970s
and in Mexico in the 1990s.
18
Why the Market Fails to Avoid Overshooting
On the whole, the authorities took a procyclical approach in both
Latin America and Asia, allowing capital surges to be transmitted domesti-
cally. They therefore fell into a financierist trap, from which it is extremely
difficult to escape without a traumatic adjustment, involving outlier ex-
change or interest rates and considerable liquidity constraints that together
generate a very unfriendly macroeconomic environment for firms and
16. IMF (1998); Radelet and Sachs (1998).
17. See Krugman (1999).
18. Ffrench-Davis and Tapia (in this volume); Ros (in this volume).
labor. Most authorities (as well as observers) took the view that nothing
could or should be done during the expansive stages, or they preferred to
benefit a little longer from the capital boom. The ex post consensus among
observers, however, was that disequilibria had accumulated.
Given that voluntary flows cannot take place without the willing con-
sent of both debtors and creditors, why did neither agent act in due time
to curb the inflows well before a crisis? Some specialists recognized and
warned of the growing vulnerability in all three episodes examined. Why,
then, did the market fail to avoid major crises?
19
Asymmetries of informa-
tion among creditors, lack of adequate internalization of the negative ex-
ternalities that each agent generates (through growing vulnerability), and
other imperfections of international capital underlie the cycles of abun-
dance and shortage of external financing.
20
These factors contribute to
herd behavior, cross-border contagion, and multiple equilibria.
Over and above these features, however, the particular nature of the
agents acting on the creditor side is crucially important. Short-term hori-
zons were a significant factor in the 1990s, as reflected in the volatility of
flows that characterized the boom-bust cycles. The gradual spread of in-
formation on investment opportunities is another key influence. Investors
from different segments of the financial market were steadily drawn into
international markets as they took notice of the profitable opportunities
offered by emerging economies. This explains why the three surges of
flows to emerging economies were processes that went on for several years,
rather than one-shot changes in supply.
On the domestic side, many Latin American economies were experi-
encing recession, depressed stock and real estate markets, high interest rates,
and initially undervalued exchange rates; capital surges directed to these
economies therefore could expect potentially high rates of return.
21
Indeed,
in the early 1990s, prices of equity stocks and real estate were extremely
depressed in Latin America. That allowed for a 300 percent average capital
gain in the Latin American stock markets between late 1990 and September
1994, with rapidly rising price-to-earnings ratios (see table 1-2).
22
Average
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio ::
19. For instance, see a warning advice on Latin America, as early as in mid-1992, reproduced in
Ffrench-Davis (2000, chap. 9).
20. McKinnon (1991); Rodrik (1998); Stiglitz (2000); Wyplosz (1998).
21. A similar outcome tends to result in an emerging economy moving from a closed to an open
capital account. The rate of return is naturally higher in the productive sectors of capital-scarce emerg-
ing economies than in mature markets that are capital rich.
22. Figures in current dollars; the real rate of return is obtained by adding distributed profits and
subtracting dollar inflation.

T
a
b
l
e

1
-
2
.
L
a
t
i
n

A
m
e
r
i
c
a

a
n
d

E
a
s
t

A
s
i
a
:

S
t
o
c
k

E
x
c
h
a
n
g
e

P
r
i
c
e
s
,

1
9
9
0

2
0
0
0
a
I
n
d
e
x
e
s

(
J
u
l
y

1
9
9
7

=
1
0
0
)
R
e
g
i
o
n

a
n
d

c
o
u
n
t
r
y
D
e
c

1
9
9
0
S
e
p

1
9
9
2
S
e
p

1
9
9
4
M
a
r

1
9
9
5
J
u
l
y

1
9
9
7
A
u
g

1
9
9
8
D
e
c

1
9
9
9
N
o
v

2
0
0
0
L
a
t
i
n

A
m
e
r
i
c
a

(
7

c
o
u
n
t
r
i
e
s
)
2
1
.
7
4
4
.
6
9
2
.
5
5
2
.
3
1
0
0
.
0
4
7
.
2
8
4
.
0
6
7
.
9
A
r
g
e
n
t
i
n
a
1
3
.
4
4
6
.
9
7
8
.
2
5
3
.
5
1
0
0
.
0
5
3
.
4
8
6
.
7
6
4
.
6
B
r
a
z
i
l
8
.
0
2
2
.
1
7
1
.
8
4
2
.
8
1
0
0
.
0
4
4
.
4
7
3
.
5
6
1
.
0
C
h
i
l
e
2
4
.
5
5
1
.
4
9
3
.
1
8
9
.
4
1
0
0
.
0
4
8
.
0
7
5
.
3
6
4
.
6
C
o
l
o
m
b
i
a
1
6
.
6
6
5
.
0
1
1
3
.
1
9
6
.
3
1
0
0
.
0
4
9
.
9
4
4
.
9
2
6
.
0
M
e
x
i
c
o
3
8
.
6
7
2
.
7
1
3
2
.
1
4
5
.
9
1
0
0
.
0
4
9
.
7
1
1
0
.
8
9
3
.
0
P
e
r
u
n
.
a
.
n
.
a
.
7
2
.
9
5
6
.
4
1
0
0
.
0
5
7
.
3
6
6
.
0
4
7
.
7
V
e
n
e
z
u
e
l
a
8
4
.
9
8
2
.
2
5
0
.
8
3
7
.
9
1
0
0
.
0
2
6
.
2
3
7
.
4
4
4
.
7
E
a
s
t

A
s
i
a

(
6

c
o
u
n
t
r
i
e
s
)
n
.
a
.
4
9
.
9
1
1
0
.
0
9
7
.
9
1
0
0
.
0
3
7
.
0
1
0
1
.
6
5
9
.
5
I
n
d
o
n
e
s
i
a
n
.
a
.
5
3
.
7
8
4
.
2
7
1
.
6
1
0
0
.
0
1
1
.
1
3
7
.
0
1
5
.
8
K
o
r
e
a
n
.
a
.
8
7
.
6
1
8
7
.
2
1
6
1
.
9
1
0
0
.
0
3
0
.
2
1
3
5
.
5
6
2
.
9
M
a
l
a
y
s
i
a
n
.
a
.
6
3
.
7
1
1
9
.
0
1
0
3
.
5
1
0
0
.
0
1
6
.
8
4
9
.
9
4
2
.
1
P
h
i
l
i
p
p
i
n
e
s
n
.
a
.
6
7
.
1
1
3
4
.
6
1
0
8
.
6
1
0
0
.
0
3
0
.
4
6
1
.
0
3
2
.
9
T
a
i
w
a
n
n
.
a
.
3
7
.
1
8
0
.
9
7
3
.
5
1
0
0
.
0
4
7
.
6
8
4
.
1
5
0
.
0
T
h
a
i
l
a
n
d
n
.
a
.
1
3
3
.
9
2
7
9
.
8
2
3
6
.
3
1
0
0
.
0
1
9
.
0
5
9
.
1
2
7
.
5
S
o
u
r
c
e
:

A
u
t
h
o
r
s

c
a
l
c
u
l
a
t
i
o
n
s
,

b
a
s
e
d

o
n

d
a
t
a

f
r
o
m

t
h
e

I
n
t
e
r
n
a
t
i
o
n
a
l

F
i
n
a
n
c
i
a
l

C
o
r
p
o
r
a
t
i
o
n

a
n
d

S
t
a
n
d
a
r
d

&

P
o
o
r

s
,

E
m
e
r
g
i
n
g

S
t
o
c
k

M
a
r
k
e
t

R
e
v
i
e
w
,

s
e
v
e
r
a
l
i
s
s
u
e
s
.
a
.

T
h
e

a
v
e
r
a
g
e
s

f
o
r

L
a
t
i
n

A
m
e
r
i
c
a

a
n
d

E
a
s
t

A
s
i
a

a
r
e

w
e
i
g
h
t
e
d

b
y

a
m
o
u
n
t

o
f

t
r
a
n
s
a
c
t
i
o
n
s
.

V
a
l
u
e
s

c
o
r
r
e
s
p
o
n
d

t
o

t
h
e

e
n
d

o
f

e
a
c
h

p
e
r
i
o
d
,

e
x
p
r
e
s
s
e
d

i
n

c
u
r
r
e
n
t
d
o
l
l
a
r
s
,

e
x
c
l
u
d
i
n
g

d
i
s
t
r
i
b
u
t
e
d

e
a
r
n
i
n
g
s
.

S
e
l
e
c
t
e
d

d
a
t
e
s

c
o
r
r
e
s
p
o
n
d

t
o

p
e
a
k
s

a
n
d

m
i
n
i
m
u
m

l
e
v
e
l
s

f
o
r

t
h
e

L
a
t
i
n

A
m
e
r
i
c
a
n

a
v
e
r
a
g
e
.
prices dropped sharply around the time of the tequila crisis, which produced
contagion to all Latin American stock markets, and then nearly doubled
between March 1995 and June 1997, pushed up directly by portfolio in-
flows.
Domestic interest rates tended to be high at the outset of surge epi-
sodes, reflecting the binding external constraint faced by most countries
during periods of sharp reductions in capital inflows, the restrictive mon-
etary policies then in place, and the short-term bias of the financial re-
forms implemented in Latin America.
23
Finally, the increased supply of
external financing generated a process of exchange rate appreciation in
most Latin American countries (see table 1-3). This encouraged additional
inflows from dealers operating with maturity horizons within the expected
appreciation of the domestic currency.
The interaction between the two sets of factorsthe nature of inves-
tors and the adjustment processexplains the dynamics of capital flows
over time. When creditors discover an emerging market, their initial expo-
sure is nonexistent. They then generate a series of consecutive flows, which
result in rapidly increasing stocks of financial assets in the emerging mar-
ket. The creditors sensitivity to negative news, at some point, is likely to
suddenly increase remarkably with the stock of assets held in a country (or
region) and with the debtors degree of dependence on additional flows,
which is associated with the magnitude of the current account deficit, the
refinancing of maturing liabilities, and the amount of liquid liabilities
likely to flow out of the country in the face of a crisis. The probability of
a reversal of expectations about future trends therefore grows steeply after
a significant increase in asset prices and exchange rates accompanied by ris-
ing stocks of external liabilities.
The accumulation of stocks and a subsequent reversal of flows can
both be considered to be rational responses on the part of individual sup-
pliers, given the short-term horizon of the principal agents on the supply
side. Investors with short horizons are little concerned with whether cap-
ital surges improve or worsen long-term fundamentals while they continue
to bring inflows. What is relevant to these investors is whether the crucial
indicators from their point of viewexchange rates and real estate, bond,
and stock pricescan continue providing them with profits in the near
term and, obviously, that liquid markets allow them to quickly reverse
decisions. They will continue pouring money in until expectations of an
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio :
23. See Ffrench-Davis (2000, chap. 2).
imminent near reversal start to build. For the most influential financial
operators, the more relevant variables are not related to long-term funda-
mentals at all, but rather to short-term profitability. This explains why
they may suddenly display a radical change of opinion about the economic
situation of a country whose fundamentals, other than liquidity in foreign
currency, remain essentially unchanged during a shift from overoptimism
to overpessimism.
The widespread belief that withholding of information prevented the
Mexican crisis of 1994 being foreseen is therefore mistaken. While the
provision of official information on international reserves was admittedly
only sporadic, the key datanamely, real exchange rate appreciation, the
high current account deficit and its financing with volatile resources, and
low GDP growth despite booming flowswere available on a regular
basis. Data were also available on the significant crowding out of domes-
tic savings. However, Mexican policies were widely praised in 1993 by
financial institutions, the media, and risk evaluators.
24
The incorporation
of Mexico into the North American Free Trade Agreement (NAFTA) and
the Organization for Economic Cooperation and Development (OECD)
in 1994 served to intensify the trend. The crucial problem was that nei-
cionaiizariox oi iixaxciai voiariiir\ :,
24. Gurra (1995, p. 281).
Table 1-3. Latin America: Real Exchange Rate, 19872000
a
Indexes (198790 = 100)
Country 198790 1994:4 1995:1 1997:3 1999:1 2000:3
Argentina 100.0 63.1 63.8 65.6 56.6 56.9
Brazil 100.0 74.3 71.5 64.4 102.8 91.4
Chile 100.0 96.4 97.1 80.2 85.8 91.7
Colombia 100.0 80.4 79.7 71.2 80.3 97.4
Mexico 100.0 75.9 116.3 81.4 81.9 69.2
Peru 100.0 59.3 59.9 56.0 64.5 62.0
Weighted average
(18 countries) 100.0 75.9 83.9 71.3 85.3 78.7
Source: Authors calculations, based on official ECLAC figures.
a. Quarterly averages of real exchange rate indexes for each country with respect to the currencies
of their main trading partners, weighted by the share of exports to those countries; inflated by
external CPI and deflated by domestic CPI. For Brazil we weighted the Rio CPI index (two-thirds)
and the new official series of inflation (one-third). Selected quarters correspond to peaks and
minimum levels for the Latin American average.
ther those on the supply side nor those on the demand side paid sufficient
attention to the available information until after the crisis erupted.
It is no coincidence that in all three surges, loan spreads underwent a
sustained fall while the stock of liabilities rose sharply, lasting for five to
six years in the 1970s, for three to four years before the tequila crisis, and
for over a couple of years after that crisis. This implies a downward slop-
ing medium-run supply during the expansive phase of the cycle, which is
a highly destabilizing feature indeed (see figure 1-3).
One particularly relevant issue is that firms specializing in microfi-
nance, which may be highly efficient in their field but which are short-
sighted by training and by reward, have come to play a determining role in
generating macroeconomic conditions and policy design. This leads, unsur-
prisingly, to unsustainable macroeconomic imbalances, outlier macroeco-
nomic prices, and an undermined environment for productive investment,
particularly in tradables. Authorities should obviously be making decisions
with a long-term view, yet they become seduced by the lobbying and policy
recipes of microfinance experts and the financial press, which leads to irra-
tional exuberance (to use Alan Greenspans expression) despite evidence of
inefficiency in resource allocation and total factor productivity. Macroeco-
nomic authorities need to undertake the responsibility of giving priority to
fundamentals in order to achieve macroeconomic balances that are both
sustainable and functional for long-term growth. This requires that they
avoid entering vulnerability zones during economic booms based on capital
surges, since policy design is otherwise prone to being caught in the finan-
cierist trap.
The Economic and Social Costs of the Tequila and Asian
Crises in Selected Countries
The economic literature commonly asserts that the only correct way to
conduct policy is with an open capital account, as Mexico and Korea did in
the 1990s. In reality, however, there is significant room for policy diversity.
Chile and Taiwan provide two striking examples of policy diversity and suc-
cessful prudential macroeconomic management of the capital account.
Mexico and the Tequila Crisis
The Mexican crisis that exploded in 1994 illustrates the harm that can
be caused when the absorption of an excessive volume of capital inflows
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio :o
cionaiizariox oi iixaxciai voiariiir\ :;
25. Ros (in this volume); Sachs, Tornell, and Velasco (1996).
Figure 1-3. Latin America: International Bond Issues, 19922000
Source: ECLAC; World Bank; IMF.
a. Total annualized cost is equal to average differential on bond issues plus long-term U.S. Treasury bonds.
18
15
12
9
6
3
Billions of dollars per quarter
1992:1 1992:3 1993:1 1993:3 1994:1 1994:3 1995:1 1995:3 1996:1 1996:3 1997:1 1997:3 1998:1 1998:3 1999:1 1999:3 2000:1 2000:3
17
16
15
14
13
12
11
10
9
17
14
11
8
5
Cost (percent) Maturity (no. of years)
1992:1 1992:3 1993:1 1993:3 1994:1 1994:3 1995:1 1995:3 1996:1 1996:3 1997:1 1997:3 1998:1 1998:3 1999:1 1999:3 2000:1 2000:3
Maturities
Costs
a
Costs and Maturities
Issues
leads to the accumulation of a large stock of mostly short-term or liquid ex-
ternal liabilities and to a domestic credit boom.
25
Between 1990 and 1994,
producers and consumers accommodated to a level of overall expenditure
that rapidly outstripped potential GDP: expenditures exceeded effective
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio :
GDP by 8 percent in 199294. The real exchange rate appreciated signifi-
cantly, contributing to the sharp rise of the external deficit. Since the public
sector was in balance, the disequilibrium was located in the private sector.
When expectations of profitability were reversed, the amounts involved be-
came unsustainable. Creditors cut financing sharply, forcing Mexico into a
highly contractionary adjustment and a huge devaluation after the author-
ities adopted the flexible exchange rate. Despite the large package of inter-
national support that Mexico received in 1995,
26
GDP dropped that year by
6.1 percent and capital formation by nearly 30 percent.
GDP recovered strongly shortly thereafter, but the overall rise in GDP
averaged only 3.5 percent in 19952000. The growth trend continued to be
slow despite the fact that Mexico received a positive shock from the U.S.
boom: export volume expanded vigorously by 17 percent a year in 1995
2000, four-fifths of which was directed to the U.S. markets. The investment
ratio, which dropped sharply in 1995, did not fully recover until 1998. Real
wages decreased substantially during the crisis and had not recovered by
2000. This was also true of poverty, which increased from 45 to 52 percent
of the population between 1994 and 1996. Though the indicator fell back
to 47 percent in 1998, it had still not returned to pre-crisis levels by 2000.
27
The Mexican crisis did not trigger a widespread contagion effect
throughout the region in 1995, in contrast to 1982. The most notable ex-
ception was Argentina, which experienced significant drops in GDP, em-
ployment, and investment in 1995. Nonetheless, many countries recorded
negative flows in several segments of the supply of financing, particularly
bonds, deposits, and flows to stock markets. Subsequently, the flow of funds
became extremely abundant again. GDP recovery in Argentina and Mexico
was particularly vigorous, since the sharp drop in both countries in 1995
had generated a large gap between effective GDP and productive capacity.
This facilitated a significant degree of reactivation, which led to compla-
cencynot only in those countries, but also in international financing in-
stitutions and elsewhereregarding the effects of crises and the capacity to
recover from them. However, as said above, growth in Mexico was slow for
the six-year period 19952000 as a whole, and it was even slower in Argen-
tina (1.7 percent), which entered into another severe recession in
19992000.
Significant differences also emerged between Argentina and Mexico.
Mexico moved from a quasi-fixed nominal exchange rate to a flexible rate in
26. Lustig (1997).
27. ECLAC (2000c).
cionaiizariox oi iixaxciai voiariiir\ :,
1995, which facilitated the adjustment to the financial shocks generated by
the Asian crisis. Also, the country experienced a sizable positive shock
associated with the rapid growth of the U.S. economy in 19982000. Ar-
gentina, meanwhile, was tied to the currency board; this limited the coun-
trys ability to confront the negative shocks from Asia, from the devaluation
of the Brazilian currency, and from the revaluation of the dollar. Argentinas
inability to facilitate the correction of relative prices through the active use
of the nominal exchange rate was intensified by the strong international ap-
preciation of the dollar and the persistent dryness in the supply of external
funding. Indeed, the international capital markets underwent lengthy dry
periods during the recent crisis, despite the boom in the U.S. economy and
the acceleration in Europe in 2000. This indicates that financial markets
may tend to recover slowly and incompletely, as was clearly the case in the
1980s.
Korea and the Asian Crisis
The East Asian countries suffered deep recessions in 1998, after
decades of sustained annual GDP growth of around 8 percent. Indonesia
exhibited a 13 percent recession, similar to the spectacular drop of Chi-
lean GDP during the 1982 recession. Korea, Malaysia, and Thailand re-
corded reductions of 7 to 10 percent. The nature of the crises varied across
countries in Asia. The Korean and Thai experiences, however, appeared to
be associated with the capital surge of the 1990s and the resulting liquid-
ity. During 19992000, Korea (and Malaysia) recovered faster than the
other countries of the region. Notwithstanding Koreas impressive GDP
growth in 1999 and 2000, the costs have been significant: in the period
19982000, GDP was about 12 percent below what it would have been
had the historical trend continued, and investment fell by over one-fifth
in 199899 as compared to 199396 (28.9 percent and 36.4 percent, re-
spectively). Since the country also achieved an impressive external surplus
of 9 percent of GDP in 199899, versus a deficit of almost 5 percent in
1996, disposable income was well below output.
The recessions in East Asia are comparable to those of Latin America in
198283, with drops in productive investment, banking crises, and social
regression.
28
Apart from the intrinsic strengths of the Korean economic
28. For interpretations of the Asian crisis, see Krugman (1999); Akyz (1998); Furman and Stiglitz
(1998); Perry and Lederman (1998); Radelet and Sachs (1998); Reisen (1998); Stiglitz (1998); Wang
(2000); Wyplosz (1998).
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio :c
structure and of several other Asian economies (excluding Indonesia), four
features of the international economic environment largely explain why the
shift from recession to recovery came sooner in Asia than in the Latin
American debt crisis: the plentiful supply of official external financing;
rapid action spearheaded by the United States to refinance private credits,
particularly interbank lending; significantly lower interest rates in the ad-
vanced economies; and higher growth rates, especially in the United States.
The countercyclical policy implemented by the Korean public sector also
played a significant role in the recovery; consequently, the fiscal balance
swung from a surplus of 0.3 percent of GDP in the period 199396 to a
deficit of 4.7 percent in 199899.
Until the early 1990s, Korea had extensive capital account regulations,
based on a combination of market forces and state guidance.
29
In 1991, the
country began implementing a broad range of measures aimed at liberaliz-
ing the capital account. Contrary to what is commonly assumed by observ-
ers, greenfield FDInot acquisitionswas deregulated. Local firms and
banks were allowed to issue securities abroad, and foreigners were author-
ized to purchase stocks in Korean companies subject to limits that were
raised progressively starting in 1992. Foreign currency loans to local firms,
trade credit, and short-term financing were also liberalized. Only long-
term borrowing and acquisitions remained restricted. Under the new regu-
lations, Korean banks and firms were permitted to engage in arbitrage
between international lenders and local markets by borrowing short abroad
and, in some cases, lending long at home. This practice would not have
been allowed under the regulations prevailing before the liberalization drive
in the early 1990s. In addition, Koreas sound creditworthiness afforded
local firms lower spreads and more expeditious access to funding, which
they used, in part, to borrow for financing investment and lending in other
Asian markets.
30
Capital inflows expanded hugely after liberalization and included pur-
chases of stock shares, bond issues, and private loans to banks and nonfi-
nancial firms. Liabilities became highly liquid, with short-term debt
reaching twice the level of international reserves in 1996. This was not ac-
companied, however, by strengthened prudential regulation and supervi-
sion, in a replication of the negative Latin American experience.
31
29. See Agosin (in this volume); Furman and Stiglitz (1998); Wang (2000).
30. In May 1995, one of the large international risk-rating agencies upgraded the sovereign credit
rating of Korea (see Wang, 2000). In June 1997, the World Economic Forum classified Korea as the
fifth most secure place to invest in the world (see Dean, 1998).
31. See ECLAC (1998, chap. 12). For an English version, see ECLAC (1995).
cionaiizariox oi iixaxciai voiariiir\ ::
The process of liberalizing the capital account caused the exchange
rate to appreciate with respect to the currencies in which borrowing took
place, which encouraged further borrowing. Korea accommodated the
capital surplus through import liberalization and currency appreciation,
together with a relaxation of domestic liquidity constraints. The combined
effect of increased imports and worsening export prices explain the rise in
the current account deficit to 5 percent of GDP in 1996. Net inflows rose
from U.S.$7 billion in 1992 to U.S.$24 billion in 1996, but gross inflows
amounted to U.S.$49 billion.
The opening of the capital account represented a source of vulnerabil-
ity, exacerbated by poor bank regulation and supervision. It left Korea
prone to contagion, even though the fundamentals were generally sound.
In fact, Agosin (in this volume) concludes that the main factor behind the
Korean debacle was the liberalization of the capital account, without the
concurrent adoption of adequate bank regulation and prudential macro-
economic measures to discourage excessive capital inflow.
Chile and Taiwan: Going against the Fashion
Is it possible to forge ahead with policies that go against contempo-
rary economic ideology? Two casesChile in the first half of the 1990s
and Taiwan throughout the decadeprovide evidence that it is indeed
possible and can be an efficient and cheap way to avoid costly crises.
Chile deployed three types of capital account policy in the last quar-
ter of a century.
32
The first was the plain neoliberal experiment of the
1970s, which culminated in a major crisis in 1982. That episode featured
a drop in GDP of 14 percent, followed by an increase in open unemploy-
ment to 30 percent in 1983. The second approach was that taken from
1990 to 1995. In that period, authorities went against prevailing fashion
by pursuing a set of active macroeconomic policies that included the pru-
dential regulation of financial inflows. As a result, Chile was practically un-
affected by contagion from the tequila crisis. The third type of policy,
which was implemented after 1995, involved a relative relaxation of mac-
roeconomic prudential policies. This gave way to a significant appreciation
and allowed the external deficit to double from 2.5 percent of GDP in
199095 to 5.7 percent in 199697.
Chiles performance in 1995 was diametrically opposed to that of
Mexico, despite numerous similarities in the two economies in the preced-
32. Ffrench-Davis and Tapia (in this volume).
ing years. The difference in results is attributable to the distinct divergence
in macroeconomic policies in the first half of the 1990s. Toward the end
of the 1980s, both countries had already liberalized their trade consider-
ably, they had substantially improved their fiscal budgets, privatization was
well underway, the annual rate of inflation was around 2030 percent, and
they showed similar domestic savings rates. In 199094, however, Chile
and Mexico chose divergent approaches with respect to the management
of capital inflows, exchange rate policy, and the prudential regulation and
supervision of the domestic financial system. Chiles main advantage over
Mexico in 1995 is that it responded to the abundance of external funds
from 1990 to 1994 with a deliberate policy of active prudential macroec-
onomic regulation.
Instead of allowing in and spending all the available external supply,
which would have led to significant appreciation of the peso and a ris-
ing current account deficit, the Chilean authorities chose to discourage
short-term capital inflows. In 1991 a tax was imposed, and substantial
non-interest-bearing reserves were required for external credit. The reserve
requirement was subsequently extended to foreign currency deposits and
investment in second-hand stocks, while primary issues of American de-
pository receipts (ADRs) and FDI venture capital were kept exempted.
FDI had to be held in Chile for at least one year. The financial system was
subject to relatively strict prudential regulation, including a selective
supervision of bank assets and required provisioning, as well as restrictions
and drastic penalties on operations with related parties. This set of meas-
ures effectively discouraged speculative capital inflows.
33
Most empirical
research shows that these regulations had a significant effect on the vol-
ume of short-term inflows, and, contrary to common belief, several stud-
ies also show an effect on total inflows.
34
The smooth transition to democracy, an increasing international ap-
proval of Chilean economic policies, and high domestic interest rates com-
bined to boost capital inflows to Chile after mid-1990; the process was
earlier and relatively stronger than in other emerging economies. As a con-
sequence of its prudential macroeconomic policies, however, Chile had
only a moderate external deficit by late 1994, with high international re-
serves, a manageable short-term debt, a domestic savings rate that was ris-
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio ::
33. Agosin and Ffrench-Davis (2001); Le Fort and Lehmann (2000).
34. See several references in the case of Chile in this volume. The classical paper arguing against
the effectiveness of the Chilean reserve requirement is Valds-Prieto and Soto (1998).
ing instead of falling (whereas the opposite was the case in Mexico), and a
level of domestic investment that far exceeded historical records from 1993
onward. The exchange rate in 199094 was comparatively closer to equi-
librium than in most Latin American countries, as reflected by a moder-
ate current account deficit over that period.
Policy was effective in achieving its targets for most of the 1990s, but
in 199697 the policy mix and the intensity with which it was applied re-
mained unchanged despite a new vigorous capital surge to most countries
in the region. Chile, in particular, was a target of these flows, since it had
remained immune to the tequila contagion. This surge should have been
met with increased restrictions on rising inflows. In the absence of such
measures, inflows became excessive as investors were willing to pay the in-
sufficiently restrictive cost of the reserve requirement, with no evidence of
significant evasion; also, as Le Fort and Lehmann emphasize, some inflows
that ought to have been made subject to regulation remained exempt.
35
Although the Central Bank intervened heavily in foreign exchange
markets, a sharp real exchange rate appreciation and a rise of the current ac-
count deficit were observed over the two-year period, which pushed Chile
into a vulnerability zone. Nonetheless, the active regulation implemented
up to the mid-1990s had left large international reserves, a low stock of for-
eign liabilities, and a small share of volatile flows. Unfortunately, those
strengths were partially undermined by the excessive exchange rate appreci-
ation and high current account deficit recorded in 199697, as well as the
Central Banks delay in reacting to the deterioration of the external envi-
ronment. In particular, the Bank resisted market pressures for devaluation,
concerned that exchange rate depreciation would increase inflation and
worsen the balance sheets of the large domestic firms that became highly
indebted in foreign currency liabilities in the period. The outcome was a
sort of automatic adjustment featuring a sharp loss of reserves, a 10 percent
fall in aggregate demand, a 1 percent drop in GDP, a 3.5 percentage point
increase in the unemployment rate, and a marked drop in capital formation
in 1999. Despite this recent recession, however, Chile achieved an average
growth rate of over 6 percent for the 1990s, which was its best performance
ever recorded in a decade.
The case of Taiwan is more straightforward. Taiwan similarly pursued
a policy that went against the global trend, but its policy was notably
tougher than Chiles. It kept in place a variety of direct capital controls that
cionaiizariox oi iixaxciai voiariiir\ :,
35. Le Fort and Lehmann (2000).
had been initiated in previous years.
36
The Taiwanese dollar remained non-
convertible. The authorities restricted inflows to banks, firms, and the
stock market. Acquisitions remained forbidden (as they were in Korea until
the crisis). Taiwan held quantitative restrictions on the stock shares that for-
eign persons and institutional investors could own. Domestic banks were
also strictly supervised by the Central Bank. Consequently, external liabil-
ities were low at the outbreak of the Asian crisis, because they had been re-
stricted during the preceding boom in financing to emerging economies.
The Taiwanese economy exhibited a notably stable dynamism, simi-
lar to that experienced in Korea for three decades. Data on Taiwan are
rather limited, but the available information is quite revealing. For the
years immediately before the crisis (199396), the available information
shows a GDP growth of 6.7 percent, with negligible capital inflows and a
current account surplus. Taiwan was clearly outside the capital surge. An-
other remarkable feature of Taiwan is that its fiscal performance was not-
ably different from that of most emerging economies: the government ran
a significant budget deficit, with public investment financed in the domes-
tic private market. In all, domestic private savings financed a fiscal deficit
of 2.7 percent of GDP and an external surplus of 3.2 percent in 199296.
The fiscal deficit coexisted with a stable real exchange rate and an infla-
tion rate that was lower than in Chile, Korea, or Mexico.
Taiwan experienced some contagion and a few attacks on the Taiwa-
nese dollar despite the palpably safe macroeconomic environment, which
was further bolstered by huge international reserves (U.S.$88 billion in
1996). The attack quickly receded, however, and an initial devaluation of
20 percent quickly recovered by half. In 19982000 only a slight down-
ward adjustment took place in economic growth, which fell to 5.9 percent
(as compared to 4.2 percent in Korea). The strong adherence to real mac-
roeconomic sustainability appears to have been extremely rewarding for
Taiwan.
Misleading Recipes and Misconceptions in Fashion
A series of widely accepted hypothesis or beliefs form part of the con-
ventional wisdom of the financial world, including international financial
institutions, although these have undergone some positive changes in their
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio :
36. Agosin (in this volume).
perceptions in the wake of the Asian crisis. This section reviews five com-
mon assertions that we believe have significant policy and welfare implica-
tions. These beliefs are not validated by the actual performance of
international financial markets and still less by the behavior of domestic
markets in emerging economies.
Recovery from Recent Crises Has Been Rapid
A first, common assertion is that the recovery from crises has proved
to be fast. This is fundamental for claims that authorities should allow the
free, self-correcting operation of markets, since it is assumed that attempts
to adopt policies to counter booms or accelerate recoveries tend to gener-
ate additional instability. Given that financial markets are the major source
of economic instability for emerging economies, this line of reasoning im-
plies that instability is inevitable but not excessively costly.
This assumption is inconsistent with the evidence. Crises generate
medium- and long-term effects on financial markets. The most significant
case in recent decades was the effect of the Latin American debt crisis on
long-term syndicated bank loans, which were then the principal mecha-
nism of financing. This form of lending disappeared after the crisis, and
bank lending never returned to being a major source of financing for Latin
American countries in the 1990s. Equally important, the Asian crisis had
negative effects on bond financing, which have still not been surmounted.
Issues have remained volatile, costs high, and average maturities low. The
usual assertion that markets have recovered rapidly is based on a very par-
tial view of external market conditions, which takes into account the abil-
ity of emerging economies to return to the market, but not the conditions
of such access in terms of stability, costs, and maturity of financing.
As shown above, a rapid return to positive GDP growth rates is by no
means a good basis for asserting that the effects of crises are short-lived. All
countries that have undergone severe crisesincluding Korea, where re-
covery was very strongdisplay evidence that they were pushed into a
lower GDP path. Three channels through which these medium-term or
even permanent effects on GDP are transmitted are particularly important.
First, investment falls sharply during a crisis, which affects the path of pro-
ductive capacity. Second, the bankruptcy of firms generates a loss of capac-
ity, as well as a permanent loss of the goodwill, productive and commercial
networks, and social capital of those firms. Even firms that do not go bank-
rupt may pass through a long period of debt restructuring, in which prop-
cionaiizariox oi iixaxciai voiariiir\ :,
erty rights are indeterminate; this uncertainty may affect their perfor-
mance. Finally, domestic financial crises may ensue if the portfolios of do-
mestic financial institutions deteriorate severely. The experience of
emerging economies (and Japan) indicates that restoring a healthy financial
system takes several years and generates adverse effects throughout the pe-
riod in which it is rebuilt.
A growing body of evidence indicates that boom-bust cycles also have
ratchet effects on social variables.
37
The deterioration of the labor market
(through open unemployment, a worsening in the quality of jobs, or a fall
in real wages) is generally very rapid, whereas the recovery is painfully slow
and incomplete. This is reflected in the long-lasting worsening of Argen-
tine unemployment and Mexican real wages after the tequila crisis. It is
also reflected in Brazilian and Chilean joblessness following the 1999 re-
cession, despite a fairly positive recovery in both cases.
38
The evidence on
poverty similarly indicates that a ratchet effect is present. As mentioned
above, poverty in Mexico had not abated to 1994 levels despite economic
recovery.
Opening the Capital Account Discourages Macroeconomic Disequilibria
It is commonly argued that fully opening the capital account deters
domestic macroeconomic mismanagement and encourages good macroec-
onomic fundamentals. This is partly true for domestic sources of instabil-
ity, such as large fiscal deficits, permissive monetary policy, and arbitrary
exchange rate overvaluation. The volatility in market perceptions makes
this type of control highly unreliable, however: financial markets tend to
encourage lax demand policies and exchange rate overvaluation during
booms, whereas excessive punishment during crises may actually force
authorities to adopt overly contractionary policies (so-called irrational
overkill). Contrary to the usual argument, this is not associated with inap-
propriate information. Even well-informed market actors, such as credit
rating agencies or investment banks, usually operate procyclically.
39
The opening of the capital account may actually lead emerging econo-
mies to import external financial instability, with capital inflows engender-
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio :o
37. Lustig (2000); Rodrik (2001).
38. ECLAC (2000a).
39. On rating agencies, see Larran, Reisen, and von Maltzan (2000). Market discipline can also
pose obstacles to necessary social reform (for instance, to higher taxes to finance efficient human cap-
ital investment) or to the ability to capture economic rents from natural resources.
ing a worsening in macroeconomic fundamentals. Thus while market disci-
pline can serve as a check on domestic sources of instabilityalbeit not a
very efficient one, given the whims of opinion and expectations charac-
teristic of financial marketsit is a source of externally generated instabil-
ity. Not only will the market inaccurately perceive some domestic policies as
inadequate, it may actually induce key variables to deviate from sustainable
levels. In other words, it is the market itself that generates incentives for
emerging economies to enter a vulnerability zone during the booms.
Financial operators evidently fulfill a useful microeconomic function
as intermediaries between savers and users of funds, as hedgers of risk, and
as providers of liquidity. In practice, however, they have perhaps unwit-
tingly come to play a role that has significant macroeconomic implications.
Their herd-prone expectations have contributed to intensifying financial
flows toward successful countries during capital surges, thus facilitating
rapid increases in financial assets and real estate prices and sharp exchange
rate appreciation in the recipient markets. Apart from the poor quality of
prudential regulation and supervision in these markets, these macroeco-
nomic signals contribute to prolonging a process that appears, mislead-
ingly, to be efficient and sustainable (with good profits and loan guarantees,
supported by high stock prices and the low value in domestic currency of
dollar-denominated debt). In fact, bubbles are being generated with outlier
macroeconomic prices, and sooner or later the bubbles tend to burst.
The general practice has been to encourage the recipient countries to
accept the increasing supply of resources from the international financial
institutions and financial specialists, and to praise them for doing so. The
cost of external financing typically falls during boom periods, which im-
plies that the market actually operates with a downward-sloping mid-run
supply of funds. Excessive indebtedness and periods of massive outflows
ensue, often prompting admonishment by the very agents who praised the
economic performance of the emerging economies during the boom.
There is an obvious contradiction between these two attitudes.
Fundamentals are clearly essential. There is wide misunderstanding,
however, about what constitutes sound fundamentals. The inappropriate
conventional definition, together with irrational exuberance, led to high
positive grades being given to Chile just before the crisis of 1982, to Korea
and Thailand in 1996, and to Mexico and Argentina in 1994. Something
fundamental was thus missing in markets evaluation of market fundamen-
tals! The severe crises of these five countries cannot all be due to bad luck or
contagion alone. Rather, certain crucial components of a comprehensive
cionaiizariox oi iixaxciai voiariiir\ :;
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio :
set of fundamentals deteriorated, led by massive capital inflows. A satisfac-
tory definition of fundamentals should thus include not only low inflation,
sound fiscal accounts, and dynamic exports, but also sustainable external
deficit and net debt, low net liquid liabilities, non-outlier real exchange
rate, and strong prudential regulation, supervision, and transparency of the
financial system.
Corner Exchange Rate Regimes Are the Only Right Policy
Alternatives Today
In todays open developing countries, the exchange rate regime is sub-
ject to two conflicting demands, which reflect the limited degrees of free-
dom that authorities face in a world of weaker policy instruments and
reduced policy effectiveness. The first demand comes from trade: with the
dismantling of traditional trade policies, the real exchange rate has become
a key determinant of international competitiveness. The second is from
the capital account. Boom-bust cycles in international financial markets
generate a demand for flexible macroeconomic variables to absorb, in the
short run, the positive and negative shocks generated during the cycles.
Given the reduced effectiveness of traditional policy instruments, par-
ticularly monetary policy, the exchange rate can play an essential role in
helping to absorb these shocks. This objective, which is associated with
short-term macroeconomic management, is not easily reconcilable with
the trade-related goals of exchange rate policy.
Many analysts call for limiting the alternatives to the two polar ex-
change rate regimes of a totally flexible exchange rate versus a currency
board (or outright dollarization). The relevance of the dual demand is not
captured in this approach, however, whereas intermediate regimes featur-
ing managed exchange rate flexibility, such as crawling pegs and bands and
dirty floats, attempt to reconcile these conflicting demands.
40
Currency boards certainly introduce built-in institutional arrange-
ments that provide for fiscal and monetary discipline, but they reduce or
even eliminate any room for stabilizing monetary, credit, and fiscal poli-
cies, which are all necessary for preventing a crisis and facilitating recov-
ery in a post-crisis environment. Currency boards thus allow the domestic
transmission of shocks originating in international capital markets. In so
doing, they generate strong swings in economic activity and asset prices,
with the corresponding accumulation of domestic financial vulnerability.
40. See Frankel (1999) and Williamson (2000) on intermediate exchange rate policies.
cionaiizariox oi iixaxciai voiariiir\ :,
At the other extreme, the volatility that is characteristic of freely float-
ing exchange rate regimes is not a problem when market fluctuations are
short-lived, but it becomes a major concern when there are longer waves,
as has been typical of the access of emerging economies to capital markets
in recent decades. In this case, volatility tends to generate perverse effects
on resource allocation. Moreover, under freely floating regimes with open
capital accounts, anticyclical monetary or credit policies exacerbate cycli-
cal exchange rate fluctuations, with their associated allocative, wealth, and
income effects. In fact, authorities have to adopt procyclical monetary and
credit policies if they want to smooth out real exchange fluctuations under
these conditions.
The ability of a flexible exchange rate regime to smooth out the effects
of externally induced boom-bust cycles thus depends on the authorities ca-
pacity to effectively manage a countercyclical monetary and credit policy
without enhancing procyclical exchange rate patterns. This is only possible
under intermediate exchange rate regimes with capital account regulation.
Because these intermediate regimes of managed flexibility grant a degree of
effective, albeit limited, monetary autonomy, they provide the best oppor-
tunity to respond to the dual demands on exchange rate policy. Such
regimes obviously have shortcomings and may generate costs. First, all
intermediate regimes are subject to speculative pressures if they do not gen-
erate credibility in markets. The costs of defending the exchange rate under
these conditions are very high; at critical conjunctures it may be advisable
to move temporarily to full flexibility. Second, sterilized reserve accumula-
tion during booms may also be costly. Finally, the capital account regula-
tions that are necessary for managing intermediate regimes efficiently are
only partially effective. All things considered, however, intermediate re-
gimes offer a sound alternative to costly volatility.
External Savings Tend to Complement Domestic Savings
There is a strong correlation between investment and growth rates,
with a significant interaction between capital accumulation and technical
progress.
41
The Latin American experience in recent decades also provides
compelling evidence that the way investment is financed is not irrelevant,
because external savings are unstable and may crowd out domestic savings.
Moreover, the induced real exchange rate effects of unsustainable external
financing may lead to a misallocation of resources.
41. Schmidt-Hebbel, Servn, and Solimano (1996).
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio ,c
In the 1950s and 1960s capital flows to developing countries were
mostly tied to particular investments and to public users, financing real as-
sets mostly through direct investment and official project lending. The na-
ture of financing tended to generate strong complementarities between
external and domestic savings. Commercial bank lending in the 1970s and
private portfolio investments in the 1990s, however, made the link between
foreign savings and domestic investment weaker and less direct.
42
This dis-
association between capital flows and actual investment has three implica-
tions: inflows may increase consumption rather than investment; inflows
do not necessarily enhance the recipient countrys ability to earn foreign ex-
change through expansion of capacity in the tradables sector; and foreign
investment is easily reversed, insofar as the acquisition of securities is essen-
tially a short-term commitment. The first of these effects is associated with
the empirical fact that consumers and financial asset markets tend to re-
spond more quickly to released liquidity constraints than do productive in-
vestors, who have a longer response lag because of the irreversibility and
long maturity of their decisions. Nonetheless, construction generally re-
sponds faster than other forms of capital formation and is more prone to in-
vestment overshooting, given the effects of liquidity on asset prices.
The diverse behavior of markets and participants explains why the
stability of flows has a significant effect on the relation between foreign and
domestic savings. Uthoff and Titelman distinguished between the trend
and deviations from the trend for foreign and domestic savings.
43
They find
a strong crowding out of domestic by foreign savings when capital inflows
are temporary (above or below the trend). Moreover, the effects of flows on
the real exchange rate distort the allocation of investment. Real apprecia-
tion during booms may lead to a suboptimal allocation of funds to the ex-
pansion of productive capacity in tradables sectors (that is, it may generate
so-called Dutch disease effects). This seriously weakens the mid-term ob-
jective of penetrating external markets with new exports and strengthens
the negative vis--vis the positive effects of trade reform.
44
Prudential Regulation and Supervision of Banks is Sufficient for Facing
External Volatility
Lax or poor prudential regulation and supervision of domestic finan-
cial institutions obviously reinforces disequilibria. Strong regulation and
42. Turner (1996).
43. Uthoff and Titelman (1998).
44. ECLAC (1995, 1998, 2000b); Ffrench-Davis (2000, chap. 3); World Bank (1998).
cionaiizariox oi iixaxciai voiariiir\ ,:
supervision per se does not solve the problem, however. First, a significant
share of capital inflows usually is not intermediated by domestic financial
institutions; cases in point include Chile before the 1982 crisis and Korea
and Mexico in the 1990s. Nonbank to nonbank flows have become increas-
ingly important as a result of the type of diversification that has taken place
in supply. Second, expectations are in themselves volatile. This applies not
only to expectations of financial intermediaries, but also to those of na-
tional authorities and international financial institutions. Regulation and
supervision may influence the transmission mechanism, but it can hardly
affect the source of instability (volatility in expectations). Third, normal
regulatory practices (including Basel Committee criteria) have procyclical
features.
45
In particular, there is a lag between loan-loss provisions and over-
due loans; and in turn, these lag behind the level of economic activity. Reg-
ulations are generally ineffective in dampening the strong incentive to lend
during booms because overdue loans and provisions are low, prices and
guarantees are biased upward, and high profit levels facilitate meeting the
capital requirements necessary to increase lending. During crises, the level
of overdue loans increases, and provisions must therefore be increased as
well. This reduces the capacity to expand lending by biting into capital re-
quirements, at the same time that low profits also reduce the funds available
to increase the capital base of financial intermediaries.
Robust Policy Lessons
Several of the policy lessons that we developed in earlier works are re-
inforced by the research in this volume on the recent experiences in Asia
and Latin America.
46
We group these lessons into five areas of action.
Maintain a Sustainable Volume, Composition, and Use of
Capital Inflows
The volume of inflows must be consistent with the absorptive capac-
ity of the host country. The failure to address this point is at the core of
recent macroeconomic instability in emerging economies. Absorption
capacity refers, of course, to both the use of existing productive capacity
and the creation of new capacity. The composition of flows is relevant for
three reasons. First, FDI (excluding acquisitions of existing assets) feeds di-
45. Griffith-Jones (in this volume); Ocampo (2001).
46. See also Ffrench-Davis (2000); Ocampo (1999).
rectly into capital formation, as do long-term loans to importers of capi-
tal goods. Second, volatile flows tend to have a direct impact on foreign
exchange and stock markets and a relatively weak impact on capital for-
mation, which requires long-term financing. Third, temporary capital
surges tend to leak into consumption, since consumers can respond faster
than firms undertaking irreversible productive investment.
Allowing an excessively large share of capital inflows to drain off into
the stock exchange and the consumption of imported goods will usually
create bubbles in asset markets and imbalances in the external sector, which
tend to be unsustainable. Fast rising stocks of net liquid foreign liabilities,
in particular, generate deep vulnerabilities. Consequently, higher ratios of
long-term flows and productive investment imply a higher capacity for
efficient absorption. This means not only that the economy will efficiently
absorb a higher volume of capital flows, but that a higher flow will be sus-
tainable.
Recent experience offers a dramatic demonstration that recipient
emerging economies can pay a high cost for allowing the volume and com-
position of capital flows to be determined by the markets dominated by
agents with short horizons. The microeconomic costs associated with the
use of regulations on capital inflows should therefore be balanced against
the social benefits in terms of macroeconomic stability, investment, and
growth. Effective and efficient regulation can result in higher and sus-
tained GDP growth, as occurred in Chile and Taiwan in the 1990s.
Avoid Outlier Prices and Ratios
Economic authorities must ensure that capital flows do not generate
outlier prices or significant distortions of basic macroeconomic indicators,
such as interest rates, exchange rates, aggregate demand, the composition
of expenditure in terms of consumption and investment, and the produc-
tion of tradables. An artificial increase in absorptive capacity characterized
by outlier appreciation and reduced interest rates usually leads to costly ad-
justment. First, real appreciation during booms tends to distort the allo-
cation of investment, seriously weakening the structural goal of increasing
competitive export capacity. Second, if productive investment capacity re-
acts with a lag and domestic financial markets remain incomplete and
poorly supervised, capital surges cannot be absorbed efficiently in the do-
mestic economy and instead either leak to inefficient investment or crowd
out domestic savings.
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio ,:
Capital surges should not be used to target a single domestic economic
variable, such as inflation. This tends to throw other major variables off
balance. It is risky to remain bound to a fixed nominal rate or to dollarize
permanently, unless the economy shares an optimum currency area with
the United States. As discussed above, intermediate exchange rate regimes,
which cover a broad diversity, are generally preferable, as they are better
adapted to managing the dual demands they face in emerging economies.
Adopt Flexible and Comprehensive Macroeconomic Regulation
Across-the-board opening-up of the capital account was premature.
Countries should have postponed the process, proceeding selectively until
other major reforms had been consolidated and new equilibrium prices es-
tablished.
47
During structural adjustment, open capital accounts tend to
allow capital flows to increase too fast, especially when international fi-
nancing is abundant. This produces destabilizing macroeconomic and sec-
toral effects.
It is therefore unwise to make an inflexible commitment to fully open-
ing the capital account, particularly in the light of the crucial importance of
macroeconomic stability and the disproportionate volume of the interna-
tional capital markets compared to the small size of emerging economies.
As long as flows depend on short-term horizons and domestic securities
markets remain shallow, this new modality of linkages with the global
economy will carry the risk of severe instability. The recent experiences of
Mexico, Korea, and Thailand attest to the wisdom of discouraging the ac-
cumulation of large short-term financial liabilities. Domestic prudential
macroeconomic regulations offer the best defense given the present inter-
national financial architecture. If this approach makes use of a market-
based set of policies, including, for instance, the Chilean-style reserve
requirement, its level must be adjusted to the intensity of the supply of
funds.
Sustaining economic growth in the face of volatile capital flows re-
quires the deployment of a battery of policy instruments, including pru-
dential price-based capital account regulations to deter speculative inflows
and improve their maturity structure; an exchange rate regime based on a
crawling band with intramarginal intervention or managed flexibility; the
sterilization of the monetary effects of capital inflows; strong prudential
cionaiizariox oi iixaxciai voiariiir\ ,,
47. Williamson (1993).
regulation and supervision of the financial system, with countercyclical de-
vices; and strong fiscal accounts, also with countercyclical mechanisms.
Reserve requirements alone (or any other policy that increases the cost of
external borrowing), while clearly useful, are insufficient for deterring
speculative attacks when large exchange rate fluctuations are anticipated.
Obviously, this implies that authorities must make an effort to avoid cu-
mulative exchange rate disequilibria and sharp changes in the macroeco-
nomic environment.
Reform the International Environment for a More Efficient and
Balanced Globalization
The governance of domestic and international financial markets is key
to the future of the world economy. A common factor in recent crises has
been the great volatility of the most rapidly growing segment of interna-
tional financial markets: namely, short-term and speculative funds. Succes-
sive waves of overexpansion, followed by financial panic, indicate that the
market tends first to grow and then to contract more than is justified by
economic fundamentals. These features are inconsistent with a balanced
and efficient globalization. More energy is being spent on resolving crises
than on avoiding them. While appropriate prudential regulation of do-
mestic financial markets has obviously been lacking in most of the emerg-
ing economies affected by the crises, the lack of appropriate international
and regional institutions to monitor such a sophisticated, but unstable, fi-
nancial market is even more notorious.
48
Focus on Crisis-Prevention Policy, Based on Prudential Management
of Booms
The focus of attention for international and domestic institutions
should be the management of booms, rather than the resolution of crises,
which are usually the consequence of badly managed booms. Given that
existing international institutions and instruments have been ineffective in
warning of impending turbulence and instead have actually encouraged
unsustainable booms, it is particularly relevant to design domestic pruden-
tial macroeconomic policies and appropriate domestic regulatory frame-
works, aimed at controlling booms before they become unsustainable.
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio ,
48. United Nations (1999); Griffith-Jones (in this volume); Ocampo (1999, 2001).
This principle also applies to the design of the international institutions
required for a more balanced and stable globalization.
References
Agosin, M., and R. Ffrench-Davis. 2001. Managing Capital Inflows in Chile. In Short-
Term Capital Flows and Economic Crises, edited by S. Griffith-Jones, M. F. Montes, and
A. Nasution. WIDER/Oxford University Press.
Akyz, Y. 1998. The East Asian Financial Crisis: Back to the Future. In Tigers in Trou-
ble: Financial Governance, Liberalisation and Crises in East Asia, edited by K. S. Jomo.
London: Zed Books.
Ariff, M., and O. G. Ean. 2000. East Asian Response to the Instability of Financial Mar-
kets. In The Management of Global Financial Markets, edited by J. J. Teunissen. The
Hague: Forum on Debt and Development (FONDAD).
Calvo, G. 1998. Varieties of Capital-Market Crises. In The Debt Burden and its Conse-
quences for Monetary Policy, edited by G. Calvo and M. King. Macmillan.
Culpeper, R. 1995. Resurgence of Private Flows to Latin America: The Role of North
American Investors. In Coping with Capital Surges, edited by R. Ffrench-Davis and S.
Griffith-Jones. Lynne Rienner.
Dean, J. 1998. Why Left-Wing Moralists and Right-Wing Academics Are Wrong about
Asia. Challenge 41(2).
Devlin, R. 1989. Debt and Crisis in Latin America: The Supply Side of the Story. Princeton
University Press.
Daz-Alejandro, C. F. 1985. Goodbye Financial Repression, Hello Financial Crash. Jour-
nal of Development Economics 19(1/2).
ECLAC (United Nations Economic Commission for Latin America and the Caribbean).
1995. Latin America and the Caribbean: Policies to Improve Linkages with the Global Econ-
omy. Santiago.
. 1998. Amrica Latina y el Caribe: polticas para mejorar la insercin en la economa
mundial, 2d ed. Santiago: Fondo de Cultura Econmica and ECLAC.
. 2000a. Economic Survey of Latin America and the Caribbean, 19992000. San-
tiago.
. 2000b. Growth, Equity and Citizenship. Santiago.
. 2000c. Social Panorama of Latin America, 19992000. Santiago.
Ffrench-Davis, R. 2000. Reforming the Reforms in Latin America: Macroeconomics, Trade,
Finance. St Martins Press and Palgrave.
Ffrench-Davis, R., and H. Reisen, eds. 1998. Capital Flows and Investment Performance:
Lessons from Latin America. Paris: Organization for Economic Cooperation and Devel-
opment (OECD), Development Center.
Frankel, J. A. 1999. No Single Currency Regime Is Right for All Countries or at All Times.
Working Paper 7338. Cambridge, Mass.: National Bureau of Economic Research.
Furman, J., and J. Stiglitz. 1998. Economic Crises: Evidence and Insights from East Asia.
BPEA 2:1998.
Griffith-Jones, S. 1998. Global Capital Flows. Macmillan.
cionaiizariox oi iixaxciai voiariiir\ ,,
Gurra, J. A. 1995. Capital Flows: The Mexican Case. In Coping with Capital Surges, ed-
ited by R. Ffrench-Davis and S. Griffith-Jones. Lynne Rienner.
Jomo, K. S., ed. 1998. Tigers in Trouble: Financial Governance, Liberalisation and Crises in
East Asia. London: Zed Books.
IMF (International Monetary Fund). 1998. World Economic Outlook. Washington.
Krugman, P. 1999. Balance Sheets, the Transfer Problem, and Financial Crises. In Interna-
tional Finance and Financial Crises, edited by P. Izard, A. Razin, and A. Rose. Dordrecht:
Kluwer.
Larran, F., ed. 2000. Capital Flows, Capital Controls and Currency Crises. University of
Michigan Press.
Larran, G., H. Reisen, and J. von Maltzan. 2000. Emerging Market Risk and Sovereign
Credit Ratings. In Pensions, Savings and Capital Flows, edited by H. Reisen. Northamp-
ton, Mass.: Edward Elgar.
Le Fort, G., and S. Lehmann. 2000. El encaje, los flujos de capitales y el gasto: una eval-
uacin emprica. Working Paper 64. Santiago: Central Bank of Chile.
Lustig, N. 1997. The United States to the Rescue: Financial Assistance to Mexico in 1982
and 1995. CEPAL Review 61 (April). Santiago.
. 2000. Crises and the Poor: Socially Responsible Macroeconomics. Econo-
ma 1(1).
McKinnon, R. 1991. The Order of Economic Liberalization: Financial Control in the Tran-
sition to a Market Economy. Johns Hopkins University Press.
Ocampo, J. A. 1999. International Financial Reform: The Broad Agenda. CEPAL Re-
view 69. Santiago.
. 2001. Reforming the International Financial Architecture: Consensus and Di-
vergence. In New Roles and Functions of the UN and the Bretton Woods Institutions, ed-
ited by D. Nayyar. WIDER/Oxford University Press (forthcoming).
Perry, G., and D. Lederman. 1998. Financial Vulnerability, Spillover Effects and Conta-
gion: Lessons from the Asian Crises for Latin America. World Bank Latin American and
Caribbean Studies Viewpoints. Washington.
Radelet, S., and J. Sachs. 1998. The East Asian Financial Crisis: Diagnosis, Remedies,
Prospects. BPEA 1:1998.
Reisen, H. 1998. Domestic Causes of Currency Crises: Policy Lessons for Crisis Avoid-
ance. Technical Paper 136. Paris: Organization for Economic Cooperation and Devel-
opment (OECD), Development Center.
Rodrik, D. 1998. Who Needs Capital Account Convertibility? In Should the IMF Pur-
sue Capital Account Convertibility?, edited by P. Kenen. Princeton Essays in International
Finance 207. Princeton University, International Economics Section.
. 2001. Why Is There So Much Economic Insecurity in Latin America? CEPAL
Review 73 (April). Santiago.
Rodrik, D., and A. Velasco. 1999. Short-Term Capital Flows. Working Paper 7364.
Cambridge, Mass.: National Bureau of Economic Research.
Sachs, J., A. Tornell, and A. Velasco. 1996. Financial Crises in Emerging Markets: The
Lessons from 1995. BPEA 1:1996.
Schmidt-Hebbel, K., L. Servn, and A. Solimano. 1996. Savings and Investment: Para-
digms Puzzles, Policies. World Bank Research Observer 11(1).
iicaioo iiiixcu-oavis axo ;osi axroxio ocaxio ,o
Stiglitz, J. 1998. The Role of the Financial System in Development. Paper prepared for
the Fourth Annual Conference on Development in Latin America and the Caribbean,
San Salvador. World Bank (June).
. 2000. Capital Market Liberalization, Economic Growth and Instability. World
Development 28(6).
Turner, P. 1996. Comments on Reisen. In Promoting Savings in Latin America, edited by
R. Hausmann and H. Reisen. Inter-American Development Bank (IDB) and Organiza-
tion for Economic Cooperation and Development (OECD), Development Center.
. 2000. Procyclicality of Regulatory Ratios? In Global Finance at Risk: The Case
for International Regulation, edited by J. Eatwell and L. Taylor. New York: New Press.
United Nations. 1999. Towards a New International Financial Architecture. Report of the
Task Force of the Executive Committee on Economic and Social Affairs. Santiago: Economic
Commission for Latin America and the Caribbean (ECLAC).
Uthoff, A., and D. Titelman. 1998. The Relation between Foreign and National Savings
under Financial Liberalization. In Capital Flows and Investment Performance: Lessons
from Latin America, edited by R. Ffrench-Davis and H. Reisen. Paris: Organization for
Economic Cooperation and Development (OECD), Development Center.
Valds-Prieto, S., and M. Soto. 1998. New Selective Capital Controls in Chile: Are They
Effective? Emprica 25(2).
Wang, Y. 2000. Getting the Sequencing Right: Lessons from the Korean Experience with
Capital Market Liberalization. Seoul: Korea Institute for Economic Policy.
Williamson, J. 1993. A Cost-Benefit Analysis of Capital Account Liberalization. In Fi-
nancial Opening, edited by H. Reisen and B. Fischer. Paris: Organization for Economic
Cooperation and Development (OECD).
. 2000. Exchange Rate Regimes for Emerging Markets: Reviving the Intermedi-
ate Option. Policy Analyses in International Economics 60. Washington: Institute for
International Economics (IIE).
World Bank. 1998. Global Economic Prospects and the Developing Countries,199899.
Wyplosz, C. 1998. International Financial Instability. In The Policy Challenges of Global
Financial Integration, edited by J. J. Teunissen. The Hague: Forum on Debt and Devel-
opment (FONDAD).
cionaiizariox oi iixaxciai voiariiir\ ,;
,
2
K
orea and Taiwan fared very differently in the in-
ternational financial crisis that started in July 1997
with the crash of the Thai baht. While Taiwan came through the crisis rel-
atively unscathed, Korea experienced a major economic downturn bor-
dering on a depression, the countrys first since entering a period of fast
growth following the Korean War. Prior to the onset of the crisis, both of
these economies were viewed by commentators, international financial in-
stitutions, and academics as among the most successful developing coun-
tries in the postwar era. The crisis in Korea was certainly unexpected,
perhaps more so than in most other stricken East Asian countries. As late
as June 1997, the World Economic Forum had classified Korea as the fifth
most secure place to invest in the world.
1
The similarities between Taiwan and Korea prior to Koreas descent into
financial distress in November 1997 are striking. Both economies had man-
aged to sustain per capita gross domestic product (GDP) growth rates of
xaxuii i. acosix
*
Korea and Taiwan in
the Financial Crisis
*I wish to thank the many people in Korea and Taiwan who helped with research for the prepara-
tion of this paper, especially Inkyo Cheong, Yophy Huang, Shin-Yuan Lai, Kenneth Lin, Chao-Chen
Mai, Jae-Joon Park, Yung Chul Park, Julius Caesar Parreas, Shi Schive, James Tsuen-Hua Shih, Sang-
dal Shim, Lee-Rong Wang, Yunjong Wang, and Rong-I Wu. Thanks also to Ricardo Ffrench-Davis
and Jaime Ros for extensive comments on earlier drafts.
1. Dean (1998).
xoiia axo raiwax ix rui iixaxciai ciisis ,,
around 7 percent a year for over three decades.
2
Both started out at similarly
low levels of income per capita in the mid-1950s. Both based their almost
miraculous growth rates on export-oriented industrialization and the cre-
ation of new comparative advantages. Saving rates were high in both
economies and especially so in Korea, whose domestic saving-to-GDP ratio
was above 30 percent. Neither country relied on foreign capital: both had
repeatedly recorded current account surpluses or negligible deficits (in
Korea, up to 1995). Furthermore, both countries had a track record of ex-
cellent macroeconomic management, emphasizing prudent monetary poli-
cies, stable real exchange rates, and low inflation rates. Until the early 1990s,
both economies had extensive controls over the capital account.
3
In sum,
both Korea and Taiwan had achieved strong, sustained economic growth
through the harmonious blending of market forces and state guidance.
The two economies also demonstrate important differences, of course.
While Korean policy favored the formation of large conglomerates (the
chaebol), Taiwanese development relied on small- and medium-size com-
panies. Korean investment rates were consistently higher than Taiwanese
rates, but the growth of Korean output was concentrated in more capital-
intensive sectors than Taiwans.
What is surprising, then, is the sudden unraveling of the Korean mir-
acle. Korea was devastated by the financial crisis that engulfed the
economies of the clearly more troubled countries of Southeast Asia such
as Indonesia, Malaysia, and Thailand. In Taiwan, in contrast, the main
manifestations of the crisis were a marked slowdown in export volume
growth, a mild deceleration of the countrys fast growth rate, and a tem-
porary depreciation of the real exchange rate of about 20 percent.
This paper tries to explain the disparity in economic performance
since the onset of the financial crisis in mid-1997. It finds that the main
factor behind the Korean debacle was the liberalization of the capital ac-
count, without the concurrent adoption of adequate bank regulation and
prudential macroeconomic measures to discourage excessive capital inflow.
In Taiwan, the currency is still nonconvertible, and the monetary author-
ities continue to impose a variety of capital controls, which were very suc-
cessful in preventing a build-up of external debt or a large inflow of
2. See World Bank (1993).
3. There is extensive literature on the development of these economies and the causes of their suc-
cess. The two classic texts are Amsden (1989) and Wade (1990).
portfolio capital. Other factors include major differences in the structure
of the economy, industrial organization, and bank regulation.
Economic Performance up to 1997
From 1992 to 1997, both economies grew rapidly, Korea at an aver-
age annual rate of 6.6 percent and Taiwan at 6.8 percent. Both countries
recorded investment rates in line with their historical performance. The
growth rates of export volumes were considerably higher in Korea, but
Taiwans export growth figures were nonetheless quite respectable (see
table 2-1).
These economies were also similar in other respects. Inflation was low
in both countries. With regard to the public budget, Korea had a more fa-
vorable performance than Taiwan. Korea was able to maintain its public
finances practically in balance, while Taiwan ran deficits. The red ink in
Taiwan, however, was due exclusively to investment expenditures. Taiwan
also consistently ran public surpluses in current expenditure and income.
The deficit on the capital account was financed largely with recourse to
domestic borrowing.
Where these economies differ is in the balance of payments (see table
2-2). Korea tended to run current account deficits in the 1990s until the
onset of the crisis, while Taiwan had a current account surplus equivalent
to a couple of points of GDP. Up to 1996, however, the Korean deficit was
very moderate. In 1996, the worst year, the deficit was almost 5 percent
of GDP, declining to less than 2 percent in 1997.
The behavior of the real exchange rate also differed (see table 2-3).
Koreas real exchange rate appreciated consistently but moderately during
the period 199396, owing to large capital inflows. In contrast, Taiwans
real exchange rate was fairly steady in this period.
4
It would thus have been difficult to predict the depth of the crisis by
looking only at real exchange rate appreciation or the cumulative current
xaxuii i. acosix c
4. The degree of real effective overvaluation in both Taiwan and Korea was much greater than sug-
gested by the calculation in terms of the U.S. dollar. The 32 percent depreciation of the yen vis--vis
the U.S. dollar between the second quarter of 1995 and the first quarter of 1997 spelled overvalua-
tion for all the currencies in East Asia, which were either pegged or quasi-pegged to the U.S. dollar.
Taiwan and Korea both employed a quasi-peg. Prior to the crisis, Korea operated a crawling band ex-
change rate regime whose reference currency was the U.S. dollar, whereas Taiwan allowed the N.T.
dollar to fluctuate within a narrow band around a reference price for the U.S. dollar.
account deficit. However, Korea did experience large capital inflows over
a short period of time (199396). Net capital inflows were consistently
larger than the current account deficit during this period, and the Bank of
Korea accumulated considerable reserves (see table 2-4). By 1997 there
were signs of increasing financial fragility. Short-term debt was rising very
xoiia axo raiwax ix rui iixaxciai ciisis :
Table 2-1. Korea and Taiwan: Indicators of Economic Performance,
199299
Percent, except as indicated
Indicator 1992 1993 1994 1995 1996 1997 1998 1999
Growth rate of GDP
Korea 5.4 5.5 8.3 8.9 6.7 5.0 -6.7 10.7
Taiwan 7.5 7.0 7.1 6.4 6.1 6.7 4.6 5.7
Gross fixed investment
(% of GDP)
Korea 37.0 36.2 36.0 36.7 36.8 35.1 29.8 28.0
Taiwan 24.1 25.2 24.6 24.9 22.5 22.8 23.5 22.9
Export volume growth
Korea 8.3 6.7 14.9 23.9 19.8 24.9 16.9 9.2
Taiwan 5.3 7.2 5.5 12.8 7.1 8.7 2.4 9.6
Export price changes
Korea 4.8 3.2 1.8 0.9 9.7 -16.9 -2.2
Taiwan -5.4 5.2 0.6 6.9 1.7 2.1 5.6 1.1
Inflation (CPI)
Korea 6.2 4.8 6.2 4.5 4.9 4.4 7.5 0.9
Taiwan 4.4 3.0 4.1 3.6 3.1 0.9 1.7 0.2
Fiscal surplus
(% of GDP)
Korea -0.5 0.6 0.3 0.6 0.3 -1.5 -3.8 -4.6
Taiwan -5.4 -3.9 -1.7 -1.1 -1.3 -1.6 0.1 -1.3
Source: For Korea: Asian Development Bank, Asian Development Outlook, various issues;
International Monetary Fund (IMF), International Financial Statistics, Washington, various issues;
Korea Development Institute, Major Indicators of the Korean Economy, Seoul, various issues; The
European Union Chamber of Commerce in Korea, EU Chamber Monthly Bulletin, Seoul, various
issues; Sangdal Shim, Recent Trends and Macroeconomic Forecast for 1998/1999, Korea
Development Institute, Seoul, 15 December 1998, unpublished. For Taiwan: Chung-Hua
Institution for Economic Research, Major Economic Indicators for Taiwan, Taipei, February 1999;
Asian Development Bank, Asian Development Outlook, various issues; Central Bank of Taiwan,
Balance of Payments Quarterly, Taipei, various issues; Central Bank of Taiwan, Financial Statistics
Monthly, Taipei, various issues; Council for Economic Planning and Development, Taiwan
Statistical Data Book 1997, Taipei; Central Bank of Taiwan, Domestic Key Economic and Financial
Indicators, Taipei, 1 March 1999. Data for 1998 and 1999 for both economies were partly drawn
from Pacific Economic Cooperation Council, Pacific Economic Outlook, 200001, Asia Pacific Press
at the Australia National University, Canberra, 2000.
rapidly: at more than twice the value of reserves, it had reached a level that
made the economy very vulnerable to a change in investor and creditor
sentiment. As shown in table 2-5, up to 1996 the rise in the short-term
debt-to-reserves ratio was due not to a fall in reserves, which doubled be-
tween 1992 and 1996, but rather to the much faster increase in debt. In
the second half of 1997, as the economy began to look increasingly frag-
ile, capital outflows led to rapidly falling reserves.
5
The Impact of the Crisis on the Real Economy, 199798
Economic performance was very dissimilar in the aftermath of the cri-
sis. In Korea, the withdrawal of external capital that began in October
1997 caused a contraction in GDP of 6.7 percent in 1998. Unemploy-
ment, which had typically held at about 2 percent of the labor force,
jumped to 8.6 percent by February 1999 and remained at over 5 percent
at the end of that year.
6
Although inflation was originally forecast to rise
well above 10 percent, the slowdown in the economy kept the rate of in-
crease in consumer prices at 7.5 percent in 1998. A sharp correction in the
won (which had initially overshot in the direction of depreciation) inten-
sified the effects of the slow economy, resulting in an inflation rate of just
0.9 percent in 1999, far below its historical average (see table 2-1).
xaxuii i. acosix :
5. The first signs of trouble in the real economy appeared in early 1997 with the bankruptcy of
Hanbo Steel and Kia Motors. While growth was still fairly strong in 1997, it was due mostly to over-
expansion of exports by chaebol attempting to beat back the increased competitiveness of Japanese
firms, which had benefited from the sharp fall in the yen beginning in mid-1995.
6. PECC (2000, p. 31); see also Shim (1998).
Table 2-2. Korea and Taiwan: Current Account Balance, 199299
Current
account balance 1992 1993 1994 1995 1996 1997 1998 1999
Billions of dollars
Korea -3.9 1.0 -3.9 -8.5 -23.0 -8.2 40.4 24.5
Taiwan 8.6 7.0 6.5 5.5 10.0 7.1 3.4 5.9
Percent of GDP
Korea -1.3 0.3 -1.0 -1.7 -4.7 -1.7 12.7 6.0
Taiwan 4.0 3.1 2.7 2.1 3.9 2.4 1.3 2.0
Source: See table 2-1.

