. _ an. amine
umwwnwthmm ma.

v...
1.

ﬂ . . ‘ m.
awayviwninﬂﬁ

ﬂy

 

 

2......
.4) .
.4... an”?
3.x. .
#5 .. .
523...»..an 1:. ‘
.1 3.2.3.59? , , ,.
112%. .
hint} rﬂﬁm
. 3 r , .. ..
.. I. .
3.3%.? .

i
Hunks}...

«v.»

 

Iﬂdel

 

5mg

549MH53

 

 

 

LIBRARY
MiChigan State This is to certify that the
U nive rsity dissertation entitled

 

FINANCIAL DOLLARIZATION
IN EMERGING MARKET AND TRANSITION ECONOMIES

presented by

ALINA C. LUCA

has been accepted towards fulﬁllment
of the requirements for the

 

 

 

 
 

 

 

Doctoral gree in -' ' ' ’ ; Economics
#1, #1 r, 'AAW
/ *Majo rofessor’s Signature
/
/8 ﬁgmf M32
Date

MSU is an Afﬁnnative Action/Equal Opportunity Institution

' v ‘ —.— ._

PLACE IN RErURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

EH 31 30M

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DatoDuo.p65-p.15

 

FINANCIAL DOLLARIZATION
IN EMERGING MARKET AND TRANSITION ECONOMIES

By

Alina C. Luca

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Economics

2003

and 1
coun

C8US

eval
and

onf

:Xct
hue
0pc
10:1:
her
the
am
ba:

p0

ABSTRACT

FINANCIAL DOLLARIZATION
IN EMERGING MARKET AND TRANSITION ECONOMIES

By
Alina C. Luca

The dollarization of bank deposits and loans is widespread in emerging market
and transition economies. Since levels of dollar deposits and loans vary widely across
countries and over time, and dollarization is generally associated with potential costs, the
causes and economic consequences of ﬁnancial dollarization deserve thorough analysis.

This dissertation (1) examines the sources of dollarization of bank credit, (2)
evaluates the economic effects of policy measures that attempt to restrict dollarization,
and (3) assesses the effects of the regulation of foreign exchange and capital transactions
on ﬁnancial intermediation and capital outﬂows.

The ﬁrst chapter, “Credit Dollarization, Bank Currency Matching, and Real
Activity” models the link between ﬁnancial dollarization and integration in the
international goods market. This chapter develops a general equilibrium model of a small
open economy with costly banking and ﬁrms that use both domestic currency and dollar
loans to ﬁnance production. Dollar loans are used to pay for a foreign intermediate good,
hence the link to economic integration. Credit dollarization in equilibrium depends on
the characteristics of the production process and the relative returns on domestic currency
and dollar instruments. Moreover, deposit dollarization can cause credit dollarization, as
banks match by currency their deposits and loans. Finally, this chapter shows that

policies that attempt to reduce dollarization will cause ﬁnancial disintermediation.

TC

do
an:

ﬁr]

evaj
don‘
excl
don]
Enu]
incre
in dc

then

To date, the contributions of ﬁrms and banks to the dollarization of credit remain
relatively unexplained. The second chapter, “Credit Dollarization: Is It Firms’ or
Banks’ “Fault”?”, based on joint work with Iva Petrova, is an empirical study of credit
dollarization in transition economies. It separates the respective contributions of banks
and ﬁrms to the dollarization process by grouping potential determinants into bank- and
ﬁrm-speciﬁc (ﬁnancial and mainly real) factors. Empirical results provide evidence that
bank currency matching is the main driving force of credit dollarization. In addition,
there is limited evidence that real dollarization causes ﬁnancial dollarization. However,
currency mismatches tend to be concentrated in the real sector.

The third chapter, “Regulation, Financial Development, and Capital Flows”,
evaluates the effects of regulation of foreign exchange and capital transactions on the
domestic and international activities of banks. It separates the effects of foreign
exchange liberalization from capital outﬂows liberalization. Will [banks provide more
domestic credit and “go abroad” less when allowed to freely lend in dollars domestically?
Empirical results provide evidence that the liberalization of foreign exchange operations
increases the level of domestic credit. Surprisingly, the liberalization of domestic lending
in dollars also increases banks’ holdings of foreign assets. In addition, banks increase

their holdings of foreign assets when capital outﬂows are liberalized.

C0pyright by
Alina C. Luca
2003

T 0 Mom and Dad, and Brosco, with love

 

IOC

and

und

anaf

lildt

Lip!

Cer

met

and

the

Pro

eat

the

 

l0]

hel]

BOC

 

 

ACKNOWLEDGEMENTS

Over the years, many people have contributed to my education, which enabled me
to complete this dissertation. Today, I have the opportunity today to thank a few of them.

First, there were many years of study in my home country, Romania. Dedicated
and very competent professors guided my steps from elementary school up to my
undergraduate studies. They taught me mathematics and physics, structured my
analytical and logical thinking, and helped me become who I am now. I am truly
indebted to Professors Zoe Muresan, Piroi Eugenia, Tiberiu Superceanu, Gheorghe
Lipovan, and Vasile Bivolaru. At Universitatea de Vest Timisoara, Professor Silviu
Cema taught my ﬁrst course in macroeconomics and monetary policy. He was a true
mentor and a role model for me as both an economist and a teacher.

After my undergraduate years in Romania, I came to the US to explore the world
and the economics science. While working as a research assistant in the Clinical Center,
the Institute for Managed Care, I met two great people, Professor Bill Given and
Professor Cathy Bradley. Our regular discussions over lunches helped me stay sane and
eat healthy, and provided me advice and guidance throughout my graduate studies.

Foremost, I am truly grateful to my advisor, Professor Rowena Pecchenino, for all
the time and effort she spent with me. She has always been there for me, always willing
to read one more draft of my papers, and provide valuable suggestions. Her friendship
helped me get through the hard times and stay on the right track.

I would like to thank Professors Susan J. Linz, Ana Maria Herrera, and Geoffrey

Booth, for serving on my dissertation committee, and for their advice. Susan and Ana

vi

havc

OVCI

Iﬂda
achh
cons
midi
“ant

CHICII

forth
With
Papers

myon

from n
dreanl

SuppOr

 

 

have always provided suggestions and comments on my work, and had the patience to go
over numerous drafts. Professor Booth provided valuable suggestions ﬁ‘om a ﬁnance
perspective.

My fellow graduate students, and in particular Iva Petrova, Linda Bailey, and
Elda Pema, have always represented a support network for me. We shared our
achievements, complained about the hard and frustrating life of a PhD student, and
constantly helped each other. I used to spend hours during coffee breaks and lunches
with Iva, working on joint papers and talking about economics or just life in general. I
want to thank all my other friends for making life in East Lansing enjoyable and
entertaining.

My boyfriend Radu has been a constant source of encouragement and challenge
for me. I have always wanted to be as good as him and, if possible, just slightly better.
With love and patience, he helped me ﬁgure out what grad school is about, read my
papers, listened to my practice talks, and tolerated my moodiness. He has always been
my only family far away ﬁom home.

If I am who I am today, it is because of the lifetime support and encouragement
from my parents, Georgeta and George. They have always encouraged me to follow my
dreams, to challenge myself, and to aspire to more and better. For all the love and

support over the years, and for believing in me, thank you!

vii

REFI
APPE

CHA)
CREE
BANK

 

TABLE OF CONTENTS

LIST OF TABLES .............................................................................................................. x
LIST OF FIGURES .......................................................................................................... xii
INTRODUCTION .............................................................................................................. 1
CHAPTER 1

CREDIT DOLLARIZATION, BANK CURRENCY MATCHING, AND REAL
ACTIVITY

1. Introduction ............................................................................................................... 6
2. A Model of Financial Dollarization ........................................................................ 10
2.1. The Economy ................................................................................................. 11
2.2. Households ..................................................................................................... 13
2.3. Firms .............................................................................................................. 16
2.4. Banks .............................................................................................................. 19
2.5. Government .................................................................................................... 23
2.4. Competitive Equilibrium ............................................................................... 24
3. Impact of Domestic and External Shocks on Financial Dollarization .................... 25
3.1. Impact of a Change in Households’ Preference for Dollar Deposits ............. 26
3.2. Impact of a Change in the Rate of Devaluation ............................................. 29
3.3. Impact of a Change in Restrictions on Dollar Instruments ............................ 32
~ 3.4. Impact of an External Supply Shock ............................................................. 33
3.5. Determinants of Credit Dollarization ............................................................. 35
4. Policies that Aim to Reduce Financial Dollarization .............................................. 37
5. Conclusions ............................................................................................................. 38
REFERENCES ................................................................................................................. 41
APPENDIX ....................................................................................................................... 44
CHAPTER 2
CREDIT DOLLARIZATION 1N TRANSITION ECONOMIES: IS IT FIRMS’ OR
BANKS’ “FAULT”?
1. Introduction ............................................................................................................. 59
2. Conceptual Framework and Main Hypotheses ....................................................... 64
2.1. Bank-speciﬁc Factors ..................................................................................... 65
2.1.1. Asset and Liability Management .......................................................... 65
2.1.2. Proﬁtability and Risk-taking Behavior ................................................. 66
2.1 .3. Concentration ........................................................................................ 67
2.2. F inn-speciﬁc Factors ..................................................................................... 68
2.2.1. Liability Management ........................................................................... 68
2.2.2. Hedging Behavior ................................................................................. 68
2.2.3. Proﬁtability and Risk-taking Behavior ................................................. 70
2.3. Macroeconomic Factors and Other Controls ................................................. 71

viii

2.3.1. Overall Hedging Opportunities ............................................................. 71

2.3.2. Restrictions on Foreign Exchange Operations ...................................... 71
2.3.3. Level of Financial Development ........................................................... 72
2.3.4. Level of Economic Development ......................................................... 73
2.3.5. Overall Uncertainty, Lack of Credibility of Domestic Policies, and
Persistence Effects ................................................................................ 73
3. Econometric Model and Estimation Methods ......................................................... 74
4. Data ......................................................................................................................... 77
4.1. Credit Dollarization Data ............................................................................... 77
4.2. Bank-speciﬁc Variables ................................................................................. 77
4.3. F inn-speciﬁc Variables .................................................................................. 78
4.4. Other Variables .............................................................................................. 80
5. Empirical Analysis .................................................................................................. 81
5.1. Pooled OLS (POLS) Estimation .................................................................... 81
5.2. First Difference (FD) Estimation ................................................................... 84
5.3. Fixed Effects (FE) Estimation ....................................................................... 86
6. Robustness Tests ..................................................................................................... 87
6.1. Endogeneity Issues ......................................................................................... 87
6.2. Alternative Speciﬁcations .............................................................................. 89
7. Conclusions and Future Extensions ........................................................................ 92
REFERENCES ................................................................................................................. 95
APPENDIX ....................................................................................................................... 98
CHAPTER 3
REGULATION, FINANCIAL DEVELOPMENT, AND CAPITAL FLOWS
1. Introduction ........................................................................................................... 1 15
2. A Simple International Portfolio Choice Model ....................................... . ............ 121
3. Empirical Mode] and Estimation Methods ........................................................... 126
4. Data ....................................................................................................................... 130
4.1. Regulatory Measures ................................................................................... 130
4.2. Financial Intermediation and Capital Flows ................................................ 134
4.3. Controls ........................................................................................................ 134
5. Empirical Analysis ................................................................................................ 137
5.1. Effects on Financial Intermediation ............................................................. 138
5.2. Effects on Capital Outﬂows ......................................................................... 140
6. Alternative Measures and Possible Biases ............................................................ 142
7. Conclusions ........................................................................................................... 145
REFERENCES ............................................................................................................... 147

 

T2

T3
T3
Tat
Tab
Tabl
Tabl
Table
Table

CHAI
Table
Table ,‘
Table 3
Table 3

Table 3i

LIST OF TABLES

CHAPTER 2
Table 1. Country and Dollarization Data Coverage ........................................................ 103
Table 2. Summary Statistics ........................................................................................... 104
Table 3a. Pooled OLS

Dependent Variable: Credit Dollarization (% of Total Credit) ............................... 107
Table 3b. Pooled OLS

Dependent Variable: Credit Dollarization (% of Total Bank Assets) ...................... 108
Table 4a. First Difference Estimator

Dependent Variable: Credit Dollarization (% of Total Credit) ............................... 109
Table 4b. First Difference Estimator

Dependent Variable: Credit Dollarization (% of Total Bank Assets) ...................... 110
Table 5a. Fixed Effects Estimator