T
a
b
l
e

2
-
3
.
K
o
r
e
a

a
n
d

T
a
i
w
a
n
:

B
e
h
a
v
i
o
r

o
f

t
h
e

E
x
c
h
a
n
g
e

R
a
t
e

a
n
d

S
h
o
r
t
-
T
e
r
m

I
n
t
e
r
e
s
t

R
a
t
e
s
,

1
9
9
2

9
9
I
n
d
i
c
a
t
o
r
1
9
9
2
1
9
9
3
1
9
9
4
1
9
9
5
1
9
9
6
1
9
9
7
1
9
9
8
1
9
9
9
S
h
o
r
t
-
t
e
r
m

i
n
t
e
r
e
s
t

r
a
t
e
s

(
l
e
n
d
i
n
g
)
K
o
r
e
a
1
0
.
0
8
.
6
8
.
5
9
.
0
8
.
8
1
1
.
8
1
5
.
3
9
.
4
T
a
i
w
a
n
8
.
3
8
.
2
7
.
9
7
.
9
7
.
7
7
.
6
8
.
0
n
.
a
.
N
o
m
i
n
a
l

e
x
c
h
a
n
g
e

r
a
t
e
a
(
n
a
t
i
o
n
a
l

c
u
r
r
e
n
c
y

p
e
r

U
.
S
.

d
o
l
l
a
r
)
K
o
r
e
a
7
8
1
8
0
3
8
0
3
7
7
1
8
0
4
9
5
2
1
,
4
0
1
1
,
1
8
9
T
a
i
w
a
n
2
5
.
2
2
6
.
4
2
6
.
5
2
6
.
5
2
7
.
5
2
8
.
7
3
3
.
5
3
2
.
3
R
e
a
l

e
x
c
h
a
n
g
e

r
a
t
e

(
1
9
9
2

=
1
0
0
)
b
K
o
r
e
a
1
0
0
1
0
0
9
5
9
0
9
2
1
0
4
1
3
9
1
1
8
T
a
i
w
a
n
1
0
0
1
0
3
1
0
1
1
0
1
1
0
4
1
0
9
1
2
0
n
.
a
.
S
o
u
r
c
e
:

S
e
e

t
a
b
l
e

1
.
a
.

P
e
r
i
o
d

a
v
e
r
a
g
e
.
b
.

N
o
m
i
n
a
l

p
r
i
c
e

f
o
r

t
h
e

U
.
S
.

d
o
l
l
a
r
,

d
e
f
l
a
t
e
d

b
y

c
o
n
s
u
m
e
r

p
r
i
c
e

i
n
d
e
x
,

m
u
l
t
i
p
l
i
e
d

b
y

t
h
e

U
.
S
.

p
r
o
d
u
c
e
r

p
r
i
c
e

i
n
d
e
x
.
The Korean government was unable to control the outflow of capital,
and it called in the IMF in late November 1997. The IMF put together a
package of financial resources consisting of U.S.$21 billion to be disbursed
in eleven installments over a three-year period. In addition, it secured
commitments totaling U.S.$36 billion from bilateral and multilateral
sources.
7
Disbursements were subject to the usual IMF conditions of tight
money and the achievement of a fiscal surplus, despite the facts that the
fiscal deficit was small and the economy had entered a steep recession. Sur-
prisingly, it also required the Korean authorities to introduce the follow-
ing far-reaching structural reforms:
The complete opening of the capital account of the balance of pay-
ments, including the elimination of any remaining limits on mergers and
acquisitions (M&A);
The transformation of labor markets in the direction of Western in-
stitutional arrangements by abolishing lifetime employment practices and
obstacles to firing;
The restructuring of the chaebol to promote better corporate gov-
ernance, greater transparency, and an end to what was viewed as crony cap-
xaxuii i. acosix
7. For an analysis of the IMF program, see Y. C. Park (1998); Feldstein (1998); Radelet and Sachs
(1998); Wang and Zang (1999); Furman and Stiglitz (1998).
Table 2-4. Korea: Capital Inflow, Current Account Balance, and Change
in Reserves, 199299
Billions of U.S. dollars
Current account Change in Errors and
Year Capital inflow surplus reserves
a
omissions
1992 6.4 -3.9 -3.7 1.1
1993 2.7 1.0 -3.0 -0.7
1994 10.7 -3.9 -4.6 -1.8
1995 17.3 -8.5 -7.0 -1.2
1996 23.9 -23.0 -1.4 1.1
1997 -9.2 -8.2 23.0 -5.0
1998 -8.3 40.4 -25.9 -6.2
1999 12.7 24.5 -33.3 -3.5
Source: International Monetary Fund (IMF), International Financial Statistics, Washington,
several issues.
a. A negative sign indicates an increase.

T
a
b
l
e

2
-
5
.
K
o
r
e
a

a
n
d

T
a
i
w
a
n
:

I
n
d
i
c
a
t
o
r
s

o
f

F
i
n
a
n
c
i
a
l

F
r
a
g
i
l
i
t
y
,

1
9
9
2

9
8
B
i
l
l
i
o
n
s

o
f

U
.
S
.

d
o
l
l
a
r
s
,

e
x
c
e
p
t

a
s

i
n
d
i
c
a
t
e
d
I
n
d
i
c
a
t
o
r
1
9
9
2
1
9
9
3
1
9
9
4
1
9
9
5
1
9
9
6
1
9
9
7
1
9
9
8
I
n
t
e
r
n
a
t
i
o
n
a
l

r
e
s
e
r
v
e
s
a
K
o
r
e
a
1
7
.
2
2
0
.
4
2
5
.
8
3
2
.
8
3
4
.
2
2
0
.
5
5
2
.
1
T
a
i
w
a
n
8
2
.
3
8
3
.
6
9
2
.
5
9
0
.
3
8
8
.
0
8
3
.
5
9
0
.
3
F
o
r
e
i
g
n

d
e
b
t
a
K
o
r
e
a
4
4
.
1
4
7
.
2
7
2
.
4
8
5
.
8
1
5
5
.
8
1
3
7
.
0
1
3
9
.
1
T
a
i
w
a
n
1
9
.
9
2
3
.
2
2
6
.
2
2
7
.
1
2
7
.
5
3
8
.
5
3
0
.
0
S
h
o
r
t
-
t
e
r
m

d
e
b
t
a
K
o
r
e
a
1
1
.
9
1
2
.
2
3
1
.
6
4
6
.
6
6
6
.
6
5
3
.
8
2
8
.
1
T
a
i
w
a
n
1
7
.
3
1
9
.
4
2
0
.
0
1
9
.
5
1
8
.
8
2
2
.
2
1
9
.
4
R
a
t
i
o

o
f

s
h
o
r
t
-
t
e
r
m

d
e
b
t

t
o

r
e
s
e
r
v
e
s

(
%
)
K
o
r
e
a
6
9
.
2
5
9
.
9
1
2
2
.
7
1
4
2
.
1
1
9
4
.
9
2
6
8
.
8
5
4
.
0
T
a
i
w
a
n
2
1
.
0
2
3
.
3
2
1
.
6
2
1
.
6
2
1
.
3
2
6
.
5
2
1
.
4
S
o
u
r
c
e
:

F
o
r

K
o
r
e
a
,

W
o
r
l
d

B
a
n
k
,

G
l
o
b
a
l

D
e
v
e
l
o
p
m
e
n
t

F
i
n
a
n
c
e
,

W
a
s
h
i
n
g
t
o
n
,

2
0
0
0
;

f
o
r

T
a
i
w
a
n
,

A
s
i
a
n

D
e
v
e
l
o
p
m
e
n
t

B
a
n
k
,

A
s
i
a
n

D
e
v
e
l
o
p
m
e
n
t

O
u
t
l
o
o
k
,
v
a
r
i
o
u
s
i
s
s
u
e
s
.
a
.

E
n
d

o
f

p
e
r
i
o
d
.
xaxuii i. acosix o
italism (basically, opaque relations between different affiliates and between
the conglomerates and the government); and
The closure of troubled banks, the promotion of bank mergers, the
opening of the banking sector to purchases by foreign entities, and an in-
crease in capital requirements and bank reserves to reflect Basel Commit-
tee standards.
The IMF adjustment program was bound to make matters worse in the
short term. Higher interest rates were not able to contain the run on the
won, which traded as high as W 1,960 to the dollar in December 1997, up
from about W 830 at the end of 1996. If investors perceive a sharp increase
in interest rates as reducing the likelihood of loan repayment, the increase
will not only fail to shore up the value of the currency but could wind up
fuelling further depreciation.
8
The combination of high interest rates and
sharp depreciation caused widespread bankruptcies, as firms were both
highly leveraged and heavily indebted in foreign currency. Company bank-
ruptcies had extremely adverse effects on banks, many of which had already
been weakened by currency and maturity mismatches between assets
(which were mostly domestic currency denominated and long term) and li-
abilities (which were mostly foreign currency denominated and short
term). Many banks failed and had to be rescued or restructured by the Bank
of Korea.
As shown in figure 2-1, the sharp rise in both interest rates and the nom-
inal exchange rate eventually subsided as a result of the evolving recession.
The Korean adjustment was so severe that the current account swung from
a deficit of U.S.$8.2 billion in 1997 (less than 2 percent of GDP) to a sur-
plus of U.S.$40 billion in 1998 (12.7 percent of GDP). This was achieved
basically through a fall in import volumes, since export earnings were almost
flat. The IMF itself recognized that it had erred in imposing sharp fiscal re-
straints, and it relaxed its fiscal targets at the beginning of 1998.
Several authors criticize the conditions related to structural reform
that the IMF imposed in Korea. Radelet and Sachs rightly point out that
it is difficult and counterproductive to initiate reforms in the midst of a
crisis.
9
Whether the reforms are even necessary is debatable. I argue below
that the liberalization of the capital account without taking any precau-
tion for the potential adverse effects of capital surges was one of the causes
of the Korean crisis. This implies that further liberalization is unnecessary
and probably counterproductive. Other reforms were ill-timed. Requiring
8. See Furman and Stiglitz (1998) for a full development of this argument.
9. Radelet and Sachs (1998).
xoiia axo raiwax ix rui iixaxciai ciisis ;
Figure 2-1. Republic of Korea: Exchange Rate and Short-Term Interest
Rates, 199798
Source: IMF, International Financial Statistics.
18
16
14
12
10
8
1,600
1,400
1,200
1,000
800
Won per U.S. dollar Percent
March
1997
June
1997
Sept.
1997
Dec.
1997
March
1998
June
1998
Sept.
1998
Dec.
1998
Short-term interest rates
(right axis)
Exchange rate
(left axis)
banks to increase their capital-asset and reserve-asset ratios during a bank-
ing crisis increased the intensity of the credit crunch and led to additional
bankruptcies.
Feldstein argues that the IMF program for Korea constituted an un-
warranted intrusion into issues that are far removed from the countrys bal-
ance-of-payments crisis.
10
He holds that the changes sought in the IMF
program, which should be left to national political processes and not im-
posed from outside under duress in times of crisis, were derived from long-
standing demands on Korea to open its markets to Japanese goods and U.S.
capital.
The impact of the crisis was much milder in Taiwan. There was a sharp
deceleration in export growth and a modest decline in GDP growth, since
domestic demand compensated the slack. As the crisis in Southeast Asia
unfolded, the New Taiwan (N.T.) dollar came under strong attack. The ini-
tial response of the Central Bank was to defend the N.T. dollar and to ac-
cept a reduction in its comfortable level of foreign exchange reserves. When
foreign reserves began to decline, however, it switched policies and stopped
defending the currency in October 1997. The N.T. dollar quickly lost
10. Feldstein (1998).
xaxuii i. acosix
Figure 2-2. Taiwan: Exchange Rate and Interest Rates, 199798
Source: Central, Bank of China, Domestic Key Economic and Financial Indicators, March 1999; Chung-Hua
Institution for Economic Research, Major Economic Indicators for Taiwan, February 1999.
8.0
7.5
7.0
6.5
6.0
5.5
5.0
4.5
34
32
30
28
26
24
22
N.T. dollars per U.S. dollar Percent
March
1997
June
1997
Sept.
1997
Dec.
1997
March
1998
June
1998
Sept.
1998
Dec.
1998
Interbank
money market rate
(right axis)
Exchange rate
(left axis)
about 20 percent of its value in relation to the U.S. dollar (from N.T.$28 to
N.T.$34 to the U.S. dollar). In late 1998, the N.T. dollar began to appreci-
ate again.
11
At the same time, nominal interest rates began an upward trend
in mid-1997 but fell again in the first quarter of 1998. By December 1998
they were lower than before the crisis broke (figure 2-2).
Thus the growth rate of the Taiwanese economy dropped temporarily
by about 2 percentage points between 1997 and 1998, mainly as a result
of the deceleration of export growth. Its unemployment rate remained
largely unchanged at about 2.7 percent. Consumer price inflation contin-
ued to fall toward a zero rate. Taiwan was poised to resume rapid growth
as demand for its exports recovered in 1999. In fact, Taiwanese firms may
have strengthened their competitiveness by pursuing a policy of moving
the export-oriented assembly of light manufactures offshore to locations
in Southeast Asia, since the effects of the financial crisis have made those
countries very attractive sites for export processing.
12
11. In November 2000, the exchange rate stood at N.T.$31 to the U.S. dollar.
12. Huang (1998).
xoiia axo raiwax ix rui iixaxciai ciisis ,
The Role of Economic Structure
Why Taiwan was able to stave off the crisis should be clear from the
above analysis. What is much less obvious is why Korea was hit so badly.
Korea was not a mismanaged economy, and it had high investment rates
financed basically by domestic saving. A serious problem became apparent
when the crisis broke, however. Korean firms had developed excess capac-
ity, partly as a result of the depreciation of the yen beginning in the third
quarter of 1995.
13
Rather than pulling back, Korean manufacturers re-
sponded with even larger increases in production and exports, which
caused a sharp worsening of export prices in 1996.
14
This was particularly
the case in the semiconductor industry, where the Korean response to
gains in the competitiveness of Japanese firms contributed to the develop-
ment of considerable excess capacity globally.
Government action to coordinate decisionmaking by individual firms
had been on the wane for some time, as the government gradually handed
over to the market most resource allocation decisions. Excess capacity
problems had previously been dealt with through rationalization efforts
led by planning agencies, but by the mid-1990s that restraint on the chae-
bol dominating Korean industry was simply no longer there.
15
These
moves toward the liberalization of economic policy and the dismantling of
planning instruments received impetus from international agreements on
subjects such as export subsidies and financial services, under the aegis of
the World Trade Organization (WTO). The chaebol were basically left to
their own devices to pursue whatever course of action they perceived to be
in their own interests.
16
In Taiwan, increases in export volume were much more modest prior
to the crisis, and average export prices continued to rise. Like Korea, Tai-
wan exports semiconductors, but firms were able to weather the global
semiconductor slump of 1996 by reducing costs and switching their prod-
uct mix away from those in excess supply. Strategic alliances between Tai-
wanese producers and leading firms in the United States, Japan, and
Europe contributed to avoiding excess capacity.
17
Another important characteristic of the chaebol is that they were
13. Of all the East Asian economies, Korea is the most competitive with the Japanese.
14. Y. C. Park (1998).
15. Chang (1998).
16. Y. C. Park (1998).
17. Yanagihara (1999).
xaxuii i. acosix ,c
highly leveraged. In 1996, the thirty largest chaebol had debt-equity ratios
of 380 percent, compared to less than 100 percent for Taiwanese firms.
18
High leverage is good when the economy and profits are expanding briskly
and when credits are renewed. In such situations, high leverage actually in-
creases a firms profitability, since interest rates are normally much lower
than the rates of return on investment. What is a blessing in an expand-
ing economy, however, becomes a curse in a shrinking one. When credits
are not renewed or when interest rates rise sharply, companies go bankrupt
more rapidly if they are highly leveraged than if their debt-capital ratios
are more reasonable. Since Korean companies had never experienced a
downdraft like the one that hit them in late 1997, their high debt-equity
ratios seemed rational to them.
By contrast, Taiwanese companies are much smaller than their Korean
counterparts, and they are consequently highly risk averse and extremely
conservative in taking on debt. Investments are usually financed with re-
tained profits and capital drawn from the extended family rather than
from banks. This explains why the average debt-equity ratio in Taiwan is
much lower than in Korea.
The Structure and Regulation of the Banking Industry
Many Korean conglomerates encompass a finance company, and they
obtain funding from financial markets through these financial affiliates.
Others have extensive dealings with a single bank, much in the manner
of the relationship between the Japanese keiretsu and their main bank. In
the years leading up to the crisis, the finance companies had been allowed
to become full-fledged merchant banks, which improved the Korean con-
glomerates access to international debt markets. Since merchant banks do
not accept deposits from the public, it was generally believed that they
did not need regulation. Commercial banks were, on paper, more tightly
regulated. In practice, however, regulators turned a blind eye to many of
their activities, in the mistaken belief that regulations were somehow con-
trary to the spirit of liberalization, particularly in view of the countrys ac-
cession to the Organization for Economic Cooperation and Development
(OECD).
19
18. Y. C. Park (1998, p. 32).
19. J.-J. Park (1998) and personal communication.
xoiia axo raiwax ix rui iixaxciai ciisis ,:
Korean banks were accustomed to strong government guidance. Until
the late 1980s, a large proportion of their assets was made up of so-called
policy loans, which were loans made at preferential interest rates to firms in
sectors that were favored by government policies for selective intervention.
In the mid-1980s, however, the authorities began implementing a policy of
broad liberalization of financial markets, and policy loanswhich had rep-
resented 60 percent of bank lending in the late-1970swere gradually
phased out.
20
Until the mid-1980s, then, commercial banks were either state-owned
or under strong guidance by the planning authorities. Domestic banks
were therefore unaccustomed to operating in a market environment,
despite the recent liberalization. They had not yet developed expertise in
credit analysis, risk management, or due diligence. Their ability to deal
with the risks of foreign financial transactions was particularly inade-
quate.
21
Banking authorities performed poorly in the area of regulating
and monitoring foreign financial activity. Prior to 1985, the economic au-
thorities practiced extensive control over all items in the capital account,
such that banking supervision and regulation in this area were redundant.
This may explain why Korea had not developed a strong regulatory frame-
work for financial institutions.
Taiwanese banks are very prudent in their lending policies, and they
are also closely regulated by the Central Bank. Taiwanese banks were tra-
ditionally state-owned. The establishment of private banks was very grad-
ual: even as late as 1990, twelve of the twenty-four banks in operation
were state-owned. These banks accounted for 75 percent of all bank
branches, 86 percent of all deposits, and 87 percent of all loans.
22
In
1991, rules governing the establishment of private banks were eased, and
sixteen new private banks were opened. In addition, the Central Bank al-
lowed the licensing of new foreign banks. Both public and private banks
in Taiwan have long been subject to stringent prudential regulations.
23
In September 1998, for example, the capital adequacy ratio of Taiwanese
banks was 11.4 percentquite a bit above the Basel Committee norm
of 8 percent. Taiwanese banks also face a number of restrictions on open
20. The liberalization of financial intermediation through formal banks caused the curb market to
shrink considerably (Noland, 1996).
21. Y. C. Park (1998).
22. APEC (1998, chap. 9).
23. Yang (1998).
positions in foreign exchange, which must be approved by the Central
Bank.
24
These differences in banking structure and regulation show up in the
behavior of nonperforming loans. At the outbreak of the crisis in 1997,
nonperforming loans in Taiwan were only 4 percent of all loans, whereas
they were 14 percent in Korea. This ratio inevitably increased in both
countries as a consequence of the crisis, but in Taiwan they remained at a
manageable level (5.5 percent in 2000). In Korea, in contrast, they came
close to crippling the banking sector, peaking in 1998 at 25 percent of all
loans (or 34 percent of GDP) and then gradually falling back to an esti-
mated 18 percent in 2000.
25
This forced the Bank of Korea to intervene
by liquidating banks, merging weak banks with stronger ones, injecting
more capital into banks deemed to be salvageable, and selling others off to
foreign banks. Nonperforming loans are still a major stumbling block in
the ability of the banking sector to grant fresh credit to the private sector.
Policies on Capital Inflows
In globalized financial markets, it is difficult for a country to run into
an international financial crisis without first having experienced large cap-
ital inflows. A country can have a balance-of-payments crisis simply be-
cause of bad domestic policies, of course, but the IMF is prepared to deal
with this type of crisis by means of conditional lending. Another possible
source of crisis in the absence of large capital inflows might be a currency
crisis produced entirely by contagion. If a crisis elsewhere induces fears of
a possible depreciation of the currency and if international financial trans-
actions are reasonably free, then national and foreign financial investors
could cause a crisis by selling the domestic currency and buying foreign
exchange. Such contagion was a factor explaining why countries that had
been quite well managed experienced difficult times during the recent cri-
sis regardless of past policy. Examples include Chile and Hong Kong.
26
Despite the enormous differences between these two economies, they both
were pursuing rather prudent policies before the financial crisis. Nonethe-
less, they both suffered significant contagion.
xaxuii i. acosix ,:
24. Personal communication with Central Bank officials.
25. Data on nonperforming loans are taken from Goldman Sachs, March 1998; J.P. Morgan, April
1998; Asiaweek, The Asiaweek Financial 500, September 19, 1997, and September 11, 1998.
26. On Chile, see Ffrench-Davis and Tapia (in this volume).
However, the new breed of crisis associated with the globalization of
finance is generally preceded by a strong upsurge in capital inflow. This
was the case in Mexico prior to 1994.
27
It was universally the case in the
countries that were most badly hit by the recent financial crisis: Brazil, In-
donesia, Korea, Malaysia, Philippines, Russia, and Thailand. Large capital
inflows can occur for a variety of reasons, including the perception by for-
eign financial investors that a country is doing things well or simply that
its currency will appreciate because they believe others think it will.
28
Since
interest rates are normally higher in emerging markets than in mature
ones, expectations of currency appreciation can produce very large inflows.
Such inflows are often not marginal for an individual recipient. They
can therefore have important negative externalities: they can generate cur-
rent account deficits, appreciate the exchange rate, and unleash asset price
bubbles. They also often increase the level of short-term debt relative to
international reserves. At the same time, the short-term nature of the in-
flows makes it easy for investors to flee and for creditors to not renew their
lending when they sense trouble.
29
Consequently, any worsening of fun-
damentals that generates perceptions of vulnerability will cause capital in-
flows to slow down, usually quite abruptly, and then to reverse direction.
When the country has lost sufficient reserves, foreign and domestic finan-
cial investors discover exchange rate risk. As the fear of depreciation gath-
ers momentum, everybody seeks to liquidate positions in domestic
currency, accelerating the loss of reserves and precipitating a full-blown
balance-of-payments crisis.
These stylized facts fit the case of Korea fairly well. Korean banks in-
termediated huge capital inflows and financial resources in the 1990s. Even
so, the current account was not out of control despite being in deficit (it
had been in surplus in the second half of the 1980s). Although the real ex-
change rate appreciated as a result of capital inflow, the degree of apprecia-
tion was not excessive by recent standards. Financial fragility rose to very
dangerous levels, however: the ratio of short-term debt to international re-
serves exceeded 200 percent in 1997, making the economy very vulnerable
to capital flight by scared investors and creditors (see table 2-5).
xoiia axo raiwax ix rui iixaxciai ciisis ,,
27. See Ros (in this volume).
28. This is a peculiar international version of Keyness beauty contests (see Eatwell, 1997, p. 243).
Some investors are more concerned with what other investors are doing than with the underlying
worth of assets. When these participants predominate over the so-called fundamentalists, financial
markets can be very volatile indeed.
29. Ffrench-Davis and Ocampo (in this volume).
Korea introduced a broad range of measures in 199193 to liberalize
the capital account. In 1991, all remaining restrictions on greenfield for-
eign direct investment (FDI) were scrapped, and resident companies and
banks were allowed to issue securities abroad. Beginning in January 1992,
foreigners were allowed to purchase stocks in individual Korean compa-
nies, with the limitation that a maximum of 10 percent of a companys eq-
uity could be owned by foreigners as a group.
30
Both the aggregate and
individual investors ceilings were gradually enlarged thereafter.
Short-term flows, however, were liberalized ahead of most long-term
flows. Ceilings on loans in foreign exchange to domestic firms were lifted
in 1994, but ceilings on banks long- and medium-term borrowing from
international financial markets were not. Only trade credits and other
short-term borrowing abroad were liberalized. This forced banks to raise
short-term credit to finance long-term loans to the chaebol.
31
Other types
of long-term capital movements were also excluded from the trend toward
liberalizing the capital account. These included purchases of domestic
bonds by foreigners, as well as a virtual ban on M&A owing to restrictions
on foreign ownership of existing Korean companies. The crisis and the
conditions attached to the IMF program eliminated most of the remain-
ing restrictions. By the end of 1997, the aggregate ceiling for foreign port-
folio investment had been raised to 55 percent of a companys equity. All
restrictions on M&A were eliminated in May 1998.
32
The short end of the financial market was thus opened up ahead of
bond markets, long-term bank finance, and the market for corporate con-
trol. The authorities seemed to hold the erroneous view that short-term
finance (most of which was assumed to be trade related) is self-liquidating
and, as such, unlikely to generate overindebtedness. However, Korean
banks and firms were borrowing large sums of short-term funds to finance
long-term investments. Even when the authorities became aware of the
deteriorating quality of bank assets and the increasing mismatches in the
term structure of assets and liabilities, they failed to intervene, perhaps be-
cause they lacked experience and resources, but also, to a significant degree,
because they were influenced by the atmosphere of euphoria predominat-
ing in international financial markets.
33
xaxuii i. acosix ,
30. Park (1996). Until that date, foreigners were allowed to invest in Koreas stock market only
through authorized open-end mutual funds.
31. Y. C. Park (1998).
32. Wang and Zang (1999, chap. 9).
33. Y. C. Park (1998).
The flows that followed liberalization in the early to mid-1990s were
enormous (see table 2-6). Curiously, the big increases were not in FDI
(perhaps because of restrictions on M&A), but in portfolio transactions
and bank lending. Purchases of shares on the Seoul market jumped from
U.S.$2.5 billion in 1992 to U.S.$6.6 billion in 1993. Since Korean com-
panies are among the worlds most competitive, they were undoubtedly
very attractive to foreign portfolio investors. Thus much of the equity
portfolio inflows that took place after 1992 was a stock adjustment phe-
nomenon that should have been fully anticipated by the authorities. Un-
fortunately, it was not. This lack of foresight fueled the capital surge.
Debt-creating portfolio inflows rose spectacularly, from U.S.$2.5 bil-
lion in 1992 to U.S.$15.2 billion in 1996. These were mostly securitized
forms of bank lending. International banks were acting as middlemen in
the placement of securities by Korean banks (and also by banks from
other East Asian countries) in international financial markets.
34
Private
credit to both banks and nonfinancial firms also rose dramatically. At the
same time, Korea experienced large outflows in the form of debt-creating
portfolio flows and lending by banks and nonbanks. This is an indication
that Korean banks and firms were engaging in interest rate arbitrage be-
tween international capital markets, on the one hand, and Korean or
other markets, on the other. As shown above in table 2-4, net capital in-
flows consistently exceeded the current account deficit plus reserve accu-
mulation between 1993 and 1996. Large negative errors and omissions
demonstrate that Korean banks and chaebol were engaging in financial in-
vestments in other countries, such as Indonesia and Thailand. There are
even some reports that Korean banks were purchasing Russian govern-
ment bonds.
Financial liberalization, then, led to large capital inflows and outflows.
Korean banks borrowed short abroad in order to lend long to the chaebol,
which borrowed to finance aggressive investment plans. In some instances,
the borrowers on international markets were the chaebol themselves or
their foreign subsidiaries. This mismatch between the maturity structure
of assets and liabilities was accompanied by an equally dangerous currency
mismatch: Korean banks took on debt in foreign exchange and lent in won
(and also in baht, Indonesian rupiahs, and other currencies). Banks be-
came totally illiquid during the crisis when foreign creditors refused to roll
over short-term loans. The first to flee were the foreign equity investors,
xoiia axo raiwax ix rui iixaxciai ciisis ,,
34. For a description, see Kregel (1998a, 1998b); Taylor (1999).