Dependent Variable: Credit Dollarization (% of Total Credit) ............................... 111
Table 5b. Fixed Effects Estimator

Dependent Variable: Credit Dollarization (% of Total Bank Assets) ...................... 112
Table 6a. IV Regressions

Dependent Variable: Credit Dollarization (% of Total Credit) ............................... 113
Table 6b. IV Regressions

Dependent Variable: Credit Dollarization (% of Total Bank Assets) ...................... 114
CHAPTER 3
Table 1. Country and Regulatory Measures Coverage ................................................... 150
Table 2a. Coding and Data Sources ................................................................................ 151
Table 2b. Summary Statistics ......................................................................................... 153
Table 3a. Correlations among Alternative Regulatory Measures ................................... 154
Table 3a. Correlations among Main Independent Variables .......................................... 155

Table 4a. Dependent Variable: Domestic Credit (% of GDP) ....................................... 156

Table 4b. Dependent Variable: Domestic Credit to Private Sector (% of GDP) ........... 157
Table 4c. Dependent Variable: Domestic Credit (% of Total Bank Assets) ................. 158
Table 5a. Dependent Variable: Banks’ Foreign Assets (% of GDP) ............................. 159

Table 5b. Dependent Variable: Banks’ Foreign Assets (% of Total Bank Assets) ....... 160

xi

LIST OF FIGURES

Figure 1. Credit and Deposit Dollarization in Emerging Market and Transition
Economies ......................................................................................................................... 58

xii

 

 

 

III

tre

INTRODUCTION

Foreign currencies, such as the US dollar or the Euro are widely used in emerging
market and transition economies1 instead of or in addition to domestic currencies, for
transaction, saving, or accounting purposes. The dollarization of the banking sector,
denoted below by ﬁnancial dollarization, occurs when a signiﬁcant share of residents’
deposits and loans with the domestic banks are denominated in foreign currencies.2 The
main two components of this phenomenon are deposit and credit dollarization.

Previous work on ﬁnancial dollarization, as well as on currency substitution and
external dollar debt highlights at least four potential costs associated with this
phenomenon. First, domestic policies become much more complex and their effects on
the economy weaker. The effects of monetary policy in a dollarized economy, for
example, are transmitted through the balance sheet channel in addition to the traditional
interest rate channel. Second, there is more macroeconomic volatility as changes in the
exchange rates affect the value of the dollar assets, liabilities, costs, and returns (the
balance sheet effects). Third, policy makers have fewer available policy instruments as
dollar instruments (deposits and loans) introduce new effects that might work in opposite
directions. Central bankers in highly-dollarized economies cannot choose between
increasing the interest rate and devaluating the domestic currency, since the devaluation
will negatively impact domestic agents with dollar liabilities. Fourth and most
importantly, there is increased exposure to ﬁnancial and currency crises when there are

balance sheet mismatches in the economy in either the real or the ﬁnancial sector.

 

' “Emerging markets” refers mainly to Latin American economies, which are in general highly-dollarized,
while “transition economies” denotes the Central and Eastern European and Central Asian economics.

2 “Dollar” denotes foreign currencies and not necessarily the US dollar. Thus, dollarization refers to the
use of foreign currencies in general.

 

 

 

 

int
wh

trar

pred
econ
one

hnern

mamh
banking
good a]

ﬁnancia

 

Potential beneﬁts, such as increased ﬁnancial intermediation and more saving and
investment instruments available to domestic agents are often neglected. Contrary to the
previous work which highlights the costs, this dissertation focuses on the beneﬁts of
ﬁnancial dollarization. It also examines, theoretically and empirically, the sources of
dollar bank credit and discusses the resulting policy implications for emerging market
and transition economies. Furthermore, it examines the effects of the regulation of
foreign exchange and capital transactions on ﬁnancial intermediation and capital
outﬂows.

The ﬁrst chapter, “Credit Dollarization, Bank Currency Matching, and Real
Activity”, explores the link between ﬁnancial dollarization and integration in the
international intermediate goods market. It addresses the following two questions. First,
what can explain the high levels of dollar credit observed in emerging market and
transition economies? Second, what are the policy implications of this phenomenon?

This chapter develops a general equilibrium model of a small open economy with
predetermined exchange rate and no restrictions on goods and capital ﬂows. The
economy is composed of households, ﬁrms, banks, and the government. There is only
one ﬁnal good that is produced and consumed in the economy, which is also
internationally traded.

The main economic agents are banks and ﬁrms. Banks derive a beneﬁt when they
match by currency their deposits and loans. Hence, there is currency matching in the
banking sector. Firms use dollar loans to ﬁnance the imports of a foreign intermediate
good and domestic currency loans to pay for a domestic intermediate good. Thus,

ﬁnancial dollarization is directly linked to the real sector.

 

 

 

prc
for

CUI

inte
Furt

disir

prod l
dollar

and IC

based
transiti
dollan'g

(ﬁnanc

 

Process
Stall fr;
dOIian/i

Policy 1'

 

Credit dollarization in equilibrium depends on the characteristics of the
production process: the relative price and the elasticity of substitution of the domestic and
foreign inputs in the production function, as well as the relative returns on domestic
currency and dollar instruments.

Results from the model demonstrate that the higher is the integration in the
international intermediate goods market, the higher are the dollarization levels.
Furthermore, policies that attempt to reduce ﬁnancial dollarization will cause ﬁnancial
disintermediation. In addition, factors that affect the demand side of the economy, such
as a change in household preference for dollar deposits, will be transmitted to the
production side and thus change the level of credit dollarization. Hence, deposit
dollarization can create credit dollarization as banks match the currency of their deposits
and loans.

The second chapter, “Credit Dollarization: Is It Firms’ or Banks’ “Fault”?”,
based on joint work with Iva Petrova, is an empirical study of credit dollarization in
transition economies. It separates the contributions of banks and ﬁrms to the
dollarization of credit by grouping potential determinants into bank- and ﬁrm-speciﬁc
(ﬁnancial and mainly real) factors.

To date, the respective contributions of banks and ﬁrms to the dollarization
process remain relatively unexplored. Most studies that analyze ﬁnancial dollarization
start from the presumption that credit dollarization is bank-determined. Whether
dollarization comes mainly from the ﬁnancial sector or from the real sector has important

policy implications. First, it indicates if there are balance sheet mismatches in the

 

 

6C0

\th

ecor
the

and

factr
Mac
indic
exch

effec

dﬁvil
dollar
the a

indire

economy. Second, it shows in which sector of the economy they are concentrated and
which sector (if any) policy makers should target in order to contain dollarization.

The empirical analysis uses a newly-assembled dataset for twenty-two transition
economies for the period 1990 -— 2001, with data mostly available for the second part of
the period. Bank-speciﬁc factors include indicators of asset and liability management
and currency matching, proﬁtability, concentration, and risk management. Firm-speciﬁc
factors are the measures of real dollarization3 and access to alternative ﬁnancing sources.
Macroeconomic controls include factors that affect both ﬁrms and banks, such as
indicators of overall hedging opportunities, liberalization and deregulation of the foreign
exchange market, uncertainty and lack of credibility of domestic policies, and persistence
effects.

Empirical results provide evidence that bank currency matching is the main
driving force of credit dollarization. In addition, there is limited evidence that real
dollarization causes ﬁnancial dollarization. However, among the countries included in
the analysis, currency mismatches tend to be concentrated in the real sector, which
indirectly exposes the economy to ﬁnancial and currency crises.

The third chapter, “Regulation, Financial Development, and Capital Flows”,
evaluates the effects of regulation of foreign exchange and capital transactions on the
domestic and international activities of banks. It separates the effects of foreign
exchange liberalization from the effects of capital outﬂows liberalization. The main

question of the chapter is if allowing banks to freely lend in dollars domestically will

 

3 Real dollarization occurs when frrrns have returns and/or costs of production that are denominated in
foreign currencies. That is, ﬁrms export the ﬁnal good and/or use imported intermediate goods in the
production of the ﬁnal good.

 

 

 

 

 

cu
cu

ins
cos

dau

oper
caph
theli
In ad
libera

banks'

dISInte
Policy
the her

 

make them provide more credit and “go abroad” less, and thus increase ﬁnancial
intermediation and reduce capital outﬂows. In addition, the chapter examines the
determinants of foreign assets for banks in transition economies.

This chapter develops a simple theoretical minimum variance portfolio allocation
model for a domestic bank. Bank liabilities are composed of deposits in domestic
currency and dollars, while assets are composed of loans to residents in domestic
currency and dollars, and foreign assets. There are separate regulation costs on dollar
instruments and on holdings of foreign assets. The effects of changes in these regulation
costs on bank lending and holdings of foreign assets are estimated using the same panel
dataset employed in the second chapter.

Empirical results provide evidence that the liberalization of foreign exchange
operations increases the level of domestic credit, possibly because of keeping domestic
capital in the economy and discouraging capital ﬂight. Limited evidence suggests that
the liberalization of capital outﬂows, on the other hand, reduces ﬁnancial intermediation.
In addition, banks increase their holdings of foreign assets when capital outﬂows are
liberalized. Surprisingly, the liberalization of domestic lending in dollars also increases
banks’ holdings of foreign assets.

Is a dollarized ﬁnancial system better than none? This chapter suggests there is a
disintermediation cost associated with restrictions on foreign exchange operations.
Policy makers might have to weigh the potential costs of ﬁnancial dollarization against
the beneﬁt of higher ﬁnancial intermediation, and decide if they prefer a dollarized

ﬁnancial system to none.

 

 

of

“'31
epis
and

How

Conti.‘

Shi (2

 

 

CHAPTER 1
CREDIT DOLLARIZATION, BANK CURRENCY MATCHING,

AND REAL ACTIVITY

1. Introduction

Financial dollarization, deﬁned as the holding by residents of a signiﬁcant share
of bank deposits and loans denominated in foreign currencies, is a phenomenon present in
most emerging market and transition economies (see Figure 1). Originally, this process
was associated with macroeconomic instability: high inﬂation and sudden depreciation
episodes. As a result, residents’ conﬁdence in the domestic currency was undermined,
and they abandoned their domestic currency in preference for dollar instruments.l
However, macroeconomic stability does not reverse dollarization. Financial dollarization
continues to increase in most emerging economies (see Balii’io et a1. (1999), Honohan and
Shi (2002), and Arteta (2002)), even after stabilization policies are implemented.2

There are several possible explanations for this increasing trend. First, the
stabilization policies are not credible. Even if inﬂation has been curbed, people expect it

to increase again in the future. Second, residents become used to holding dollar

 

’ Following the current literature, I use the term “dollar” to refer to any foreign currency. I use both
“ﬁnancial dollarization” and “dollarization” to refer to partial dollarization of ﬁnancial intermediation. The
chapter does not address issues raised by the “full dollarization” case that is, the adoption of a foreign
currency as a legal tender.

2 Exceptions are several transition economies, where there were sharp declines in the level and share of
dollar deposits and loans once the economy was stabilized.

 

ir

ﬁr

dcr
SIUI
pan

sub:

credi

ﬁnan

3Real
Either a
tfadablc
‘ II IS v
inSII'Um
levels 0
HOOD

increagc

 

 

instruments, which gives rise to persistence/hysteresis effects. Third, the process of
ﬁnancial dollarization is connected to international economic integration, and thus real
sector dollarization3 and trade dependence. Therefore, ﬁnancial dollarization is a natural
result of higher integration of the emerging economies in international goods and
ﬁnancial markets. I explore the last two explanations further in the chapter and abstract
from any credibility issues.

A growing body of literature documents the risks imposed by ﬁnancial
dollarization and policy options in highly-dollarized economies.4 However, only a few
studies evaluate the determinants of dollarization of ﬁnancial intermediation. In
particular, deposit dollarization has been extensively studied in the context of currency
substitution, while credit dollarization has only recently started to receive attention.5

Findings from three studies suggest that other factors, besides macroeconomic
variables, have to be considered in order to explain ﬁnancial dollarization in general, and
credit dollarization in particular. Ize and Levy Yeyati (2003) evaluate the link between

ﬁnancial dollarization and macroeconomic uncertainty. They use a minimum variance

 

3 Real sector dollarization denotes a high pass-through from the exchange rate to the price level, due to
either a high degree of openness of the economy (large tradable goods sector) or dollar pricing in the non-
tradable goods sector (see Ize and Levy Yeyati (2003)).