T
a
b
l
e

2
-
6
.
K
o
r
e
a
:

C
a
p
i
t
a
l

i
n
f
l
o
w
s
,

1
9
9
2

9
9
B
i
l
l
i
o
n
s

o
f

U
.
S
.

d
o
l
l
a
r
s
I
t
e
m
1
9
9
2
1
9
9
3
1
9
9
4
1
9
9
5
1
9
9
6
1
9
9
7
1
9
9
8
1
9
9
9
b
F
o
r
e
i
g
n

d
i
r
e
c
t

i
n
v
e
s
t
m
e
n
t
I
n
f
l
o
w
s

0
.
7
0
.
6
0
.
8
1
.
8
2
.
3
2
.
8
5
.
4
9
.
3
O
u
t
f
l
o
w
s
-
1
.
2
-
1
.
3
-
2
.
5
-
3
.
5
-
4
.
7
-
4
.
4
-
4
.
7
-
4
.
2
P
o
r
t
f
o
l
i
o

I
n
f
l
o
w
s
:

e
q
u
i
t
y
2
.
5
6
.
6
3
.
6
4
.
2
6
.
0
2
.
5
3
.
9
1
2
.
1
I
n
f
l
o
w
s
:

d
e
b
t
2
.
5
3
.
9
4
.
5
9
.
7
1
5
.
2
9
.
8
-
4
.
1
-
5
.
1
O
u
t
f
l
o
w
s
:

e
q
u
i
t
y

-
0
.
2
-
0
.
4
-
0
.
2
-
0
.
7
-
0
.
3
0
.
0
-
0
.
3
O
u
t
f
l
o
w
s
:

d
e
b
t
0
.
8
-
0
.
3
-
1
.
6
-
2
.
0
-
5
.
3
2
.
3
-
1
.
6
2
.
0
P
r
i
v
a
t
e

c
r
e
d
i
t
s

t
o

K
o
r
e
a
B
a
n
k
s
1
.
8
0
.
7
7
.
4
1
1
.
4
1
0
.
0
-
9
.
8
-
6
.
2
1
.
4
O
t
h
e
r
3
.
8
-
0
.
3
6
.
6
1
0
.
7
1
5
.
1
-
3
.
1
-
1
2
.
3
-
3
.
4
P
r
i
v
a
t
e

c
r
e
d
i
t
s

f
r
o
m

K
o
r
e
a
B
a
n
k
s

-
3
.
2
-
4
.
0
-
5
.
1
-
9
.
2
-
8
.
2
-
8
.
3
7
.
0
-
0
.
2
O
t
h
e
r
0
.
2
0
.
1
-
1
.
9
-
4
.
6
-
4
.
7
-
5
.
0
-
0
.
2
-
2
.
1
N
e
t

p
u
b
l
i
c

t
r
a
n
s
a
c
t
i
o
n
s
-
0
.
9
-
2
.
6
-
5
.
6
-
1
.
0
-
1
.
0
4
.
5
4
.
6
3
.
1
T
o
t
a
l
a
7
.
0
3
.
2
1
0
.
7
1
7
.
3
2
3
.
9
-
9
.
2
-
8
.
4
1
2
.
7
S
o
u
r
c
e
:

I
n
t
e
r
n
a
t
i
o
n
a
l

M
o
n
e
t
a
r
y

F
u
n
d

(
I
M
F
)
,

I
n
t
e
r
n
a
t
i
o
n
a
l

F
i
n
a
n
c
i
a
l

S
t
a
t
i
s
t
i
c
s
,
W
a
s
h
i
n
g
t
o
n
,

N
o
v
e
m
b
e
r

2
0
0
0
.
a
.

E
x
c
l
u
d
e
s

c
a
p
i
t
a
l

t
r
a
n
s
f
e
r
s
.
b
.

P
r
o
v
i
s
i
o
n
a
l

f
i
g
u
r
e
s
.
who dumped Korean stock even when doing so meant taking a big hit. As
the stock market fell, so did the international value of the won. The ex-
change rate depreciation, in turn, caused many banks and firms to become
technically bankrupt, since their assets were insufficient to cover their lia-
bilities.
The case of Taiwan is quite different. Taiwan did not experience large
capital inflows, not because foreign investors and lenders were not inter-
ested in Taiwan, but because the authorities did not allow it to happen.
Although Taiwan underwent a process of domestic financial liberalization
and limited opening of the capital account in the 1990s, that process was
extremely cautious. Capital inflows are still strictly regulated. Given Tai-
wans long-standing current account surplus, the authorities gave priority
to liberalizing capital outflow.
Taiwan limits the value of shares that can be held by foreigners: any
single foreigner may hold up to 15 percent of a listed companys equity,
and all foreigners together may hold up to 30 percent. Acquisitions of Tai-
wanese companies by foreigners are forbidden.
35
On the other hand, there
are no restrictions on greenfield FDI investments, profit remittances, or
the repatriation of capital. Foreign institutional investors have a ceiling of
U.S.$600 million on the investments they can make in Taiwan. Natural
foreign persons can purchase up to U.S.$5 million in Taiwanese assets, and
firms that are not listed institutional investors face a ceiling of U.S.$50
million. As already noted, Taiwanese banks are under the strict guidance
and supervision of the Central Bank as regards their taking on foreign li-
abilities for conversion into N.T. dollardenominated assets. The use of
N.T. dollar derivatives is authorized by the Central Bank only for genuine
hedging purposes. For example, cross-currency swaps between the N.T.
dollar and the U.S. dollar may be used to cover bank loan obligations or
interest payments from one currency to another in order to hedge currency
and interest rate risks.
With regard to capital outflows, the use of foreign exchange for over-
seas investment is allowed, but such investments must be approved by the
Ministry of Economic Affairs and the Central Bank. Natural persons may
invest up to U.S.$5 million abroad without permission. The limit for a
domestic firm is U.S.$50 million.
It might be objected that the current account surplus is evidence that
greater capital account liberalization would not have led to stronger capi-
xoiia axo raiwax ix rui iixaxciai ciisis ,;
35. Judging from press reports, this norm appears to have been liberalized in 1999.
xaxuii i. acosix ,
tal inflows. Stronger capital inflows are certainly compatible with a surplus
on current account. Had they taken place, foreign investments in the Tai-
wanese economy would have been even greater and the outflows of capi-
tal from Taiwan even larger. Larger capital inflows, however, would have
made Taiwan more vulnerable to the type of events that took place in
Korea. Had Taiwan been more open to capital inflows, the effects of the
crisis undoubtedly would have been far worse. Taiwan is one of the most
successful economies in the world. It has strong companies that might be
of interest to many transnational corporations. Its stock market is un-
doubtedly attractive for institutional investors and individuals looking for
plays in emerging markets. Many Taiwanese banks might have borrowed
more had the monetary authorities allowed them to do so.
36
They might
have used the good credit rating of Taiwan to borrow and lend in other
countries, much as the Korean banks did.
The fact is that none of the damaging sequence of events that took
place in Korea happened in Taiwan. The main cause for Taiwans relative
success is clearly its much more cautious policy on the kinds of capital in-
flow that are likely to lead to those situations. At the time of the crisis, Tai-
wans external debt was notably low at about U.S.$25 billion, a figure that
is insignificant in comparison to its international reserves (see table 2-5).
Most of its debt was in the form of trade credit. In other words, Taiwan
had low levels of external indebtedness because it was cautious about bor-
rowing excessively. Equity markets had little foreign investment, so port-
folio outflows were commensurately moderate after the crisis struck.
Korea, on the other hand, found itself in a situation of extreme fi-
nancial fragility shortly after liberalizing short-term capital inflows. By
1996, just three years after the implementation of the most ambitious
measures, the ratio of short-term debt to reserves had reached a level that
was not macroeconomically sustainable, which should have alerted the au-
thorities. The increase in the ratio was not the result of a fall in reserves,
which had in fact risen; it was due entirely to a very rapid accumulation
of short-term debt. Large inflows of portfolio capital also meant that the
country had made itself very vulnerable to a change in sentiment. And that
is what happened: before the crisis became full-blown, most of the foreign
investors in Korean stock had already exited.
36. Taiwanese interest rates are lower than those in Korea, however, so the possibilities for arbitrage
are more limited.
xoiia axo raiwax ix rui iixaxciai ciisis ,,
The Role of Contagion
Contagion clearly had a leading role in the Korean debacle. Would the
crisis have occurred in the absence of contagion? Probably not. If measured
on the basis of pre-crisis levels for the exchange rate and interest rates,
major Korean firms and banks were not insolvent. Banks were certainly
vulnerable to a liquidity squeeze, and the chaebol were vulnerable to a
slowdown in world demand for their products, but both banks and firms
would have been able to resolve their problems if foreign investors had
been willing to keep their money in Korea and creditors had rolled over
expiring loans, even at higher interest rates. The proper role of the IMF
should have been to coordinate expectations around the fact that Korean
institutions were solvent. By insisting that major structural changes were
required, it reinforced the view of market participants that something was
fundamentally wrong with the economy, its firms, and its banks.
Korea clearly had better fundamentals than, say, Thailand or Malaysia.
As noted, Koreas current account deficit was significant only in 1996 (4.4
percent of GDP). For the period 199296, the average Korean deficit was
a mere 1.6 percent. And even in 1996, Koreas deficit was well below Thai-
lands (7.9 percent of GDP). Thailand ran an average current account
deficit of 6.5 percent of GDP for the period 199296. The average deficit
for Malaysia in the same period was 5.8 percent of GDP, with a peak of
9.7 percent in 1995.
37
International financial markets worried about Korea only after the
onset of the financial crisis elsewhere. Before the crisis struck other East
Asian countries, international portfolio investors and international banks
were pouring finance into the Korean economy. After the Thai, Malaysian,
and Indonesian crises, they reassessed their commitments to Korea and
pulled out en masse, with all the adverse consequences described above.
The almost unrestricted opening of the capital account (following over
three decades of universally acclaimed economic success) made the coun-
try highly vulnerable to just such a sequence of events.
Korea experienced several forms of contagion, starting with the effect
on trade. Generally speaking, when a crisis affects real economies, other
countries experience a drop in export volumes or export prices or both. In
the Korean case, export prices had worsened before the onset of the crisis
37. Data drawn from International Monetary Fund, International Financial Statistics, various issues.
as a consequence of the insufficient demand vis--vis a huge increase in
supply, partly because the chaebols insisted on going ahead with their in-
vestment plans regardless of world market conditions.
The second important form of contagion was financial contagion. A
financial crisis in one country can spread to other economies if the ensu-
ing panic causes investors to flee and creditors to refuse to roll over credit.
This was the major form of contagion in the Korean case, and the one that
produced the most damage to the domestic economy.
The third common manifestation of contagion is in the exchange rate,
which occurs when foreign and domestic agents dump the domestic cur-
rency and purchase foreign exchange in response to a financial crisis else-
where. Korea experienced this type of contagion, but it was relatively
short-lived. The foreign exchange market witnessed enormous overshoot-
ing. The permanent effect was a depreciation of the won from about W
888 to the U.S. dollar before the crisis (at the end of the second quarter
of 1997) to about W 1,100 in October 2000. This represents a permanent
nominal depreciation of about 25 percent. In real terms, the depreciation
was less than 20 percent. This, of course, reflects the inefficiency of fi-
nancial markets. The Korean won was overvalued in mid-1997, but it did
not need to go through the wild gyrations that it experienced between the
fourth quarter of 1997 and the end of 1998.
Taiwan similarly experienced strong trade contagion, which was ex-
pressed in a sharp fall in the growth rate of export volumes. Some degree
of exchange rate contagion was also felt, but it was much milder than in
Korea. As already noted, most of the nominal exchange rate depreciation
that occurred in Taiwan in late 1997 was quickly corrected. Where Taiwan
really differs from Korea is in the area of financial contagion: Taiwan ex-
perienced no such phenomenon. The country had little volatile foreign
capital within its borders, so little fled. If Taiwanese individuals and com-
panies wished to leave, their ability to do so was limited by financial reg-
ulations.
Concluding Remarks
The main lesson to be drawn from the differing experiences of Korea
and Taiwan is that liberalization of the capital account of the balance of
payments needs to be very cautious, particularly when it comes to short-
xaxuii i. acosix oc
xoiia axo raiwax ix rui iixaxciai ciisis o:
term flows and portfolio capital. Countries that choose to liberalize must,
at the very least, ensure that banking supervision and regulation are ade-
quate. Since domestic interest rates in emerging economies are bound to
be higher than international rates, domestic banks operating in economies
with a high degree of integration into international financial markets are
under constant temptation to borrow abroad. Preventing major mis-
matches requires strong, well-trained supervisory authorities. This is eas-
ier said than done. Good banking regulation and supervision develop
gradually and are highly correlated with the level of development. Their
evolution is often speeded, however, as a consequence of a financial crisis.
Korea is likely to come out of the crisis with a much improved domestic
financial system, much as Chile did after its 1982 banking crisis.
But better banking supervision is not a panacea. In financially open
economies where banks are not allowed to borrow freely, domestic firms
that can borrow directly from international markets may take on exces-
sive international debt.
38
When times are good and the real exchange rate
is appreciating, even rational firms may simply ignore the possibility of a
damaging exchange rate depreciation. Unhedged borrowing by firms was
a component in the Korean and other financial crises of the late 1990s.
Since the borrowing by each firm inflicts an adverse externality on the
economy as a whole (by exposing it to the risk that changes in creditor
and investor sentiment will force many firms into bankruptcy), there is
a good argument for monitoring such borrowings and limiting it when
necessary.
39
Large capital inflows may have other adverse effects on the economy, as
well. Large inflows unleash asset price bubbles that could attract even more
foreign capital, which is extremely volatile and can thus cause excessive
fluctuations in the real exchange rate. Such fluctuations are highly damag-
ing in countries that are attempting to integrate into international markets
through trade. A good part of these flows is temporary, but they can have
permanent effects on recipient economies. It may therefore be advisable for
the authorities to limit them. This is precisely what Taiwan has doneand
what Korea used to do before 1992. Korea has now dismantled most of its
controls on capital inflows, however. Much of this process was carried out
38. This point is made forcefully by Furman and Stiglitz (1998).
39. Fischer (1998), a fervent advocate of capital account liberalization, suggests that prudential con-
trols ought to include matching requirements on nonbank firms assets and liabilities in foreign ex-
change.
xaxuii i. acosix o:
before the crisis struck, and the authorities have essentially completed that
task as a result of the crisis, partly under pressure from the IMF. The policy
apparatus that allowed the country to change its economic structure dra-
matically and to build strong, internationally competitive firms is now
gone.
The firms are still there, however, and the Korean economy is recov-
ering from the crisis much faster than anyone expected. The growth of the
Korean economy was very strong in 1999 (almost 11 percent), and despite
a slowdown in the fourth quarter, it appears to have exceeded 8 percent in
2000. Much of the growth in 1999 and 2000 represents a recovery of con-
sumption and equipment investment. The Central Bank was able to re-
duce interest rates as international reserves recovered and capital began to
flow back in, which was very beneficial to the highly-leveraged Korean cor-
porate sector. Strong export growth has also assisted the recovery.
Nonetheless, the crisis is forcing major changes on the Korean econ-
omy. The banking sector is undergoing a profound transformation, and it
is to be hoped that banking regulation and supervision will advance sig-
nificantly. On the other hand, the Korean authorities have been forced to
open the economy to the purchase of domestic firms by foreign interests.
There are reports of large inflows of FDI, mainly in the form of M&A.
Many financial and nonfinancial firms are being bought up at bargain-
basement prices by investment fundsand often not by transnational cor-
porations in the same field as the acquired companies, which could be
expected to contribute new management or technology. The purchasers
are portfolio investors rather than long-term direct investors, and they are
likely to hold on to their Korean assets only until their prices rise with the
consolidation of economic recovery.
Recovery has thus stimulated renewed capital inflows. Korean firms
are already regaining their access to international financial markets. For-
eign funds are returning to the stock market. Korean banks will take longer
to regain international creditworthiness, since the bank restructuring
process is far from complete.
40
It is to be hoped that the Korean authori-
ties will soon have in place the banking regulation and supervision that
can control unnecessary and potentially damaging capital inflow. It is also
to be hoped that they will find substitutes for the now discarded capital
account controls of the past.
40. Park (1999).
References
Amsden, A. H. 1989. Asias Next Giant: South Korea and Late Industrialization. Oxford
University Press.
APEC (Asia-Pacific Economic Cooperation). 1998. The Impact of Investment Liberalization
in APEC: Policy Reviews and Case Studies. Report by the Economic Committee. Sydney.
Chang, H.-J. 1998. South Korea: The Misunderstood Crisis. In Tigers in Trouble: Fi-
nancial Governance, Liberalisation and Crises in East Asia, edited by K. S. Jomo. London:
Zed Books.
Dean, J. 1998. Why Left-Wing Moralists and Right-Wing Academics Are Wrong about
Asia. Challenge (March-April): 4460.
Eatwell, J. 1997. International Financial Liberalization: The Impact on World Develop-
ment. ODS Discussion Paper D-12. New York: United Nations Development Program
(UNDP), Office of Development Studies.
Feldstein, M. 1998. Refocusing the IMF. Foreign Affairs (March-April).
Fischer, S. 1998. Capital Account Liberalization and the Role of the IMF. In Should the
IMF Pursue Capital-Account Convertibility?, edited by S. Fischer and others. Essays in In-
ternational Finance 207. Princeton University, International Economics Section.
Furman, J., and J. E. Stiglitz. 1998. Economic Crises: Evidence and Insights from East
Asia. BPEA 2:1998.
Huang. Y. 1998. Reflections on Taiwans Economic Investment Policy Following the
Asian Currency Crisis. In The Asian Currency Crisis: The Taiwan Experience, edited by
C. Mai. Taipei: Chung-Hua Institution for Economic Research.
Kregel, J. 1998a. Derivatives and Global Capital Flows: Applications to Asia. Integracin
y Desarrollo (February-March): 2630. Lima.
. 1998b. East Asia Is Not Mexico: The Difference between Balance of Payments
Crises and Debt Deflation. In Tigers in Trouble: Financial Governance, Liberalisation
and Crises in East Asia, edited by K. S. Jomo. London: Zed Books.
Noland, M. 1996. Restructuring Koreas Financial Sector for Greater Competitiveness.
APEC Working Paper 96-14. Washington: Institute for International Economics (IIE).
Park, J.-J. 1998. Financial Crisis in Korea: Why It Happened and How It Can Be Over-
come. Seoul: Bank of Korea. Mimeographed.
Park, Y. C. 1996. The Republic of Koreas Experience with Managing Foreign Capital
Flows. In The Tobin Tax: Coping with Financial Volatility, edited by M. ul Haq, I. Kaul,
and I. Grunberg. Oxford University Press.
. 1998. The Financial Crisis in Korea and Its Lessons for Reform of the Interna-
tional Financial System. In Regulatory and Supervisory Challenges in a New Era of Global
Finance, edited by J. J. Teunissen. The Hague: Forum on Debt and Development (FON-
DAD).
. 1999. The Banking Reform in Korea: Issues and Challenges. Korea University
at Seoul. Mimeographed.
PECC (Pacific Economic Cooperation Council). 2000. Pacific Economic Outlook,
200001. Canberra: Asia Pacific Press at Australian National University.
Radelet, S., and J. Sachs. 1998. The East Asian Financial Crisis: Diagnosis, Remedies,
Prospects. BPEA 1:1998, 190.
xoiia axo raiwax ix rui iixaxciai ciisis o,
xaxuii i. acosix o
Shim, S. 1998. Recent Trends and Macroeconomic Forecast for 1998/99. Seoul: Korea
Development Institute. Mimeographed.
Taylor, L. 1999. Lax Public Sector, Destabilizing Private Sector: Origins of Capital Mar-
ket Crises. In International Monetary and Financial Issues for the 1990s, vol. 10, edited
by the United Nations Conference on Trade and Development (UNCTAD), 13155.
New York.
Wade, R. 1990. Governing the Market: Economic Theory and the Role of Government in East
Asian Industrialization. Princeton University Press.
Wang, Y., and H. Zang. 1999. Adjustment Reforms in Korea Since the Financial Crisis. Seoul:
Korea Institute for International Economic Policy.
World Bank. 1993. The Asian Miracle: Economic Growth and Public Policy. Oxford Uni-
versity Press.
Yanagihara, T. 1999. Regional Policy Coordination in Asia. Paper prepared for the work-
shop, Entering the 21st Century: The Changing Development Landscape, Tokyo. World
Bank and Foundation for Advanced Studies in Development (FASID) (2627 May).
Yang, Y. 1998. Reflections on the ROCs Financial Policies in Response to the Asian Cur-
rency Crisis. In The Asian Currency Crisis: The Taiwan Experience, edited by C. Mai.
Taipei: Chung-Hua Institution for Economic Research.
3
o,
C
hiles recent history includes three interesting epi-
sodes that shed light on the policies designed to
confront surges of capital inflows and their effect on the national econ-
omy. The first episode began in the mid-1970s and ended with the 1982
debt crisis; the second began in the early 1990s and continued through the
so-called Mexican tequila crisis; and the third consisted of the two years
preceding the Asian crisis. The three have features in common: they all
originated in an overabundance of foreign funds; the capital flew into the
private sector; in all cases Chile exhibited some strong fundamentals, such
as a fiscal surplus, and held an outstanding image as a successful country;
and each financial boom was followed by an international financial crisis.
Notable differences, however, are found in both the policy approach used
to face the capital flood in the three cases and the results obtained. The
three cases thus offer a valuable opportunity to study the relation between
the economic policies in question and their consequences.
1
This paper focuses on the characteristics of the three booms and their
impact on the sustainability of macroeconomic balances. It is during this
iicaioo iiiixcu-oavis
uiiiniiro raiia
*
Three Varieties of Capital
Surge Management
in Chile
*We are grateful to Guillermo Le Fort, Manuel Marfn, Carlos Massad, Jaime Ros, Gonzalo San-
hueza, John Williamson, Felipe Jimnez, Daniel Titelman, Mara ngela Parra, and Andras Uthoff for
substantive comments and criticisms.
1. See Ffrench-Davis (2001, chaps. 5, 6, and 10).
iicaioo iiiixcu-oavis axo uiiiniiro raiia oo
stage of the cycle that future crises are seeded or avoided. We only margi-
nally consider the post-boom period, when additional shocks (such as on
the terms of trade) may affect the adjustment path of the economy.
The impact of capital flows on emerging economies can be analyzed in
two dimensions. The first refers to the capacity for domestic absorption,
which largely determines how flows affect fundamentals in the host coun-
try. If absorption is efficient, capital inflows will contribute to a sustainable
growth process. If, on the contrary, the new funds distort key macroeco-
nomic prices, then wrong resource allocation results, and external sav-
ing will crowd out national saving. The central issue here is how to extract
the benefits from a higher availability of capital. The second dimension
has to do with external vulnerability. It involves examining the economys
exposure as it encounters volatile financial markets in order to determine
the cost of instability. Both dimensions are crucial for assessing the manage-
ment of capital flows. Dealing with the first is a necessarybut not suffi-
cientcondition for sound financial integration, while neglecting
the second may generate considerable short-term benefits at the cost of
medium-term risks stemming from higher foreign debt and exchange rate
and current account imbalances. These factors send the wrong signals to
the productive sector and ultimately make the economy the victim of its
own success.
Chile was the recipient of huge capital flows in each of the three peri-
ods under consideration. Private sector capital predominated in all three
cases, and it was directed to Chiles private sector. However, indicators of
vulnerability, such as the current account deficit and an appreciated ex-
change rate, and indicators of sound domestic absorption, such as national
saving and investment rates and the growth rate of potential gross domes-
tic product (GDP), behaved differently in each case. The first case in the
1970s coincided with low domestic investment and saving. Foreign saving
sharply crowded out national saving, and the macroeconomic environment
with an appreciated exchange rate and extremely high interest rates
discouraged the channeling of resources to productive investment. GDP
growth was high for several years, however, because the large underutilized
capacity resulting from the heavy 1975 recession (when GDP fell 13 per-
cent) left an important margin for easy economic recovery. In other words,
GDP was able to grow without requiring new capital formation. Rapid ex-
pansion of aggregate demand, stimulated by financial capital inflows, thus
moved the economy toward its productive frontier, closing the gap by
198081.
2
The large indebtedness made the economy highly vulnerable to
external shifts, but it was also in itself unsustainable, given a current ac-
count deficit of 14.5 percent of GDP in 1981 (at current prices; the figure
is 21 percent with a normalized dollar) and growing amounts coming due.
Pronounced inefficiencies thus characterized both the absorption of flows
(which was mainly limited to contributing to the recovery of economic ac-
tivity) and management of the external vulnerability. The result was that in
1982 Chile suffered the deepest debt crisis of all Latin American countries.
The situation was entirely different in the first half of the 1990s, re-
flecting the lessons learned. GDP grew strongly and was accompanied by
high investment and saving. The pre-tequila phase in Chile, unlike the pe-
riod preceding the debt crisis, began with high utilization of productive ca-
pacity. Policy designed to maintain the stability of macroeconomic prices
allowed aggregate demand to increase along with saving and investment.
The high growth rates were therefore sustainable over time. In other words,
capital inflows were absorbed efficiently. In the external sector, the current
account deficit was carefully controlled, and the Central Bank pursued an
active policy of preventing excessive exchange rate appreciation. With the
Mexican crisis of 199495, the Chilean economy showed that its domestic
achievements were accompanied by great strengths on the external front. A
positive trade shock reinforced this solid position.
Chile was highly regarded in the international community at the start
of the third episode. The economy had been almost immune to contagion
from the tequila effect, and it enjoyed full utilization of capacity and solid
domestic fundamentals. In 199697 Chile recorded vigorous growth that
could be judged sustainable thanks to high domestic investment rates, but
the economy was becoming vulnerable to changes in the international en-
vironment. When such changes occurred in 1998, driven by the large trade
and financial shocks brought on by the Asian crisis, Chile had a high exter-
nal deficit and an appreciated exchange rate. The country was forced to un-
dergo an intense process of macroeconomic adjustment, which led to a
recession (though with a moderate drop in GDP of 1 percent in 1999).
The specific policies and approaches used in each of the three episodes
varied, evolving from the extreme naivet of the 1970s into the more prag-
matic approach of the early 1990s. The end of the century saw a turning
away from macroeconomic sustainability as authorities gave in to the temp-
caiirai suici xaxacixixr ix cuiii o;
2. Ffrench-Davis (2001, chap. 1).
tation to move toward financial globalization without properly respecting
underlying risks.
Liberalization of the Capital Account: 197582
Shortly after the military coup of 1973, Chile began its transition to
a free-market economy with a series of institutional reforms of great depth
and breadth. The implementation of the neoliberal model meant substan-
tial changes in the public sectors role in the national economy. An intense
program of import liberalization was initiated in 1973, and the financial
system was radically reformed in 1975. Almost all the banks that had been
nationalized by the previous government were privatized. Interest rates
were abruptly deregulated in April 1975, and the rules governing lending
(which encouraged its channeling to production rather than consump-
tion) were eliminated. Authorities underrated prudential supervision, be-
lieving that the market would do the job.
3
Volume, Source, and Use of Flows
Capital flows into Chile rose fast starting in 1977 as a result of the
boom in lending to developing countries by international banks, which was
triggered by the capital surplus of oil-producing countries and was further
stimulated by domestic financial reforms. Chile exhibited a significant ac-
cumulation of international reserves up to 1981, although shortly thereaf-
ter the external sector showed a high and rising current account deficit. The
capital flow was overwhelmingly concentrated in the private sector without
guarantees by the State, with less than one-fifth of inflows taking the form
of foreign direct investment (FDI) and loans to the public sector.
Flows were sizeable in the second half of the 1970s, as shown in table
3-1. Foreign saving (the current account deficit) and debt service also grew
as a share of exports, despite very strong export activity in the initial years
of the neoliberal experiment. Foreign debt service in 1982 equaled 88 per-
cent of goods and services exports, that is, three times the debt-service co-
efficient for 197074. The magnitude of capital movements is reflected in
the fact that for 198081, gross loans received climbed to 24 percent of
iicaioo iiiixcu-oavis axo uiiiniiro raiia o
3. See Ffrench-Davis and Arellano (1981); Daz-Alejandro (1985); Edwards and Cox Edwards
(1987); Ffrench-Davis (2001, chap. 5); Harberger (1985); Morand and Schmidt-Hebbel (1988);
Reinstein and Rosende (2000); Tapia (1979); Valds-Prieto (1992).

T
a
b
l
e

3
-
1
.
C
h
i
l
e
,

N
e
t

C
a
p
i
t
a
l

I
n
f
l
o
w
s
,

a
n
d

G
D
P

G
r
o
w
t
h
,

1
9
7
7

2
0
0
0
P
e
r
c
e
n
t

o
f

G
D
P
F
i
x
e
d
A
c
t
u
a
l

G
D
P
i
n
v
e
s
t
m
e
n
t
g
r
o
w
t
h
P
o
t
e
n
t
i
a
l

P
e
r
i
o
d
C
u
r
r
e
n
t

p
r
i
c
e
s
1
9
8
6

p
r
i
c
e
s
a
N
a
t
i
o
n
a
l
E
x
t
e
r
n
a
l
(
1
9
8
6

p
r
i
c
e
s
)
(
1
9
8
6

p
r
i
c
e
s
)
G
D
P

g
r
o
w
t
h
b
1
9
7
7

8
1
1
0
.
8
1
7
.
2
1
1
.
6
7
.
2
1
8
.
7
7
.
5
3
.
0
1
9
8
2

8
9
5
.
5
6
.
4
1
1
.
5
6
.
2
1
8
.
2
2
.
6
3
.
0
1
9
9
0

9
5
6
.
9
6
.
9
2
2
.
1
2
.
5
2
6
.
1
7
.
8
7
.
0
1
9
9
6

9
7
8
.
0
9
.
2
2
1
.
2
5
.
7
3
1
.
6
7
.
4
7
.
1
1
9
9
8

2
0
0
0
1
.
2
1
.
3
2
1
.
6
2
.
7
2
8
.
6
2
.
7
4
.
8
S
o
u
r
c
e
:

C
e
n
t
r
a
l

B
a
n
k

o
f

C
h
i
l
e
.

P
o
t
e
n
t
i
a
l

G
D
P

b
a
s
e
d

o
n

R
.

F
f
r
e
n
c
h
-
D
a
v
i
s
,

E
c
o
n
o
m
i
c

R
e
f
o
r
m
s

i
n

C
h
i
l
e
:

F
r
o
m

D
i
c
t
a
t
o
r
s
h
i
p

t
o

D
e
m
o
c
r
a
c
y
,

c
h
a
p
.

1
,

U
n
i
v
e
r
s
i
t
y

o
f
M
i
c
h
i
g
a
n

P
r
e
s
s
,

2
0
0
1
.
a
.

T
h
e

c
o
n
s
t
a
n
t

p
r
i
c
e

s
e
r
i
e
s

w
a
s

d
e
r
i
v
e
d

b
y

d
e
f
l
a
t
i
n
g

t
h
e

d
o
l
l
a
r

s
e
r
i
e
s

u
s
i
n
g

a
n

i
n
d
e
x

o
f

f
o
r
e
i
g
n

p
r
i
c
e
s

f
a
c
e
d

b
y

t
h
e

C
h
i
l
e
a
n

e
c
o
n
o
m
y
.

F
o
r

t
h
e

d
e
n
o
m
i
n
a
t
o
r
,
G
D
P

a
t

c
o
n
s
t
a
n
t

p
r
i
c
e
s

w
a
s

c
o
n
v
e
r
t
e
d

i
n
t
o

1
9
8
6

d
o
l
l
a
r
s

u
s
i
n
g

t
h
e

1
9
8
6

p
e
s
o
-
d
o
l
l
a
r

e
x
c
h
a
n
g
e

r
a
t
e
.
b
.

P
r
o
d
u
c
t
i
v
e

c
a
p
a
c
i
t
y

c
r
e
a
t
e
d

i
n

e
v
e
r
y

p
e
r
i
o
d
.
S
a
v
i
n
g
s
(
c
u
r
r
e
n
t

p
r
i
c
e
s
)
N
e
t

c
a
p
i
t
a
l

i
n
f
l
o
w
s
iicaioo iiiixcu-oavis axo uiiiniiro raiia ;c
GDP. All of these indicators were substantially greater than the Latin
American averages for the 1970s.
4
The composition of agents participating in capital flows changed sig-
nificantly in this period. On the lending side, 84 percent of the foreign
debt in 1981 was to banks, compared to only 19 percent at the end of
1974. Chiles bank debt grew 57 percent a year between 1977 and 1981,
while the average for developing countries was 28 percent. On the borrow-
ing side, the growing net inflows were directed primarily at the private sec-
tor from 1975 on, while the State moved toward a budget surplus. A
deliberate policy of reducing the role of the State was facilitated by changes
in international markets, where official institutions lending to govern-
ments lost relevance while private capital markets emerged strongly as a
source of access for both public and private borrowers.
The great majority of private debt was contracted with no explicit
guarantee by the State. Fully two-thirds of Chiles total debt lacked such
guarantee in 1981. Through 1977, the private sector obtained a signifi-
cant part of its loans directly, as a result of quantitative restrictions then
applying to local banks with respect to foreign financial credits.
5
From
1978 on, domestic banks were allowed gradually greater importance in the
direct intermediation of foreign financing. It was only in April 1980, how-
ever, that domestic banks were allowed access to foreign credit on the same
conditions as the nonbanking sector. Lending through the banking sector
grew dramatically from then on, yet flows to the nonfinancial sector con-
tinued to grow at a significant rate as well. The private sector received 94
percent of the gross flow in 197681.
6
Debt and Macroeconomic Adjustment
Chile took on massive quantities of debt between 1977 and 1981,
which had significant effects on aggregate demand and its composition. It
also contributed to both economic recovery and the concentration of
wealth, crowded out domestic savings, and had a decisive effect on mon-
etary and exchange rate policies. Generally speaking, the initial impact of
foreign borrowing is an increase in the supply of foreign currency, which
can lead to two outlets: an increase in reserves or exchange rate apprecia-
tion and a growing current account deficit. Chile experienced both effects
4. Ffrench-Davis (1984).
5. A large share of these loans was backed by bank guarantees at the beginning of the borrowing
boom. See Ffrench-Davis (2001, table 5.4).
6. See Ffrench-Davis (2001, chap. 5).
caiirai suici xaxacixixr ix cuiii ;:
up to 1981. Net capital inflows were greater than the capacity of the do-
mestic economy to absorb foreign funding, despite the growing expansion
of that capacity as a result of economic recovery, import liberalization, and
exchange rate revaluations driven by inflows.
The rapid accumulation of reserves, in turn, had substantial effects on
the money supply. The large capital flowsboth those used to finance in-
creased imports and those that went to expanding international reserves
led to a high proportion of the economys domestic credit being based on
foreign funding. Despite this volume, substantial differences between do-
mestic and external interest rates persisted.
xoxirai\ ioiic\ axo caiirai coxriois Starting in 1975, net pur-
chases of foreign currency by the Central Bank were the main source of
money supply expansion.
7
They represented over 100 percent of money
supply changes in the three-year period 197880. As already mentioned,
the overwhelming share corresponded to private borrowing; operations
with the public sector actually had a contractive effect in some years. The
severe macroeconomic disequilibria that had been accumulating finally
emerged in 1981, when international reserves began to fall. The monetary
effect of foreign exchange operations correspondingly became strongly
contractive.
Some direct restrictions were applied to capital inflows as a part of the
monetary program. One such restriction was designed specifically to mod-
erate the monetary effects of flows by limiting the amount that banks
could exchange each month at the Central Bank. This measure was in ef-
fect from 1977 to 1979, with a series of modifications in the maximum
amount allowed.
8
The restrictions were not sufficient, however, to hold
foreign loans to the private sector at a volume consistent with the rate of
monetary expansion considered desirable by economic authorities. Do-
mestic credit by the Central Bank (rediscounts or loans to the banking sec-
tor) was therefore tightened in the face of increased money supply
stemming from foreign exchange dealings. Nevertheless, the increase in ag-
gregate demand was much more rapid than GDP growth. A strong real ex-
change rate revaluation took place despite the accumulation of reserves,
which had the effect of crowding out domestic producers of tradables.
Targets for expanding liquidity were in effect while a closed-economy
approach predominated in policymaking. In 1979, authorities adopted an
7. Ffrench-Davis and Arellano (1981, table 13).
8. Ffrench-Davis and Arellano (1981).
iicaioo iiiixcu-oavis axo uiiiniiro raiia ;:
open-economy monetary approach to the balance of payments. The nomi-
nal exchange rate was frozen in June, and a supposedly neutral monetary
policy was established. Variations in the level of international reserves were
to automatically determine the economys level of liquidity, within the con-
text of a fiscal budget in surplus and low reserve ratios in the banking sector.
In April 1979, an unremunerated reserve requirement was established
for foreign loans (under Article 14), with percentages varying according to
the term of the loan. The prohibition on loans with an average maturity
of less than twenty-four months was maintained (with a 100 percent re-
serve requirement), and reserve requirements were based on a structure of
decreasing rates according to the loan term. Loans over sixty-six months
were exempt. In August 1979 and July 1980, rates were lowered for loans
with a maturity of over twenty-four months. The modifications were not
significant for two reasons, however. First, no loan of less than two years
could be contracted throughout this period, except for trade credits gov-
erned by limits that had survived the deregulation of April 1980. Second,
the strong segment of the international market at the time was syndicated
loans, whose average term varied from six to ten years in 197680.
9
The monetary approach to the balance of payments prevailed from
mid-1979 to mid-1982, with a neutral monetary policy based on the dollar
standard. At the end of these three years, a contractionary automatic adjust-
ment was in effect, with disastrous results on employment and output.
10
ixcuaxci iari ioiic\ Exchange rate policy evolved significantly in the
197482 period.
11
Up to June 1976, there was a crawling-peg regime, with
adjustments made one to four times a month. The subsequent regime
combined daily devaluations determined each month in advance with four
abrupt changes in the value of foreign currency (two devaluations and two
revaluations). In early 1978 and 1979, a table of daily adjustments was es-
tablished for the remainder of each year. In June 1979, this system was
interrupted with a devaluation that brought the price of the dollar to the
level scheduled for the end of the year (39 pesos). This rate was maintained
until June 1982, when there was an 18 percent devaluation and an an-
nouncement of a table of smaller devaluations. This was soon replaced,
however, by a short-lived freely floating exchange rate, which was followed
by a return to a crawling-peg regime. The evolution of the real exchange
rate is shown in figure 3-1.
9. Ffrench-Davis (1984, table 11).
10. See Arellano and Cortzar (1982); Ffrench-Davis (2001, chaps. 1 and 4).
11. See Ffrench-Davis (2001, chap. 4).
caiirai suici xaxacixixr ix cuiii ;,
Figure 3-1. Real Exchange Rate, 19742000
a
Source: United Nations Economic Commission for Latin America and the Caribbean (ECLAC); Central Bank of
Chile; Ffrench-Davis (2001).
160
140
120
100
80
60
40
20
Index, 1974 100
1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000
ECLAC
Adjusted
Central
Bank
Net inflows were a crucial factor in making it possible for the author-
ities to manage the exchange rate with goals other than efficient resource
allocation until 1981. The use of the exchange rate to shape expectations
(in 197679) or to anchor domestic prices to external prices (197982)
resulted in lower inflation, but it also produced significant real exchange
rate appreciation (45 percent between 1975 and 1981). When combined
with import liberalization (the average tariff was reduced from 94 percent
to 10 percent) and the economic recovery of 197781, this caused the not-
able current account deficit mentioned above.
12
Furthermore, the gradual
appreciation brought the observed real cost of foreign borrowing down to
negative levels in 1979 and 1980, which encouraged inflows.
ixriiisr iari oiiiiiixriais The government expected liberalization
to cause a strong trend toward the reduction of spreads, with an equalizing
of domestic and foreign interest rates. Large differentials between lending
12. The increase in the deficit was also associated with high interest payments and falling copper
prices. The loss in government revenue in 1981compared to the 196570 averagerepresented 1.6
percent of GDP. The sharp fall in copper prices was partially offset by the governments capturing the
economic rent following the nationalization of copper mining in successive stages in 1966, 1969, and
1971.
and borrowing rates in the domestic financial market persisted throughout
the period, however, and both rates were quite high. In addition, domestic
rates continued to exceed international rates by a wide margin despite the
large capital inflows, especially from 1979 on (see table 3-2).
13
The distributive impact of interest rate differentials is illustrated by
the fact that between 1976 and 1982 the average domestic borrower paid
interest exceeding a normal annual real rate of 8 percent, equivalent to
300 percent of the principal borrowed. In other words, a borrower who
paid the lender 8 percent, renewed principal, and capitalized interest over
8 percent would, by the end of 1982, have had a debt four times the in-
itial one, in currency of constant purchasing power. By contrast, the debtor
borrowing abroad on the same terms would have owed 44 percent less be-
iicaioo iiiixcu-oavis axo uiiiniiro raiia ;
13. One recurring explanation is that the rate differentials derived from expectations of devaluation.
In 1981, however, many local borrowers still moved from peso-denominated to foreign-denominated
debt. Ffrench-Davis (2001, chap. 5), and Harberger (1985) examine the causes of the real interest rate
differentials in pesos. These causes were reinforced by the rediscount constraints imposed by the Cen-
tral Bank.
Table 3-2. Chile: Domestic and External Real Interest Rates, 197582
a
Annual percent in pesos
Year Domestic External Differential
1975
b
121.0 . . . . . .
1976 51.2 -21.1 72.3
1977 39.4 0.2 39.2
1978 35.1 3.8 31.3
1979 16.9 -0.9 17.8
1980 12.2 -8.0 20.2
1981 38.8 12.4 26.4
1982 35.2 45.0 -9.8
Sources: Authors calculations, based on data from the Central Bank of Chile; Instituto Nacional
de Estadsticas (INE); R. Cortzar and J. Marshall, Indice de precios al consumidor en Chile:
197078, Coleccin Estudios CIEPLAN 4 (November), 1980; R. Ffrench-Davis and J. P. Arellano,
Apertura financiera externa: la experiencia chilena en 197380, Coleccin Estudios CIEPLAN 5
(July), 1981.
a. The domestic rate refers to the predominant segment of the market, covering transactions for
terms of thirty to ninety-one days. The external interest rate is that paid for bank credits obtained
under Article 14, plus the cost in respect of compulsory deposits and the financial spread, all
converted into their peso equivalents. In 1982 the preferential exchange rate was used to calculate
the external interest rate.
b. Second semester, after interest rate liberalization.
fore the 1982 devaluation than the original debt contracted in 1976. The
massive real devaluation of 80 percent then raised this borrowers debt to
its original level (equivalent to only one-fourth of the first borrowers peso-
denominated debt). The figures vary, obviously, according to the date of
borrowing. For example, late borrowers who took loans denominated in
foreign currency in 1981 lost dramatically, given the real devaluations that
took place in 1982.
These differentials had significant effects, from both the allocative and
distributive points of view.
14
Small and medium-size enterprises (SMEs)
were largely relegated to the high-interest segment of the market. Firms as-
sociated with the management of financial institutions, as well as the larger
economic groups, enjoyed expeditious access to foreign credit, whether
directly or through the intermediation of Chilean banks. The low rates of
saving and investment during this period show that the environment pro-
vided more incentives for financial activity and speculation than for
productive activities.
15
oiriiioiario ioirioiio quaiir\ axo iiuoixriai suiiivisiox For-
eign borrowing gave rise to a domestic lending boom. The increase in
credit took several forms: direct foreign loans not backed by guarantees,
direct foreign loans with guarantees, foreign loans intermediated by local
banks, and loans made by these banks in domestic currency on the basis
of the liquidity generated by foreign exchange operations. The lend-
ing boom took place in an atmosphere of very lax prudential supervision.
Related-party lending rose rapidly, often without guarantees or with ficti-
tious guarantees. Regulations were circumvented via cross-lending and
loans to paper firms, as well as loans through off-shore institutions. Faced
with high interest rates, the banks renewed loans (often on a thirty-day
basis) and financed interest payments with new loans. Meanwhile, the
stock of loans grew 38 percent a year in real terms between 1976 and
1981. Nonperforming loans appeared low and the banks profits high.
Many loans were backed by stock and real estate, but the prices of such
collateral were inflated as a result of the financial boom and the mistaken
belief that the Chilean economy would continue to grow at around 8 per-
cent a year.
Required provisions were low. In 1979, general requirements were re-
caiirai suici xaxacixixr ix cuiii ;,
14. Zahler (1980).
15. See Mizala (1992).
duced from 2 percent of loans to 0.75 percent. Priority was to be given to
individual provisions, although this was not implemented. After the ex-
plosion of the banking crisis in January 1983, observers realized that 19
percent of loans in late 1982 were with related parties and represented 249
percent of the capital and reserves of private banks. The severe weakness
of the financial system, which resulted from typical neoliberal reforms, had
enormous fiscal costs in the following years.
16
It should be emphasized that around 40 percent of external lending to
private firms in the domestic market (Article 14) was handled directly,
with no intermediary between borrowers and foreign lenders. Conse-
quently, not even the strictest and most effective prudential supervision of
the local financial system could have fully prevented the development of
macroeconomic imbalances as a result of the boom in foreign funding.
17
Inefficient Domestic Absorption and External Vulnerability
The world economy and the domestic market underwent a process of
increasing indebtedness, which led to a gradual exchange rate appreciation
and great vulnerability for Chiles external sector. The belief that the appre-
ciation would prove sustainable made foreign loans more attractive. Stock
and real estate prices were rising at a dizzying rate.
18
Such feedback rein-
forced the cycle, increasing the inflow of capital. This, in turn, increased ag-
gregate demand and supported the persistent exchange rate appreciation
and a notable rise in asset prices. The economy became increasingly accom-
modated to a massive financial inflow, and aggregate demand became
highly intensive in importables.
19
The form that foreign funding took and the incentives provided by
the economic model led to a decrease in investment, and, more so, in na-
tional savings (see table 3-1 above).
20
Gross fixed capital formation aver-
iicaioo iiiixcu-oavis axo uiiiniiro raiia ;o
16. Held and Jimnez (2001); Sanhueza (1999).
17. See Valds-Prieto (1992) on the weaknesses of prudential supervision on the supply side and
the importance of nonbank flows.
18. The General Stock Price Index grew by 1,000 percent between 1976 and 1980. Real estate
prices peaked in 1981 at more than 100 percent above their 1976 prices (Morand, 1992).
19. Ffrench-Davis and De Gregorio (1987).
20. One theoretical interpretation of the crowding out of national savings is that a sharp rise in pro-
ductivity generated the expectation that GDP capacity would continue to grow by 8 percent a year
even given a low investment ratio. Anticipating the consumption of future increased income was there-
fore seen as an equilibrating intertemporal adjustment. Since the expectation was wrongcapacity
was growing very slowlythe adjustment was actually destabilizing (Ffrench-Davis and Reisen, 1998,
chap. 1).
aged only 18.7 percent of GDP in 197781, significantly under the 21.2
percent average of the 1960s.
21
A growing share of funds was channeled
to the consumption of imported goods, crowding out spending on do-
mestic tradables and national savings. This phenomenon intensified in
1981, as the process of recovery came up against the economys produc-
tive capacity. National savings fell to 8 percent of GDP in 1981 and to 2
percent in 1982. Excessively high domestic interest rates, the disman-
tling of public mechanisms to support productive activities, the drop in
public investment, and an abrupt import liberalization with an appre-
ciated and unstable exchange rate all combined to discourage productive
investment.
It was obvious that the external deficit could not be sustained for long,
even if the international environment were to remain unchanged. Never-
theless, the government argued that the process was in balance because
private lenders were financing it and that it would be self-regulating.
Demand for imports would soon reach the saturation point, spontaneously
checking the rise. A currency crisis could not occur against a background of
budget surpluses, international reserves larger than high-powered money,
and a supposedly neutral monetary policy. When international financial
problems began to emerge in 1981, however, the current account deficit
was 14.5 percent.
22
There was a clear and compelling need to reduce the ex-
ternal imbalance and correct the appreciated exchange rate.
Chiles difficulty in obtaining foreign loans in 1982 coincided, then,
with a great domestic need for fresh funds to pay growing interest and
amortization costs, as well as to cover the trade deficit. The composition
of the foreign debt had some strongly negative dynamics. First, given the
prevalence of variable international interest rates, net interest payments
quadrupled between 1978 and 1982 (reaching 7 percent of GDP), after the
contraction of international financial markets. Second, short-term debt
grew from what was considered a normal level, associated with a sustainable
volume of trade, to double its share, reaching 20 percent of all debt by 1982
caiirai suici xaxacixixr ix cuiii ;;
21. Figures in 1986 currency. The authorities reiterated that increased productivity made higher in-
vestment unnecessary. Our calculations show, however, that of the actual GDP increase of 6.5 percent
in the 197680 period, only 2 percent a year was due to increased capacity, while the rest reflected ex-
ploitation of existing capacity that had been underused since the deep recession of 1975 (Ffrench-
Davis, 2001, chap. 1).
22. This figure, in current prices, underestimates the importance of the deficit as a result of the ex-
cessive exchange rate appreciation in 1981. The deficit climbs to 21 percent of GDP if measured using
the exchange rate of 197678. For a discussion of alternative measures of this coefficient, see Ffrench-
Davis (2001, chap. 5).
(13 percent of GDP). Third, amortization of private medium-term debt
grew rapidly.
In brief, the Government was relying on the automatic functioning of
the dollar standard, and the productive apparatus was weak and suffering
from overindebtedness. The external shocks of the early 1980s thus found
the Chilean economy in a very vulnerable position. The effects of the fi-
nancial and terms-of-trade shocks were multiplied in the domestic market,
with a reduction in aggregate spending of 30 percent and a GDP drop of
17 percent in 198283 (see table 3-3). In the midst of this general crisis,
many observers failed to analyze the specific policies for capital flow man-
agement used in this period. The tools were clearly inadequate for coun-
teracting the crisis. This was not so much the fault of the instruments
themselves, however, as of the fact that a small economy was confronting
an enormous supply of funds and that very high differentials between do-
mestic and foreign interest rates (about 20 to 30 percentage points) made
any sort of inflow profitable.
Active Regulation of Capital Inflows: 199095
Latin America began to see renewed inflows of private capital in the
late 1980s. Chile was one of the first to attract new funds and was among
iicaioo iiiixcu-oavis axo uiiiniiro raiia ;
Table 3-3. Debt Crisis in Chile, 197783
Indicator 1977 1978 1979 1980 1981 1982 1983
GDP
a
8.3 7.8 7.1 7.7 6.7 -13.4 -3.5
Aggregate expenditure
a
12.9 9.7 10.4 10.5 12.4 -23.8 -8.6
Fiscal expenditure
a
8.8 -18.8 -7.4 -3.2 7.6 16.7 -4.0
Exports
a
12.0 11.8 14.2 14.5 -9.0 4.5 0.1
Imports
a
32.4 18.1 23.9 20.7 13.3 -34.0 -17.7
Current account deficit
b
3.7 5.2 5.4 7.1 14.5 9.2 5.4
Real exchage rate
c
62.8 69.7 69.7 60.9 51.8 60.1 72.1
Terms-of-trade effect
d
-0.2 -0.9 1.8 -0.6 -2.2 -0.8 1.6
Source: Based on data from the Central Bank of Chile; F. Larran, Public Sector Behavior in a
Highly Indebted Country: The Contrasting Chilean Experience, 197085, in The Public Sector and
the Latin American Crisis, edited by F. Larran and M. Selowsky, San Francisco: ICS Press, 1991;
Ffrench-Davis (2001).
a. Official figures at 1986 constant prices, annual growth (percent).
b. Current prices, percent of GDP.
c. Index, 1986 = 100.
d. Current prices, percent of GDP.
the countries facing the greatest quantity of inflows in relation to the size
of its economy. Recipient countries generally experienced macroeconomic
problems stemming from the magnitude of the inflows and their compo-
sition, which was prone to volatility.
23
Chilean policy in the first half of the 1990s represented a significant
step toward a more pragmatic approach to macroeconomic management.
In brief, policymakers responded to the massive availability of foreign cap-
ital by attempting to moderate short-term inflows while keeping the door
open to long-term flows. Specifically, an unremunerated reserve require-
ment was established to raise the cost of bringing in short-term capital;
this is a market-based instrument that affects relative costs. Authorities also
used exchange rate intervention and monetary sterilization to hold down
the appreciation of the real exchange rate in the face of those flows that
surpassed the reserve barrier and gave way to the monetary effects of for-
eign operations. These tools were used in support of a development strat-
egy that encouraged export growth and diversification.
The policy was highly successful, in the sense that in 199194 the cur-
rent account deficit was moderate, the currency appreciated less than in
most of the regions countries, and the total short-term external debt was
held to a fairly low magnitude. When the Mexican exchange rate crisis ex-
ploded in late 1994, the Chilean economy proved immune to contagion
(see table 3-4). The country had significantly reduced its vulnerability.
Capital Inflows from 1990 to 1995
The return to democracy in 1990 coincided approximately with the
beginning of an episode of abundant foreign capital flowing to emerging
economies. Effective private inflows to Chile achieved some importance in
1989; after that date, FDI increased steadily and represented a solid ma-
jority of capital inflows in the 1990s (see figure 3-2).
24
A set of active mac-
roeconomic policies was adopted to regulate the capital surge. During this
period, productive capacity expanded vigorously, and the economy was
running close to full capacity. This was a determining factor in creating a
virtuous circle of rapid capital formation.
25
caiirai suici xaxacixixr ix cuiii ;,
23. See, for instance, Calvo, Leiderman, and Reinhart (1993); Ffrench-Davis and Griffith-Jones
(1995).
24. Significant movements began in 1986 in relation to a debt conversion program, but this activ-
ity involved debt paper instead of cash flows. See Ffrench-Davis (2001, chap. 7).
25. See Agosin (1998); Ffrench-Davis and Reisen (1998).
Private short-term lending also figured prominently in the capital
surge into Latin America in the 1990s. For a country to be the target of
interest rate arbitrage, domestic interest rates must exceed international
rates by a margin that more than offsets the currencys expected deprecia-
tion and the country risk. Such conditions obtained in Chile starting in
the early 1990s. On the one hand, in 1992 and 1993, international rates
on dollar loans were at their lowest level in thirty years, and though they
later rose, they continued to be moderate and remained far below their lev-
els in the 1980s. On the other hand, Chile is a capital-scarce country, with
a stock consistent with a GDP of only U.S.$5,000 per capita. Because the
scarcity entails a higher price, the interest rate in emerging economies
tends to be higher than in developed countries.
26
Monetary policy must
keep average real interest rates above international rates to ensure sustain-
able macroeconomic balances.
27
Other conditions for interest rate arbitrage also proved favorable to in-
flows. After a cumulative real depreciation of 130 percent in the 1980s,
iicaioo iiiixcu-oavis axo uiiiniiro raiia c
26. Our policy implications are consistent with the idea that the fastest growing economies have
higher interest rates associated with their GDP growth rates.
27. Chilean authorities increased the structural interest rate differential through a macroeconomic
adjustment in 1990, based unilaterally on a substantial rise in the interest rate by the Central Bank
(see Ffrench-Davis, 2000, chap. 7).
Table 3-4. Tequila Crisis in Chile, 199195
Indicator 1991 1992 1993 1994 1995
GDP
a
8.0 12.3 7.0 5.7 10.6
Aggregate expenditure
a
6.2 15.0 10.8 5.5 16.2
Fiscal expenditure
a
10.4 12.3 7.4 5.6 6.2
Exports
a
12.4 13.9 3.5 11.6 11.0
Imports
a
7.0 21.8 14.2 10.1 25.0
Current account deficit
b
0.3 2.3 5.6 3.0 2.0
Real exchage rate
c
106.4 97.6 96.9 94.3 88.9
Terms-of-trade effect
d
0.1 -0.3 -1.9 2.8 2.5
Source: Based on data from the Central Bank of Chile; Direccin de Presupuestos del Ministerio
de Hacienda (DIPRES), Estadsticas de las finanzas pblicas: 19901999, Santiago, March 2000;
Ffrench-Davis (2001).
a. Official figures at 1986 constant prices, annual growth (percent).
b. Current prices, percent of GDP.
c. Index, 1986 = 100.
d. Current prices, percent of GDP.
Figure 3-2. Composition of Capital Flows, 19802000
Source: Central Bank of Chile.
8
6
4
2
0
-2
-4
-6
-8
Billions of dollars
Other
Portfolio
FDI outflow
FDI inflow
1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
the exchange rate began to appreciate gradually in the 1990s. As in the case
of other countries in the region, Chiles country risk premium fell. An at-
mosphere of emerging markets mania ( la Kindleberger) on the part of
international investors generated a spectacular drop in perceived country
risk.
28
Private short-term capital inflows were heavy well into 1992, after
which they began to fall as a result of policy measures taken to check them.
Portfolio inflows took two forms: investment through large interna-
tional mutual funds and the issuing of American depositary receipts
(ADRs) by large Chilean firms. Primary issues of ADRs represent an op-
portunity for a firm to expand its capital at a relatively low cost, since costs
in international markets naturally tend to be lower than in the Chilean fi-
nancial market. Secondary issues of ADRs occur when foreigners purchase
securities available on the Chilean stock market and subsequently convert
them into ADRs.
29
This operation constitutes a change of ownership from
nationals to foreigners, without a direct financial effect on the firm. Chiles
relatively developed domestic securities market, together with the growing
use of ADRs to reach the U.S. securities market, made Chilean securities
caiirai suici xaxacixixr ix cuiii :
28. For instance, the Euromoney index shows an improving country risk from 47.7 in 1990 to 72.9
in 1995.
29. See Ffrench-Davis, Agosin, and Uthoff (1995).
attractive for investors seeking new financial vehicles. These changes of
ownership exposed the economy to an additional degree of uncertainty and
volatility, since foreign investors can easily withdraw their investments.
30
Such flows clearly played a destabilizing role in the economy. They contrib-
uted to the inflationary behavior of the securities market in 1994 and 1997
and depressed the market in 1995 and 1998, operating procyclically.
While private sector flows increased, public debt fell. The Chilean
government reduced new credits from international financial institutions
and made sizable prepayments, particularly in 199596.
The Response of Macroeconomic Policy and Its Effects
Monetary authorities in Chile deployed a wide range of measures to
regulate the surge in financial capital in 199095. The Central Bank took
steps to discourage inflows of short-term, liquid capital by means of the
unremunerated reserve requirement. At the same time, it kept the doors
open for FDI risk capital. It moderated the impact of those waves of cap-
ital by intervening in the foreign exchange market to prevent an overabun-
dance of foreign currency from appreciating the real exchange rate too
much and by sterilizing the monetary effects of the rapid accumulation of
international reserves.
31
Three other policies contributed to the success in managing capital in-
flows. First, fiscal policy was very cautious. Increases in social spending were
financed with new taxes. Consequently, Chile had a significant nonfinan-
cial public sector surplus of 1 or 2 percent of GDP. This prudential ap-
proach, which included observing the regulations of a stabilization fund for
public copper revenues, facilitated the monetary authorities task of regulat-
ing capital inflows, and it contributed to preventing excessive exchange rate
appreciation in the first half of the decade. Of course, running a fiscal sur-
plus does not guarantee financial stability. The great 1982 crisis occurred
despite Chiles having had several years of budget surpluses. The same thing
happened in Mexico with tequila crisis and in Korea with the Asian crisis. In
all three cases, the external deficit was led by the private sector.
iicaioo iiiixcu-oavis axo uiiiniiro raiia :
30. Some analysts maintain that foreigners who become pessimistic about a country generally sell
their ADRs on the U.S. stock market, and the sale therefore has no impact on domestic stock and for-
eign exchange markets. This argument disregards the fact that when firms issue ADRs, the prices of
securities on the domestic and U.S. markets tend toward equality through arbitrage. In fact, equity
price movements of Chilean firms that have issued ADRs on U.S. markets are closely linked with
movements on the Santiago exchange.
31. See Ffrench-Davis, Agosin, and Uthoff (1995).
Second, prudential banking regulations had been introduced in re-
sponse to the banking crisis of 198283, and the regulatory system had
then been improved over the years. Authorities effectively resisted pres-
sures to weaken supervision when lobbying sectors argued that the system
was mature enough to self-regulate; in fact, prudential supervision was
strongly intensified. This made it difficult for capital inflows to trigger an-
other credit boom in the commercial banks, which in turn helped keep the
current account deficit and exchange rate within sustainable limits until
1995. It is important to note that a significant share of financial flows is
not channeled through local banks. This phenomenon became decidedly
more pronounced in the 1990s, with the development of portfolio invest-
ment funds and derivatives markets. Nevertheless, fiscal responsibility and
effective prudential supervision of the financial system are two very impor-
tant complements in regulating capital flows.
Third, authorities continually monitored aggregate demand and its
consistency with productive capacity. Consequently, macroeconomic dis-
equilibria were not allowed to accumulate. Some overheating occurred in
1993, but the authorities conducted a downward adjustment. When the
tequila crisis exploded, Chile had a moderate external deficit, large re-
serves, and space for increasing economic activity. Those conditions would
not have been feasible without regulating capital inflows, managing the ex-
change rate flexibly, and pursuing an active monetary policy.
32
ioiic\ ioi xaxacixc caiirai ixiiows Strategic features of the policies
used went against the fashion of capital account liberalization. The two
main targets of exchange rate and inflows management policies were, first,
achieving sustained macroeconomic stability in an economy prone to huge
cycles (Chile experienced the sharpest recessions in all Latin America in
1975 and 1982) and, second, supporting the growth model adopted by
the authorities, which gave the expansion and diversification of exports a
crucial role.
In the face of a plentiful supply of foreign funds, Chilean authorities
opted to regulate the foreign currency market in order to prevent large
misalignments in the real exchange rate relative to its medium-term trend.
The natural short-term horizon of financial markets can lead to exchange
rate values that are inconsistent with medium- and long-run trends, which
caiirai suici xaxacixixr ix cuiii ,
32. Good luck also played a role, with a sharp improvement in the terms of trade in 1995. Even so,
the other factors provided strong macroeconomic insurance.
negatively affects decisions in the productive sector.
33
The authorities
sought to preserve the predominance of medium-term fundamentals over
short-term factors influencing the exchange rate.
In June 1991, the financial environment featured accelerating growth
in available external funds and a high copper pricefactors considered to
have a significant transitory component. The authorities reacted by estab-
lishing an unremunerated reserve requirement of 20 percent on foreign
credit (covering the entire spectrum of foreign credit, from that associated
with FDI to trade credit). The reserve was to be on deposit at the Central
Bank for a minimum of 90 days and a maximum of one year, according
to the time frame of the operation. At the same time, a tax on domestic
loans of 1.2 percent for operations of up to one year was extended to for-
eign loans. The reserve requirement, an option of paying its financial cost,
and the tax on foreign credits all involved zero marginal cost for the loans
outstanding after a one-year term. The first two, however, were heavily on-
erous for very short-term inflows.
Since capital inflows persisted, the reserve requirement was tightened
and extended to almost all international financial transactions. In May 1992
the requirement was raised to 30 percent and was extended to cover time de-
posits in foreign currency, and in July 1995 it was extended to the purchase
of Chilean stocks (secondary ADRs) by foreigners.
34
The term of the deposit
was raised to one year, independent of the maturity of the loan. With some
lag, authorities took measures to eliminate a loophole that made it possible
to circumvent the reserve requirement by means of FDI (since risk capital
was exempt). This was accomplished by scrutinizing FDI applications. Per-
mission for FDI exemptions from reserve requirements was denied when it
was determined that the inflow was disguised financial capital. In these
cases, foreign investors had to register their funds at the Central Bank as fi-
nancial investments subject to the reserve requirement.
Beginning in 1991, measures were put in place to facilitate capital out-
flows as a way of lightening pressure on the exchange rate.
35
Specifically, a
series of incremental changes allowed Chilean pension funds to invest up
iicaioo iiiixcu-oavis axo uiiiniiro raiia
33. Ffrench-Davis (2000, chap. 6); Ffrench-Davis and Ocampo (in this volume); Zahler (1998).
34. It is not difficult to impose reserve requirements on foreign portfolio investments. If funds to
be invested are deposited in a Chilean bank, they remain subject to the reserve requirement there. For
funds not using a Chilean bank as an intermediary, the reserve requirement can be enforced at the
time the asset is registered to a party with a foreign address. Registration with the Central Bank is also
necessary when converting securities into ADRs.
35. See Ffrench-Davis, Agosin, and Uthoff (1995).
to 6 percent of their total assets abroad as of 1995.
36
Expeditious access to
the formal foreign currency market was provided for residents wanting to
invest abroad (chapter XII of the exchange regulations), which had the ad-
ditional benefit of encouraging the development of Chilean firms. This
was effective in stimulating considerable flows of FDI, as well as acquisi-
tions by Chilean firms in neighboring countries.
37
Still, the greater profit-
ability of financial assets in Chile compared to financial investments
abroad, along with expectations of an appreciating exchange rate, discou-
raged foreign investments by pension funds and by the mutual funds cre-
ated at that time.
The immediate effect of deregulating outflows was probably to en-
courage new inflows thanks to the greater certainty that it gave potential
investors coming into Chile. This scenario may actually produce the op-
posite of the desired effect, because the market takes advantage of the
opportunity to move foreign currency abroad when expectations of appre-
ciation are replaced by expectations of depreciation, which is precisely
when the economy is most vulnerable and likely to suffer from speculative
attack.
38
The progressive deregulation of outflows can thus imply a risk of
capital flight not only from the sudden exit of capital that previously came
in, but also from domestic funds seeking to speculate against the peso.
This procyclical feature makes it particularly dangerous to facilitate out-
flows by domestic investors, since it weakens the effect of measures such
as the reserve requirement on inflows, as was proved in 199899.
ixcuaxci iari ioiic\ Exchange rate policy also underwent a substan-
tial shift after the 198283 crisis. From 1984 on, authorities used a crawl-
ing band, which was widened to 5 percent in 1989. The official rate was
devalued daily, according to an estimate of net inflation. Discrete nominal
devaluations were added at various junctures, serving to achieve the not-
able real depreciation of 130 percent between 1982 and 1988.
Chile was thus coming out of the depths of a debt crisis faced with de-
preciation. The real exchange rate had reached historic highs, such that
there was room for some appreciation. However, the economy was mov-
ing from a shortage to a very abundant supply of foreign savings, and au-
thorities wanted to avoid an excessive and overly rapid adjustment in the
caiirai suici xaxacixixr ix cuiii ,
36. It was argued that this would contribute to diversifying risk and increasing the profitability of
the private pension system.
37. Caldern and Griffith-Jones (1995).
38. Williamson (1992); Labn and Larran (1997).
exchange rate.
39
One particularly problematic aspect of the situation in-
volved foreign expectations: as pessimism turns to optimism, foreign in-
vestors tend to rapidly define a new stock of desired investments in the
emerging market; this generates excessive inflows of capital, whose high
levels are naturally transitory, not permanent.
The crawling band was broadened to 10 percent in January 1992.
This produced a wave of revaluatory expectations fed by capital inflows.
The flows were stimulated by the knowledge that the Central Bank would
not intervene within the set band. For many months, a proposal had cir-
culated at the Central Bank to initiate a dirty or regulated float within the
band. Proponents of such intramarginal intervention argued that given an
increasingly active informal foreign exchange market and a more porous
formal market, the current pure band led to an observed exchange rate
with a tendency toward the extremes of the band (the ceiling in 198990,
the floor shortly thereafter). The sudden revaluation of almost 10 percent
in the observed rate between January and February 1992 contributed to
the Banks initiating the dirty float in March. The observed rate then fluc-
tuated within a range of 1 to 8 points above the floor of the bandin
other words, generally off the bottom of the band and always with active
purchases by the Bank, though also with sales from time to time.
The initial broadening of the band had created expectations that the
Bank had renounced the attempt to avoid revaluatory pressures and sus-
tain the export strategy, instead giving the marketwhich was dominated
by short-term participantsthe job of determining the observed rate
within a very wide range. By establishing the regulated float, the Central
Bank regained a greater role, which allowed it to strengthen the long-term
variables that determine the exchange rate facing producers of exportables
and importables.
In subsequent months, U.S. interest rates continued to fall, putting
pressure on the Central Bank. Nevertheless, the Chilean economy had a
notable boom, with a GDP increase firmly in the double digits. The Bank
therefore wanted to raise rather than lower domestic interest rates to pro-
mote macroeconomic balance. To deter arbitrage, it decided to increase the
reserve requirement on capital inflows. The effectiveness of the reserve re-
quirement and its flexible application at that juncture facilitated monetary
policy and avoided accumulation of macroeconomic imbalances.
Finally, in July 1992, the dollar was replaced by a basket of currencies
iicaioo iiiixcu-oavis axo uiiiniiro raiia o
39. See Zahler (1998).
as the currency standard for the exchange rate. This measure was intended
to introduce greater exchange rate uncertainty in short-term dollar-based
operations, thereby reducing incentives for interest rate arbitrage, given
the daily instability governing the international values of the currencies in-
cluded (namely, the dollar, the deutsche mark, and the yen). At the same
time, replacing the dollar with the basket meant greater stability for aver-
age peso values of returns on exports. Unlike financial operations, which
are denominated largely in dollars, trade is geographically diversified and
thus operates with a more diverse mix of currencies.
The actual market performance strongly indicates that the exchange
rate appreciation during this period represented equilibrating movements.
This is consistent both with the disappearance of the effects of the 1980s
crisis, allowing the elimination of the overdevaluation then required, and
with net improvements of productivity for Chilean tradables. A revealed
proof is the fact that the current account deficit was very moderate: 2.5
percent GDP in 199095.
srioxcii naxxixc suiiivisiox The Chilean banking crisis of 1981
86, which followed on the heels of a massive capital surge in the late 1970s,
carried a number of valuable lessons that were reflected in Chilean legisla-
tion.
40
Some of the elements of prudential supervision adopted include the
continuous monitoring of the quality of bank assets; strict limits on banks
lending to related agents; automatic mechanisms to adjust banks capital
when its market value falls beneath thresholds set by regulators; and the
authority to freeze bank operations, prevent troubled banks from transfer-
ring funds to third parties, and restrict dividend payments by institutions
not complying with capital requirements. Chiles financial markets had
also become deeper; this allowed the orderly infusion of new funds, as well
as their withdrawal, without significantly affecting the quality of bank
portfolio.
41
Despite the quality of prudential supervision, however, macroeco-
nomic imbalances that suddenly lead to massive devaluations and very
high interest rates can unexpectedly affect the quality of banks portfolios,
as can exploding bubbles in asset markets. Sustainable macroeconomic
balances are an essential partner of sustainable prudential supervision.
caiirai suici xaxacixixr ix cuiii ;
40. See Daz-Alejandro (1985); Held and Jimnez (2001); Reinstein and Rosende (2000); Valds-
Prieto (1992).
41. C. Larran (1995); Aninat and Larran (1996).
Effectiveness of Policies for Stabilizing Macroeconomic Variables
The Chilean mechanism for prudential macroeconomic regulation at-
tracted considerable international attention, and many studies attempt to
measure its effectiveness.
42
Various tests seek to determine how the reserve
requirement affects the composition and volume of flows, as well as their
impact on the exchange rate, the authorities ability to make monetary pol-
icy, and the aggregate demand. There is ample evidence that Chiles regu-
lations on foreign capital changed the maturity structure of inflows,
reducing the short-term component.
43
This evidence points to a very pos-
itive feature of the instrument, since the liquidity of foreign liabilities is a
major factor in the probability and severity of crises.
44
Disagreement arises, however, on the effect of the overall volume of
flows, since some econometric studies fail to find that these instruments
affected the total volume or exchange rates, though they do find an effect
on the composition of flows.
45
The implication here is a high substitution
between short- and long-term flows. Part of this compensatory phenome-
non between flows of different maturities is to be expected, in the sense
that long-term investors prefer a more stable country (as results from the
reserve requirement). In fact, FDI was predominant in Chile, whereas in
Latin America as a whole it accounted for only one-fourth of total flows
in the first half of the 1990s.
46
The flows, however, do not encompass the
same money or the same investor; it is a different investor with a different
behavior, closely connected to productive investment. On the one hand,
FDI creates new capacity, which contributes to capital formation and also
increases importation, in particular of capital goods. Given the size of the
inflows, therefore, the foreign currency market experiences a smaller excess
supply than in the case of financial inflows. On the other hand, FDI in
Chile has behaved as a permanent variable, while other flows have acted
iicaioo iiiixcu-oavis axo uiiiniiro raiia
42. References to Chilean-style policy for dealing with capital flow instability are frequent in the
main circles where these issues are discussed, for example, by people such as Andrew Crockett, Paul
Krugman, Dani Rodrik, and Lawrence Summers. See, for example, Stiglitz (1998), then vice presi-
dent of the World Bank; see also El Diario, Santiago, March 30, 2000, which quotes Horst Koehler,
the recently appointed managing director of the IMF.
43. De Gregorio, Edwards, and Valds (2000); Edwards (1999b); Le Fort and Lehman (2000);
Schmidt-Hebbel, Hernndez, and Gallego (1999).
44. See, for instance, Rodrik and Velasco (1999); Tapia (2000).
45. De Gregorio, Edwards, and Valds (2000); Edwards, (1999b); Valds-Prieto and Soto (1998).
46. It should be pointed out that loans associated with FDI remained subject to the reserve re-
quirement. Since the average maturity of these loans was around seven years, the cost incidence was
small, while it did minimize the risk that short-term loans would be disguised as long-term flows.
as transitory disturbances.
47
Because FDI is much less volatile than other
types of incoming capital, macroeconomic regulatory policies such as re-
serve requirements that are based on short-term or volatile inflows will en-
courage long-term stability over continued volatility.
Some observers have stated that the effectiveness of measures to dis-
courage capital inflows is only temporary, since private sector agents gen-
erally find ways around such measures.
48
In principle, a number of
loopholes facilitate such evasion. One is to underbill imports or overbill
exports. Another is to delay payment for imports or accelerate export re-
ceipts. A third possibility is to bring in funds through the informal cur-
rency market. Fourth, short-term funds can be registered as FDI, although
this was a costly option since Chilean law required FDI to remain in the
country for at least a year. Nonetheless, this method was becoming an im-
portant loophole, so the authorities agreed to eliminate it. Fifth, business
partners may arrange back-to-back operations in which, for instance, one
party pays for imports with a local bank deposit in Chilean pesos rather
than foreign currency, while the exporter is then paid in foreign currency
through a bank in his or her country. All of these tactics (and others) are
possible, but they all imply costs, and some may have undesirable reper-
cussions on the tax liabilities of those employing them to circumvent re-
serve requirements. Though a certain level of evasion is inevitable, there is
no evidence suggesting large-scale evasion of measures to discourage short-
term capital, as shown by the reserve deposits actually made and the col-
lection of the equivalent fee by the Central Bank.
The most recent studies tend to confirm that the reserve requirement
also reduced total inflows and moderated exchange rate appreciation.
49
Qualitative analysis reinforces the conclusion: Chile confronted a supply
of foreign funds that was greater (in relation to its GDP) than other Latin
American countries, owing to its more attractive economic performance
and its greater political stability in the early 1990s. Nevertheless, exchange
rate appreciation and the current account deficit (as a fraction of GDP or
exports) were smaller, on average, than in the other countries in the region
that received large amounts of foreign capital.
50
All of these studies coincide in their assessment that the reserve re-
quirements maintained an adequate spread between domestic and inter-
caiirai suici xaxacixixr ix cuiii ,
47. Agosin and Ffrench-Davis (2001); Le Fort and Lehmann (2000).
48. See, for example, Valds-Prieto and Soto (1998).
49. Le Fort and Lehmann (2000); Schmidt-Hebbel, Hernndez, and Gallego (1999).
50. See Ffrench-Davis (2000, chap. 10).
national interest rates and thus provided room for monetary policy. This
factor was important in the process of sustained growth seen throughout
the decade, since frequent mini-adjustments by the Central Bank pre-
vented the need for maxi-adjustments and allowed the economy to remain
persistently at or near its productive capacity or production frontier. Ac-
tual output thus coincided with potential output. The resulting perception
of real sustainable stability stimulated capital formation and the growth of
productive capacity and employment.
51
The combination of policies used involved financial costs for the
monetary authority, since accumulation of large volumes of foreign cur-
rency implies a significant cost. This also generates a social cost for the
economy, since the profitability of these assets abroad is naturally less than
interest payments on the Central Bank liabilities issued to sterilize the
monetary effects of accumulating reserves. The losses for the Central Bank
are estimated at 0.5 percent of GDP annually. Evidence shows that disin-
centives to short-term, liquid capital inflows tended to reduce the magni-
tude of the financial costs of sterilization and generated substantial
macroeconomic benefits. Even more flexible and timely management of
the intensity and coverage of the reserve requirement and additional mech-
anisms by the monetary authority would undoubtedly have kept the fi-
nancial costs lower.
52
Furthermore, the existence of high levels of reserves
has frequently been a very important stabilizing factor in emerging econ-
omies facing crisis situations, as was proved in Chile in the next episode.
Domestic Strengths and External Vulnerability: 199697
The previous section examined how prudential policy in the first half of
the 1990s was effective in achieving objectives of macroeconomic sustain-
ability during a capital surge. In 1995, the year in which the tequila crisis
erupted, GDP dropped in Mexico and Argentina by 6.1 percent and 3 per-
cent, respectively, whereas output rose 10 percent in Chile in a great show of
external strength (helped by a positive terms-of-trade shock).
53
Paradoxi-
cally, satisfactory performance was one of the causes of the disequilibria
iicaioo iiiixcu-oavis axo uiiiniiro raiia ,c
51. Agosin (1998); Ffrench-Davis and Reisen (1998).
52. Another potential source of real compensation for the Central Bank losses is the exchange rate
band, which allows the exchange rate board to buy cheap (near the floor) and to sell expensive (near
the top). This would generate profits only if there is not an excessively strong revaluation of the band.
53. A negative impact was felt only on the stock market. See Ffrench-Davis (2000, chap. 10).
built in 199697. In the first place, foreign investors confidence in the
strength of the Chilean economy encouraged them to invest massively,
which created additional pressure on the exchange rate and tested the econ-
omys capacity for efficient absorption. Second, a generalized optimism
characterized the world financial environment, based on the notion that the
international community had been learning how to handle international
crises and that virulent crises were gone forever. One reflection of this
was the attempt by the IMF, at its annual meeting in Hong Kong in Octo-
ber 1997, to achieve a mandate to promote capital account liberalization
among member countries. Lobbying pressure and a strong fashion toward
financial liberalization were enormous factors that were well received in
Chile, where many actors felt themselves immune from the effects of finan-
cial crises. Third, some officials understandably let themselves be swayed by
this euphoric atmosphere, accepting as sustainable a rising external deficit
and the sharp appreciation of the real exchange rate.
Chile thus gave in to the pressures of the dominant international and
domestic environment at the time. It indeed managed to reduce inflation
quickly, but it paid a price in the form of the imbalances this generated.
54
The general policy was kept in place, against the fashion, but the author-
ities failed to strengthen measures in the face of the very abundant supply
of capital during 199697. The combination of policies and the intensity
with which they were applied remained constant. The surge clearly weak-
ened the fundamentals of the Chilean economy: the current account def-
icit rose, the exchange rate appreciated, and the stock of liquid foreign
liabilities grew somewhat. The deterioration certainly could have been
checked during the boom by means of higher reserve requirements and
other measures. Nevertheless, a notorious complacency reigned while the
new boom made the country increasingly vulnerable to external shocks.
The effects of contagion from the Asian crisis began in late 1997 and
were felt strongly in 199899. This time, Chile experienced a sharp drop
in the terms of trade, equivalent to 3 percent of GDP per year. At the same
time, the large 199697 capital inflows gave way to outflows of domestic
and foreign funds, and the nominal exchange rate began to depreciate.
This correction process interrupted the most successful period of eco-
nomic growth recorded in Chiles history, which lasted from 1991 to 1998.
caiirai suici xaxacixixr ix cuiii ,:
54. An overwhelming proportion of operators pushed for greater deregulation and maintained that
larger external deficits were sustainable. Following the crisis, most of them blamed what they called
excessive fiscal spending as the source of the subsequent recessive adjustment. This hypothesis has no
empirical support, however, as shown below.
iicaioo iiiixcu-oavis axo uiiiniiro raiia ,:
The New Wave of Capital
As shown in figure 3-2 above, the tequila crisis led to a considerable re-
duction in net capital inflows to Chile. This is mostly explained by the be-
havior of portfolio flows, which entirely disappeared in Chile, as in the rest
of Latin America. Despite the reduction in the most volatile flows, the con-
fidence of long-term investors was striking, as reflected in increased FDI.
Capital flows returned to the region shortly after the tequila crisis. Net
inflows to Latin America grew from U.S.$30 billion in 1995 to U.S.$80
billion in 1997. Of this, a larger share than in the preceding years went to
Chile. The image of a successful country on solid footing attracted foreign
investors in massive quantities. In 1997, net inflows climbed to more than
10 percent of GDP.
The foreign capital boom of 199697 involved some changes in com-
position. FDI remained high. Portfolio flows increased sharply, however,
despite the fact that these were subject to the reserve requirement. The
progressive increase of investment abroad, which started at the beginning
of the decade, reached considerable levels. Some of this investment was in-
creasingly funded by borrowing abroad. At the same time, the main insti-
tutional investorsthe private pension fundshad placed only slightly
over 0.5 percent of their assets abroad by mid-1997.
In 1998, for the third time in sixteen years, financial flows to Latin
America underwent a drastic reversal and actually became negative. The
crisis was ameliorated by the presence of FDI, however. In 1982, most of
the loans were bank loans with variable interest rates. This time, the bulk
of flows were from FDI, which tends to be a more friendly component in
periods of crisis, and these remained high in Chile. The behavior of FDI
also tends to be anticyclical, reducing the amount of profit that can be sent
abroad. In fact, FDI remittances fell from U.S.$1.9 billion in 1997 to
U.S.$1.1 billion in 1999.
Investment abroad intensified even more. Devaluatory expectations
(starting in late 1997) triggered a massive exit of funds through those open
channels. Pension fund outflows, for example, increased notably starting
in late 1997, when expectations shifted from appreciation to depreciation.
Outflows from January 1998 to June 1999 peaked at the equivalent of 4.8
percent of 1998 GDP, or 12 percent of the funds.
55
The 10 percent drop
55. A hurried financial deregulation runs the risk of leaving too many exit doors open, which can
be a massive oversight when the market is nervous and expectations are shifting to the devaluatory side
(as was duly advised in Ffrench-Davis, Agosin, and Uthoff, 1995). This tends to make it more diffi-
in aggregate demand, and the recession it generated, were partly associated
with the monetary contraction resulting from additional devaluating pres-
sures caused by these outflows.
Policy after the Success of 1995
The set of policies available to authorities in the second half of the
1990s was relatively unchanged, including responsible surplus-based bud-
get management. However, the use of such tools had strongly different fea-
tures compared with the previous episode, and the economic approach
thus constituted a different variety of policy. In particular, authorities
demonstrated a marked passivity at the time that the imbalances were gen-
erated, and their highly risky nature was underestimated. A proof is that
the target for a sustainable current account deficit was raised from 3.5 per-
cent to 5 percent of GDP. This contrasts with the active management of
the previous capital boom of 199194, and it highlights the fact that the
survival of macroeconomic balances requires a highly pragmatic and pru-
dent approach to episodes of financial euphoria.
As in the previous episodes, the cycle closed with negative external
shocks. This time, the shock came from Asia and was very intense, but the
breadth of its repercussions was closely linked to the Chilean economys
degree of exposure.
ioiiciis ioi iicuiarixc caiirai ixiiows An examination of capital
inflow regulation does not reveal significant changes, despite strong pres-
sures by lobbyists pushing for full liberalization of the capital account. Not
bending to those pressures deserves clear recognition, given the world en-
vironment of financial euphoria. The reserve requirement stayed at 30 per-
cent with a one-year term for much of the period, with only minor
adjustments.
56
Still, there was a radical shift in how the problem was ap-
proached. As analyzed above, the authorities made explicit efforts in the
caiirai suici xaxacixixr ix cuiii ,,
cult to sustain exchange rate and macroeconomic stability, and it makes international financial crises
more painful. Such was the case in Chile in 199899, when the pension funds and chapter XII in-
vestors were responsible for the capital flight that worsened the external imbalance. Curiously, analysts
have practically ignored this highly significant fact.
56. In June 1996, the possibility of renewing a trade credit more than once a year was eliminated
to close a loophole. A threshold (U.S.$200,000) under which small transactions were not subject to
reserve requirements was set in December 1996 to help reduce the operating costs of the instrument.
This exemption became a source of evasion, however, and then the threshold was lowered to
U.S.$100,000 in March 1997.
first half of the 1990s to address the stability of key macroeconomic prices
and not to create vulnerability to eventual external shocks; these efforts in-
cluded limiting the growth of the current account deficit and restraining
currency revaluation.
57
The external accounts and the exchange rate were
monitored continuously, and there was ample willingness to refine the
tools by increasing their coverage and evaluating their effects. The various
instruments used were not ends in themselves, but only means to achiev-
ing the objectives mentioned.
The approach was rather different after the tequila crisis. Chiles
strength in the tequila crisis sent a signal of invulnerability to many. The
perception of Chile as solid and different from the rest of Latin America
provoked doubt about the need to use the prudential macroeconomic in-
struments, which many analysts considered inefficient or unnecessary for
a modern country. The result was a complacent attitude on the part of the
authorities, who failed to prevent the development of evident imbalances
in 199697, on account of the abundant inflows, and then took no meas-
ures to solve them when it was possible to do so, namely, during the boom.
On the contrary, the prevailing trend was to modernize the management
of the capital account and to take advantage of currency appreciation in
the fight against inflation, as had been done two decades earlier.
As a part of the move toward financial integration, measures were taken
to facilitate the investment of Chilean capital abroad. The percentage of for-
eign investment authorized to pension funds was raised from 6 percent in
1995 to 12 percent in April 1997, and new alternatives for such investment
were opened up.
58
Rather than discouraging entry, the changes encouraged
exit. This was supposed to diversify risk. Given the extremely optimistic ex-
pectations about the national economy and the opportunities for profit that
the domestic markets offered in 199697, the outflows of funds were mod-
est until expectations shifted to devaluation (see figure 3-3). While the
changes did allow the funds to diversify risk as intended, they clearly in-
creased risk for the economy as a whole, as demonstrated in 199899.
Throughout this period, the Central Bank did not dismantle policies.
The government, for its part, forcefully defended the reserve requirement
in negotiations of a free trade agreement with Canada. When the Asian
crisis began to be strongly felt, portfolio and other short-term inflows
plummeted. The Central Bank confronted this phase of the cycle, com-
iicaioo iiiixcu-oavis axo uiiiniiro raiia ,
57. See Ffrench-Davis, Agosin, and Uthoff (1995) for more detail.
58. The limit was raised to 16 percent in January 1999.
Figure 3-3. Chilean Pension Fund Outflows and Real Exchange Rate
Source: Central Bank of Chile; Superintendencia de Administradoras de Fondos de Pensiones de Chile (SAFP).
104
99
94
89
84
79
74
16
14
12
10
8
6
4
2
Pension funds external portfolio
(percent of total)
Real exchange rate
(1986 100)
June
1994
Dec.
1994
June
1995
Dec.
1995
June
1996
Dec.
1996
June
1997
Dec.
1997
June
1998
Dec.
1998
June
1999
Dec.
1999
Devaluation zone Appreciation zone
Real
exchange
rate
(right axis)
Limit to
external
portfolio
External
portfolio
(left axis)
bined with a depressed external capital supply, by lowering the reserve re-
quirement to 10 percent in June 1998 and then to 0 percent in Septem-
ber. Authorities stated, however, that this was only a resetting of the
parameters and that its use would continue to be relevant for confronting
renewed capital surges.
59
ixcuaxci iari ioiic\ The plentiful supply of capital flows after 1995
kept the exchange rate close to the floor of the band until the end of
1997. Given the overwhelming expectations of appreciation following the
tequila shock, the large spread between the expected profitability of peso-
denominated and dollar-denominated funds (along with a good outlook
for large Chilean firms) offered foreign portfolio and short-term investors
a potentially very lucrative opportunity despite the reserve requirement
they paid to enter the domestic financial market.
Although the Central Bank formally adhered to a crawling band in
199697, the flexibility of the exchange rate was, in fact, increasing. Thus
the Bank, as well as most observers, gradually moved toward one of the
caiirai suici xaxacixixr ix cuiii ,,
59. Massad (2000).
two corner solutions that was increasingly in fashion: full foreign exchange
flexibility. For instance, to lower the floor of the band (beyond a formal
broadening of the band to 12.5 percent in early 1997), authorities made
contradictory adjustments in the weights assigned to the currency basket,
causing a loss of credibility for the basket.
60
The external inflation used to
calculate the referential exchange rate was overestimated, which generated
considerable additional revaluation of 10 percentage points between 1995
and 1997. Furthermore, an annual 2 percent appreciation of the reference
rate had been set in November 1995, based on the assumption that Chi-
lean productivity would grow more rapidly than that of its main trading
partners. At the same time, however, the Bank accumulated significant
amounts of international reserves.
When exchange rate expectations shifted in Chile in late 1997, the
Central Bank operated asymmetrically. It now strongly held back depre-
ciating pressures by selling abundant reserves, thus preventing a correction
of the exchange rate value. The authorities argued that accepting a sharp
devaluation would be significantly inflationary in an economy operating
on the productive frontier and with a high external deficit. In our view, an
ambitious anti-inflationary objective was again predominant. In mid-
1998, the band was drastically shortened right at the moment of greatest
uncertainty, and the macroeconomic adjustment process was led by inter-
est rate hikes; this intensified the monetary effects of outflows. The Central
Bank was sending a signal that it would not give in to devaluatory pressures
in the market. The band was widened again at the end of 1998 and then
suspended in September 1999 to allow the exchange rate to adjust, now in
the context of strongly depressed domestic absorption (see table 3-5).
The Consequences of a Prone-to-Risk Policy
Macroeconomic policy in 199697 proved unable to prevent the crea-
tion of imbalances in the external sector. As discussed above, the financial
environment was characterized by an overoptimistic mood of success and
invulnerability. Most participants seemingly ignored the danger of a finan-
cial crisis, which is always present in todays volatile world.
The economy took in large inflows in the form of FDI, which sur-
iicaioo iiiixcu-oavis axo uiiiniiro raiia ,o
60. In November 1994, the weight of the dollar was reduced from 50 percent to 45 percent, re-
flecting the declining influence of that currency in Chile trade. It was arbitrarily raised to 80 percent
in January 1997. For a comparative analysis of bands in Chile, Israel, and Mexico, see Helpman, Lei-
derman, and Bufman (1994). On Chile, Colombia, and Israel, see Williamson (1996).
passed all previous records when they reached 6.8 percent of GDP in
199697. The signs of solvency and stability that Chile exhibited in the
tequila crisis no doubt contributed to attracting these funds. The supply
to emerging economies also underwent a significant positive shift that
heightened the country-specific trend. At the same time, portfolio flows
experienced unusual growth despite the cost of the reserve requirement.
The evidence thus indicates that entry to the market was considered cheap
in comparison with the very positive economic parameters and a strong
likelihood that the real exchange rate would appreciate as a result of the
large supply of foreign capital.
61
This is a classical case of an imbalance
creating bubbles in the exchange rate market and in aggregate demand.
The Central Bank made heavy purchases of foreign currency, but it
was unable to avoid a very significant real exchange rate appreciation (16
percent between the 1995 average and October 1997). The appreciation
and the creation of liquidity by the inflows, in turn, sharply stimulated ag-
gregate demand and biased it towards tradables. The current account def-
icit thus rose to 5.7 percent of GDP. Key macroeconomic prices had
clearly become outliers, and instruments for dealing with excessive finan-
cial flows needed to be strengthened. The Central Bank, however, declared
itself unable to fulfil its commitment to defending a more stable and com-
caiirai suici xaxacixixr ix cuiii ,;
61. See Herrera and Valds (1997). Country risk was again reduced; the Euromoney index, for in-
stance, improved from 72.9 in 1995 to 79.0 in 1997.
Table 3-5. Asian crisis in Chile, 1996-99
Indicator 1996 1997 1998 1999
GDP
a
7.4 7.4 3.9 -1.1
Aggregate expenditure
a
7.9 9.1 3.9 -10.0
Fiscal expenditure
a
9.6 6.1 6.3 5.7
Exports
a
11.8 9.4 5.9 6.9
Imports
a
11.8 12.9 5.4 -14.3
Current account deficit
b
5.8 5.7 6.2 0.2
Real exchage rate
c
84.7 78.2 78.0 82.4
Terms-of-trade effect
d -
3.4 0.6 -3.0 0.2
Source: Based on data from the Central Bank of Chile; DIPRES (2000); Ffrench-Davis (2001).
a. Official figures at 1986 constant prices, annual growth (percent).
b. Current prices, percent of GDP.
c. Index, 1986 = 100.
d. Current prices, percent of GDP.
petitive long-term exchange rate.
62
Instead, it gave priority to the inflation-
ary goal, which was facilitated by the appreciating trend.
The situation called for a demand-reducing adjustment focused on
the control of the external imbalance. Given that the large supply of cap-
ital inflows was the source of the disequilibrium, measures were needed to
increase the cost of inflows, to devalue the real exchange rate, and not only
to raise the cost of domestic credit (that is, the interest rate). The mainte-
nance of a constant rate and coverage of the reserve requirement, without
the implementation of additional measures, became insufficient for deal-
ing with the massive capital inflow in 199697.
63
If the prudential regu-
lation of inflows had been strengthened in 199697, it would have
accomplished the triple goal of regulating the growth of the stock of liq-
uid external liabilities, softening appreciation, and moderating the increase
in aggregate expenditure.
A controversial issue involves the level of fiscal responsibility for the
excess aggregate demand of 199697. An expansionary fiscal policy was
recorded in the period, with government expenditure rising somewhat
faster than GDP (7.9 versus 7.4 percent a year). However, several points
have to be taken into account to fairly assess this fact. First, the fiscal ex-
penditure with macroeconomic effects represents only one-fifth of the
economy. The large majority of the pulls behind the 8.5 percent growth
in domestic aggregate demand in the period were in the private sector,
which accounted for 90 percent of the expenditure increase. A fiscal con-
tribution to moderating total expenditure would thus have been insuffi-
cient. Second, the main components in the increased fiscal expenditure
were education, justice, and infrastructure. All three areas underwent
major and widely demanded transformations, with a political consensus
around increasing expenditure on them. Third, the fiscal budget showed
a surplus over 2 percent GDP and, consequently, overfinanced its expen-
diture. Moreover, the government stopped borrowing from the World
Bank and the Inter-American Development Bank (IDB) and prepaid debt.
Fourth, the measured fiscal surplus does not include resources accumu-
lated in the Copper Buffer Fund.
Here again, imbalances were externally generated and overwhelmingly
private. The governments responsibility in this case lay in its failure to
iicaioo iiiixcu-oavis axo uiiiniiro raiia ,
62. See, for example, the president of the Central Banks statements in the Chilean business news-
paper Estrategia, October 16, 1997.
63. Le Fort and Lehmann (2000, p. 33) offer a similar assessment. Le Fort was at that time the di-
rector of the Central Banks International Division.
enforce coordination between the Central Bank and the Ministry of
Finance.
64
This shortcoming was related to the Central Banks autonomy.
Authorities need to recognized that there is not a single form of autonomy
in the world, but rather several alternative ones.
As a result of the lack of timely and sufficiently strong measures, the
Asian crisis found Chile with a significantly appreciated exchange rate, an
overstimulated aggregate demand, and a high current account deficit that
was double the 199095 average. This all added up to a level of vulner-
ability that was unprecedented in the decade, and Chile received a strong
trade shock. Nevertheless, if the reserve requirement had not been main-
tained, portfolio inflows would have been much greater, the overall flow
more significant, and the exchange rate appreciation more marked.
65
Financial capital began to flow out in late 1997 and accelerated in
199899. The nominal exchange rate depreciated to correct for misalign-
ment.
66
This time, the impact of the exit of foreign capital was aggravated
by the outflows associated with pension funds, which intensified the de-
valuating pressures on the exchange rate. The channels that had been pro-
gressively opened up over the course of the decade, under the argument
they would moderate the abundance of foreign currency in boom periods
and diversify risk, were effectively used only during the bust. Thus the ob-
jective was not achieved during the boom. Rather, the mechanism caused
a significant loss of international reserves during the crisis, a monetary
contraction, and a sharpening of the recessionary adjustment in 199899.
The cumulative current account deficit was moderate during the
decade thanks to the active management of inflows in the first half of the
1990s and the persistence of regulations in the following years with only
gradual liberalization. The stock of foreign liabilities was relatively low,
and volatile funds played only a minor role, creating external imbalances
only in 1996 and 1997. These conditions, together with the countrys con-
siderable international reserves, put Chile on a better footing than in the
caiirai suici xaxacixixr ix cuiii ,,
64. The lack of coordination between the Central Bank and the government was evident. As men-
tioned, the Central Bank authorities expressed no concern about imbalances in the external sector,
while the Minister of Economics, for instance, held that it is necessary to intensify and strengthen
policies such as the reserve requirement to reduce exchange rate appreciation. Estrategia, September
26, 1997.
65. Schmidt-Hebbel, Hernndez, and Gallego (1999); Le Fort and Lehmann (2000).
66. The real exchange rate reached its peak for 1999 in November. This depreciation represented a
movement toward equilibrium, in correction of the disequilibrium of 199697. The real exchange rate
actually recovered its 1995 average level; this real rate tended to prevail in the second semester of 2000,
after strong swings during the first semester.
previous crises for confronting the hardships of trade and financial shock
caused by the Asian recession. The domestic financial system did not suf-
fer radically, as a result of the strictness of the Chilean banking commis-
sion. Nonperforming loans as a percentage of total loans rose from 0.97
percent in December 1997 to 1.8 percent at its worst moment in April
1999a level comparable to 1992, a crisis-free year.
67
This is remarkable
given that aggregate demand fell by 10 percent.
Still, the social and economic costs were significant. Production under-
utilization in 1999 fell by about U.S.$7 billion, opening a gap of approxi-
mately 10 percent between actual GDP and productive capacity that year.
Together with the increase in interest rates, this gap caused gross fixed in-
vestment to fall 17 percent, jeopardizing the potential GDP growth of 7
percent on average. Unemployment exceeded 10 percent of the labor force
in 19992000, which reversed some of the progress that had been made in
reducing poverty and inequality since 1990.
68
Concluding Remarks
The rich experience of Chilean policy in managing foreign capital
surges throws light on key principles and specific instruments for exploit-
ing the benefits and minimizing the costs of financial globalization. De-
spite the extent to which globalization reduces the space for action, there
is still room for each country to foster its objectives. In this sense, a coun-
try that falls victim to a financial crisis of external origin is never blame-
less. We have solid evidence that market imbalances in the exchange rate,
high current account deficits, and excessive shares of short-term or liquid
funds make a small, open economy vulnerable in a crisis. The challenge,
then, is to fight worsening in these parameters.
Exchange rate policy has proved to be one of the key elements in the
development of recent crises in Chile. Freezing the exchange rate as part of
a price stabilization policy in the late 1970s and keeping it frozen for three
years entailed a very significant outlier real price of the dollar, based on an
unsustainable level of capital inflows. The discrepancy stimulated and fi-
nanced a high current account deficit up to 1981, and a correction was to
be expected. When it occurred in 1982, the economic and social results
iicaioo iiiixcu-oavis axo uiiiniiro raiia :cc
67. SBIF (2000).
68. See Ffrench-Davis (2001, chap. 9).
were very negative: GDP fell 17 percent, and the unemployment rate
jumped to over 30 percent. In the 1990s, policy aimed at preventing a re-
peat performance. The use of a managed crawling band reduced exchange
rate volatility while moderating its appreciation up to the mid-1990s. This
supported the target of maintaining an exchange rate that was consistent
with the export model, at the expense of the objective of reducing inflation
more quickly. The approach was successful in the first-half of the 1990s.
69
After the tequila crisis, however, the Central Bank did not check the large
appreciation. Monetary authorities gave precedence to achieving more am-
bitious inflation goals despite an increased current account deficit and an
excessive aggregate demand, thus repeating risky behavior of the past.
A curious fact of the Chilean case is that the three financial boom epi-
sodes generated different outcomes with the consistent use of certain tools
to regulate capital inflows, in particular, an unremunerated reserve require-
ment on short-term capital. The main lesson is that an instrument that al-
ters relative prices in the market, as does the reserve requirement, cannot
remain unchanged in the face of a changing supply of foreign funds. Fur-
thermore, a policy of sustainable macroeconomic balances must not rest
solely on one inflexible instrument, excellent though the instrument may
be. A reserve requirement was clearly ineffective in saving the economy
from the debt crisis in 1982, while in 199697 its impact was insufficient
for staving off contagion from the Asian crisis. However, one might ask
what would have happened if it had not been applied in those cases. Em-
pirical evidence indicates that the imbalances would have been worse, es-
pecially with regard to the Asian crisis. The effectiveness of this sort of
instrument depends, first, on the values set for its parameters and the de-
gree of coverage, and, second, on complementary policies that support the
strategy, as in the first half of the 1990s. These policies include fiscal re-
sponsibility and effective prudential supervision of the financial system.
Regulatory instruments such as a reserve requirement have costs.
70
Quasi-fiscal losses stemming from excessive accumulation of reserves and a
lower GDP growth rate owing to higher interest rates are those most often
mentioned. A serious assessment requires weighing these effects against
their macroeconomic benefits. The benefits are extensive, deriving from
caiirai suici xaxacixixr ix cuiii :c:
69. The economys performance in that period was quite satisfactory, with dynamic exports and a
gradual but large reduction of inflation. Sebastin Edwards summarizes this point well: Given the rel-
ative success of the Chilean band system, it is surprising that more countries have not adopted this
type of regime (1999b, p. 37).
70. Schmidt-Hebbel, Hernndez, and Gallego (1999) summarize and measure these costs.
the notable imperfections of international financial markets (including
high volatility and information asymmetry). By helping to counterbalance
these market imperfections, measures to drive capital flows at a sustainable
level and moderate exchange rate instability reduce both the likelihood of a
crisis and its potential costs. They are also associated with gains in growth
because of their positive impact on resource allocation and export develop-
ment. Finally, economic growth benefits enormously when monetary au-
thorities are given room to regulate spending and smooth the cyclical
behavior of the economy. These benefits are difficult to quantify, since the
moderation of cyclical behavior not only creates a better environment for
productive investment and growth, but also favors social development by
improving conditions of poverty, income distribution, and welfare pro-
grams, a dimension which is highly sensitive to economic fluctuations.
71
In liberalizing the capital account, a dose of prudence is also in order
with regard to outflows. Observers and lobbyists argue that the funds are
too big for Chile and that profitability is increased by investing abroad.
Chile is a capital-scarce country, however, and the return of capital tends to
be higher in Chile than in developed countries. Hence, it is questionable
whether funds saved by Chileans and held by pension funds should be en-
couraged to move abroad rather than preferentially invested domestically.
The issue is particularly controversial when macroeconomic risk diversifi-
cation is achieved at the expense of macroeconomic sustainability because
of its procyclical nature. This issue requires further discussion and analysis.
Another of the dimensions that make Chile an interesting case study
is the development of its domestic financial institutional structures. The
deregulation of banking activity in the absence of a proper framework of
prudential regulation in the 1970s was, without a doubt, a destabilizing
element that contributed to an overborrowing syndrome.
72
Moreover,
progress made to address the weakness contributed to the development of
one of the healthiest financial markets among emerging countries and fos-
tered macroeconomic stability throughout the 1990s. It would be a mis-
take, however, to emphasize this dimension as the sole explanation of
financial crisis in Chile. In the three episodes examined here, the domes-
tic financial market operated as the channel for only a portion of the funds
coming into the country. This was particularly relevant in the 1990s, when
bank loans constituted a minority of inflows. As the events of 199899
iicaioo iiiixcu-oavis axo uiiiniiro raiia :c:
71. See Rodrik (2001).
72. McKinnon and Pill (1997).
showed, the high quality of Chiles domestic financial market did not pre-
vent the country from becoming quite vulnerable to external financial
shocks. This point must be stressed, given the persistence of the danger-
ous hypothesis that efficient banking supervision is enough to impede the
absorption of excessive inflows.
The fact that fiscal policy did not play a leading role in managing the
Chilean crises may seem unusual, especially considering the broad litera-
ture that explains external sector crises as a result of unsustainable budget
deficits. In the three Chilean episodes discussed here, however, the fiscal
situation was solid, with surpluses each year.
73
Deficits were registered only
in 1982 and 1999, and these were the result of the crisis conditions, not
the cause. They were thus associated with external disequilibria during the
previous booms. Once more, the lesson is not to trust in a single funda-
mental of the economy as adequate protection.
The challenge of making good use of foreign capital in emerging econ-
omies, ensuring the healthy domestic absorption of funds, and keeping
vulnerability to external factors low requires a comprehensive set of meas-
ures. It is indispensable to take an active, pragmatic approach to checking
the development of imbalances, to respond to shifts in domestic and ex-
ternal variables, to work in close coordination with a solidly structured
financial sector, to implement a foreign exchange policy that prevents ex-
cessive misalignment of the real rate, to design fiscal and monetary poli-
cies that keep a handle on spending, and to take a flexible approach to
managing the capital account so that measures taken are consistent with
the realities of a developing country where capital is not only scarce in ab-
solute terms but minuscule in comparison with the volume of capital that
can flood a country through an imperfect and volatile international finan-
cial market. Chiles experience shows that all of this is possible and that the
challenge is ongoing. New problems will always arise, and dangerous
temptations can exact a heavy price tomorrow for todays success.
References
Agosin, M. R. 1998. Capital Inflows and Investment Performance: Chile in the 1990s.
In Capital Flows and Investment Performance: Lessons from Latin America, edited by
R. Ffrench-Davis and H. Reisen. Paris: Organization for Economic Cooperation and
Development (OECD), Development Center.
caiirai suici xaxacixixr ix cuiii :c,
73. Larran (1991); DIPRES (2000).
Agosin, M. R., and R. Ffrench-Davis. 2001. Managing Capital Inflows in Chile. In
Short-Term Capital Flows and Economic Crises, edited by S. Griffith-Jones, M. F. Montes,
and A. Nasution. WIDER/Oxford University Press.
Aninat, E., and C. Larran. 1996. Flujos de capitales: lecciones a partir de la experiencia
chilena. Revista de la CEPAL 60 (December).
Arellano, J. P., and R. Cortzar. 1982. Del milagro a la crisis: algunas reflexiones sobre el
momento econmico. Coleccin Estudios CIEPLAN 8 (July).
Caldern, A., and S. Griffith-Jones. 1995. Los nuevos flujos de capital extranjero en la
economa chilena: acceso renovado y nuevos usos. Serie Desarrollo Productivo 24. San-
tiago: United Nations Economic Commission for Latin America and the Caribbean.
Calvo, G., L. Leiderman, and C. Reinhart. 1993. Capital Inflows and Real Exchange Ap-
preciation in Latin America: The Role of External Factors. IMF Staff Papers 40(1).
Cortzar, R., and J. Marshall. 1980. Indice de precios al consumidor en Chile: 197078.
Coleccin Estudios CIEPLAN 4 (November).
De Gregorio, J., S. Edwards, and R. Valds. 2000. Controls on Capital Inflows: Do They
Work? Working Paper 7645. Cambridge, Mass.: National Bureau of Economic Research.
Daz-Alejandro, C. F. 1985. Goodbye Financial Repression, Hello Financial Crash. Jour-
nal of Development Economics 19(1/2).
DIPRES (Direccin de Presupuestos del Ministerio de Hacienda). 2000. Estadsticas de las
finanzas pblicas: 19901999. Santiago.
Edwards, S. 1999a. Capital Flows to Latin America. In International Capital Flows, ed-
ited by M. Feldstein. University of Chicago Press for the National Bureau of Economic
Research.
. 1999b. How Effective Are Capital Controls? Journal of Economic Perspectives
13(4): 6584.
Edwards, S., and A. Cox Edwards. 1987. Monetarism and Liberalization: The Chilean Ex-
periment. Cambridge, Mass.: Ballinger.
Ffrench-Davis, R. 1984. International Private Lending and Borrowing Strategies of De-
veloping Countries. Journal of Development Planning 14.
. 2000. Reforming the Reforms in Latin America: Macroeconomics, Trade, Finance.
Macmillan/Palgrave.
. 2001. Economic Reforms in Chile: From Dictatorship to Democracy. University of
Michigan Press.
Ffrench-Davis, R., M. R. Agosin, and A. Uthoff. 1995. Capital Movements, Export Strat-
egy and Macroeconomic Stability in Chile. In Coping with Capital Surges. The Return
of Finance to Latin America, edited by R. Ffrench-Davis and S. Griffith-Jones. Lynne
Rienner.
Ffrench-Davis, R., and J. P. Arellano. 1981. Apertura financiera externa: la experiencia
chilena en 197380. Coleccin Estudios CIEPLAN 5 (July).
Ffrench-Davis, R., and J. De Gregorio. 1987. Orgenes y efectos del endeudamiento ex-
terno en Chile: antes y despus de la crisis. El Trimestre Econmico 213 and 214 (Janu-
ary-March). Mexico City.
Ffrench-Davis, R., and S. Griffith-Jones, eds. 1995.Coping with Capital Surges. The Return
of Finance to Latin America. Lynne Rienner.
iicaioo iiiixcu-oavis axo uiiiniiro raiia :c
caiirai suici xaxacixixr ix cuiii :c,
Ffrench-Davis, R., and H. Reisen, eds. 1998. Capital Flows and Investment Performance:
Lessons from Latin America. Paris: Organization for Economic Cooperation and Devel-
opment (OECD), Development Center.
Harberger, A. 1985. Observations on the Chilean Economy, 197383. Economic Devel-
opment and Cultural Change 33 (April).
Held, G., and F. Jimnez. 2001. Liberalizacin, crisis y reforma del sistema financiero chi-
leno. Ford Foundation.
Helpman, E., L. Leiderman, and G. Bufman. 1994. A New Breed of Exchange Rate
Bands: Chile, Israel and Mexico. Economic Policy 9(19).
Herrera, L. O., and R. Valds. 1997. Encaje y autonoma monetaria en Chile. Santiago:
Central Bank of Chile. Unpublished.
Labn, R., and F. Larran. 1997. Can a Liberalization of Capital Outflows Increase Net
Capital Inflows? Journal of International Money and Finance 16(3): 41531.
Larran, C. 1995. Internacionalizacin y supervisin de la banca en Chile. Estudios Pbli-
cos 60 (Spring). Santiago.
Larran, F. 1991. Public Sector Behavior in a Highly Indebted Country: The Contrasting
Chilean Experience, 197085. In The Public Sector and the Latin American Crisis, ed-
ited by F. Larran and M. Selowsky. San Francisco: ICS Press.
Le Fort, G., and S. Lehmann. 2000. El encaje, los flujos de capitales y el gasto: una eval-
uacin emprica. Working Paper 64. Santiago: Central Bank of Chile.
Massad, C. 2000. Capital Flows in Chile: Changes and Policies in the Nineties. In Fi-
nancial Globalization and the Emerging Economies. Santiago: United Nations Economic
Commission for Latin America and the Caribbean (ECLAC) and Jacques Maritain
International Institute.
McKinnon, R., and H. Pill. 1997. Credible Economic Liberalization and Overborrow-
ing. American Economic Review 87(2).
Mizala, A. 1992. Las reformas econmicas de los aos setenta y la industria manufactu-
rera. Coleccin Estudios CIEPLAN 35 (September).
Morand, F. 1992. The Dynamic of Real Asset Prices, the Real Exchange Rate, Trade Re-
forms and Foreign Capital Inflows: Chile, 197689. Journal of Development Economics
39(1):11139.
Morand, F., and K. Schmidt-Hebbel, eds. 1998. Del auge a la crisis de 1982: ensayos sobre
liberacin financiera y endeudamiento en Chile. Santiago: Instituto Interamericano de
Mercados de Capital and Georgetown University/ILADES.
Reinstein A., and F. Rosende. 2000. Reforma financiera en Chile. In La transformacin
econmica de Chile, edited by F. Larran and R. Vergara. Santiago: CEP.
Rodrik, D.2001. Why Is There So Much Economic Insecurity in Latin America? CEPAL
Review 73 (April).
Rodrik, D., and A. Velasco. 1999. Short-Term Capital Flows. Working Paper 7364. Cam-
bridge, Mass.: National Bureau of Economic Research.
Sanhueza, G. 1999. La crisis financiera en los aos ochenta en Chile: anlisis de sus sol-
uciones y su costo. Economa Chilena 2(1). Santiago: Central Bank of Chile.
SBIF (Superintendencia de Bancos e Instituciones Financieras de Chile). 2000. Evolucin
de la banca en el tercer trimestre de 2000. Santiago. Mimeographed.
Schmidt-Hebbel, K., L. Hernndez, and F. Gallego. 1999. Capital Controls in Chile: Ef-
fective? Efficient? Working Paper 59. Santiago: Central Bank of Chile.
Stiglitz, J. 1998. The Role of the Financial System in Development. Paper Prepared for
the Fourth Annual Bank Conference on Development in Latin America and the Carib-
bean, San Salvador. World Bank (June).
Tapia, D. 1979. Apertura al mercado financiero internacional. In Institucionalidad
econmica e integracin financiera con el exterior. Santiago: Instituto de Estudios Banca-
rios.
Tapia, H. 2000. Liquidez de los capitales externos y probabilidad de crisis. Unpublished.
Valds-Prieto, S. 1992. Financial Liberalization and the Capital Account: Chile,
197484. In Financial Reform: Theory and Experience, edited by E. Caprio and J. Han-
son. Cambridge University Press.
Valds-Prieto, S., and M. Soto. 1998. New Selective Capital Controls in Chile: Are They
Effective? Empirica 25(2).
Williamson, J. 1992. Acerca de la liberalizacin de la cuenta de capitales. Estudios de
Economa 19(2). Santiago: Universidad de Chile.
. 1996. The Crawling Band as an Exchange Rate Regime: Lessons from Israel, Chile,
and Colombia. Washington: Institute for International Economics (IIE).
Zahler, R. 1980. Repercusiones monetarias y reales de la apertura financiera: el caso chi-
leno, 197578. Revista de la CEPAL 10 (April).
. 1998. El Banco Central y la poltica macroeconmica de Chile en los aos no-
venta. Revista de la CEPAL 64 (April).
iicaioo iiiixcu-oavis axo uiiiniiro raiia :co
4
:c;
I
n the early 1990s, after almost a decade of economic
decline and high inflation, the Mexican economy
appeared to be on its way to recovering economic growth and price stabil-
ity. Privatization revenues and a Brady agreement had facilitated a large re-
duction in domestic and external public debt, and a successful stabilization
program was expected to finally allow the economy to take advantage of
the structural reforms undertaken since the mid-1980s. Or so it appeared
to the many observers who believed that Mexico, a model reformer and
successful emerging market, was to become a Latin American economic
miracle. Optimistic expectations became even more widespread when the
North American Free Trade Agreement (NAFTA) was approved in 1993.
The optimistic view was further reinforced by the fact that the Mexican
government had complemented wide-ranging economic reforms with an
anti-poverty program aimed at meeting the countrys pressing needs in
health and education infrastructure. To many, then, Mexico appeared to
be on a firm path toward economic and social modernization.
By the end of 1994, however, the Mexican economy was in a finan-
cial crisis and entering its worst recession since the Great Depression of the
;aixi ios
*
From the Capital Surge to
the Financial Crisis and
Beyond: The Mexican
Economy in the 1990s
*I am grateful to Maiju Perala for research assistance and to Ricardo Ffrench-Davis and participants
at seminars held at the United Nations Economic Commission for Latin America and the Caribbean
(ECLAC) in Santiago, Chile, for helpful comments.
1930s. The Mexican government was unable to roll over its debt follow-
ing the December devaluation of the peso. Amid surging fears of default,
an unprecedented international rescue package was implemented in early
1995 to stop panic selling of Mexican debt. The country also experienced
political turmoil and violence throughout 1994: an armed uprising by the
Zapatistas in January (on the day NAFTA came into effect), the assassina-
tion of the presidential candidate in March and of the Secretary General
of the Partido Revolucionario Institutional (PRI) in September, and the
resignation in November of the Assistant Attorney General who had been
in charge of investigating the latter assassination.
The country was considered one of the most successful emerging mar-
kets, and it was supposed to be entering a period of sustained prosperity,
yet in the mid-1990s it became immersed in the worst economic crisis of
the last seventy years. How could the Mexican government find itself on
the brink of default just four years after the first Brady agreement, despite
having significantly reduced its external and domestic debt? These are the
issues addressed in this paper. The following section reviews the episode of
massive capital inflows of the early 1990s and discusses the financial re-
forms that preceded it. The paper then examines how the success of the
Mexican economy in attracting foreign capital itself undermined the fun-
damentals of the economy by causing a large appreciation of the peso, a
misallocation of resources, and increasing financial fragility. The next sec-
tion analyzes how these three consequences of the capital surge combined
and interacted to produce the crash landing that was to follow, and com-
pares our interpretation to other explanations of the crisis. The final sec-
tion looks at the performance of the economy after the crisis and discusses
its medium-term prospects.
Financial Liberalization, Debt Reduction, and
the Capital Surge
While the capital surge of the early 1990s extended to most countries
in Latin America and other developing regions, Mexico had a prominent
place in this episode: nearly half of the capital inflows to Latin America in
the period 199093 went to Mexico (see table 4-1). These massive inflows
had their origin in a diversity of circumstances: almost all of the various
internal and external determinants of capital movements had a positive ef-
fect on inflows. The external environment was characterized by the con-
;aixi ios :c
tinuous reduction of foreign interest rates through early 1994, a U.S. re-
cession in the early 1990s, and regulatory changes introduced by the U.S.
Securities Exchange Commission, all of which were favorable to an in-
crease in U.S. investment abroad.
1
In addition, improved terms of trade in
the late 1980s and the debt relief agreement with creditor banks led to a
reduction in Mexicos risk premium.
2
The favorable domestic factors arose out of the new institutional en-
vironment that characterized the Mexican economy as a result of the struc-
tural reforms initiated in the mid-1980s. The reforms that were most
directly relevant to the capital surge include the liberalization of domestic
financial markets and their opening to foreign investors, as well as the der-
egulation of foreign direct investments (FDI). Starting in 1988, a succes-
rui xixicax icoxox\ ix :,,cs :c,
1. The regulatory changesregulation S and rule 144Astimulated portfolio investment in de-
veloping countries through the issue of American depository receipts (ADRs) on the New York stock
market and the development of country funds in U.S. financial markets. See ECLAC (1995, 1998);
Ffrench-Davis (2000, chap. 5). See also Banco de Mexico, Informe Anual, 1993, available at www.
banixco.org.mx.
2. The Brady Plan was announced in March 1989 and an agreement in principle was signed in July
1989. Mexico became the first country to sign a debt agreement with commercial banks under the
United States-sponsored plan in February 1990.
Table 4-1. Capital Inflows in Mexico and Latin America, 197795
Billions of dollars, except as indicated
Capital inflows 197781 198389 199093 199495
Total annual capital inflows (net)
Latin America 28.86 8.15 45.89 38.55
Mexico 8.15 -0.66 22.50 3.60
Mexico as percent of total 28.3 -8.1 49.0 9.3
Foreign direct investment
Latin America n.a. n.a. 11.43 17.28
Mexico n.a. n.a. 4.05 6.99
Mexico as percent of total n.a. n.a. 35.4 40.4
Other capital inflows (net)
Latin America n.a. n.a. 34.46 21.27
Mexico n.a. n.a. 18.45 -3.39
Mexico as percent of total n.a. n.a. 53.5 -15.9
Source: United Nations Economic Commission for Latin America and the Caribbean (ECLAC),
Amrica Latina y el Caribe: polticas para mejorar la insercin en la economa mundial, 2d ed.,
Santiago: Fondo de Cultura Econmica and ECLAC, 1998, tables 9.1 and 9.4.
;aixi ios ::c
sion of measures relaxed reserve requirements, credit quotas to high prior-
ity sectors, and controls on interest rates; unified the controlled and free
exchange rates; and abolished the exchange controls that were adopted
during the 1982 crisis. Restrictions on foreign investments in the domes-
tic bond and stock markets were eliminated in 1989 and 1990.
3
Through-
out the 1980s, the De La Madrid and Salinas de Gortari administrations
liberally interpreted the Foreign Investment Law of 1973 and adopted a
series of regulatory changes to facilitate the inflow of FDI. These changes
culminated with a new Foreign Investment Law in 1993. The new Law
opened areas that had previously been closed to foreign companies and
eliminated the clause that restricted foreign participation in local firms to
49 percent. Other relevant policy changes include the privatization of the
banking system and telecommunications in 199192. In addition to dem-
onstrating the governments commitment to market-oriented reforms, pri-
vatization provided the government with large revenues that were used to
sharply reduce the burden of domestic and external debt in the fiscal ac-
counts (see table 4-2).
These domestic reforms and positive external shocks, together with the
initiation of NAFTA negotiations in 1990, contributed to the surge of cap-
ital inflows through three channels. The first was a sharp reduction in the
countrys risk premiumthe improved image of Mexico as a good place to
invest. The debt relief agreement, the fall in foreign interest rates, and the
use of large privatization revenues to repay external debt in 199192 pro-
duced a substantial reduction of interest payments both as a percent of ex-
ports and as a percent of gross domestic product (GDP) (see table 4-3).
These indicators, which are conventionally used by national and foreign in-
vestors to evaluate country risk, fell in the 1990s to levels below those
achieved in the mid-1970s before the onset of the balance-of-payments cri-
sis that was under gestation in the period 197476.
4
Moreover, privatiza-
tion revenues also financed a sharp reduction of domestic, as well as
external, government debt. By the first quarter of 1993, domestic debt was
down to half its real value in mid-1990. Total government debt had there-
fore fallen from 55 percent of GDP in 1990 to 35 percent in 1993, which
3. A decree in May 1989 liberalized the neutral-investment regime to promote the entry of foreign
investors into the local stock market. By late 1990, restrictions on the foreign purchase of domestic
bonds (largely government bonds) had been eliminated.
4. From 1990 to 1993, Mexicos foreign debt was upgraded by foreign risk evaluators, and in the
case of two Mexican development banks, it was given an investment grade rating (see Gurra, 1995).
This occurred despite the emergence of the first symptoms of serious financial vulnerability, as dis-
cussed below.
rui xixicax icoxox\ ix :,,cs :::
is about half the average for the member countries of the Organization for
Economic Cooperation and Development (OECD) and less than a third of
that for Italy or Belgium.
A second channel was the opening of domestic financial markets, which
had the same effect as a shift in the portfolio preferences of foreign investors
between domestic and foreign assets. In an earlier study of the determinants
of capital inflows, I find that the opening of the bond market accounts for
most of the shift in asset preferences during the period.
5
Only after 1994
the year NAFTA went into effect and a year after Congress approved the
new Foreign Investment Lawdid FDI become significantly more impor-
tant than in the 1960s and 1970s. This partly accounts for the heavy con-
centration of capital inflows in portfolio investments through late 1993.
The third channel, which interacted with the reduction in the risk pre-
mium, was the real appreciation of the peso and the very high interest rates
5. Ros (1994b).
Table 4-2. Mexico: Use of Privatization Revenues, 199192
Destination of revenues 1991 1992 1991 1992 199192 199192
Total revenues from
privatization
a
3.3 3.3 9.42 10.74 20.16 100
Financial deficit
b
1.4 0.1 4.02 0.28 4.30 21.3
Net reduction of debt 1.9 3.2 5.39 10.46 15.85 78.6
External 1.5 0.8 4.39 2.55 6.94 34.4
Central Bank 1.9 0.2 5.49 0.70 6.19 30.7
Private sector total
c
-1.6 2.2 -4.48 7.21 2.73 13.5
Residents
d
1.8 1.5 5.03 4.94 9.97 49.5
Nonresidents
e
-1.2 -2.7 -3.54 -8.77 -12.31 -61.1
Commercial banks -2.1 3.4 -5.97 11.04 5.07 25.1
Source: J. Ros, Mercados financieros y flujos de capital en Mxico, in Los capitales externos en
las economas latinoamericanas, edited by J. A. Ocampo, Bogot: Fedesarrollo and Inter-American
Development Bank (IDB), 1994. Calculations based on data from Banco de Mxico, Informe
Anual, 1991 and 1992, available at www.banixco.org.mx.
a. Includes cash revenues (or amortization of the banking privatization bonds) and reduction of
the domestic or external public sector debt as a result of the reclassification of Mexican
telecommunications as a private enterprise.
b. Accrued public financial deficit. Includes the accrued nonpaid interests that are only
incorporated to the financial deficit at the time of liquidation.
c. Banking and nonbanking (includes nonresidents).
d. Nonbanking. Obtained as a residual of private sector debt reduction.
e. Foreign portfolio inflows to the money market (displayed with a negative sign).
Percent
of total Billions of dollars
Percent of
GDP