’ It is well accepted that ﬁnancial dollarization complicates monetary policy and limits available policy
instruments. For papers on the optimal exchange rate regime and monetary policy in economies with high
levels of corporate debt dollarization, see Chang and Velasco (2001), Céspedes et al. (2000), Devereux,
(2001) and (2002), and Ize and Levy Yeyati (2003). Furthermore, frnancial dollarization is believed to
increase the fragility of the domestic banking sector, if ﬁrms and banks have unhedged dollar positions. A
currency crisis might trigger a banking crisis as ﬁrms with unhedged dollar credit cannot repay their debt
and banks become insolvent. For models of currency crises, based on balance sheet effects, see Krugrnan
(1999), McKinnon and Pill (1999), Chinn and Kletzer (2001), and Aghion et a1. (2001).

Surprisingly enough, most existent studies examine the currency composition of external credit, even if
this is almost completely denominated in foreign currencies. Domestic credit, however, shows much more
cross-country and temporal variation, with respect to the currency of denomination. Domestic agents’
incentives to borrow in dollars from abroad are examined by Schneider and Tomell (2001), Burnside et al.
(2000), Martinez and Werner (2002), Caballero and Krishnamurthy (2002), and Jeanne (1999a), (1999b)
and (2001), among others. The external borrowing in foreign currencies is also referred to as “the original
sin” that is, the inability of emerging economies to borrow in their own currency (see Hausmann et al.
(2001)).

 

 

3C?

Ho

dep

afi-

by C
bank
van a

bank

intern.
Open .
domes;
800d 5
Producg
c051 to
deaniI

indusk}

 

portfolio (MVP) allocation model, in which agents compare the hedge beneﬁts of
domestic currency and dollar deposits and loans. Financial dollarization is caused by real
sector dollarization, as well as by volatility of inﬂation and the real exchange rate. The
levels of dollarization predicted by the MVP allocation model only partially match the
actual dollarization levels for a sample of developed and developing economies.
However, only data on deposit dollarization are used in the empirical analysis.

Arteta (2002) empirically evaluates the effect of the exchange rate regime on both
deposit and credit dollarization. While deposit dollarization is higher in a ﬂoating than in
a ﬁxed exchange rate regime, credit dollarization does not seem to vary across regimes.

The importance of bank-speciﬁc factors for ﬁnancial dollarization is highlighted
by Catao and Terrones (2000). They evaluate the link between ﬁnancial dollarization and
banking cost, market structure, and regulatory parameters, in addition to macroeconomic
variables. However, they model only the banking sector, and ignore the demand side of
bank credit and deposits.

This chapter models the link between ﬁnancial dollarization and integration in the
international intermediate goods market. I use a general equilibrium model of a small
open economy with ﬁrms, banks, households, and a government. Firms use both
domestic currency and dollar loans to ﬁnance production. They operate in the traded-
good sector and need a foreign intermediate good in addition to a domestic input to
produce. Hence, there is real sector dollarization and trade dependence. Banks incur a
cost to produce deposits and loans and derive a beneﬁt when matching the currency of
denomination of their “products”. This generates currency matching in the banking

industry.

 

 

 

 

in

di

tha

inc:

CUI'l

stuc

doll.

The link that I model between ﬁnancial dollarization and international economic
integration can explain the increasing dollarization trend observed in practice and
discussed above. As emerging market and transition economies become more integrated
in the international goods markets, ﬁnancial dollarization increases.6 Moreover, I show
that deposit dollarization, caused by a higher preference of residents for dollar deposits
than domestic currency deposits, can create credit dollarization. Thus, factors that
increase deposit dollarization end up increasing credit dollarization as banks match the
currency of denomination of their deposits and loans.7 To date, no other theoretical
studies link bank currency matching to ﬁnancial dollarization, thus allowing deposit
dollarization to cause credit dollarization.8

I consider the effects of several domestic and external shocks on ﬁnancial
dollarization. I show that ﬁnancial dollarization depends on macroeconomic variables, as
well as bank- and ﬁrm-speciﬁc factors. Bank regulations and restrictions on dollar
instruments do matter, as well as the characteristics of the production sector. To the
extent that they vary across countries, so do the levels of credit dollarization. This
chapter complements the existing literature, which explains ﬁnancial dollarization based

on macroeconomic variables only (see Ize and Levy Yeyati (2003)), or on both

 

6 There is empirical evidence of a link between ﬁnancial and real sector dollarization. Honohan and Shi
(2002) and Ize and Levy Yeyati (2003) ﬁnd that deposit dollarization is associated with a high pass-through
from exchange rate changes to the price level. Studies using ﬁrm-level data show that it is exporting ﬁrms
and the ﬁrms in the traded goods sector, thus ﬁrms with returns indexed to the exchange rate, that borrow
(mostly) in dollars (see Keloharju and Niskanen (2001) and Aguiar (2002)).

The hypothesis that banks match the currency of denomination of deposits and credit receives empirical

support as well. Using a panel of fourteen Latin American countries, Baraj as and Morales (2003) conclude
that dollar deposits are the main driving force of dollar credit. Table 1 shows a similar trend in deposit and
credit dollarization ratios for selected transition economies. Arteta (2002), however, ﬁnds that there are
currency mismatches in the banking sector, and that they are ampliﬁed by ﬂoating exchange rate regimes.
8 Calvo (2001) and (2002) refers to currency matching in the banking sector in connection to credit
dollarization (which he denotes by “liability dollarization”), and suggests that bank deposit dollarization
may induce bank loan dollarization, since banks try to match the currency composition of their assets and
liabilities. However, he does not model the link formally.

9

macroeconomic and microeconomic variables, but in partial equilibrium (see Catao and
Terrones (2000)).

Similarly to Ize and Levy Yeyati (2003), I ﬁnd that restrictive measures on
dollarization are counterproductive as they reduce the total domestic credit and thus
output in the economy. However, I do not imply that ﬁnancial dollarization should be
encouraged, nor that measures should not be taken to control it. This process deserves
further analysis, both on theoretical and empirical grounds. Therefore, I conclude that
any restrictive policy measures must look beyond reducing dollarization as a goal in itself
and evaluate ﬁnancial dollarization in connection to the real activity.

The rest of the chapter is organized as follows. Section 2 presents the theoretical
model. Section 3 examines the effects of changes in deposit dollarization (and other
domestic and external shocks) on credit dollarization, and the transmission mechanism to
the overall economy. In section 4 I discuss the economic effects of two policy measures
that aim to reduce ﬁnancial dollarization. Section 5 summarizes the results and presents

possible extensions of the analysis.