T
a
b
l
e

4
-
3
.
M
e
x
i
c
o
:

M
a
c
r
o
e
c
o
n
o
m
i
c

P
e
r
f
o
r
m
a
n
c
e

I
n
d
i
c
a
t
o
r
s
,

1
9
8
8

9
8
P
e
r
c
e
n
t
I
n
d
i
c
a
t
o
r
1
9
8
8
1
9
8
9
1
9
9
0
1
9
9
1
1
9
9
2
1
9
9
3
1
9
9
4
1
9
9
5
1
9
9
6
1
9
9
7
1
9
9
8
G
D
P

g
r
o
w
t
h
1
.
2
4
.
2
5
.
1
4
.
2
3
.
6
2
.
0
4
.
4
-
6
.
2
5
.
2
6
.
7
4
.
8
I
n
f
l
a
t
i
o
n
5
1
.
7
1
9
.
7
2
9
.
9
1
8
.
8
1
1
.
9
8
.
0
7
.
1
5
2
.
0
2
7
.
7
1
5
.
7
1
8
.
6
R
e
a
l

e
x
c
h
a
n
g
e

r
a
t
e

(
i
n
d
e
x
)
1
0
0
.
0
9
4
.
6
8
9
.
9
8
2
.
0
7
5
.
0
7
0
.
8
7
3
.
6
1
0
6
.
6
9
6
.
7
8
5
.
5
8
6
.
4
R
e
a
l

i
n
t
e
r
e
s
t

r
a
t
e
a
7
.
3

1
3
.
8

1
.
0

-
1
.
3
3
.
4

1
4
.
0

1
3
.
3

6
.
6
9
.
2
8
.
8
1
0
.
1
P
r
i
v
a
t
e

s
a
v
i
n
g
s

r
a
t
e
b
1
7
.
5
1
4
.
8
1
3
.
2
1
0
.
6
1
0
.
5
1
1
.
6
1
1
.
4
1
5
.
8
1
6
.
7
D
o
m
e
s
t
i
c

s
a
v
i
n
g
s

r
a
t
e
b
2
1
.
4
1
9
.
5
2
0
.
3
1
9
.
0
1
8
.
1
1
8
.
2
1
8
.
0
1
8
.
5
2
2
.
5
c
2
4
.
1
c
2
0
.
5
F
i
x
e
d

p
r
i
v
a
t
e

i
n
v
e
s
t
m
e
n
t
b
1
1
.
7
1
1
.
8
1
2
.
8
1
4
.
1
1
5
.
6
1
4
.
8
1
4
.
3
1
1
.
0
1
3
.
2
1
5
.
3
1
6
.
8
F
i
x
e
d

p
u
b
l
i
c

i
n
v
e
s
t
m
e
n
t
b
3
.
9
4
.
0
4
.
2
4
.
1
3
.
8
3
.
8
4
.
9
3
.
6
2
.
9
3
.
0
2
.
5
U
r
b
a
n

u
n
e
m
p
l
o
y
m
e
n
t
3
.
6
3
.
0
2
.
8
2
.
6
2
.
8
3
.
4
3
.
7
6
.
2
5
.
5
3
.
8
3
.
2
A
v
e
r
a
g
e

r
e
a
l

w
a
g
e

(
i
n
d
e
x
)
1
0
0
.
0
1
0
5
.
9
1
0
7
.
4
1
1
4
.
2
1
2
2
.
8
1
3
0
.
6
1
3
6
.
9
1
1
8
.
8
1
0
9
.
1
1
1
2
.
5
1
1
5
.
5
T
r
a
d
e
b
E
x
p
o
r
t
s
1
3
.
9
1
4
.
0
1
4
.
1
1
4
.
2
1
4
.
4
1
5
.
2
1
7
.
2
2
3
.
9
2
6
.
8
2
7
.
9
2
9
.
2
I
m
p
o
r
t
s
1
1
.
7
1
3
.
2
1
5
.
0
1
6
.
6
1
9
.
2
1
9
.
2
2
2
.
3
2
0
.
2
2
3
.
6
2
7
.
1
2
9
.
5
C
a
p
i
t
a
l

a
c
c
o
u
n
t
b
D
i
r
e
c
t

i
n
v
e
s
t
m
e
n
t
1
.
6
1
.
5
1
.
1
1
.
8
1
.
6
1
.
5
3
.
5
3
.
1
2
.
9
3
.
7
2
.
9
P
o
r
t
f
o
l
i
o

i
n
f
l
o
w
s
0
.
5
0
.
2
1
.
4
4
.
9
6
.
6
1
0
.
1
2
.
6
-
3
.
2
4
.
2
1
.
5
0
.
4
L
o
a
n
s
-
1
.
8
0
.
4
4
.
7
3
.
1
-
0
.
6
1
.
0
0
.
4
7
.
5
-
3
.
8
-
2
.
6
1
.
1
D
e
b
t
b
G
r
o
s
s

e
x
t
e
r
n
a
l

d
e
b
t
4
3
.
7
3
6
.
6
3
2
.
1
3
0
.
1
2
8
.
4
2
7
.
5
2
7
.
1
3
1
.
0
3
0
.
8
2
6
.
9
2
4
.
9
G
r
o
s
s

d
o
m
e
s
t
i
c

d
e
b
t
2
2
.
4
2
3
.
4
2
0
.
9
1
6
.
8
1
1
.
8
1
0
.
2
1
0
.
8
9
.
5
6
.
8
7
.
4
8
.
9

P
u
b
l
i
c

f
i
n
a
n
c
e
b
T
o
t
a
l

e
x
p
e
n
d
i
t
u
r
e

3
6
.
4
3
0
.
6
2
7
.
5
2
3
.
9
2
2
.
2
2
2
.
5
2
3
.
1
2
3
.
0
2
3
.
3
2
3
.
7
2
1
.
5
O
p
e
r
a
t
i
n
g

e
x
p
e
n
d
i
t
u
r
e
1
2
.
5
1
1
.
2
1
0
.
5
9
.
9
8
.
8
9
.
1
8
.
9
8
.
2
8
.
1
8
.
3
7
.
2
I
n
v
e
s
t
m
e
n
t
2
.
9
2
.
4
2
.
7
2
.
9
2
.
7
2
.
5
3
.
0
2
.
5
2
.
9
2
.
5
1
.
8
D
o
m
e
s
t
i
c

i
n
t
e
r
e
s
t

p
a
y
m
e
n
t
s
1
2
.
5
9
.
0
6
.
8
3
.
1
2
.
2
1
.
5
1
.
1
2
.
6
2
.
3
2
.
5
1
.
5
E
x
t
e
r
n
a
l

i
n
t
e
r
e
s
t

p
a
y
m
e
n
t
s
3
.
6
3
.
3
2
.
3
1
.
9
1
.
5
1
.
2
1
.
2
2
.
0
2
.
1
1
.
6
1
.
4
O
t
h
e
r
4
.
9
4
.
7
5
.
2
6
.
1
7
.
0
8
.
2
8
.
9
7
.
7
7
.
9
8
.
8
9
.
6
T
o
t
a
l

r
e
v
e
n
u
e

2
7
.
7
2
5
.
8
2
5
.
3
2
3
.
5
2
3
.
7
2
3
.
1
2
2
.
8
2
2
.
8
2
3
.
2
2
3
.
0
2
0
.
3
F
e
d
e
r
a
l

g
o
v
e
r
n
m
e
n
t

(
t
o
t
a
l
)
1
6
.
3
1
6
.
4
1
5
.
9
1
5
.
5
1
6
.
0
1
5
.
5
1
5
.
2
1
5
.
3
1
5
.
7
1
5
.
8
1
4
.
1
T
a
x
e
s
1
1
.
4
1
1
.
1
1
0
.
7
1
0
.
7
1
1
.
3
1
1
.
4
1
1
.
3
9
.
3
9
.
0
9
.
8
1
0
.
4
N
o
n
t
a
x

f
e
d
e
r
a
l

r
e
v
e
n
u
e
5
.
0
5
.
3
5
.
2
4
.
8
4
.
8
4
.
1
3
.
9
6
.
0
6
.
7
6
.
0
3
.
7
I
n
c
o
m
e

s
h
a
r
e
s
P
r
o
f
i
t
s

a
n
d

i
n
t
e
r
e
s
t
d
5
6
.
5
5
6
.
0
5
4
.
5
5
2
.
9
5
1
.
8
5
0
.
4
4
9
.
8
5
4
.
6
5
5
.
9
n
.
a
.
n
.
a
.
T
r
a
d
a
b
l
e

g
o
o
d
s

s
e
c
t
o
r
s
2
5
.
1
2
1
.
3
1
9
.
1
1
7
.
5
1
6
.
0
1
4
.
2
1
3
.
8
1
7
.
0
1
9
.
2
n
.
a
.
n
.
a
.
N
o
n
t
r
a
d
a
b
l
e

g
o
o
d
s

s
e
c
t
o
r
s
3
1
.
4
3
4
.
7
3
5
.
4
3
5
.
4
3
5
.
8
3
6
.
2
3
6
.
0
3
7
.
6
3
6
.
7
n
.
a
.
n
.
a
.
S
k
i
l
l
e
d

l
a
b
o
r
e
9
.
8
1
1
.
7
1
2
.
3
1
4
.
0
1
5
.
6
1
7
.
2
1
8
.
2
1
6
.
6
1
5
.
5
n
.
a
.
n
.
a
.
O
t
h
e
r
f
3
3
.
8
3
2
.
3
3
3
.
2
3
3
.
0
3
2
.
6
3
2
.
4
3
1
.
9
2
8
.
7
2
8
.
6
n
.
a
.
n
.
a
.
S
o
u
r
c
e
:

I
n
s
t
i
t
u
t
o

N
a
c
i
o
n
a
l

d
e

E
s
t
a
d

s
t
i
c
a

G
e
o
g
r
a
f

a

e

I
n
f
o
r
m

t
i
c
a

(
I
N
E
G
I
)
,

N
a
t
i
o
n
a
l

A
c
c
o
u
n
t
s
,

I
n
d
i
c
a
t
o
r
s

o
f

C
o
m
p
e
t
i
t
i
v
i
t
y
,

v
a
r
i
o
u
s

y
e
a
r
s
;

B
a
n
c
o

d
e

M

x
i
c
o
,
E
c
o
n
o
m
i
c

I
n
f
o
r
m
a
t
i
o
n
;

O
r
g
a
n
i
z
a
t
i
o
n

f
o
r

E
c
o
n
o
m
i
c

C
o
o
p
e
r
a
t
i
o
n

a
n
d

D
e
v
e
l
o
p
m
e
n
t

(
O
E
C
D
)
,
O
E
C
D

E
c
o
n
o
m
i
c

S
u
r
v
e
y
s
,

v
a
r
i
o
u
s

y
e
a
r
s
;

S
e
c
r
e
t
a
r

a

d
e

H
a
c
i
e
n
d
a

y
C
r

d
i
t
o

P

b
l
i
c
o

d
e

M

x
i
c
o
,

h
i
s
t
o
r
i
c
a
l

d
a
t
a
;

I
n
t
e
r
n
a
t
i
o
n
a
l

M
o
n
e
t
a
r
y

F
u
n
d

(
I
M
F
)
,

I
n
t
e
r
n
a
t
i
o
n
a
l

F
i
n
a
n
c
i
a
l

S
t
a
t
i
s
t
i
c
s
,

v
a
r
i
o
u
s

y
e
a
r
s
;

U
n
i
t
e
d

N
a
t
i
o
n
s

E
c
o
n
o
m
i
c
C
o
m
m
i
s
s
i
o
n

f
o
r

L
a
t
i
n

A
m
e
r
i
c
a

a
n
d

t
h
e

C
a
r
i
b
b
e
a
n

(
E
C
L
A
C
)
,

E
c
o
n
o
m
i
c

S
u
r
v
e
y

o
f

L
a
t
i
n

A
m
e
r
i
c
a

a
n
d

t
h
e

C
a
r
i
b
b
e
a
n
,

1
9
9
8

9
9
.
a
.

D
i
s
c
o
n
t
i
n
u
o
u
s

s
e
r
i
e
s
:

1
9
8
8

9
2

c
a
l
c
u
l
a
t
i
o
n
s

b
a
s
e
d

o
n

E
C
L
A
C

d
a
t
a
;

1
9
9
3

9
8

c
a
l
c
u
l
a
t
i
o
n
s

b
a
s
e
d

o
n

I
M
F

s
t
a
t
i
s
t
i
c
s
.
b
.

P
e
r
c
e
n
t

o
f

G
D
P
.
c
.

V
a
l
u
e

n
o
t

c
o
m
p
a
r
a
b
l
e

w
i
t
h

e
a
r
l
i
e
r

s
e
r
i
e
s
.
d
.

O
p
e
r
a
t
i
n
g

s
u
r
p
l
u
s

(
e
x
c
l
u
d
i
n
g

o
i
l
,

a
g
r
i
c
u
l
t
u
r
e
,

c
o
m
m
e
r
c
e
,

a
n
d

o
t
h
e
r

s
e
r
v
i
c
e
s
)

m
i
n
u
s

p
r
o
p
e
r
t
y

i
n
c
o
m
e

p
a
i
d

a
b
r
o
a
d

p
l
u
s

g
o
v
e
r
n
m
e
n
t

i
n
t
e
r
e
s
t

p
a
y
m
e
n
t
s

t
o

t
h
e
p
r
i
v
a
t
e

s
e
c
t
o
r
.
e
.

E
s
t
i
m
a
t
e
d

u
s
i
n
g

t
h
e

a
v
e
r
a
g
e

w
a
g
e

i
n

c
o
n
s
t
r
u
c
t
i
o
n

a
s

a
n

u
p
p
e
r

l
i
m
i
t

t
o

t
h
e

w
a
g
e

o
f

u
n
s
k
i
l
l
e
d

l
a
b
o
r
.
f
.

U
n
s
k
i
l
l
e
d

l
a
b
o
r

a
n
d

s
e
l
f
-
e
m
p
l
o
y
e
d

e
a
r
n
i
n
g
s
.

E
s
t
i
m
a
t
e
d

a
s

a

r
e
s
i
d
u
a
l
.
;aixi ios ::
that prevailed during the initial stages of the disinflation program of late
1987. The annual real interest rate on three-month Treasury bills (CETES)
in mid-1989 was still on the order of 19 percent to compensate for the in-
flation uncertainty that accompanied the disinflation process. The rapid
success of the counterinflation program, which required a significant ap-
preciation of the peso in real terms, implied extremely high ex post dollar
interest rates on short-term domestic securities.
The magnitude and composition of the inflows were key elements for
the macroeconomic consequences of the capital surge. Gross inflows in-
creased tenfold from U.S.$3.5 billion in 1989 to U.S.$33.3 billion in
1993, before falling sharply in 1994.
6
Net inflows rose from negligible
numbers to 12.6 percent of GDP at their peak in 1993. During the capital
surge, the inflows were strongly biased toward highly liquid assets. The
composition of these short-term investments changed over time. In the in-
itial stages of the surge (1989 and 1990), inflows involved foreign loans to
the private sector and the acquisition of bank deposits, but the liberaliza-
tion of the domestic money and stock markets redirected inflows toward
government bonds and stocks. Throughout the period, portfolio invest-
ments in the stock market increased rapidly and represented the largest
component at the peak of the surge in 1993. From the end of 1990 to the
end of 1993, the market value of the accumulated portfolio investments
in the stock market jumped from U.S.$4.5 billion to U.S.$54.6 billion.
In 1993 alone, the flow of portfolio investments in the stock market
amounted to U.S.$28.4 billion.
7
Short-term outflows were dominant dur-
ing the contraction after 1994: average portfolio investments fell to around
1.7 percent of GDP from 1994 to 1998, well below FDI (3.3 percent of
GDP on average and about 20 percent of private domestic investment).
The cycle of capital movements appears to be especially marked in the
case of Mexico, which has a history of being the main beneficiary of inflows
during booms and the most vulnerable country during contractions. Dur-
ing periods of expanding capital inflows, the Mexican economy has tended
to absorb a large share of the portfolio inflows into Latin America, much
larger than its fraction of the less volatile FDI flows. This concentration of
inflows in liquid assets increases Mexicos financial fragility relative to other
countries, such that the downturn catches Mexico in a very vulnerable po-
sition. The result is that the country is generally one of the most severely hit
by the reversal of the portfolio inflows and receives a very small share dur-
6. Banco de Mexico, Informe Anual, 1994, available at www.banixco.org.mx.
7. See Gurra (1995).
rui xixicax icoxox\ ix :,,cs ::,
ing the period of contraction. The recent past has been an exception to this
pattern. Financial turbulence started in East Asia and later in Russia and
Brazil, while Mexico maintained a relatively solid position.
How Success Undermined the Fundamentals
The magnitude and composition of capital inflows should have raised
three legitimate sources of concern for economic policy: the continuous
appreciation of the peso in the midst of a rather radical trade liberalization
process; an allocation of resources biased toward consumption rather than
investment and toward the production of nontradables, which originated
in the difficulties of adequately intermediating such massive inflows and
resulted in slow economic expansion; and the increasing financial fragility
that resulted from the concentration of inflows in highly liquid assets and
the progressive deterioration of the banks balance sheets. This section dis-
cusses these three consequences of the capital surge.
Capital Inflows, Financial Markets, and the Real Appreciation
of the Peso
The capital surge episode featured a large real appreciation of the peso:
from 1988 to 1993, the real value of the peso increased by more than 40
percent (the real exchange rate fell by 30 percent; see table 4-3). The pro-
cess of appreciation appears to have been even more dramatic if measured
by the ratio of tradable to nontradable goods prices: the ratio fell from
140.9 in 1988 to 74.3 in 1993 (Gurra, 1995).
8
This section presents a
formal analysis of the mechanisms at work behind the real appreciation
and of the macroeconomic adjustment problem it generated. The analyt-
ical framework is based on a model of the goods and financial markets in
which the exchange rate and the interest rate have a dual role, as key prices
in the allocation of resources and as prices of financial assets.
9
First, they
8. In 1981, in the eve of the 1982 currency devaluations, the ratio was 98.1.
9. The model draws on Branson (1985), with elements borrowed from Ros (1994b, 1995a) and
Ros and Skott (1998). It differs from Branson in two major areas. First, relative prices respond asym-
metrically to excess demand and supply. While an increase in the demand for nontraded goods leads
to a real appreciation of the currency (an increase in the relative price of nontraded goods), an excess
supply leads to a downward output adjustment (leaving the real exchange rate, at least temporarily, at
its initial level). Second, the model attributes an important role to changes in the position of the mar-
ginal investor in explaining the behavior of asset prices.
Figure 4-1. Financial and Goods Markets
e
r
IX
1
2
FM
clear the goods market, that is, they guarantee the flow equilibrium of the
economy, through their effects on desired investment, savings, and net ex-
ports. Second, they ensure financial equilibriumthe stock equilibrium
of the economythrough their role as prices of financial assets.
Figure 4-1 shows the flow and stock equilibria of the economy at the
intersection of two curves (the IX and FM curves in Bransons terminol-
ogy).
10
The first curve (IX) is a locus of real exchange rate and real inter-
est rate combinations along which desired investment is equal to savings
(national plus foreign). Investment depends positively on the real ex-
change rate (on account of its positive effects on profitability) and in-
versely on the real interest rate. Given total employment, this locus is one
along which the nontraded goods market clears through changes in the
real exchange rate. It therefore has a positive slope in figure 4-1: a fall in
the interest rate raises investment and creates excess demand for goods; the
relative price of nontraded goods increases to clear the market for non-
traded goods. Moving along the curve, lower values of the real exchange
rate cause the current account deficit to increase, as the fall in the real ex-
change rate (appreciation of the domestic currency) leads to a larger gap
between the demand for and domestic supply of traded goods.
;aixi ios ::o
10. See the appendix for a full exposition of the model.
The second curve (FM) is a locus of financial market equilibrium. In
portfolio balance, the interest rate differential between domestic and for-
eign short-term securities, adjusted for exchange rate expectations, is equal
to the risk premium on domestic securities. This premium is assumed to
be a function of an exogenous component reflecting investors subjective
perceptions of country risk and the stock of external liabilities. With re-
gressive expectations about the real exchange rate (see appendix), the locus
slopes downward. An appreciation of the domestic currency tends to gen-
erate expectations of depreciation, which requires an increase in the do-
mestic interest rate to maintain financial market equilibrium.
Consider now the situation in mid-1989. The current account deficit
was on the order of 2.6 percent of GDP (roughly the average of the past
three decades), and the real exchange rate was at its 1978 value (the aver-
age of the 197094 period). There were no significant expectations of real
depreciation, and a high risk premium accounted for the large real inter-
est rate differential between domestic and foreign securities.
11
Let these
conditions be described by point 1 in figure 4-1. Starting with the debt re-
lief agreement and the beginning of NAFTA negotiations in mid-1990, a
series of positive shocks disturbed the initial stock equilibrium. These
shocks reduced both the real interest rate and the real exchange rate; they
also reduced the risk premium and shifted the FM curve down and to the
left. As the FM curve shifts downward, the short-term equilibrium value
of the domestic currency appreciates along the IX curve. The anticipation
by financial markets of a fall in interest rates (and thus in the real exchange
rate), as a result of the downward shifts in the FM curve, generates short-
term expectations of real appreciation. Falling interest rates, real apprecia-
tion of the peso, and continuous expectations of real appreciation were
indeed key features of Mexicos financial markets from mid-1990 to mid-
1992.
12
By mid-1992 the process of reducing domestic and external govern-
ment liabilities ended, and the FM curve stopped shifting down. Exchange
rate expectations then came to be dominated by the increased gap between
rui xixicax icoxox\ ix :,,cs ::;
11. See Ros (1994b). The expected rate of real depreciation is obtained from the interest rate dif-
ferential between CETES and the dollar-indexed Tesobonos and from the inflation expectations im-
plicit in the differential between CETES and indexed Ajustabonos. Because the differential between
Ajustabonos and CETES includes an inflation-uncertainty premium, the risk premium also reflects
an inflation-uncertainty premium. Uncertainty about inflation was large in 1989, since the heterodox
stabilization program launched in December 1987 had yet to prove its success. This contributed to
the high measured risk premium between domestic and foreign securities.
12. Ros (1994b).
the exchange rate and its expected long term value.
13
At this point, the real
interest rate would have to increase again to compensate for expectations
of depreciation, which is what happened between the second and third
quarters of 1992 when the expected rate of depreciation turned from neg-
ative to positive.
Subsequently, however, real interest rates remained surprisingly stable
and low (until the political shocks of 1994), and the persistent gap be-
tween the current and long term real exchange rates failed to be reflected
in higher interest rates. This appears to have been the consequence of a
change in the position of the marginal investor (at the margin between do-
mestic and foreign short term securities), as the stock of domestic govern-
ment bonds changed hands between the bearish domestic wealth holders
and the bullish foreign investors.
14
Adjustments in the Goods and Labor Markets
Consider now the evolution of the goods and labor markets. The styl-
ized facts here are, first, the rapid and unprecedented contraction of
manufacturing employment and the expansion of underemployment in
commerce and agriculture. These trends are shown in table 4-4, together
with output and productivity performance during the crisis and post-crisis
periods (199397). As shown in the table, the reduction in manufactur-
ing employment is the result of a very high growth rate of labor produc-
tivity (6.0 percent per year) in the face of a rather slow rate of output
growth (4.2 percent per year).
Second, wage inequality in the labor market increased in the early
1990s. Figure 4-2 shows the decline since 1988 in the relative labor earn-
ings of the low wage sectors (namely, agriculture, construction, commerce,
and services). Wage inequality also increased within sectors as the wage pre-
mium for skilled labor rose. Figure 4-3 illustrates the rising trend of white-
collar wages relative to blue-collar wages in all manufacturing industries,
while table 4-3 (above) tracks the redistribution between skilled and un-
skilled labor earnings that took place from 1988 to 1994. Other evidence
reported in a number of studies similarly documents the increase in wage
dispersion and shows that it largely revolved around the increase in skilled
;aixi ios ::
13. Formally, this is equivalent to an upward shift in the FM curve.
14. Bulls and bears are defined here according to their perceptions about the future of the exchange
rate. See Ros (1994b); Ibarra (1997).
relative to unskilled labor incomes.
15
As a result of these trends, the Gini co-
efficient of wage inequality increased from 0.46 in 1989 to 0.49 in 1992,
and then to 0.53 in 1994.
16
The increase in wage inequality is, in turn, the
single major factor behind the worsening of income distribution during the
period.
17
These stylized facts can be explained as consequences of trade liberal-
ization and the real overvaluation of the peso, which caused intensified
competition from imports and thus accelerated the rate of technology
adoption and reduced the demand for low skilled workers in manufactur-
ing. This hypothesis can be elaborated in the framework of a two-sector
model (tradables and nontradables) in which tradable goods are produced
under imperfect competition with unskilled labor as the variable factor,
while skilled labor is largely fixed in the short run, since it is complemen-
tary to physical capital. Trade liberalization implied greater competition in
rui xixicax icoxox\ ix :,,cs ::,
15. See Hanson and Harrison (1995); Cragg and Epelbaum (1996); Alarcn and McKinley (1997).
16. Lustig and Szkely (1998), based on income and expenditure household surveys.
17. Ros and Lustig (1999).
Table 4-4. Mexico: Output, Employment, and Productivity Growth,
198897
Average annual growth rates
Sector 198893 199497 198893 199497 198893 199497
Total 3.1 3.2 3.7 2.4 0.6 -0.8
Mining
a
-7.4 8.8 2.0 4.1 9.4 4.7
Manufacturing -1.8 4.9 4.2 4.7 6.0 -0.2
Agriculture 5.8 0.5 1.9 1.7 -3.9 1.2
Nontradable goods 3.6
c
3.9
c
3.1 1.8 -0.5 -2.1
Construction 4.1 -1.7 4.9 0.1 0.8 1.8
Transport and
communication 5.0 2.7 4.1 5.0 -0.9 2.3
Commerce
b
5.1 3.4 4.3 1.0 -0.8 -2.4
Services 3.0 5.5
d
3.9 1.6
d
0.9 -3.9
Government 0.8 n.a. 0.9 n.a. 0.1 n.a.
Source: INEGI, National Accounts, various years; INEGI, National Survey of Employment, 1997.
a. Includes electricity.
b. Includes restaurants and hotels.
c. Includes unspecified activities.
d. Includes government.
Productivity growth Output growth Employment growth