2. A Model of Financial Dollarization

I use a rather standard model of an open economy with costly banking. I assume
that banks use tradable resources in order to produce deposits and credit, and derive a
beneﬁt from matching the currency of denomination of their assets and liabilities. The
model is an extension of Edwards and Végh (1997). I introduce dollar deposits and
loans, in addition to those denominated in domestic currency, in order to examine

ﬁnancial dollarization.

10

All distortions in the economy are concentrated in the banking sector.
Furthermore, this is an inside money economy. That is, households and ﬁrms do not hold
cash; they need demand deposits to consume and loans to produce. By modeling an
inside money economy, I can focus on the substitution between domestic currency and
dollar deposits, as well as domestic currency and dollar loans, and abstract from the
substitution between cash holdings and bank deposits, as well as internal funds and bank

loans.

2.1. The Economy

Consider a small open economy with no restrictions on goods or capital ﬂows.

There is only one good produced and consumed, which is non-storable and

internationally traded. The domestic price of this good, P,, is given by the law of one
price: P, = E t P: , where E t is the nominal exchange rate, expressed in units of domestic

currency per dollar, and P: is the dollar price of the good. Real variables will be

expressed in terms of this good.

There is no uncertainty in the economy. All agents have perfect foresight,

therefore the rate of devaluation of the exchange rate, a,( E t /E, ) , is predetermined. I do

not model expectations about the exchange rate or the rate of change in the exchange rate
explicitly. However, I assume that deposit dollarization depends on the rate of

devaluation 6,. Therefore, an increase in the rate of devaluation will increase

dollarization of deposits as households demand more dollar deposits relative to domestic

11

 

 

 

 

 

CUl’

CCO

C01“:

fort

.
7:, I
5311

gen

loal
that

this

Cun
IS i]
fOre

f0”(

HOU

currency deposits. This implicitly captures the higher desirability of dollar deposits in an
economy with a higher rate of devaluation as a way to preserve the value of deposits.

I assume perfect capital mobility. Consequently, the uncovered interest parity

a o . .* . o n a o .* -
condition holds: 1, =1, + a, , where 1, rs the domestrc nominal interest rate and 1, rs the

foreign nominal interest rate. Domestic inﬂation 7r,( Pt /P, ) is equal to a, + 7r: , where

7r:( P,*/ P: ) is foreign inﬂation. By Fisher’s equation, the real interest rate, r , is the
same across the world, and I assume it constant: r = i, — 7r, :1: — 7r: . Without loss of

generality, I assume that foreign inﬂation equals zero: 7r: = 0.

There are no nominal rigidities in the model. I introduce the role for deposits and
loans through two deposit-in-advance and two credit-in—advance constraints. I assume
that households pay a ﬁxed proportion of their consumption with dollar deposits, and that
this proportion depends on the rate of devaluation and a “preference for dollar deposits”
variable. On the production side of the economy, I assume that ﬁrms produce the ﬁnal
good using a constant elasticity of substitution (CES) technology, and both domestic
labor and a foreign good as factors of production. Wages are paid with domestic
currency credit, while imports of the intermediate good are paid with dollar credit. Labor
is inelastically supplied, thus the domestic factor of production is scarce relative to the
foreign. The currency composition of domestic debt is endogenously determined and
follows from the assumptions of credit segmentation and trade dependence.

The economy is composed of households, ﬁrms, banks, and a government.

Households own the ﬁrms and banks, consume the ﬁnal good, and provide labor. In

12

 

a."

I
I,“ us" a.

 

 

 

 

OI

\V

dc

It?

C35

sta'

3"d d

 

(DIA ,

order to consume, households need to keep bank deposits. Firms produce the ﬁnal good
using domestic labor and a foreign input. They need bank credit to pay the cost of the
working capital, both the domestic and foreign factors of production. Banks produce
deposits and loans in both domestic currency and dollars, and can borrow or lend abroad.
The government plays a passive role. It chooses the rate of devaluation and the reserve
requirement ratios for both dollar and domestic currency deposits.

Any agent in the economy can hold internationally traded bonds. This
assumption introduces capital mobility and allows for instantaneous adjustment in the
case of a shock. There are no dynamics in this setup. The economy is always in steady

state equilibrium.

2.2. Households
Households consume the ﬁnal good and provide labor inelastically (they have a
time endowment assumed ﬁxed and equal to 1). Their lifetime utility is

(2.1) Ilog c, exp( —- ,Bt)dt ,
0

where c, is real consumption and ,6 is the subjective discount rate.9

Households must pay for consumption with checks drawn on domestic currency
and dollar deposits. Their need for deposits is introduced through two deposit-in-advance

(DIA) constraints. A fraction of consumption, s,, has to be paid with dollar deposits

(d: ) , the rest has to be paid with domestic currency deposits (d , )

 

9 Following the current literature, I assume from now on ,6 to be equal to r, in order to avoid unnecessary
dynamics.

13

 

 

 

 

the d

 

dolla.
currer
dollar
there
with 1
Preset

8! an

(2.5)

. erg
(1997

(1999)

(2.2) d,’ (1 + e, ) 2. 5,0,,
(23) d, Z(I—s,)c,.
Condition (2.2) includes the devaluation of the exchange rate from the moment

the dollar deposits are created until they are used to pay for consumption.

The deposit dollarization ratio, calculated as the ratio of dollar deposits to total

deposits is therefore equal to s,

d:(1+8t) Stct
=—_=St'

 

(2.4) DD; = ,
d, +d,(1+e,) Ct

The existing literature on deposit dollarization shows that households shift to
dollar deposits following episodes of sudden depreciation as their conﬁdence in domestic
currency is undermined. Furthermore, as they become used to holding dollar deposits,
dollarization remains high even after stabilization policies are implemented. That is,
there are persistence/hysteresis effects.10 While I do not explicitly model the uncertainty
with respect to the exchange rate or the rate of devaluation, I try to capture the two effects

presented above by assuming that deposit dollarization depends on the rate of devaluation

a, and a “preference for dollar deposits” variable A, :

1+
DD, :3, = 8’ 1
(2'5) 1+£,+—-

I

The deposit dollarization ratio DD, (3,) is increasing in the rate of devaluation a,

and the preference for dollar deposits A,. When A, —)0, households do not “like”

 

1° Persistence/hysteresis effects are described by several studies: Guidotti and Rodriguez (1992) and Uribe
(1997) for currency holdings, Honohan and Shi (2002), Ize and Levy Yeyati (2003), and Baliﬁo et a1.
(1999) for bank deposits, and Arteta (2002) for both bank deposits and credit.