Figure 4-2. Mexico: Relative Labor Earnings in Low Wage Sectors,

198897
Source: INEGI, Cuentas Nacionales.
a. Average labor earnings in the sector relative to the ecomomys average.
1.00
0.80
0.60
0.40
0.20
Relative wage
a
1991 1990 1989 1988 1992 1993 1994 1995 1996 1997
Construction
Real estate and rental services
Other services
Agriculture, hunting, forestry, and fishing
Commerce
Figure 4-3. Mexico: Ratio of White- to Blue-Collar Earnings in
Manufacturing, 198897
Source: INEGI, Cuentas Nacionales.
3.5
3.0
2.5
2.0
1.5
1.0
0.5
Wage ratio
1991 1990 1989 1988 1992 1993 1994 1995 1996 1997
Food, beverages and tobacco
Paper and paper products, printing, and
publishing
Nonmetallic mineral products, except from
petroleum and coal
Fabricated metal products, machinery, and
equipment
Total
Wood and wood products, including
furniture
Chemicals and chemical, petroleum, coal,
rubber, and plastic products
Basic metal industries
Other manufacturing
the domestic market and increased the share of imports in domestic de-
mand. At the same time, the real value of the peso rose as a result of the
capital inflows, the increased demand for nontradables, and the chosen ex-
change rate policy. Increased competition and the appreciation of the real
currency, in turn, caused the profit markup over marginal costs in the
import-competing goods sectors to fall. The profit squeeze is shown in
table 4-3: the property income share fell continuously from 1988 to 1994,
with a cumulative decline of 7 percentage points. The impact of the profit
squeeze was very different among sectors, with a large redistribution tak-
ing place from the tradables toward the nontradables sectors. The lower
profit markup had a negative effect on unskilled employment, since this is
the variable factor that firms can adjust in the short run. This resulted in
the tendency to substitute skilled for unskilled labor, which is observable
in the industrial employment data.
Figure 4-4 shows the determination of the real wage (W/P) and the
level of employment in the import-competing goods sector at the intersec-
tion of two schedules. The NN locus is the equilibrium schedule in the
markets for unskilled labor and nontradables (which are intensive in un-
skilled labor). The schedule slopes upward: an increase in employment in
the import-competing goods sector (L
T
) reduces the labor supply to the
nontradable goods sector and increases the real wage. The TT locus is the
equilibrium schedule in the market for import-competing goods. It is
downward sloping, since an increase in wages increases costs and prices in
this sector, which reduces the demand for import-competing goods (to the
benefit of imported and nontradable goods).
The increase in capital inflows generated an expansion in the demand
for nontradable goods, which shifted the NN curve up and to the left. This
increased the real wage and crowded out employment in the tradables sec-
tor. At the same time, the effects of trade liberalization and currency over-
valuation on the goods and labor markets operated through three channels.
First, the relative price of imported goods fell, shifting the TT locus down
and to the left. Output demand and employment in the import-competing
goods sector fell in the short run. Given nominal wages in the import-com-
peting goods sector, the adjustment involved an expansion of employment
in the nontradables sector, which reduced productivity and wages in that
sector. This opened a gap between the wages of unskilled labor employed in
the two sectors, observable in the increasing wage inequality between low-
and high-wage sectors (see figure 4-2). The downward pressure on real
wages in the import-competing goods sector eventually reduced the prod-
rui xixicax icoxox\ ix :,,cs :::
Figure 4-4. Financial Flows, Exchange Rate, Wages, and Employment
W/P
NN
TT
L
T
uct wage of unskilled labor there and generated an intrasectoral increase in
the wage premium of skilled labor. The cost and relative price reductions in
the import-competing goods sector tended to reestablish the initial relative
prices, shifting the TT schedule upward. Even if the initial relative prices
are reestablished in the long run, however, the displacement of local pro-
duction by imports may have hysteresis effects which lead to a permanent
increase of the ratio of imports to domestically produced import-compet-
ing goods. Through this second channel, the downward shift of the TT
schedule may be permanent and has long-lasting effects on employment
and real wages in the import-competing and nontradable goods sectors.
These hysteresis effects are the consequence of the real currency apprecia-
tion associated with trade liberalization.
Finally, the intensified competition in the local market led to an in-
crease in the price elasticity of demand facing producers of import-com-
peting goods. The higher price elasticity reduced the firms profit margins,
which had two consequences. First, it implied a fall in the price of import-
competing goods relative to nontradables and imported goods. This in-
creased the demand for import-competing goods and shifted the TT
schedule upward. Second, the fall in profitability tended to reduce invest-
ments in the import-competing goods sector and had negative longer-run
effects on employment. This second effecttogether with the short-run
;aixi ios :::
and long-run effects of import penetration discussed aboveclearly dom-
inated in the Mexican experience.
Currency overvaluation also had adverse effects on aggregate invest-
ment, to the extent that the increased profitability of nontradables was less
than the reduced profitability of tradables. (As shown in table 4-3, the
overall share of profits in total income fell.) Together with the distributive
effect of the profit squeeze against incomes with relatively high savings
rates, these adverse effects on investment explain why the rapid expansion
of bank credit that accompanied the capital surge fuelled a consumption
boom. From 1988 to 1994, the gap between private investment and sav-
ings increased by 6.2 percentage points of GDP. The increase was fully ex-
plained by a collapse of private savings (6.1 percentage points of GDP)
rather than by any significant change in the investment rate.
The positive shock that hit the Mexican economy in the early 1990s
reduced the real interest rate and appreciated the real value of the domes-
tic currency. In figure 4-1 (above), as the real value of the peso appreciates
along the IX curve, the current account deficit increases. Thus a large cur-
rent account deficit emerges at point 2 in the figure, exacerbated by the
effects of the capital surge on domestic private savings. From then on, the
resulting accumulation of foreign liabilities should produce an upward
pressure on the domestic savings rate (or a reduction in the domestic in-
vestment rate or both), while the expectations of depreciation should gen-
erate an increase in the real interest rate. The resulting excess supply of
nontradable goods would, in a flex price economy, tend to increase the
price ratio of tradable to nontradable goods and to gradually correct the
current account deficit. The IX line shifts up in this process, bringing
about a real depreciation. Because expectations of depreciation fall as the
gap between the current and the long-term exchange rate closes, the path
also features a reduction of domestic interest rates.
This mechanism of current account correction and resource realloca-
tion will fail to operate if the excess supply of goods does not lead to a
lower relative price of nontradables. In that case, the ex ante excess of sav-
ings over investment will turn into a downward output adjustment. This
output adjustment can be seen as a downward shift of the IX curve that
prevents the correction of the real exchange rate, the fall of depreciation
expectations, and the reduction of the interest rate. Currency overvalua-
tion persists, and large current account deficits remain in the face of a de-
celeration of growth.
Such was the actual path followed by the Mexican economy: growth
rui xixicax icoxox\ ix :,,cs ::,
Figure 4-5. Mexico: Trade Balance and GDP Growth, 197099
Source: INEGI, National Accounts; Organization for Economic Cooperation and Development, OECD Economic
Surveys, various years.
0.25
0.20
0.15
0.10
0.05
0.00
0.05
0.05
0.0
0.05
0.10
0.15
0.20
0.25
GDP growth
Trade balance/GDP
(right axis)
GDP growth (left axis)
1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999
Trade Balance/GDP
slowed in 1992 and substantially decelerated in 1993. As shown in figure 4-
5, the decline in the trade balance occurred in the midst of sluggish growth
and then persisted with excess capacity. While earlier current account defi-
cits on the order of 78 percent were an exceptional feature occurring only
in periods of very fast economic expansion (with GDP growth rates above
7 percent), in the early 1990s these large deficits appeared in a semi-
stagnant economy and showed no signs of reversal as growth decelerated in
1992 and 1993.
18
Financial Fragility and the Deterioration of the Banks Balance Sheets
In the financial markets, the continuous accumulation of external
liabilities eventually caused the FM curve to shift up, putting further up-
ward pressure on the interest rate (at a given value of the real exchange rate,
as the IX curve endogenously shifted down with the level of output). As
already discussed, this increase in interest rates was initially prevented by
the foreign purchase of the stock of government bonds held by local
investors. In 1994, three additional processes came into play to hold down
;aixi ios ::
18. For a full discussion, see Ros (1994b) and Lustig and Ros (1998).
Figure 4-6. Mexican Government Short-Term Public Debt
Source: World Bank, based on Banco de Mxico, Indicadores Econmicos.
a. Three-month peso-denominated Treasury bonds.
b. Three-month dollar-indexed bonds.
120
100
80
60
40
20
Billions of new pesos
Dec. 1993 June 1994 Dec. 1994
Tesobonos
b
Cetes
a
the government debt service. All three were the consequence of the slow-
down of capital inflows that occurred as foreign investors, who had be-
come the major holders of short-term government securities by late 1993,
became increasingly fearful about the governments ability to hold to its
exchange rate policy.
The first process, which began in full in early 1994 and accelerated
with the assassination of the PRIs presidential candidate in March of that
year, was the dollarization of the stock of domestic government debt. As
peso-denominated Treasury bills (CETES) fell due, they were converted
into dollar-indexed short-term bonds (Tesobonos) (see figure 4-6). This
process represented a change in the position of the investor at the margin
between CETES and Tesobonos, and it prevented even larger increases in
the CETES rate.
The second process was the shortening of the average maturity of gov-
ernment bonds. The dollarization of debt involved not only peso-denom-
inated short-term Treasury bills, but also peso-denominated securities with
longer maturities. These long-term bonds (Bondes and inflation-adjusted
Ajustabonos) were similarly converted into Tesobonos (which is why in
figure 4-6 the stock of Tesobonos in December is larger than the stock
of CETES a year earlier). The result was a steady decline in the average
rui xixicax icoxox\ ix :,,cs ::,
maturity of the debt from a peak of 306 days in April 1994 to 206 days
in December 1994. A consequence of this process was the large amount
of debt coming due in 1995 in the form of both CETES and Tesobonos.
In the case of Tesobonos, U.S.$28.6 billion was due in 1995, of which 35
percent (U.S.$9.9 billion) was due in the first quarter.
19
The third process involved the Central Banks decision to sterilize the
effects of reserve losses on the money supply. The loss of international re-
serves stemmed from massive speculative attacks, especially in March 1994
at the time of the Colosio assassination, as well as from a more gradual
process that occurred the second half of 1994 as a result of the persistently
large current account deficit in the face of the slowdown of capital inflows.
The sterilization of reserve losses prevented the money supply from shrink-
ing and contributed to keeping interest rates relatively low.
Why did the government choose this course of action, given the high
risks involved? The next section addresses this question fully, but part of
the answer has to do with another set of problems created by the capital
surge: the increased vulnerability of a recently privatized banking system.
A large fraction of capital inflows was intermediated by commercial banks,
whether directly through bank deposits acquired by foreigners or indi-
rectly through the substitution, in the portfolio of residents, of commer-
cial bank liabilities for financial assets sold to foreigners. Trigueros
estimates that domestic banks decided the final destination of close to 60
percent of foreign inflows.
20
As a result of the intermediation of capital in-
flows, commercial bank credit to the nonfinancial private sector multi-
plied by a factor of 3.5 between the end of 1989 and the end of 1994.
Credit was misallocated, partly as the unavoidable result of the mag-
nitude of its expansion, partly in response to the profit squeeze on invest-
ment, and partly because the demand for consumer durables had been
repressed for almost a decade. The result was an imbalance between the
financing of consumption and the financing of investment, together with
a deterioration in the banks balance sheets. The magnitude of the imbal-
ance in the financing of consumption and investment is revealed in the
fact that household credit increased from 9.6 percent of total credit in De-
cember of 1989 to 26.7 percent in November 1994. The deterioration of
the banks balance sheets is revealed by the increase in nonperforming
loans, which rose from a negligible amount in 1990 to about 9 percent of
;aixi ios ::o
19. See Cole and Kehoe (1996); Ibarra (1997).
20. Trigueros (1998).
total bank loans in 1994. At the same time, dollar-denominated bank
liabilities expanded rapidly: dollar deposits (mostly in commercial banks)
captured nearly one-fourth of total capital inflows from the beginning of
1990 to the third quarter of 1994. This expansion, together with the fact
that the sources of foreign exchange necessary to service the commercial
bank liabilities in hard currency were not being created, created a vulner-
ability to exchange rate changes that had dramatic consequences after the
December 1994 devaluation.
The Crisis: Currency, Banking or Debt?
The path that the economy had taken well before the convulsions of
1994 was likely to lead, sooner or later, to a slowdown of capital inflows.
The slowdown came in 1994, and it partly took the form of a series of spec-
ulative attacks associated with the political shocks of that year. The govern-
ment was then faced essentially with three options: an early (albeit already
overdue) exchange rate realignment through a shift to a floating rate, a step
devaluation, or a faster slide of the band ceiling; a monetary contraction re-
sulting from unsterilized reserve losses, which would have led to large in-
creases in interest rates; or an expansion of domestic credit to sterilize the
effects of reserve losses on the monetary base, combined with the dollariza-
tion of domestic debt to stem capital flight and prevent further increases in
interest rates.
The first option was probably judged unnecessary and likely to send
the wrong signal concerning the Central Banks commitment to price
stability (which had recently been elevated to the sole objective of the Cen-
tral Bank). The second option was discarded, probably out of the fear on
the part of the Central Bank that such a monetary contraction would trig-
ger a major banking crisis. The possibility of a major banking crisis was
undoubtedly a real reason for concern: the speculative attack in March had
caused a loss of one-third of international reserves, and the balance sheet
of the banking system had deteriorated as a result of the earlier overexpan-
sion of bank credit.
Another major reason why the authorities pursued the third option
was that they assumed the slowdown in capital inflows was temporary and
reversible. The slowdown was not seen as the beginning of a macroeco-
nomic adjustment process that required a change in exchange rate policy
in order to proceed more or less smoothly. Policymakersand for a while
rui xixicax icoxox\ ix :,,cs ::;
many financial market participantswrongly assessed the deterioration of
the real fundamentals of the Mexican economy that began well before the
political events and hard policy dilemmas of 1994. They thus perceived
the reserve losses following the runs on the peso in January and March as
unique events that had unfortunately, but only temporarily, sent a healthy
situation off course. Furthermore, the sterilization and debt management
decisions temporarily accomplished their intended objectives of stopping
capital flight and relieving the pressure on domestic interest rates. Domes-
tic interest rates started to fall in July, and from April to mid-November
there were no new dramatic drops in international reserves. Reserves did
undergo a gradual and continuous loss, however, as the slowdown in cap-
ital inflows persisted and the large current account deficit was financed by
running down reserves.
The Currency Crisis
Interpretations of the Mexican crisis fall into several categories. A first
distinction can be made between a disequilibrium view, which emphasizes
the economys real disequilibria (that is, real overvaluation and unsustain-
able current account deficit), and an equilibrium speculative attack view,
which gives preeminence to the excessive monetary growth that caused
reserves to decline until the exchange rate regime collapsed.
21
The interpre-
tation presented here clearly falls into the real disequilibria school. The cur-
rent account deficit was indeed on an unsustainable path, and its correction
would eventually require a large real exchange rate realignment. Between
1989 and 1993, the current account deficit (excluding interest on the debt)
averaged 31.7 percent of exports. Given an export growth rate of 8.7 per-
;aixi ios ::
21. See Sachs, Velasco, and Tornell (1996) for this distinction. Within the first view, Dornbusch
and Werner (1994) explain the real appreciation of the peso as the result of the exchange ratebased
stabilization program in the face of sluggish price adjustments; Ros (1994b, 1995a) and Lustig and
Ros (1998) discuss the so-called Mexican disease, that is, the unsustainability of exchange rate and
trade policies and the misallocation effects of the capital surge, which led to the collapse of private sav-
ings and unsustainable current account deficits; and Atkeson and Ros-Rull (1996) argue that Mexico
hit a borrowing constraint in early 1994 following the collapse of domestic savings in previous years.
For examples of the second view, which emphasizes the inconsistency between exchange rate and mon-
etary policy generated by the Central Banks attempt to sterilize the reserve losses arising from the slow-
down of capital inflows, see Sachs, Velasco, and Tornell (1995) and Calvo and Mendoza (1996). The
standard speculative attack model is Krugman (1979); later generations of speculative attack models
are discussed below. Many of these interpretations go beyond the events leading to the devaluation in
December 1994, encompassing the post-devaluation crisis. Here, the focus is on their interpretations
of the economic mechanisms that led to the devaluation.
cent per year (the average for 198993), Dadush, Dhareshwar, and Jo-
hannes estimate a steady-state ratio of foreign liabilities to exports of 3.6.
22
I reach similar conclusions in earlier papers.
23
The World Bank debt tables
classify a country as severely indebted if its debt is greater than 220 percent
of exports, that is, well below the 3.6 steady state ratio in Mexico. It is thus
unlikely that foreign investors would have been willing to finance this high
debt path. The Mexican economy would therefore have faced severe debt
and balance-of-payments problems well before reaching the steady state.
24
The presence of real disequilibria, however, is not enough to explain
the currency crisis, let alone its timing. What if, for example, the Central
Bank had chosen not to finance the current account deficit by running
down reserves during most of 1994? Before looking more closely at this
question, a further distinction must be made among three different types
of currency crisis models, this time within the speculative attack hypoth-
esis.
25
Fundamentalist or first-generation models la Krugman or Flood
and Garber are based on the inconsistency between an exchange rate peg
and an expansionary monetary and fiscal policy.
26
When forced to mone-
tize a fiscal deficit, the central bank defends the peg by gradually losing re-
serves. The inconsistency is perceived by speculators, and it consequently
leads to a speculative attack and the collapse of the exchange rate. These
are fundamentalist models in the sense that the origin of the crisis has to
do with wrong fundamentals, in this case the inconsistency between the
rate of domestic credit expansion and the exchange rate regime.
Second-generation models endogenize economic policy by introducing
the costs and benefits of devaluation. Speculative attacks can then arise in
the absence of any fundamental problems.
27
The reason is that the endogen-
eity of economic policy introduces an element of indeterminacy and self-
fulfilling prophecy: a government may be willing to continue to defend the
exchange rate (and follow a consistent monetary policy) if no speculative at-
tacks occur, but it would rather shift to a floating regime if a speculative at-
rui xixicax icoxox\ ix :,,cs ::,
22. Dadush, Dhareshwar, and Johannes (1994).
23. Ros (1994a, 1994b).
24. The 3.6 ratio of foreign liabilities to exports includes not only debt, but also other forms of for-
eign investments. This qualification does little, however, to reduce the assessment of Mexicos vulner-
ability, given that most of these investments took the form of highly liquid and volatile portfolio
inflows with a rate of return that was usually higher than that of foreign debt.
25. See P. Krugman, Balance Sheets, the Transfer Problem and Financial Crises, 1999, available at
web.mit.edu/krugman/www/.
26. Krugman (1979); Flood and Garber (1984).
27. Obstfeld (1986).
tack forces it to face the alternative of monetary contraction. The private
sector is in equilibrium in both cases, and multiple equilibria exist within a
range of costs and benefits of defending the exchange rate. Self-fulfilling
changes in expectations can then shift the economy from one equilibrium to
the other. The European Monetary System (EMS) currency crises of 1992,
it has been argued, fit into this framework.
28
Sachs, Tornell, and Velasco
apply a variant of the argument to the 1994 Mexican devaluation (see
below).
29
Finally, the Asian crises inspired third-generation models that empha-
size banking problems and the effect of moral hazard on lending and in-
vestments.
30
Calvo and Mendoza bring elements of this approach into
their interpretation of the Mexican crisis.
31
Although these elements are not entirely absent from the Mexican cri-
sis, none of these views fits the Mexican experience especially well. The
inconsistency between exchange rate policy and a monetary policy that
sterilized reserve losses, for example, was certainly a feature of economic
policy in 1994, but the reserve losses were the result of the speculative at-
tacks themselves rather than, as in the Krugman model, of a fiscal deficit
financed by the Central Banks expansion of credit which then led to a spec-
ulative attack. Trying to understand the crisis through such a model leaves
unexplained the very phenomenon that these first-generation models aim
to explain (namely, the speculative attack). The more gradual reserve losses
during the second half of the year, which resulted from the slowdown of
capital inflows in the face of a huge current account deficit, were validated
by the sterilization policies followed throughout the year. Here again, how-
ever, first-generation models cannot explain why the Central Bank fol-
lowed this course of action: there was no fiscal deficit to account for the
expansion of domestic credit.
Why, then, did the Central Bank validate the reserve losses instead of
adopting a tight monetary policy to reduce the current account deficit and
increase the capital account surplus? Such a tight monetary policy prob-
ably would not have prevented new speculative attacks and the eventual
collapse of the exchange rate, even over a short horizon. The reasons that
the Central Bank pursued the policy it did are related to the second- and
;aixi ios :,c
28. Obstfeld (1994).
29. Sachs, Tornell, and Velasco (1996).
30. See P. Krugman, Balance Sheets, the Transfer Problem and Financial Crises, 1999, available at
web.mit.edu/krugman/www/.
31. Calvo and Mendoza (1996).
third-generation models, though in rather subtle ways. First, as argued ear-
lier, the Central Bank had to consider the prospect of a severe banking cri-
sis. Banking problems thus had an important role in the developments
leading to the devaluation, as in recent models inspired by the Asian cri-
sis. This factor, however, did not arise because the currency crisis was a by-
product of a bank run caused by perceptions of insolvency or even because
government guarantees and moral hazard generated bad lending and a hid-
den fiscal deficit. Banking problems were important because the fragility
of the banking system, which resulted from the nature of the capital surge,
influenced the authorities decision to sterilize the reserve losses and dol-
larize the domestic debt.
Second, a contractionary monetary policy would have generated addi-
tional high costs. The fact that the exchange rate regime was eventually
abandoned reveals a growing conflict between exchange rate stability and
other objectives of economic policy, including employment and growth.
The costs of not devaluing, as perceived by the government, eventually be-
came too high. Ibarra applies this reasoning to explain why the exchange
rate policy was abandoned just after the change in administration in
December 1994, when the costs of devaluing (such as the loss of reputation)
could be minimized.
32
Moreover, the adoption of a tight monetary policy
would have lacked credibility in the eyes of investors, given the high costs in-
volved. High interest rates are often the prelude to speculative attacks pre-
cisely because they signal to speculators the high costs of defending the peg.
The endogeneity of economic policy, as emphasized by second-gener-
ation models of currency crises, was thus clearly relevant to the Mexican
experience. This does not mean, however, that the endogeneity of eco-
nomic policy necessarily generated multiple equilibria, as argued by Sachs,
Tornell, and Velasco.
33
They hold that multiple equilibria arise when a suf-
ficiently high (but not too high) level of government debt generates par-
tial credibility in the exchange rate peg. The governments incentive to
devalue then depends on self-fulfilling expectations of devaluation: if these
are low, interest rates allow debt to continue to be serviced without deval-
uing; if they are high, they create a strong incentive for the government to
reduce the value of the debt through devaluation. The interdependence
between exchange rate policy and devaluation expectations opens the door
to self-fulfilling outcomes.
rui xixicax icoxox\ ix :,,cs :,:
32. Ibarra (1997).
33. Sachs, Tornell, and Velasco (1996).
This argument has two unsatisfactory features. First, devaluation
could hardly inflate away the real value of the debt, which by this time was
mostly in dollars or dollar-indexed bonds. It is doubtful, then, that this
could have acted as an incentive to devalue, so the incentives for the gov-
ernment action on December 20 have to be found elsewhere. Second, the
argument neglects the effects that the change in government probably had
on the net costs of devaluing. As argued by Ibarra, the reduction in these
costsas perceived by a new administration that could blame the previ-
ous government for the necessary policy changewas probably a major
factor in the timing of the devaluation.
34
If this is the case, as I believe it
is, a model with multiple equilibria is not necessary for explaining the cur-
rency crisis. The shift in the perceived costs of devaluing probably moved
the economy from a situation in which the peg had enough credibility that
a single equilibrium existed (because nobody believed that the old admin-
istration would devalue on the eve of the change in government) to one in
which a commitment by the new government to pursue the same exchange
rate policy had so little credibility, given the costs of maintaining the pol-
icy, that again a single equilibrium (devaluation) existed.
All this suggests that a hybrid modelin which worsening policy trade-
offs and shifting perceptions of the costs of devaluation interact against the
background of real disequilibria, a large current account deficit, and a slow-
down in capital inflowsis perhaps the most useful approach for under-
standing the timing of the devaluation of December 1994. In this view, the
presence of banking problems, together with policymakers confidence
about the future course of capital inflows, decisively influenced the mone-
tary and debt management decisions of 1994. The strategy of dollarizing
debt to avoid tightening monetary policy had run its course, with practically
the whole stock of CETES existing in late 1993 converted into Tesobonos
by December 1994, and capital inflows had not recovered sufficiently. In
this context, the change in administration that took place in December
abruptly modified the perceived costs of devaluing.
The Debt Crisis
So far the paper has focused on the policy actions and events leading
to the devaluation. Understanding the currency crisis, however, is only half
of the story, for the Mexican crisis involved more than the exchange rate.
;aixi ios :,:
34. Ibarra (1997).
Multiple equilibria arguments become much more relevant in understand-
ing the second half of the story than they were in the first.
After November 1991, the peso-dollar parity fluctuated within a band
based on a fixed floor and a sliding ceiling (with a daily crawling peg of
0.04 centavos after October 1992). The peso depreciated within the band
by around 10 percent in 1994, and the bands ceiling was raised 15 per-
cent on December 20, 1994. The announcement of the new band ceiling
immediately sent the value of the dollar to the new 4 peso ceiling, and it
is estimated that U.S.$5 billion left the country in the course of two days.
The Central Bank then decided to withdraw from the foreign exchange
market. A rapid and substantial depreciation followed: the dollar parity
went from 4 pesos on December 20 to 6.35 on January 30 and 7.55 on
March 9, when an emergency economic program for 1995 finally stopped
the slide. Panic selling of government debt sent the risk premium on Tes-
obonos (measured as the interest rate differential between Tesobonos and
U.S. Treasury bills) from less than two percentage points before the deval-
uation to five percentage points at the end of December and twenty per-
centage points at the end of January 1995.
35
On December 27, a Tesobono
refinancing auction was cancelled because investors were no longer willing
to hold these obligations at interest rates acceptable to the Mexican au-
thorities. In the stock market, it is likely that some of the investors who
had their funds invested in pesos interpreted the devaluation as a breach
of contract. Their reaction was to withdraw their capital as soon as pos-
sible. Toward the second week of January, Mexico was on the verge of
default, and the financial markets of Latin America and other regions in-
creasingly began to be affected.
In sum, the collapse of the peso rapidly degenerated into a debt crisis.
Why was a government that had substantially reduced its foreign and do-
mestic liabilities since 1990 unable to roll over its debt in December 1994
and January 1995? Why was a rescue package of about U.S.$50 billion
eventually required to stop the collapse of the peso? The answer to these
questions is that fears of default surged in the face of the small level of re-
serves left at the Central Bank at the time of floating and the large amount
of short-term debt, particularly dollar-linked Tesobonos, coming due
throughout 1995. Moreover, the dollarization of the debt meant that de-
valuation did nothing to reduce the high ratio of Tesobonos to reserves,
while it increased the real value of the dollar debt in pesos. Through sim-
rui xixicax icoxox\ ix :,,cs :,,
35. Cole and Kehoe (1996); Ibarra (1997).
ple arithmetic calculations, investors saw that payments coming due in
1995 (which were estimated at about U.S.$50 billion, assuming that most
of the short-term debt would not be rolled over) were far greater than the
estimated resources available: international reserves in the Bank of Mexico
were about U.S.$6 billion. Individual investors may have been willing to
roll over their holdings of Mexican short-term debt if others did the same,
but when others panicked, few remained confident of being repaid. This
probably explains why the first international rescue package of U.S.$18
billion was unsuccessful in calming the markets. The rescue package plus
the international reserves would barely have covered half of Mexicos finan-
cial obligations for 1995. Stopping the collapse of the peso required a
package of around U.S.$50 billion, equivalent to the total debt payments
coming due in 1995.
36
Elements of this story have been formalized by Ibarra, who highlights
the role of the ratio of Tesobonos to reserves in the post-devaluation spec-
ulation against the peso, while Cole and Kehoe model the governments
inability to roll over its debt as a self-fulfilling debt crisis.
37
The latter may
seem paradoxical since Mexicos government debt fell continuously and
substantially in the six years preceding the crisis. Cole and Kehoes argu-
ment rests on the maturity structure of the debt (and not only on its level).
Their model identifies a crucial interval of debt for which the government
will be willing to service the old debt provided that it can roll it over, but
will default if it cannot sell new debt. Outside this interval, debt service is
presumably either so low that the government will not want to default
even if it fails to sell new debt or so high that it overwhelms the costs of
default. When government debt is in the critical interval, which the au-
thors call the crisis zone, multiple equilibria exist, and whether default oc-
;aixi ios :,
36. The final package included $20 billion of loan guarantees from the U.S. government, $17.8 bil-
lion of credits from the IMF, $10 billion in short-term loans from central banks via the Bank for In-
ternational Settlements, and several billion dollars of loans from other governments in North and
South America. The IMF loan amounted to over seven times Mexicos IMF quota and was unprece-
dented in the history of the IMF. The Mexican government declared its intention to reduce govern-
ment spending by 1.3 percent of GDP and to cut the amount of credit granted by state development
banks. A National Accord among workers, business, and government was set up to contain the infla-
tionary pressures arising from the devaluation. The Mexican government also emphasized its com-
mitment to market-oriented reforms and pledged to propose constitutional amendments to open
previously restricted areas of the economy to private investment and to increase foreign participation
in the domestic banking system. As a source of collateral, the Mexican government agreed to have im-
porters of Mexican oil make payments through an account at the Federal Reserve Bank of New York.
For an analysis of the package, see Lustig (1997).
37. Ibarra (1997); Cole and Kehoe (1996).
curs depends on whether the government can roll over its debt. A key fea-
ture of the model is that the size of the crisis zone depends on the average
length of maturity of government debt. With a very short maturity struc-
ture, the interval is large and includes levels of debt that are not high by
international standards, as was the case in Mexico before the crisis.
38
This line of reasoning has interesting implications. First, it suggests
that the post-devaluation panic could not have taken place if the devalua-
tion had occurred in the first quarter of 1994 (after the Colosio assassina-
tion), when reserves were still close to U.S.$20 billion, Tesobonos were
about U.S.$6 billion, and the average maturity of government bonds was
306 days (compared to 206 days in December). Given these figures, the
economy was probably still outside the crisis zone, and a devaluation at
that time would have looked more like the EMS crises of 1992 than the
Mexican debacle of 1994. Second, if multiple equilibria indeed existed,
this implies that while the post-devaluation financial panic in December
was a possibility, it was not inevitable. A Mexican government default was
obviously not in the collective interest of investors. The financial panic
that was forcing the government to default was a bad equilibrium that was
clearly inferior to what collective rationality dictated. To the extent that
the self-fulfilling debt crisis reflected the lack of mechanisms for organiz-
ing investors collective action, the argument suggests that the rescue pack-
age had a good chance of succeedingas it did.
Economic Performance after the Crisis and
the Medium-Term Prospects
The argument in this paper can be summarized as follows: the critical
problems faced by the Mexican economy in the mid-1990s resulted from
the fact that trade liberalization and the opening of the domestic bond and
stock markets took place in the midst of a persistent real overvaluation of
the peso and a high volatility of capital flows. The combination of trade
and exchange rate policies was unsustainable, particularly given the vul-
nerability generated by massive portfolio capital inflows from 1990 to
1993. In this sense, Mexicos success in attracting capital from abroad con-
tained the seeds of its subsequent failure.
rui xixicax icoxox\ ix :,,cs :,,
38. On the crisis or vulnerability zone, see also Ffrench-Davis (2000); Ffrench-Davis and Ocampo
(in this volume).
The legacy of the crisis will be long lasting. The effects on living stan-
dards and income distribution were dramatic: average real wages fell by 20
percent between 1994 and 1996. The devaluation caused a deep crisis in
the domestic banking sector, which at the time had net foreign currency
liabilities and plenty of bad performing loans, and its bailout will more or
less double the domestic public debt.
Despite these adverse consequences, however, the economy emerged
from the crisis stronger than it was before in several ways. After the deep re-
cession of 1995and in contrast to the period following the 1982 debt cri-
sisthe economy recovered a growth rate of about 5 percent per year, and
open unemployment fell to about 3 percent. Can the country sustain
growth at rates higher than those observed since the early 1980s? A number
of factors lend some credibility to such a scenario. First, the international
environment is more favorable today than in the aftermath of the 1982 debt
crisis, despite increased volatility and contagion effects such as those experi-
enced in the aftermath of the East Asian and Russian financial crises. The
United States, Mexicos main trading partner, has experienced the longest
peace-time economic expansion in history. Mexicos participation in
NAFTA has therefore been a bonus, which is reflected in the fact that
among Latin American countries Mexico was one of the least affected by the
Brazilian currency crisis. While other countries in the region experienced a
recession in 1999, the Mexican economy grew at 3.7 percent, with a contin-
ued deceleration of inflation to around 12 percent. Moreover, foreign inter-
est rates have remained low compared to the early 1980s, and in contrast to
the earlier crisis, the international rescue package successfully prevented the
financial panic from degenerating into a decade-long credit run in the inter-
national credit markets. In this context, fiscal and exchange rate policies that
ensure the economy is operating at full capacity and with a high real ex-
change rate have greater chances of success now than in the 1980s.
Second, the crisis itself may have positively altered the long-term mac-
roeconomic outlook. The huge real devaluation of the peso abruptly elim-
inated the unsustainable combination of trade and exchange rate policies
that played a major role in the process that generated the crisis. The mac-
roeconomic impact of export expansion has been enhanced in the context
of a more open economy and greater integration with the U.S. economy
under NAFTA. A comparison of the performance of Mexicos exports dur-
ing the 1995 crisis with the 1983 recession illustrates the positive legacy
of a more open economy. In 1995, the exchange rate increased by 86 per-
cent. In that year, maquiladora (in-bond assembly) exports increased by 30
percent, while nonoil and non-maquiladora exports expanded by 37 per-
;aixi ios :,o
rui xixicax icoxox\ ix :,,cs :,;
cent. In 1983, following a 466 percent increase in the exchange rate in
1982, maquiladora exports increased by 29 percent while non-maquiladora
and nonoil exports fell by nearly 6 percent. Contrary to what happened in
the period 198894, net exports expanded after 1995.
39
Given its magnitude, the real devaluation of the peso also contributed
to a reversal of the factors underlying the profit squeeze in the tradables
sector and the fall of the private savings rate. The positive effects on the
profitability of the tradables sector must, in turn, have contributed to the
observed increase in FDI. These investments now represent nearly 20 per-
cent of total private investment, and between 1996 and 1998 they ac-
counted for around 60 percent of total foreign investment (including
portfolio inflows). Consequently, a much lower portion of the current ac-
count deficit is now being financed by short-term capital than was the case
in the first half of the 1990s, thus making the economy less vulnerable to
new domestic or external shocks. Private savings increased substantially in
1995 and continued to increase up to 1997. The increase was probably
due not only to the real devaluation, but also to the domestic banking cri-
sis and the associated credit crunch.
As credit constraints become less acute, however, the upward trend of
private savings may be reversed. The domestic savings rate declined signif-
icantly in 1998, which most probably reflects a fall in private savings.
Moreover, the banking crisis is far from resolved, and the bailout of the
banking system will impose a large burden on the fiscal accounts in the
years to come. With the exception of the banking sector, the economy is
more robust than in the early 1990s, but it is too early to tell whether pol-
icymakers have learned from the experience of 199094. The sources of
policy concern associated with capital surges (namely, appreciation, re-
source misallocation, and financial fragility) are not nearly as serious today
as they were in the first half of the 1990s. This is the case, however, sim-
ply because the magnitude of capital inflows did not return to their previ-
ous levels after Mexico regained access to international capital markets.
Appendix
The model of the goods and financial markets used in the paper has
two equations, corresponding to the conditions of goods market (or flow)
equilibrium and financial market (or stock) equilibrium. The equation of
39. See Ros (1999) for a full discussion.
the IX curve is a locus of real exchange rate (e) and real interest rate (r)
combinations along which desired investment (I ) is equal to national sav-
ings (S
d
) plus foreign savings (S
f
):
(1) I (r, e) = S
d
(L, D*) + S
f
(e, L, D*),
where Ir < 0, Ie > 0, S
d
L > 0, S
d
D* > 0, S
f
e < 0, S
f
L > 0, and S
f
D*
> 0, and where L is total employment, which positively affects both do-
mestic and foreign savings (in the last case, through its effects on the trade
deficit) and D* is the stock of external liabilities, which positively affects
domestic savings on account of wealth effects and positively affects the
current account deficit (S
f
) on account of the higher interest payments
abroad associated with a higher level of D*. Investment depends positively
on the real exchange rate (on account of its positive effects on profitabil-
ity) and inversely on the real interest rate. Given total employment, this
IX locus is one along which the nontraded goods market clears through
changes in the real exchange rate.
The equation of the financial market (FM) curve refers to the equal-
ity between the risk premium (u) on domestic securities and the interest
rate differential between domestic and foreign short-term securities, ad-
justed for exchange rate expectations. The risk premium is assumed to be
a function of an exogenous component (u
o
) reflecting investors subjective
perceptions on country risk and the stock of external liabilities (D*):
40
(2) r = r* + e
e
+ u and u = u (D*, u
o
),
where r* is the real foreign interest rate. Expected depreciation (e
e
) is as-
sumed to depend on both the future short-term equilibrium exchange rate
(e
+1
), which for simplicity is given by the solution of the model, and the
long-term equilibrium exchange rate (e*), which is defined as the value of
e consistent with a sustainable current account deficit and full employ-
ment:
e
e
= a (e
+1
- e) + b (e* - e),
where 0 < a < 1 and 0 < b < 1.
Exchange rate expectations are thus subject to two forces that tend to
offset each other whenever the short-term and long-term equilibrium ex-
change rates move in opposite directions. Substituting this mechanism of
;aixi ios :,
40. In principle, the domestically held stock of government securities should also be included. The
specification in equation 2 is justified if the risk of default on domestically held securities depends on D*.
expectations formation into equation 2 and solving for e gives the equa-
tion of the FM locus:
(3)
where u = u (D*, u
o
).
Equation 3 defines a downward sloping line in (e, r) space: a fall in
the exchange rate tends to generate expectations of depreciation (for given
values of e* and e
+1
), and this requires an increase in the domestic inter-
est rate to maintain financial market equilibrium.
References
Alarcn, D., and T. McKinley. 1997. The Paradox of Narrowing Wage Differentials and
Widening Wage Inequality in Mexico. Development and Change 28(3): 50530.
Atkeson, A., and J. V. Ros-Rull. 1996. The Balance of Payments and Borrowing Con-
straints: An Alternative View of the Mexican Crisis. Journal of International Economics
41 (November): 33150.
Branson, W. 1985. Causes of Appreciation and Volatility of the Dollar. In The U.S. Dol-
lar: Recent Developments, Outlook and Policy Options, edited by the Federal Reserve Bank
of Kansas City, 3363.
Calvo, G., and E. Mendoza. 1996. Mexicos Balance of Payments Crisis: A Chronicle of
a Death Foretold. Journal of International Economics 41 (November): 23564.
Cole, H., and T. Kehoe. 1996. A Self-Fulfilling Model of Mexicos 199495 Debt Crisis.
Journal of International Economics 41 (November): 30930.
Cragg, M., and M. Epelbaum. 1996. Why Has Wage Dispersion Grown in Mexico? Is It
the Incidence of Reforms or the Growing Demand for Skills? Journal of Development
Economics 51(1): 99116.
Dadush, U., A. Dhareshwar, and R. Johannes. 1994. Are Private Capital Flows to Devel-
oping Countries Sustainable? Policy Research Working Paper 1397. World Bank.
Dornbusch, R., and A. Werner. 1994. Mexico: Stabilization, Reform, and No Growth.
BPEA 1:1994, 253315.
ECLAC (United Nations Economic Commission for Latin America and the Caribbean).
1995. Latin America and the Caribbean: Policies to Improve Linkages with the Global Econ-
omy. Santiago.
. 1998. Amrica Latina y el Caribe: polticas para mejorar la insercin en la economa
mundial, 2d ed. Santiago: Fondo de Cultura Econmica and ECLAC.
Ffrench-Davis, R. 2000. Reforming the Reforms in Latin America: Macroeconomics, Trade,
Finance. Macmillan.
Flood, R., and P. Garber. 1984. Collapsing Exchange Rate Regimes: Some Linear Exam-
ples. Journal of International Economics 17(1/2): 114.
e
b
a b
e
a
a b
e
a b
r r u =
+
L
N
M
O
Q
P
+
+
L
N
M
O
Q
P
-
+
L
N
M
O
Q
P
- -
+
( )
*
( ) ( )
( * ),
1
1
rui xixicax icoxox\ ix :,,cs :,,
Gurra, J. A. 1995. Capital Flows: The Mexican Case. In Coping with Capital Surges, ed-
ited by R. Ffrench-Davis and S. Griffith-Jones. Lynne Rienner.
Hanson, G., and A. Harrison. 1995. Trade, Technology and Wage Inequality in Mexico.
Working Paper 5110. Cambridge, Mass.: National Bureau of Economic Research.
Ibarra, C. 1997. Three Essays on the 199495 Mexican Currency Crisis. Ph.D. disser-
tation, University of Notre Dame.
Krugman, P. 1979. A Model of Balance of Payments Crisis. Journal of Money, Credit and
Banking 11(3): 31125.
Lustig, N. 1997. The United States to the Rescue: Financial Assistance to Mexico in 1982
and 1995. CEPAL Review 61 (April): 4162.
Lustig, N., and J. Ros. 1998. Las reformas econmicas, las polticas de estabilizacin y el
sndrome mexicano. Desarrollo econmico 37(148). January-March: 50332.
Lustig, N., and M. Szkely. 1998. Economic Trends, Poverty and Inequality in Mexico.
Inter-American Development Bank (IDB). Mimeographed.
Obstfeld, M. 1986. Rational and Self-Fulfilling Balance of Payments Crises. American
Economic Review 76 (1): 7281.
. 1994. The Logic of Currency Crises. Working Paper. University of California,
Center for German and European Studies (January).
Ros, J. 1994a. La economa mexicana en el largo plazo. Paper prepared for the Congreso
Mexicano sobre Prospectiva, Mexico City. Fundacin Barros Sierra (September 2627).
. 1994b. Mercados financieros y flujos de capital en Mxico. In Los capitales ex-
ternos en las economas latinoamericanas, edited by J. A. Ocampo. Bogot: Fedesarrollo
and Inter-American Development Bank (IDB).
. 1995a. Mercados financieros, flujos de capital y tipo de cambio en Mxico.
Economa mexicana CIDE 4(1): 567. Mexico City: Centro de Investigacin y Docen-
cia Econmicas (CIDE).
. 1995b. Trade Liberalization with Real Appreciation and Slow Growth: Sustain-
ability Issues in Mexicos Trade Policy Reform. In Manufacturing for Export in the De-
veloping World, edited by G. Helleiner. Routledge.
. 1999. La liberalizacin de la balanza de pagos en Mxico: efectos en el creci-
miento, el empleo y la desigualdad salarial. Mimeographed.
Ros, J., and N. Lustig. 1999. Trade and Financial Liberalization with Volatile Capital In-
flows: Macroeconomic Consequences and Social Impacts in Mexico during the 1990s.
Paper prepared for the Conference on Globalization and Social Policy, New York. New
School University, Center for Economic Policy Analysis (CEPA) (January 1012).
Ros, J., and P. Skott. 1998. Dynamic Effects of Trade Liberalization and Currency Overval-
uation under Conditions of Increasing Returns. The Manchester School 66(4): 46689.
Sachs, J., A. Tornell, and A. Velasco. 1995. The Collapse of the Mexican Peso: What Have
We Learned? Working Paper 5142. Cambridge, Mass.: National Bureau of Economic
Research.
. 1996. The Mexican Peso Crisis: Sudden Death or Death Foretold? Journal of
International Economics 41 (November): 26584.
Trigueros, I. 1998. Capital Inflows and Investment Performance: Mexico. In Capital Flows
and Investment Performance, edited by R. Ffrench-Davis and H. Reisen. Paris: Organiza-
tion for Economic Cooperation and Development (OECD), Development Center.
;aixi ios :c
5
::
T
he deep integration of developing countries into
the global economy has many positive effects. In
particular, capital flows to developing countries have important bene-
fits. These benefits are especially clear in the case of foreign direct invest-
ment (FDI), which not only is fairly stable but also brings technological
know-how and access to markets. Other external flows also have impor-
tant positive microeconomic effects, such as lowering the cost of capital
for creditworthy firms. At a macroeconomic level, foreign capital flows can
complement domestic savings, leading to higher investment and growth.
However, large surges of short-term and potentially reversible capital
flows to developing countries can have very negative effects. First, these
surges pose complex policy dilemmas for macroeconomic management, as
they can initially push key macroeconomic variables, such as exchange
rates and prices of assets like property and shares, away from what could
be considered their long-term equilibrium.
1
Second, and more important,
such flows pose the risk of sharp reversals, which can result in very serious
losses of output, investment, and employment, as well as dramatic in-
creases in poverty, particularly if they lead to currency and financial crises.
This was clearly illustrated by the impact of the crisis in Asia, which
sriiuax\ ciiiiiru-;oxis
*
An International
Financial Architecture for
Crisis Prevention
I thank Ricardo Ffrench-Davis and Jose Antonio Ocampo for insightful suggestions and Jenny
Kimmis and Jacques Cailloux for their valuable inputs.
1. Ffrench-Davis and Griffith-Jones (1995).
quickly spread to many other countries, including Brazil. The East Asian
countries that were hit by the crisis saw their gross domestic product
(GDP) fall by 8 percent, on average, in 1998; Mexico saw its GDP fall by
almost 7 percent in 1995. Fortunately, output has tended to recover fairly
quickly in most cases. The negative effects of a crisis, however, are not just
the significant short-term costs, but also include the long-term negative ef-
fect on lower productive investment stemming from increased uncertainty.
Asian-style currency crisesand their extremely high development
costsraise a very serious concern about the net development benefits
that developing countries receive from large flows of potentially reversible
short-term international capital. While the high costs of reversals of those
flows are evident, the benefits are less clear. This contrasts sharply with
FDI and trade flows, in which the large developmental benefits clearly
outweigh the costs. As a result, volatile short-term capital flows emerge as
a potential Achilles heel for the globalized economy and for the market
economy in developing countries. If the international community and na-
tional authorities do not learn to manage these flows better, such volatile
flows could seriously undermine the tremendous benefits that globaliza-
tion and free markets can otherwise bring.
Causes of Currency Crises
The causes of currency crises always include domestic causes, which
differ from case to case. The East Asian countries, for example, displayed
serious problems, including important weaknesses in the regulation of
their domestic financial systems. In particular, short-term private debt in-
curred by both banks and nonfinancial firms had been poorly monitored
and regulated. Some mistakes also seem to have been made in the liberal-
ization of the capital accounts, as this was reportedly implemented in ways
that encouraged short-term flows. Furthermore, several of the East Asian
countries had fixed exchange rate policies that pegged their currencies to
the U.S. dollar.
However, an important cause of the East Asian crisis was not signifi-
cant changes in several macroeconomic fundamentals, but rather a sharp
deterioration in confidence throughout the region, which spread through
contagion effects. Indeed, the most disturbing element in the East Asian
crisis was that it affected countries with long track records of good eco-
nomic management, countries that had been remarkably successful over
sriiuax\ ciiiiiru-;oxis ::
extended periods in terms of economic growth, dynamism of their export
sectors, low rates of inflation, fiscal surpluses, and high rates of saving.
A crucial causal factor for explaining crises thus relates to the behav-
ior of international capital flows. This aspect is linked to certain imperfec-
tions of international capital markets. These imperfections almost always
featured in the financial panics of earlier times, but their impact has in-
creased significantly as a result of the speed with which capital markets can
react in todays global economy, aided by highly sophisticated information
technology. Paradoxically, this impact appears to be strongest for econo-
mies that are highly successful or that are perceived to be in the process of
becoming so. Such conditions trigger a perverse interaction between eu-
phoria in international capital markets and complacency by governments
in recipient countries.
Successful economies offer high returns by way of yields as well as cap-
ital gains. If international investors can find ways to enter these economies,
or if their entrance is facilitated by capital account liberalization, they tend
to rush in, generating a surge of capital inflows that affects key economic
variables. Exchange rates become overvalued; the prices of key assets, such
as shares or real estate, rise quickly and sharply. Both real income and per-
ceived wealth increase. Banks assume that current trends will continue and
therefore tend to relax lending standards, lifting liquidity constraints on
firms. The balance of payments deteriorates, often quite rapidly, as both
consumption and investment rise. Initially, this is not seen as a problem,
since foreign lenders and investors are willing to continue lending and in-
vesting. Economic authorities thus delay necessary adjustment, confident
that their previous success will continue and that crises happen elsewhere.
Then something changes. The change may be domestic or interna-
tional, economic or political, important or relatively small. Whatever it is,
it triggers a sharp modification in perceptions and leads to a large fall in
confidence in the economy among internationally mobile investors, that is,
both foreign and national investors who are able to take their liquid assets
out. The change of perception tends to be both large and quick. A country
that was perceived as a successful economy or a successful reformerfor
which no amount of praise was sufficientis suddenly seen as fragile, risky,
and crisis prone. The media, and in particular the specialized financial
press, contribute in an important way to creatingor magnifyingrapid
changes in perceptions: the media suddenly tend to focus on bad short-
term news and to underplay more favorable long-term fundamentals. The
rating agencies also tend to play a procyclical role, accentuating changes of
iixaxciai aicuiricruii ioi ciisis iiivixriox :,
perceptions. The change of perception tends to be far larger than the mag-
nitude of the underlying change in macroeconomic fundamentals. Fur-
thermore, any weakness in fundamentals is then discovered and magnified
by markets. There can be much overshooting. Exchange rates collapse, and
stock markets and property prices also fall sharply.
This pattern helps explain the currency and banking crises in the
Southern Cone of Latin America in the early 1980s, as well as the Mexi-
can peso crisis of 1994. It also provides important elements for under-
standing the 1997 East Asian crisis and the more recent crisis in Brazil.
There are significant differences among these crises, of course, and among
the crises in the East Asian economies, particularly with regard to domes-
tic causes. However, they demonstrate a common denominator in the
boom-bust behavior of short-term lenders and investors, driven not just
by real trends, but by dramatic changes in perceptions. The complacency
of the economic authorities in recipient countries during the period of
boom represents another common feature.
Capital and financial markets are special in that although they gener-
ally function well, they are prone to important imperfections.
2
Asymmetric
information and adverse selection play an important role in explaining
these imperfections, since financial markets are particularly information in-
tensive.
3
Furthermore, financial markets offer strong incentives for herd-
ing, as individual short-term investors, lenders, or fund managers try to
choose the investment or loan that they think is most likely to be chosen by
other investors or lenders. Structures within private financial institutions
also cause distortions by giving insufficient influence on decisionmaking to
research departments concerned with analysis and risk assessment. Further
investigation is required to explore the microeconomics of speculation car-
ried out by different categories of market actors. This could be helpful for
designing measures, for example, to be taken by the financial industry it-
self, aimed at helping to inhibit excessive speculative behavior; this would
imply designing new systems of reward and punishment that would help
prevent crises.
Another relevant area for understanding the East Asian crisis is the
analysis of self-fulfilling attacks, that is, crises that arise without obvious
current policy inconsistencies. In such cases, the attitude of speculators
and investors is crucial to whether an attack occurs. The existence of self-
sriiuax\ ciiiiiru-;oxis :
2. Griffith-Jones (1998).
3. Stiglitz (1994).
fulfilling attacks and multiple equilibria for exchange rates and other key
variables implies that good macroeconomic fundamentals are an impor-
tant and necessarybut not sufficientcondition for avoiding currency
crises. There is, at present, a limited understanding of what triggers self-
fulfilling attacks.
4
Market actors tend to give rather different explanations
for the recent crises. Policymakers in developing countries therefore face
the daunting task of playing to moving goalposts to avoid crises. However,
certain conditions of vulnerability can be identified (such as high current
account deficits as a proportion of GDP or the ratio of external liabilities
like net short-term debt, which can easily be reversed). Further investiga-
tion on conditions of vulnerability and the nature of triggering events is
clearly required to facilitate the prediction of risk and, above all, to im-
prove the prevention of currency crises.
5
Lessons for Crisis Prevention
Given the volatility and reversibility of some categories of capital flows,
the costs of these flows to countries development may be higher than their
benefits, at least during certain periods of time. Consequently, there is
a growing consensus among leading policymakersin developed as well
as developing countriesand market actors that important changes are
urgently needed both in the international monetary system as a whole and
in the policies of recipient countries. By aiming to avoid costly crises and to
manage them better if they do occur, these reforms will help limit the neg-
ative effects of volatile capital flows. Care must be taken, however, that the
measures adopted contribute to broadening access by all developing coun-
tries to capital flows, particularly long-term flows.
Actions that are urgently needed include identifying the possible
changes required for achieving these aims; ensuring that new policies are
based on an in-depth understanding of how different market actors oper-
ate; carefully evaluating the potential economic effects of such changes;
and adopting the required measures, including possible institutional de-
velopments to facilitate their implementation.
The measures discussed below can be divided into three groups.
Measures that can be taken within existing institutional arrange-
iixaxciai aicuiricruii ioi ciisis iiivixriox :,
4. Wyplosz (1998).
5. See Ffrench-Davis and Ocampo (in this volume).
ments. Two important examples are, first, the more expansive monetary
policy pursued by the U.S. and European central banks in the second half
of 1998 to deter the worsening of financial crises and, second, changes in
the capital adequacy requirements for short-term and long-term lending
that could be initiated in the context of the expanded Basel Committee on
Banking Supervision.
Measures that require some development, expansion, and adapta-
tion of existing institutions, such as the International Monetary Fund
(IMF) or the Regulatory Committees that meet under the aegis of the
Bank for International Settlements (BIS). Examples include creating new
facilities within the IMF (or adapting existing ones) to cope with capital
accountcaused currency crises (see below); expanding the Financial
Stability Forum or establishing a world financial authority to provide ap-
propriate surveillance and prudential regulation of financial intermediar-
ies, not only in developed and developing countries, but also globally; and
filling international regulatory gaps to encompass the national regulation
of mutual funds and hedge funds and their international coordination.
Measures that require more institutional radicalism, in the sense of
creating new institutions or drastically adapting existing ones. While this
is clearly desirable from the perspective of designing institutions and
mechanisms to meet the new needs of a globalized private financial sys-
tem, it is significantly more difficult to achieve. The key problem is that
there is no global government to create the new global institutions that are
needed to effectively manage a globalized private financial system, and the
political process for national governments to create global institutions will
undoubtedly be complex and slow. It seems highly desirable, however, to
develop a clear vision of an appropriate new international financial archi-
tecture that would allow an orderly global financial market to support the
development process and that would avoid developmentally and finan-
cially costly crises. Such a vision should inform current debates on a new
financial architecture.
This paper focuses on issues of crisis prevention. The measures dis-
cussed here emphasize improved information and regulation. They attempt
precisely to deal with the international capital market imperfections out-
lined above, which have played such an important part in causing recurrent
currency crises.
6
Because crises are so costly, not only to developing coun-
sriiuax\ ciiiiiru-;oxis :o
6. For an analysis of better crisis management, see, for example, Griffith-Jones, Ocampo, and Cail-
loux (1999) (also available at www.ids.ac.uk/ids/global/finance/intfin.html).
tries, but also to the international financial community, utmost priority
needs to be given to crisis prevention measures. As in medicine, so in fi-
nance: prevention is much kinder, and more efficient, than cure.
The focus of the international community in the aftermath of the
Asian crisis was on better information, on the one hand, and strengthen-
ing the financial system, on the other. This focus was clearly illustrated by
the working groups set up under the G-22, an ad hoc group made up of
the Group of Seven (G-7) and developing countries, which looked at three
areas: enhancing transparency and the disclosure of information; strength-
ening financial systems both in national economies and globally; and
appropriate burden-sharing in the event of international financial crises.
This section briefly discusses the first two areas and examines some
stronger and possibly more effective crisis prevention strategies.
Transparency and Information Disclosure
One of the areas initially defined by the G-7 countries and the IMF
as central for future crisis avoidance was enhancing transparency and the
disclosure of timely and reliable information, basically on developing
countries, and facilitating its dissemination to market actors. A flurry of
activity in improving information followed. Some of the key data gaps and
deficiencies identified were information on foreign exchange reserves, in-
cluding undisclosed forward positions; maturity and currency exposures of
the public and private sectors; and the health of the financial system, in-
cluding information on nonperforming loans.
A number of steps have already been taken. The most significant are
the publication of public information notices (PINs) by individual coun-
tries and the IMFs strengthening of the Special Data Dissemination Stan-
dard (SDSS), the information standard that the IMF established in 1996
after the Mexican peso crisis. The PINs are prepared yearly by all countries
after their Article IV consultation with the IMF, and countries are encour-
aged to release them speedily. The IMF has also started a pilot program for
voluntary release of Article IV staff reports. The strengthening of the SDSS
in the areas of international reserves and external debt has been particularly
important. Reports now incorporate full details on reserves and any claims
against them (for all countries), a practice that was instituted in April 1999.
In addition to information standards, a number of other standards
have been defined by the IMF and the BIS, in collaboration with institu-
tions like the World Bank and the Organization for Economic Cooperation
iixaxciai aicuiricruii ioi ciisis iiivixriox :;
and Development (OECD). The aim is to provide codes of good practice.
The IMF will promote the dissemination of these standards by incorporat-
ing them as conditions for IMF lending. The standards include creating
codes of good practice for fiscal transparency and for monetary and financial
policies, improving the quality of banking supervision, and addressing is-
sues such as the functioning of financial systems, bankruptcy, and corporate
governance.
Though the implementation of these standards may have very posi-
tive effects, such as strengthening financial systems, two rather serious con-
cerns need to be raised. First, is the definition of desirable standards a
sufficiently participatory process, in which the developing countries that
will be asked to implement the standards have had enough input in their
definition? Should developing countries just be encouraged to adopt these
standards, rather than making their adoption a condition of IMF support?
These related concerns can be summarized in the phrase, no standardiza-
tion without participation. Second, will implementing these standards
impose excessive administrative and other burdens on governments in
developing countries, while not always generating clear benefits for their
economies?
iixirs oi ruis aiiioacu Improved information, along the lines of the
changes described above, will clearly be very valuable and will contribute
to better market performance. By itself, however, improved information
on developing countries most probably will not avert crises. First, the in-
formation available to financial markets will never be perfect, and infor-
mation asymmetries will always exist. Second, better information alone is
not sufficient to ensure that financial markets function well, since the key
issue is how market actors process and act on the available information.
Phenomena such as euphoria and herding imply that bad news is ignored
in periods of boom and magnified in periods of bust, with the reverse
being true for good news.
7
Third, better information on developing coun-
tries has to be complemented by equally important improvements in in-
formation on international financial markets.
With regard to the first point, theoretical analysis and practical expe-
rience both clearly show that information will continue to be imperfect
and that this may cause or contribute to financial crises. A clear forerun-
ner of much of the literature on imperfect information is Keynes, who
sriiuax\ ciiiiiru-;oxis :
7. See, for example, Griffith-Jones (1998).
stresses the extreme precariousness of the basis of knowledge on which
our estimates of prospective yields have to be made.
8
More recently,
Eichengreen rather strongly summarizes the limits of improved informa-
tion for crises prevention: relying excessively on improved transparency
underestimates the extent to which information asymmetries are intrin-
sic to financial markets. The advocates of information-related initiatives
mislead when they assume the problem away.
9
Indeed, sophisticated and increasingly informed financial markets
continue to be extremely volatile. This is the case even in some of the most
developed economies in the world. Serious problems and even crises have
occurred in banking systems that had the highest ratings on transparency,
as illustrated by the banking crises in Scandinavian countries.
One very important reason for imperfect information is that much of
the relevant information to which the market reacts comes only with a lag,
and it depends on macroeconomic conditions not entirely known in ad-
vance. Increasing information that may be relevant to improving microeco-
nomic market efficiency may do little to reduce macroeconomic volatility.
10
Another source of problems has to do with the processing of informa-
tion. As pointed out above, the key issue is that investors (and lenders) are
increasingly concerned not with what an investment is really worth to a
person who buys it for keeps, but with what its market value will be in a
few hours or days. To the extent that this is true, available information on
developing countries will be less important than how the average market
participant is likely to perceive it.
Microeconomic factors, or how financial firms and banks operate, re-
inforce such problems. This may be related to both costs and firm organ-
ization. The board of a financial institution that is deciding whether to
invest or lend to a particular country may not be able (or willing) to take
account of the rich information available in the research departments of
that same institution.
11
Smaller banks, with small research departments,
generally rely even less on their internal expertise, tending instead to fol-
low the decisions of other banks. As a result, changes in the opinions of
those investors that are considered to be informed may cause noninformed
iixaxciai aicuiricruii ioi ciisis iiivixriox :,
8. Keynes (1936, chap.12).
9. Eichengreen (1999, p. 80).
10. Ocampo (1999).
11. A recent survey of banks carried out by the Bank for International Settlements shows that most
banks take decisions without considering the information available in research and other departments
within their own bank.
sriiuax\ ciiiiiru-;oxis :,c
investors to overreact, since they rely on the formers lead to make their
decisions.
A key problem in this process is that changes in opinion can occur with-
out any significant change of underlying fundamentals. Basically the same
information about a country may be interpreted totally differently at differ-
ent times, because of factors such as the mood of the markets, events in other
economies, and so forth. (This effect occurs particularly when a countrys
economy is in a vulnerable zone.) A good analogy is that a glass may be
deemed either half empty or half full, depending mainly on the mood of the
viewer and not so much on how clean and transparent the glass is.
A related issue is that of rating agencies. The important weight of sub-
jective factors in rating agencies risk analysis, together with the significant
volatility that has characterized such subjective opinions in the past, has
led to suggestions that these agencies could be regulated and that they
could improve their risk evaluations. Moreover, the role of these institu-
tions in sovereign ratings could be eliminated, substituting it with ratings
made by the supervisory agencies of the countries in which capital flows
originate, according to internationally established parameters. This would
imply that rating agencies activities would be concentrated in the evalua-
tion of private risks.
12
iiovioixc oiviioiixc couxriiis wiru aooirioxai ixioixariox ox
xaixirs As pointed out above, providing markets with better informa-
tion on developing countries has to be complemented by providing devel-
oping countries with better information on international financial
markets. During the crisis that started in Asia, in particular, policymakers
in emerging countries (and specifically central banks) encountered impor-
tant limitations in the essential information available on the functioning
of international capital and banking markets.
13
Authorities require more
information both on long-term structural changes in these markets and es-
pecially on day-to-day changes in the functioning of global and regional
marketsand their key actors.
Just as the IMF has led the process of improving the collection and dis-
semination of information on emerging market economies, which is partic-
ularly useful to markets, a parallel effort needs to be undertaken to gather
and provide timely information on market evolution to policymakers in
12. Ocampo (1999).
13. Based on interviews and personal experience.
iixaxciai aicuiricruii ioi ciisis iiivixriox :,:
emerging markets. This task should perhaps be led by the BIS and coordi-
nated by the newly created Financial Stability Forum, although inputs
from other institutions, such as the IMF and the private sector (perhaps
through the Institute of International Finance), would be very valuable.
These data requirements relate not just to better statistics on international
banks exposures, but also to better information on the international expo-
sures of investment banks, hedge funds, and other institutional investors,
including pension funds and mutual funds.
The growth of financial innovations, such as over-the-counter deriva-
tives, have made financial markets more complex and opaque. This has
created difficulties in monitoring patterns of activity in these markets and
the distribution of risks in the global financial system. The BIS has re-
sponded to this situation by drawing up a framework for the more regu-
lar collection of statistics on over-the-counter derivatives markets. Such
efforts to improve transparency, particularly in relation to derivatives and
highly leveraged institutions (such as hedge funds), are widely welcomed,
but far greater frequency is required. Major central banks and the BIS
could improve the registration of derivatives and institutions like hedge
funds by making it obligatory.
14
In addition, developing countries should
participate in the relevant working groups where information needs are
discussed and decided, so that their information needs on markets are fully
considered. In this sense, proposals by the Financial Stability Forum to im-
prove information on hedge funds and other highly leveraged institutions
are positive. Their early implementation, which must include the so-called
tax havens where access to information is particularly limited, is crucial.
Given the speed with which markets move, the frequency with which
relevant data are produced must be very high (and possibly higher in times
of market turbulence, when it becomes particularly crucial), and the in-
formation must be disseminated immediately to all countries central
banks. In this regard, the BIS could provide a special additional service in
which it would play the role of a clearing house of information. Such an
effort would require not only that the BIS disseminate the information it
gathers directly from markets, but also that the organization collect and
centralize the data that individual central banks maintain on their markets
and then use that data to compile an aggregate picture, which is outside
the scope of any individual central bank. Working via the Internet, the BIS
could standardize information requirements, collect the information, ag-
14. Based on interviews.
gregate it, and disseminate it rapidly to all central banks, as well as to other
relevant institutions.
Regulation within Capital-Receiving Countries
Though this study largely focuses on changes in international finan-
cial arrangements, establishing appropriate policies in capital-receiving
countries is clearly also of great importance. Considerable responsibility
for discouraging excessive reversible inflows lies with recipient economies.
Recipient countries have the greatest degree of freedom for policy-
making in the period of excessive surges of capital inflows. Countercycli-
cal monetary and fiscal policies are essential for reducing the excessive
growth of domestic absorption or current account deficits or both. The re-
cent experience and literature indicate that a tightening of macroeconomic
policies is particularly desirable when indicators of vulnerability to finan-
cial and currency crises start to deteriorate quickly or pass certain thresh-
olds. This includes when current account deficits start to grow rapidly,
when the proportion of easily reversible capital flows in total flows is high
and rising, and, particularly, when liquid external liabilities grow rapidly
and approach the level of or even exceed foreign exchange reserves. High
levels of foreign exchange reserves and limits on the level of short-term ex-
ternal liabilities are thus crucial for currency crisis avoidance.
A countercyclical approach should also be applied to the supervision
and regulation of the financial system, especially the banking system. In
boom times, the supervision and regulation of banksas well as banks
own credit decisionsshould not simply be based on expectations of a
continued growth scenario among borrowers. Potential downside risks, in-
cluding a slowdown or reversal of capital and a slowdown of the economy
slows, need to be considered in both credit decisions and supervisory eval-
uations of loans. The risk that the value of collateral, such as property, may
fall sharply in a possible future crisis also needs to be considered in boom
times. A possible measure for regulators to use when asset prices have gen-
erally been rising fast is to limit the value of collateral, for example, to an
average of the past three or five years or to a certain percentage of its value.
This countercyclical approach would moderate booms in domestic bank
lending, which often exacerbate the impact of excessive surges of capital
inflows.
If surges of potentially reversible capital are excessive, it may also be
appropriate for recipient countries to take measures to discourage them
sriiuax\ ciiiiiru-;oxis :,:
temporarily. Some countries (for example, Chile and Colombia) have im-
plemented measures such as taxes and nonremunerated reserve require-
ments on flows during a fixed period with this objective. Their aim has
been threefold: to change the structure of capital inflows by discouraging
short-term and potentially reversible flows; to increase the autonomy of
domestic monetary policy; and to curb large overvaluation of the exchange
rate.
15
There is growing evidence that such measures to discourage exces-
sive inflows have contributed to a relatively more successful management
of capital inflows in countries that have applied them.
16
Chiles measures,
in particular, contain two attractive features: they are market-based, rather
than quantitative, and they apply to practically all short-term flows, which
simplifies administrative procedures and reduces (though does not elimi-
nate) possibilities of evasion. Other domestic measures that can be useful
for discouraging excessive short-term borrowing include modifying taxes
on companies external borrowing.
The major international financial institutions now explicitly recognize
that while they have some limitations and minor microeconomic disad-
vantages, market measures taken by recipient governments to discourage
excessive short-term capital flows can play a positive role if they are part
of a package of policy measures that includes sound macroeconomic fun-
damentals and a strong, well-regulated domestic financial system.
International Regulatory Measures
Recipient countries clearly carry an important part of the responsibil-
ity for discouraging excessive reversible inflows, as well as managing them.
However, the large scale of international fundscompared to the small
size of developing country marketsleads to the question of whether
measures to discourage excessive short-term capital inflows by recipient
countries are sufficient for dealing with capital surges and the risk of re-
versal. Source countries must take complementary action, for three strong
reasons. First, not all major recipient countries will be willing to discou-
rage short-term capital inflows, and some may even encourage them. For
instance, the tax and regulatory measures implemented in Thailand to en-
courage the Bangkok International Banking Facility encouraged short-
term borrowing. Second, even recipient countries that have deployed a
iixaxciai aicuiricruii ioi ciisis iiivixriox :,,
15. Ffrench Davis and Griffith-Jones (1995).
16. See Ffrench-Davis and Tapia (in this volume).
battery of measures to discourage short-term capital inflows have on occa-
sion found these measures to be insufficient for stemming massive inflows.
Third, if major emerging countries experience attacks on their currencies,
which result in difficulties in servicing their debt, they will be forced to
seek large official funding. This fact creates a clear need for international
or source country regulation that will discourage excessive reversible capi-
tal inflows during booms. Without it, international private investors and
creditors might continue to assume excessive risks, based on the euphoria
during booms combined with the knowledge that they will be bailed-out
if the situation becomes critical. The latter is the classical moral hazard
problem.
The international financial crisis provoked a serious debate on how the
surveillance and supervision of the international financial system could be
strengthened to help prevent economic crises of this sort from happening
again in the future. The key issues for strengthening regulation were, first,
to significantly improve coordination and consistency of prudential regula-
tion across financial sectors and countries and, second, to fill important
regulatory gaps. The debate partly focused on whether existing arrange-
ments should be extended and improved, or whether new institutions
should be created to cope with the increasingly globalized financial system,
so as to achieve the necessary improvement of international financial regu-
lation and supervision.
rui iixaxciai sraniiir\ ioiux At the more institutionally radical end
of the scale, there have been proposals for the creation of a new international
body such as a world financial authority or a board of overseers of major
international institutions and markets.
17
Such a body would have wide-
ranging powers for overseeing global regulation and supervision. A more
moderate approach advocates expanding existing institutional arrange-
ments. Both the Canadian and British governments put forward proposals
based on this approach in 1998. In the fall of 1998, Chancellor Gordon
Brown and Secretary of State Clare Short of the United Kingdom proposed
a standing committee for global financial regulation to coordinate the mul-
tilateral surveillance of national financial systems, international capital
flows, and global systemic risk. The proposed committee would bring to-
gether the World Bank, the IMF, the BIS-based Basel Committee on Bank-
ing Supervision, and other regulatory bodies on a monthly basis to develop
sriiuax\ ciiiiiru-;oxis :,
17. Eatwell and Taylor (2000).
and implement ways to ensure that international standards for financial reg-
ulation and supervision were put in place and properly coordinated.
In October 1998, the G-7 finance ministers and central bank govern-
ors asked Hans Tietmeyer, then president of the Deutsche Bundesbank, to
develop the British proposal and more generally to consider the coopera-
tion and coordination between the various international regulatory and
supervisory bodies and to recommend new arrangements. Tietmeyers re-
port, which was released in February 1999, outlines areas in which current
arrangements should be improved, but states that sweeping institutional
changes are not needed to realize these improvements.
18
Instead, the re-
port proposes that a Financial Stability Forum be convened to regularly
discuss issues affecting the global financial system and to identify actions
needed to enhance stability. The Forum was formally endorsed by finance
ministers and central bank governors from the G-7 countries at their Feb-
ruary meeting in Bonn, and it met for the first time in April 1999.
The Financial Stability Forum is currently limited in size to forty-seven
members, in order to allow for an effective exchange of views and decision-
making. Each G-7 country has three representatives on the Forum, from
the finance ministry, the central bank, and the supervisory authority.
19
Aus-
tralia, Hong Kong, The Netherlands, and Singapore were also included in
the Forum as members. G-7 representatives have stated that while the
Forum will initially be limited to G-7 countries, they envisage that other
national authorities, including from emerging economies, will join the pro-
cess at some stage. The IMF and the World Bank have two representatives
each, as do the Basel Committee on Banking Supervision, the International
Organization of Securities Commissions (IOSCO), and the International
Association of Insurance Supervisors (IAIS). The Bank for International
Settlements (BIS), the Organization for Economic Cooperation and De-
velopment (OECD), and the two BIS-based committees (the Committee
on the Global Financial System and the Committee on Payment and Set-
tlement Systems) all have one representative on the Forum.
The Forum will be chaired by Andrew Crockett, general manager of
the BIS, for the first three years, and it will have a very small secretariat in
Basel. The Forum has initially met twice a year, beginning in 1999. One
of the key aims of the Forum is to better coordinate the responsibilities of
the main national and international authorities and supervisory bodies,
iixaxciai aicuiricruii ioi ciisis iiivixriox :,,
18. Tietmeyer (1999, p. 5).
19. The Group of Seven countries are Canada, France, Germany, Italy, Japan, the United Kingdom,
and the United States.
and to pool the information held by these various bodies, in order to im-
prove the functioning of markets and reduce systemic risk. Subsequent to
its meeting in Washington on April 14, 1999, the Financial Stability
Forum defined three ad hoc working groups, which produced their first
reports in April 2000. The objectives of the three groups are as follows:
To recommend actions for reducing the destabilizing potential of
institutions employing a high degree of leverage in the financial markets
of developed and developing economies. This group is chaired by Howard
Davies, Chairman of the U.K. Financial Services.
To evaluate measures in borrower and creditor countries that could
reduce the volatility of capital flows and the risks to financial systems of
excessive short-term external indebtedness. This group is chaired by Mario
Draghi of the Italian Treasury. Among developing countries, Chile and
Malaysia participate.
To evaluate the impact on global financial stability of financial off-
shore centers and to monitor the progress made by such centers in enforc-
ing international prudential standards and complying with cross-border
information exchange agreements. An assessment will be made of the
additional efforts required to avoid underregulation or inappropriate dis-
closure in offshore centers contributing to global financial instability. This
group is chaired by John Palmer, Superintendent of Financial Institutions
of Canada.
The working groups comprise officials of developed and developing
market economies, international financial institutions, and supervisory
groupings, and they draw on work completed or under way in various pub-
lic and private forums. It is interesting that senior officials from developing
countries have been included and that their expertise is seen as particularly
relevant. For example, the group that studies measures to cope with volatil-
ity of capital flows includes senior representatives from Chile and Malaysia,
two countries that have implemented measures to curb inflows and out-
flows (Malaysia for both and Chile for inflows).
Setting up the Financial Stability Forum is clearly a positive first step
toward increasing the coordination and cooperation of the various bodies
whose objective is to improve the way markets function in order to ensure
global stability. The question lies, however, in whether the Forum, as it has
been proposed, will be representative enough and strong enough to ad-
dress all the relevant issues.
First, the omission of any developing country authorities during the
initial years of the Forum appears to be a major error. It has increasingly
sriiuax\ ciiiiiru-;oxis :,o
been accepted, especially since the Mexican peso crisis and the interna-
tional financial crisis that started in Asia, that developing countries are im-
portant actors in this globalized financial system, and that currency crises in
developing countries pose both systemic threats to the international finan-
cial system and threats to the development prospects of the developing
countries themselves. Systematic and formal representation of developing
countries on the Forum would increase the legitimacy of the body and ex-
pand its range of expertise and perspectives. Developing countries could
easily be included without making the new Forum too large. Representa-
tives (such as central bank authorities or regulators) could be chosen on a
regional basis, for example, with one Asian, one Latin American, and one
African participant. These representatives could be appointed for a fairly
short period (for example, two years) and then rotated. This type of repre-
sentation by developing countries has been working rather well in other
contexts, as in the boards of the Bretton Woods institutions.
The Financial Stability Forum is a very important initiative that may
be able to reduce vulnerabilities in the international financial system by
promoting coordination and cooperation among G-7 regulators, central
banks, and international financial institutions. Incorporating representa-
tion from developing countries would add valuable insights to the Forums
work and also increase those countries commitment to its aims.
Second, doubts have been voiced over the institutional strength of the
new Financial Stability Forum. With a small secretariat in Basel, semian-
nual meetings, and no power of enforcement, will the Forum have the suf-
ficient institutional muscle to deal with the tasks that have been identified?
Setting up the Forum represents a significant enhancement of the system
of global regulation by agreement and peer pressure that has worked rea-
sonably well in the context of the Basel committees. International cooper-
ation at the BIS has always been based on home country control:
sovereignty remains at the level of the nation-state, and agreements are
reached through negotiation and then implemented, when necessary,
through national legislation or regulation. Countries that are not repre-
sented at the Basel Committee have also adopted some of their directives
(most notably, the capital adequacy standards). In the medium term, how-
ever, and in a world of open financial markets, an international body with
the power to make and enforce policy may well be needed.
20
This would
point toward a body more akin to some kind of world financial authority,
iixaxciai aicuiricruii ioi ciisis iiivixriox :,;
20. Eatwell (1999).
which would be endowed with executive powers along the lines of the
World Trade Organization (WTO).
In the meantime, however, the Financial Stability Forum is a very im-
portant step in the right direction. Time will tell whether this body is suf-
ficient for promoting international financial stability and to filling the
important gaps in financial regulation that undermine such stability.
iiiiixc iicuiaroi\ cais There are three categories of flows to emerg-
ing markets and their associated institutions that may require additional
regulation and supervision, whether internationally coordinated or under-
taken by the source country. All three types of flows are currently insuffi-
ciently regulated, and their surges, as well as outflows, have played a
particularly prominent role in triggering recent currency crisis, apparently
because they are reversible. The three categories are short-term bank loans,
which were particularly important in the Asian crisis; easily reversible port-
folio flows made by institutional investors such as mutual funds, which
were especially important in the Mexican peso crisis and also significant in
East Asia; and activities by hedge funds and other highly leveraged insti-
tutions.
International bank loans are already regulated in industrial countries
by the central banks; these national regulations are coordinated by the
Basel Committee. Existing regulations did not discourage excessive short-
term bank lending to several of the East Asian countries, however. Short-
term bank loans and their reversal played a major role in the East Asian
crisis. Everyone, including regulators, considered most of these East Asian
countries (and particularly countries like the Republic of Korea) as credit-
worthy until just before the crisis. Current regulatory practice also seems
to have been an important factor. For example, for non-OECD countries,
loans with a residual maturity of up to one year have a weighting of only
20 percent for capital adequacy purposes, while loans over one year have
a weighting of 100 percent. This reflects the fact that it is easier for indi-
vidual banks to pull out from renewing short-term loans. As a result of this
rule, however, short-term lending is more profitable for international
banks.
A regulatory bias that encourages short-term lending is thus added to
banks economic preference for lending short-term, especially in situations
of perceived increased risk. An overall increase in short-term loans, however,
makes countries more vulnerable to currency crises, which paradoxically
makes banks more vulnerable, as well, to the risk of nonpayment of short-
sriiuax\ ciiiiiru-;oxis :,
term loans. International authorities need to rapidly examine whether the
capital adequacy weighting differential is too large in favor of short-term
loans for non-OECD countries, resulting in excessive incentives for short-
term lending.
A narrowing of this differential clearly seems desirable. Soon after the
Asian crisis, economic authorities, such as Alan Greenspan, proposed in-
creasing the capital charge for short-term loans.
21
No progress was made
in this area for a year, however. Only in 1999 was a consultative document
issued, with proposals to address the issue. Actions have not yet been
taken, although several recommendations support the changes. The issue
thus involves not just the definition of appropriate technical measures for
a new architecture, but also the development of mechanisms for speeding
up the decisionmaking process, especially when basic agreement exists and
institutional mechanisms are in place.
With regard to portfolio flows to emerging markets, there is currently
no international framework for taking account of market or credit risks on
flows originating in institutional investors such as mutual funds (and more
broadly on flows originating in nonbank institutions). This important reg-
ulatory gap needs to be filled to protect retail investors in both developed
and developing countries from the negative effects of excessively large and
potentially volatile portfolio flows.
22
The East Asian crisis confirmed what was very clear in the Mexican
peso crisis. Institutional investors, such as mutual funds, strongly contrib-
uted to these currency crises, given the very liquid nature of their invest-
ments.
23
Regulation should therefore be introduced to discourage excessive
surges of portfolio flows. This could perhaps best be achieved by a variable
risk-weighted cash requirement for such institutional investors.
24
These
cash requirements would be placed as interest-bearing deposits in com-
mercial banks. Introducing a dynamic risk-weighted cash requirement for
mutual funds (and perhaps other institutional investors) is within the
mainstream of current regulatory thinking. It would require that standards
be provided by relevant regulatory authorities or agreed internationally.
The guidelines for macroeconomic risk, which would determine the
cash requirement, should take into account such vulnerability indicators
as the ratio of a countrys current account deficit to GDP, the level of its
iixaxciai aicuiricruii ioi ciisis iiivixriox :,,
21. Alan Greenspan, Financial Times, 28 February 1998.
22. Griffith-Jones (1998).
23. BIS (1998); Griffith-Jones (1998).
24. This proposal was elaborated in a joint study by Jane dArista and Stephany Griffith-Jones.
short-term external liabilities net of foreign exchange reserves, the fragility
of the banking system, and other relevant country risk factors.
25
The pro-
cess would require a sophisticated analysis to avoid stigmatizing countries
unnecessarily through the use of simplistic criteria. The views of the na-
tional central bank and the treasury in the source countries, together with
those of the IMF and the BIS, should be helpful in this respect. The se-
curities regulators in source countries would be the most appropriate in-
stitutions for implementing such regulations, and their actions could be
coordinated internationally by IOSCO.
The fact that the level of required cash reserves would vary with a coun-
trys perceived macroeconomic risk would make it relatively more profitable
to invest in countries with good fundamentals and relatively less profitable
to invest in countries with more problematic macroeconomic or financial
sector fundamentals. If a countrys fundamentals should deteriorate, invest-
ment would decline gradually, which should force an early correction of pol-
icy and, in turn, stimulate a resumption of flows. Although a requirement
for cash reserves on the assets of mutual funds invested in emerging markets
could increase the cost of raising foreign capital for the funds, the increased
cost would be compensated by a more stable supply of funds at a more sta-
ble cost. Furthermore, this smoothing of flows should discourage the type of
massive and sudden reversal of flows that sparked both the Mexican and the
Asian crises, making such developmentally costly crises less likely.
Given the dominant role and rapid growth of institutional investors
in countries such as the United States, the United Kingdom, and France,
the proposal for a risk-weighted cash requirement on mutual funds could
be adopted first in those countries, without creating significant competi-
tive disadvantages. Subsequently, IOSCO would need to give urgent pri-
ority to harmonizing such measures internationally, so as to prevent
mutual funds from channeling their investments through other countries
and, especially, offshore centers that did not impose the cash require-
ments.
26
International guidelines would be formulated through interna-
tional consultations similar to those employed by the Basel Committee in
developing the Core Principles for Effective Banking Supervision. For that
purpose, IOSCO could set up a working group consisting of representa-
tives of the national securities regulatory authorities in source countries,
sriiuax\ ciiiiiru-;oxis :oc
25. See Ffrench-Davis and Ocampo (in this volume).
26. The latter point draws on communication with the U.S. Federal Reserve Board, including a let-
ter from Chairman Alan Greenspan.
together with some representation from developing countries. The new
guidelines should take due account of relevant existing regulations, such
as the European Commissions Capital Adequacy Directive. The Financial
Stability Forum, in which IOSCO participates, could facilitate the imple-
mentation and coordination of such prudential guidelines and ensure their
consistency with the regulation of other institutions (such as banks) and
other potentially volatile flows.
The third category of flows, namely, those associated with hedge funds
and other highly leveraged institutions (HLIs), requires urgent study to
detect and cover existing monitoring and regulatory gaps. The Financial
Stability Forum report on this issue, mentioned above, represents a valu-
able first step. Further careful analysisboth technical and institutional
is required on how hedge funds and other highly leveraged institutions
operate and how they can best be regulated to reduce their impact on mag-
nifying the volatility of capital flows, exchange rates, and stock markets in
developing countries and thus minimize the negative effect that this vola-
tility has on development and poverty. There is a growing consensus that
HLIs can pose important risks both to direct creditors and, under certain
market conditions, to the financial system as a whole.
27
The exact nature
of the impact of HLIs on magnifying volatility in developing countries is
not yet fully understood. The role of hedge funds in the East Asian crisis,
for example, is a matter of debate owing to insufficient information and
differences in interpretation. Measures designed specifically to address the
impact of hedge funds and other HLIs on magnifying volatility in devel-
oping countries have not been given sufficient priority on the regulatory
agenda. However, policies for managing the risks posed by HLIs to credi-
tors and the financial system as a whole will also help reduce their nega-
tive impact on developing countries.
The problem does not just relate to hedge funds, but also encompasses
other highly leveraged activities and institutions, such as proprietary desks
of investment banks. HLIs can be defined as having three characteristics:
they are subject to little or no regulatory oversight, as a significant propor-
tion operate through offshore centers; they are subject to limited disclo-
sure requirements, and their operations are often very opaque; and they
take on significant leverage.
Three sets of responses can be used to address the risks posed by the
iixaxciai aicuiricruii ioi ciisis iiivixriox :o:
27. This consensus is reflected, for example, in the Basel Committees report on banks and HLIs
(Basel Committee, 1999).
HLIs. These actions are often presented as alternatives, but it would be
better to consider them as complementary. The first response is indirect,
attacking the issue through the major counterparties of HLIs (mainly
banks and securities houses). One possibility is to promote sounder prac-
tices in the way banks and securities houses assess risks when they deal
with hedge funds and other HLIs. However, further actions by supervisory
authorities also seem desirable. In particular, supervisors should impose
higher capital requirements on lending or other exposures of banks to
HLIs, to reflect the higher risks involved in such exposures as a result
of HLIs opaqueness, high leverage, and lack of regulation. Supervisors may
also need to formally or informally prohibit banks from lending to a par-
ticular class of risky counterparty. Such measures would not only protect
banks, but may also stimulate HLIs to manage risks more responsibly.
A second avenue, which is clearly complementary to the first, is to in-
crease transparency on total exposures to HLIs by all financial institutions.
For example, the concept of a credit register for bank loans could be ex-
tended along the model of the French central des risques, which provides
banks access to the aggregate amount of bank lending to each company.
Such a register would collect and centralize total exposures (both on and
off balance-sheet positions) of different financial intermediaries to single
counterparties, such as major hedge funds. Counterparties, supervisors,
and central banks in both developed and developing countries could then
get information about the total indebtedness of such institutions, which
would help them assess the risks involved more precisely. For this purpose,
the information would have to be both timely and meaningful, especially
to take account of rapid shifts in the positions of HLIs. The register should
probably be based at the BIS itself or at the Basel Committee on the Glo-
bal Financial System (formerly the Euro-currency Standing Committee),
which already has experience in gathering similar information.
A third avenue is to directly regulate hedge funds and other highly lev-
eraged institutions. Such direct regulation could take a number of forms,
including licensing requirements, minimum capital standards, and mini-
mum standards for risk management and control. In a recent report, the
Basel Committee on Banking Supervision argues that such a regulatory re-
gime should focus on the systemic risk that is potentially generated by HLI
activities owing to their excessive size and risk taking, which could endan-
ger financial stability.
28
If, as seems probable, HLIs also increase the vola-
sriiuax\ ciiiiiru-;oxis :o:
28. Basel Committee on Banking Supervision (1999).
tility of exchange rates in developing countries, this concern should also
be addressed in attempts at their regulation.
The first two forms of dealing with HLIs currently have more support
than their direct regulation, even though the latter would deal with the
problem more directly. The opposition to such direct regulation is based
on practical grounds. For example, it is argued that HLIs could restructure
themselves so as to escape any regulatory definition that may be instituted.
The most frequent argument against direct regulation of hedge funds is
that they would be able to circumvent such regulations, because they ei-
ther are located offshore or could easily move offshore.
The international community can either ignore this problem by ac-
cepting the absurd status quo (and incurring continued high costs from
the risk of major instability) or tackle it by raising the issue of extending
regulation, including possible regulation of mutual funds, to offshore cen-
ters. If global supervision and regulation is genuinely accepted as essential
in todays world of globalized financial markets, then there is no justifica-
tion for so-called no-go areas where such regulations can be evaded or
undermined. It is essential that offshore centers comply with international
standards with regard to both the provision of information and the global
regulation of institutions such as hedge funds. A political initiative in this
respect could be both effective and useful if it had the clear backing of the
G-7 countries, in particular, and support from developing countries. The
technical issues involved in applying provision-of-information standards
and regulation to offshore centers require further study so as to facilitate
political decisions for action in this sphere.
The recent report by the Financial Stability Forums working group on
HLIs and offshore centers contains positive proposals in this direction.
29
The existence of this working group within the Financial Stability Forum
provides the institutional mechanism for urgent action in this field. It is to
be hoped that the recently suggested measures will be implemented soon.
The focus has thus far been on better regulation of financial intermediaries,
which is clearly essential. It is the function of financial intermediation more
broadly, however, that really should be regulated, including financial inter-
mediation carried out by nonfinancial firms, such as suppliers (who provide
credit) or international companies making direct investment. Unfortu-
nately, such intermediation activities are far more difficult to regulate. An
essential first step is acquiring a better empirical understanding of the role
iixaxciai aicuiricruii ioi ciisis iiivixriox :o,
29. Financial Stability Forum (2000).
played by nonfinancial intermediaries in channeling private flows, particu-
larly volatile ones, to developing countries.
More generally, further empirical work is required on several issues, in-
cluding the recent changes in global credit and capital markets; the criteria
used by different categories of market actors, such as banks, mutual funds,
hedge funds, and foreign direct investors, for entering and exiting coun-
tries; and the incentives behind particular patterns of behavior that contrib-
ute to speculative pressures on individual countries and to cross-country
contagion. A better understanding of behavioral patterns and outflow
trends would help policymakers design measuresto be implemented by
individual firms, by segments of the financial industry via self-regulation,
by regulators, and by governments (for example, via tax measures)to dis-
courage market imperfections such as disaster myopia and herding, which
contribute to currency crises.
A package of measures is thus needed to make currency crises in
emerging markets far less likely and to ensure the efficient operation of the
market economy in emerging markets, which should be a basis for sus-
tained development. Avoiding crises seems to require some degree of dis-
couragement or regulation of excessive and potentially unsustainable
short-term inflows. Such measures will be most effective if they are applied
by both source and recipient countries (although the main responsibility
lies with recipient countries), if they avoid discouraging more long-term
flows (which, on the contrary, need to be encouraged), if they are simple
in design and clearly targeted at unsustainable flows, and, in particular, if
they are complemented by good policies in the emerging economies.
Conclusions
The seriousness of currency crises in developing and transition coun-
tries is a cause of major concern. No one foresaw the depth of the crisis
that started in East Asia, for example, or the speed of contagion. Conse-
quently, a sizeable proportion of the world economy entered recession for
a considerable period. This not only eroded impressive gains in poverty re-
duction in the most affected countries, but also led to major increases in
poverty.
Steps need to be taken urgently to prevent future crises and to manage
them better if they do occur. Some actions can be implemented either
within existing institutional arrangements or with some development, ex-
sriiuax\ ciiiiiru-;oxis :o
pansion, and adaptation of existing international institutions. These in-
clude improving information on international financial markets and filling
regulatory gaps, in particular in relation to mutual funds and hedge funds,
as well as improving the regulation of bank lending to help prevent excessive
surges of capital inflows. It is also important, however, to envisionand
start buildinga new financial architecture that would meet development
needs in the context of the new globalized economy. This includes moving
toward the creation of a world financial authority, which would carry out
prudential regulation of capital flows consistently across countries and fi-
nancial sectors. Such a world financial authority could build on existing in-
stitutions, especially the new Financial Stability Forum. These measures
need to be complemented by measures for crises management.
30
Though international actions are crucial, individual countries must
clearly adopt their own policies for preventing currency crises. Relevant
measures include more countercyclical macroeconomic and domestic reg-
ulatory policies, the strengthening of domestic financial systems, and more
prudent opening of the capital account, accompanied when necessary by
market-based measures to discourage excessive surges of capital inflows.
References
Basel Committee on Banking Supervision. 1999. Sound Practices for Banks Interactions
with Highly Leveraged Institutions. Basel Committee Publications 46. Basel.
BIS (Bank for International Settlements). 1998. 68th Annual Report. Basel.
Eatwell, J. 1999. Brown Achieves First Step towards Global Stability. The Observer (Feb-
ruary 28).
Eatwell, J., and L. Taylor. 2000. Global Finance at Risk. New York: The New Press.
Eichengreen, B. 1999. Toward a New International Financial Architecture: A Practical Post-
Asia Agenda. Washington: Institute for International Economics (IIE)
Financial Stability Forum. 2000. Report of the Working Group on Highly Leveraged Insti-
tutions. Basel: Bank for International Settlements. See www.fsforum.org/reports/rephli.
html.
Ffrench-Davis, R., and S. Griffith-Jones, eds. 1995. Coping with Capital Surges: The Return
of Finance to Latin America. Lynne Reinner.
Griffith-Jones, S. 1998. Global Capital Flows. Macmillan.
. 2000. Towards a New Financial Architecture: the Management of Crises. In Fi-
nancial Globalization and the Emerging Economies, edited by J. A. Ocampo and others.
Santiago: International Jacques Maritain Institute and United Nations Economic Com-
mission for Latin America and the Caribbean (ECLAC).
iixaxciai aicuiricruii ioi ciisis iiivixriox :o,
30. See Griffith-Jones (2000).
Griffith-Jones, S., J. A. Ocampo, and J. Cailloux. 1999. Proposals for a New International
Financial Architecture, with Special Reference on Needs of Poorer Countries. Institute
of Development Studies. Mimeographed.
Keynes, J. M. 1936. The General Theory of Employment, Interest and Money. Cambridge
University Press.
Ocampo, J. A. 1999. Reforming the International Financial Architecture: Consensus and
Divergence. Serie Temas de Coyuntura 1. Santiago: United Nations Economic Commis-
sion for Latin America and the Caribbean (ECLAC).
Stiglitz J. 1994. The Role of the State in Financial Markets. Proceedings of the World Bank
Annual Conference on Development Economics. Washington: International Bank for Re-
construction and Development (IBRD).
Tietmeyer, H. 1999. International Cooperation and Coordination in the Area of Financial
Market Supervision and Surveillance. Frankfurt: Deutsche Bundesbank.
Wyplosz, C. 1998. Globalized Financial Markets and Financial Crises. In Regulatory and
Supervisory Challenges in a New Era of Global Finance, edited by J. J. Teunissen. The
Hague: Forum on Debt and Development (FONDAD).
sriiuax\ ciiiiiru-;oxis :oo
Manuel R. Agosin
Universidad de Chile
and InterAmerican
Development Bank
Ricardo Ffrench-Davis
UN Economic Commission
for Latin America and
the Caribbean
Stephany Griffith-Jones
Institute of Development
Studies, University of
Sussex
Jos Antonio Ocampo
UN Economic Commission
for Latin America and
the Caribbean
Jaime Ros
University of Notre Dame.
Heriberto Tapia
UN Economic Commission
for Latin America and
the Caribbean
Contributors