14

 

 

 

di

26

dc

ar

Whe
dEpc
{CSpc
Costs

Cases
hold ,

eqUali;

 

dollar deposits and hold only domestic currency deposits. Thus, the dollarization ratio is

zero. As A, increases, the deposit dollarization ratio increases, and the effect on

dollarization of a change in the rate of devaluation decreases. Hence, hysteresis effects
arise.
Households hold demand deposits, both domestic currency and dollar

denominated, and an internationally traded bond. Their ﬁnancial wealth in real terms,

a,” , is given by
(2.6) a," =b!’ +d, +d,*(1+£,),

where b,“ denotes bond holdings.

Households’ ﬂow constraint is then

' a: 1-
07) a," =ra,” +w, +Q,f +9,” +1, —c, —(r, —z',d)d, —(i, —8, —i," )d,(1+e,),
where i,d and if” are nominal interest rates paid for domestic currency and dollar

deposits, w, is the real wage, Q! and Q? are dividends from ﬁrms and banks,

respectively, and t, denotes government transfers. Eq. (2.7) highlights the opportunity
costs of holding deposits, (1', —i,d ) and (i, —s, —i,d*) , respectively. I consider only the

cases where these costs are non-negative, i, 2 i,d and i, —£, 21?”; that is, households
hold deposits for liquidity reasons only. Thus Conditions (2.2) and (2.3) hold with

equality.

15

The household’s lifetime budget constraint is obtained by integrating forward Eq.
(2.7), taking into account Conditions (2.2) and (2.3), and imposing the transversality
condition
ag+°j{w, +r2,f +9,” +1, —c,[1+(i, —i,d)(1—s,)+(i, —r,‘” —£,)s,]}

0
exp(—rt)dt = 0.

(2.8)

Households choose consumption c, , for all t e [0, oo] , so as to maximize lifetime

utility Eq. (2.1), subject to the lifetime budget constraint Eq. (2.8), given the initial
ﬁnancial wealth a3 and the time paths of e,,i,,i,d,i,d*, w,,Q,f,!2,b, 1,.

The ﬁrst-order condition is

(2.9) -c’—= MIMI-Sam -i:’)+s.(i, —s. —i,"")].
t

where A is the time-invariant marginal utility of net wealth. Eq. (2.9) states that the
marginal utility of consumption has to be equal to the marginal utility of wealth times the
effective price of consumption. The price of consumption, in real terms, is higher than
one, due to the additional cost of holding deposits in domestic currency and dollars for

liquidity reasons.

2.3. Firms

Consider a representative ﬁrm that produces output y, according to a CBS

technology, using both domestic and foreign factors of production

I I

1—— 1-— a
(2.10) yt =("t a +1! (7)0-1,

l6

{xi

 

 

 

incl

unti.

Wh ere

TCSpeC

 

where 1, denotes domestic labor, n, denotes a foreign input, and a is the elasticity of

substitution, 0 < a < 1. I assume that ﬁrms have to use bank loans to pay for their
working capital. 11 Furthermore, I assume credit segmentation, that is, the domestic input

has to be paid with domestic currency loans 2,, and the foreign input with dollar loans

2; . The two credit-in-advance constraints are

(2.11) Zr 2 w,l,,

(2'12) z:(l+8t)2Pt"t’

where p, denotes international terms of trade (the foreign price of the intermediate good

in terms of the foreign price of the ﬁnal good) and is exogenous. Condition (2.12)
includes the devaluation of the exchange rate from the moment dollar loans are granted

until they are used to pay for the imported input.

Firm’s real ﬁnancial wealth a ,f consists of bond holdings b,f and bank credit
(2.13) a,f=b,f—z,—zf(1+e,).
Thus, the ﬂow constraint condition is

2.14 ' ' - "‘ - ’
( ) atf=ratf+yt_wtlt_ptnt_gtf_(lt1_lt)zt_[ltl '(lt-Et)]zt(1+£t)’

* . . . .
where i,’ and i,’ are nomrnal interest rates on domestic currency and dollar credrt,

respectively.

 

” The retained earnings are equal to zero, since ﬁrms are assumed to distribute in each period the proﬁts as
dividends.

l7

 

 

 

do

de:

am

ﬁor

trar

duco

time F

 

the m2:
itSma

Ihe cos

The opportunity costs, i,’ —i, for domestic currency credit, and i," —(i, —3, ) for

dollar credit, are assumed non-negative. Thus i,I 2i, and 1',” 2(2‘, —e,), and ﬁrms
demand only as much bank credit as they need to ﬁnance production. Conditions (2.11)
and (2.12) hold with equality.

F irm’s present discounted value of lifetime dividends is obtained by integrating
forward Eq. (2.14), taking into account Conditions (2.11) and (2.12), and imposing the
transversality condition

[9/ exp( -rt)dt

(2.15) 0

oo
0

The ﬁrm chooses labor, 1,, and a foreign input, n, , so as to maximize the present

discounted value of dividends Eq. (2.15), given the initial ﬁnancial wealth a5 and the

I o o o *
time paths of w,,p,,r-:,,r,,r,l,z,I .

The ﬁrst-order conditions are

(216) _l 1.]. 1_i __1_
' l, ”(n, a +1; ”)0-1=Wt[1+lt "'tja
-1 1__]. 1-1 L

Eqs. (2.16) and (2.17) state that ﬁrms equate the marginal productivity of labor to
the marginal cost of a unit of labor, and the marginal productivity of the foreign input to
its marginal cost. The costs stemming from the credit-in-advance constraints add up to

the cost of labor and the foreign input.