This page intentionally left blank

Argentina: currency board in, 19; effect of
Mexican crisis in (199495), 1819
Balance of payments: crisis in Mexico (1974
76), 110; factors precipitating crisis of, 53;
Korean crisis of, 44, 4647. See also
Capital account; Current account
Bank for International Settlements (BIS):
actions to improve transparency, 151;
potential role as information clearing
house, 15152; representation in Financial
Stability Forum, 155
Banking system, Chile: crisis in (198186),
87; prudential supervision of, 22, 8790;
role (197582), 6878
Banking system, Korea: interest arbitrage of,
55; lending and debt practices (199798),
46, 5557; regulation and government
guidance of, 5052; transformation
in, 62
Banking system, Mexico: effect of capital
surge on, 12627, 131; effect of financial
crisis on (199495), 13637; privatization
in (199192), 110, 126
Banking system, Taiwan: foreign banks in, 51;
prudential regulation of, 5152
Basel Committee: on Banking Supervision,
162; Core Principles for Effective Banking
Supervision, 160; on the Global Financial
System, 162
Brady agreement, 4, 10710
Brown, Gordon, 154
Capital account: Chilean policy (19752000),
2122; effect of open, 2627, 33;
liberalization in Chile (197582), 6878;
liberalization of Korean (199193),
2021, 3940, 4546, 5456, 6061;
price-based regulation for, 33
Capital Adequacy Directive, 161
Capital flows: associated with highly leveraged
institutions, 161; causes, effects, and costs
of, 14145; to developing countries,
14142; into East Asian economies, 23;
effect of stability of, 30; lack of controls
on surges in, 1116; needing regulation,
15864; with open capital account, 33;
policy to sustain economic growth with
volatile, 3334; volume, composition, and
use of inflows, 5, 3132. See also Foreign
direct investment (FDI); Portfolio flows
Capital flows, Chile: composition of
(19802000), 7982, 92; exemptions
from controls on, 23, 89; inflows (1990
95), 7890; inflows (1970s), 6667; net
inflows (19772000), 6870; post-1997
outflows from, 9192, 99; pre-1997
inflows, 67; regulation of flows, 22, 88,
9395, 101; restrictions on inflows, 6870
Capital flows, Korea: inflows (1990s), 2021,
41, 44, 5556, 58, 62; outflows (1990s),
42, 44, 5759
Capital flows, Mexico: inflows (198994), 4,
10811, 114, 125, 12728; outflows
related to monetary policy, 13137
Index

Capital flows, Taiwan: control into and out of

(199798), 5758
Capital formation, 3132
Capital inflows: preceding financial crises, 53
Capital markets, international: imperfections in,
143; shock transmission from, 1, 28
Central Bank, Chile: exchange rate policy
(199597), 95101; foreign exchange
intervention (199697), 23, 67;
management of capital flows (199095), 82
90; restrictions on capital inflows (1975
82), 71
Central Bank, Korea: banking sector policy
(1997), 52; monetary policy (1990s), 3942
Central Bank, Mexico: policy related to
199495 crisis, 12734; sterilization of
foreign reserve losses, 12628, 13031;
withdrawal from foreign exchange market
(199495), 133
Central Bank, Taiwan: authority in, 5152,
57; strategy to defend currency (1990s),
4748
CETES (peso-denominated bonds), Mexico,
12526, 132
Chile: capital account liberalization (197582),
6878; controls on capital flows, 22, 23;
debt crisis (1982), 67; effect of Asian
crisis in (1997), 91, 9496, 99, 101; effect
of Mexican crisis on, 23; growth rate
(1990s), 23
Contagion: from Brazilian crisis (1999), 136;
effect in Chile of Mexican crisis, 67, 7980;
effect in Korea (1997), 5960; effect in
Taiwan (1997), 60; effect of Mexican crisis,
24; from Asian crisis (1997), 91, 14142
Crockett, Andrew, 155
Currency: factors influencing Mexican peso
appreciation, 115; Mexican depreciation and
devaluation (199495), 13137
Currency board: in Argentina, 19; arguments
for and against, 28
Currency crises: Brazil (1999), 136; European
Monetary System (EMS) (1992), 130;
methods to avoid, 15253; models within
speculative attack hypothesis, 12930;
second- and third-generation models, 130;
self-fulfilling attacks, 14445
Current account: deficit and surplus in Korea,
4042, 46, 53; financing deficit in Mexico
(199495), 128; surplus in Taiwan (1990s),
4042, 5758
Davies, Howard, 156
Debt, external: of Chile (197582), 6878;
Mexicos servicing of (199192), 11011,
125; of Taiwan (1997), 58
Developing countries: uses for capital inflows,
5, 3132
Developing countries: benefits of FDI in,
14142; capital flows to (1950s, 1960s), 30;
capital surges to (1970s, 1980s), 35,
14142; impact of capital flows on, 66;
improved information to, 14752; needed
regulation of capital flows to, 15864. See
also Contagion; East Asian countries; Latin
American countries
Draghi, Mario, 156
East Asian countries: capital inflows (199297),
1016; effect of crisis (1997), 25, 3839,
142; effect of Mexican crisis on (199495),
9; GDP (19712000), 67, 911. See also
Korea; Taiwan
Emerging economies. See Developing countries
Exchange rate policy: change in Mexican
(199495), 13135; Chile (197582), 72
73; Chile (198295), 8587, 10001
Exchange rate regimes: currency board, 1, 28;
freely floating, 29; intermediate, 2829,
33
Exchange rates: Korea (199798), 46, 57, 60;
risk associated with, 53; Taiwan (199798),
4748, 60
Feldstein, M., 47
Financial crises: causes of, 142; East Asian crisis
(1997), 38, 91, 99, 14245, 159; effect in
Chile of Mexican crisis (199495), 9,
7982, 92, 97; methods to prevent, 14564;
Mexican crisis (199495), 2, 67, 10708,
12832; recent and historical Latin
American and Asian, 12. See also
Contagion; Currency crises
Financial intermediation: of capital inflows in
Mexico, 12627; of large capital inflows,
115; need for regulation of, 163
Financial markets, international: complexity
and opaqueness in, 151; effect of capital
inflows in Mexico, 11518; effect of Latin
American debt crisis on, 25; external
liabilities in Mexican, 124; imperfections in,
144; information available to, 148; need for
improved information in, 14849; opening

in Mexico, 111. See also Capital markets,
international
Financial Stability Forum: proposed functions
of, 151, 15458, 161; report on HLIs and
offshore centers, 163
Fiscal policy, Chile: effect of (199095), 8290;
expansionary (199697), 9899
Foreign direct investment (FDI): benefits to
developing countries of, 14142; as capital
inflow, 3132; in Chile, 7982, 8889,
9293, 9697; deregulation of Korean
greenfield, 20, 54; increases in (1990s), 4;
openness in Mexico to, 10910; post-1997
inflows into Korea, 62; regulation in Chile
(199095), 84; Taiwan greenfield, 57
Foreign exchange reserves: Korea, 4142, 60,
62; levels in Chile (197582), 71; levels in
Mexico (199495), 128; need for high levels
of, 152; in speculative attacks on Mexican
currency, 12631; Taiwan (199798), 4748
Foreign Investment Laws, Mexico (1973,
1993), 11011
Hedge funds: capital flows of, 161; proposed
direct regulation of, 16263; role in East
Asian crisis, 161. See also Highly leveraged
institutions (HLIs)
Highly leveraged institutions (HLIs), 161;
characteristics of, 161; proposed direct
regulation of, 16263; responses to risks
posed by, 16162
Industrial sector, Korea: chaebol foreign
investment, 55; chaebol response to East
Asian crisis (1997), 49; restructuring of
chaebol, 44, 46; unhedged borrowing by
firms in, 4951, 61
Information: about hedge funds and HLIs, 151;
BIS as potential clearing house for, 15152;
IMF collection and dissemination of, 147,
150; need for improved, 14552; public
information notices, 147
International Association of Insurance
Supervisors (IAIS), 155
International Monetary Fund (IMF): codes of
good practice, 14748; conditional lending
of, 52; financial package and structural
reforms (19972000), 44, 4647, 59;
information collection and dissemination by,
147, 150; representation in Financial
Stability Forum, 155
International Organization of Securities
Commissions (IOSCO), 155, 16061
Investors: ceiling in Taiwan for foreign, 57;
contribution to currency crises, 159; ease of
flight when crisis occurs, 53; flight from
Korea, 55, 57; in Korea, 54. See also
Portfolio flows
Korea: current account balance (199299),
4042; effect of East Asian recessions on
(1998), 19; effect of Thai baht crisis on
(1997), 3948; GDP growth (1970s
1990s), 3839; pre-1997 GDP growth rates,
3842; real exchange rate (199297),
4041, 43; structural reforms required by
IMF (19972000), 44, 46
Latin American countries: bank lending as
source of financing in, 25; capital inflows to,
34, 6, 8, 1116, 10809; effect of Mexican
crisis in (199495), 1619; GDP (1971
2000), 68; investments in (1990s), 45,
89. See also Argentina; Chile; Mexico
Mexico: Brady agreement in, 4, 10710; finan-
cial crisis and rescue (199495), 10708,
159; output, employment, and productivity
(198897), 11824; post-199495 perfor-
mance, 13537; public debt, 11011, 125
26, 13234. See also Capital flows, Mexico;
Central bank, Mexico; CETES; Tesobonos
Monetary policy. See Foreign exchange reserves;
Prudential regulation and supervision
North American Free Trade Agreement
(NAFTA): Mexican participation in, 136
Organization for Economic Cooperation and
Development (OECD), 155
Palmer, John, 156
Pension funds, Chile: authorized foreign
investments (199597), 9495; investments
(199095), 8485; post-1997 capital
outflows, 99
Portfolio flows: into Chile (199095), 8182;
increase in (1990s), 4; inflows to Latin
American countries (199297), 1116; into
Mexico (198898), 11114; into and out of
Korea (199798), 5457; post-1997 into
Korea, 62; proposed regulation of, 15960

Prudential regulation and supervision: absence
in Korea (1997), 3940; in Chile, 22, 7590,
8788; in East Asian countries (199297),
11; effect on capital flow volatility, 3031; in
emerging economies, 34; in Mexico, 22; in
Taiwan (1997), 3940, 5152
Public debt, Mexico: crisis (199495), 13334;
dollarization of, 12526, 13234; Mexicos
servicing of (199192), 11011, 125
Public information notices (PINs), 147
Radelet, S., 46
Sachs, J., 46
Short, Clare, 154
Speculative attacks: currency crisis models
within hypothesis of, 12930; on Mexican
currency (199495), 12630; on Taiwan
currency, 24
Taiwan: capital controls in, 2324, 3940;
current account balance (199297), 4042;
effect of Thai baht crisis on, 38, 3948,
4748; GDP growth (1970s1990s), 3839;
growth rate (199798), 4748; pre-1997
GDP growth rates, 3842; real exchange
rate (199297), 4041, 43
Telecommunications, Mexico, 110
Tequila crisis (199495). See Contagion;
Financial crises; Mexico
Tesobonos (dollar-indexed bonds), Mexico,
12526, 13234
Thai baht crisis (1997). See Contagion; East
Asian countries; Financial crises
Tietmeyer, Hans, 155
Titelman, D., 30
Transparency: BIS actions to improve, 151;
enhancing, 14752
Uthoff, A., 30
World Bank: codes of good practice, 14748;
representation in Financial Stability Forum,
155

icoxoxic coxxissiox ioi iarix axiiica axo rui caiinniax
ECLAC is one of five United Nations regional commissions. All Latin
American and Caribbean countries are members, as are a number of devel-
oped nations in America and Europe with strong historical, cultural, and
economic ties to the region. Since its founding in 1948, ECLAC has pro-
moted economic and social development and cooperation between nations
through training courses, technical assistance, and policy-oriented research
projects. As well as providing general economic analysis, these projects also
address the particular problems of individual Latin American and Carib-
bean countries. In recent years, ECLAC has focused its efforts on the chal-
lenges faced by the region in achieving sustained (and environmentally
sustainable) growth among pluralist democracies. These democracies face
very real demands as they find ways to develop their economies in order to
benefit the majority of the population. As a result, ECLACs main focus has
been on helping countries improve economic growth and social equity si-
multaneously. ECLAC, with headquarters in Santiago, Chile, has two sub-
regional offices in Mexico and in Trinidad and Tobago, together with
national offices in Argentina, Brazil, Colombia, Uruguay, and the United
States.