18

 

 

 

 

cred

(2.19

highe
cost q

dome.

and res

I’l'es

and d;

(2.20)

 

I deﬁne the credit dollarization ratio as the ratio of dollar bank credit to total bank

credit, and calculate it using Conditions (2.11) and (2.12)

*
Zt(1+32) = 'Ptnt
z, +z:(l+£,) Ptnt +tht

 

(2.18) CDt =

The credit dollarization ratio can be expressed as follows, using Eqs. (2.16) and

(2.17)

Cth I I 0’.
-- at
(2.19) ”[3] ”(1+4 —(i,—£,)]

 

 

Eq (2.19) shows that the level of credit dollarization in an economy is higher the
higher the relative cost of the foreign input to labor and the lower the relative opportunity
cost of dollar to domestic currency credit. In Section 3 I consider the effects of both

domestic and external shocks on the dollarization levels.

2.4. Banks
Banks take deposits from households and lend to ﬁrms. Their assets consist of

foreign bonds, b,b , domestic currency and dollar credit to ﬁrms, 2, and z; , respectively,

and reserves with the government, both for domestic currency and dollar deposits, h, and
h; , respectively. Their liabilities consist of domestic currency and dollar deposits, d,

and d; , respectively. Banks’ net wealth, a,b , is given by

(2-20) a,b =b,b +h, +h:(1+£,)+z, +z;(1+e,)—d, —d:(1+£,).

19

"V‘LSI'WW ‘Y

 

 

 

 

dep
in I

CCOl

vvher
cost

and]

 

 

andr

“Tate:

er

Banks are modeled as productive units.12 They use tradable resources to produce
deposits and loans.13 Moreover, deposits and loans are jointly produced, and an increase
in the level of deposits reduces the marginal cost of credit. Thus I assume there are
economies of scope in the banking industry. 14

The banking cost function is the following

* *
”(dt’zt)+Kt’7(dt .21).
where K, is the relative cost of dollar to domestic currency deposits and loans, and the

cost function is separable across currencies. Thus, economies of scope between deposits

and loans are assumed positive within currencies and negligible across currencies15

”dz <0, ﬂdtzt <0, ”dz. =0, ”d‘z =0.
The cost function satisﬁes

* A at
n(0,0)=0, nd(0,z,)=0, nz(d,,0)=0, ”(I-(0,2, )=0, "24d, ,0)=0,

and the usual monotonic and convexity assumptions, for all d, , d: , z, , 2: >0

 

'2 For models of costly banking, see Fisher (1983), Diaz-Gimenez et al. (1992), Edwards and Végh (1997),
Catao and Terrones (2000), Catao and Rodriguez (2000), and Burnside et a1. (2000).

'3 These are costs to service deposits and credit, to rent buildings, use ATMs, evaluate creditors, etc.

1’ The assumption of economies of scope, deposits and loans are complements in the cost function, can be
interpreted as a competitive advantage of banks as inside lenders. According to Fama (1985), the cost of
making loans to depositors is lower than for non-depositors. The ongoing history of a borrower as a
depositor can be useful in assessing credit worthiness and lowering monitoring costs. The empirical
evidence on banking economies of scope, however, is hardly conclusive. It seems to be agreement among
economists that “if economies of scope exist, they are expected to be small” (Pulley et a1. (1993), p.1).
However, most empirical studies refer to US or other developed economies (see Mester (1994), Mester et
a1. (2001), and Santos ( 1998)). Banking economies of scope are likely to be higher in emerging markets.

'5 This assumption is similar to Catao and Terrones (2000). Financial intermediaries usually adopt hedging
strategies. It might be the case that banks keep different balance sheets for deposits and loans denominated
in different currencies. Or that there are some regulations that encourage banks to match the denomination
of deposits and loans by imposing a cost on banks with mismatched portfolios.

20

”d >0, ”d. >0,I]z >0, 772: >0, ”dd >0, ”dodt >0, ﬁzz >0, ’1th. >0.

The economic interpretation for this cost function is that banks have to give up

”(db 2, )+ K1401; , z: ) resources in order to produce at, units of domestic currency

deposits, at: units of dollar deposits, 2, units of domestic currency credit and 2: units of
dollar credit. The variable K, captures any regulation of or restrictions on dollar

deposits and loans.

Banks’ ﬂow constraint is

(2.21) atb =mtb +011 —it)zt +011. _(it—£t)jz;(1+8t)+(it —i,d)d, “I‘m-51 ‘0‘”)
at , _ rt- t t
dt(1 + £,)-—r,h, ‘(lt ‘51)ht (1 +81)-'l(dt’zt)—Kt'l(dt 'zt )‘Qrb-

Banks gain the difference between the bond interest rate and the deposit and
credit interest rates. They incur the production cost and the opportunity cost for required
reserves. Assuming that the government does not pay any interest on reserves, banks will
not hold any excess reserves in either the domestic currency or in dollars
(2.22a) ht = Mb

(222!» h; =a;d,*,

where 6, and 6: are the required reserve ratios for domestic currency and dollar deposits

respectively.
Bank’s present discounted value of lifetime dividends is obtained by integrating

forward Eq. (2.21) and imposing the transversality condition

21

 

 

 

TC

 

 

00
I 9,1) exp( -rt)dt
0

2.23 °° ‘ " i
< ) zag . (my -1, )2, + [1,1 _(.-, —a,)]2,(1+81)+(i1-ird)d1 +[(iz—€:)—ird )1
0 .

d.“(1+e.)—i,h, ~01 —e.)h.'(1+s,)-n(d..z.)-Km(d£' .2? )1 expr-rodt.
Banks choose the level of domestic currency and dollar deposits, d, and d: ,
respectively, the level of reserves h, and h: , and domestic currency and dollar credit, 2,

and z: , respectively, in order to maximize the present discounted value of dividends Eq.

(2.23), subject to the minimum reserve constraints Eqs. (2.22a) and (2.22b), given the

initial assets a3 and the time paths of 8,, i,, 1,], i,1*, i,d, if”, 6,, 5:, K,.16

The ﬁrst-order conditions from the maximization problem are

it] ‘it =’72(dt»zt)a

(2.24)

.1* . :1: :1:
(2.25) 1, -(11-81)= Km; (d1 .21 ),
(2.26) it ‘itd =i15t+”d(dtrzt)a
(2.27)

. . t . t t #
('1'51)"td =(1t‘3r)5t +Kt’7d’(dt'zt)-

Eqs. (2.24) - (2.27) reﬂect the distortions from the economy. If there were no cost
to “produce” credit, competitive banks would equate the lending rates to the nominal
bond rate (taking into account the rate of devaluation, for the dollar credit). Therefore,

the lending spreads depend on the marginal cost of credit. The deposit spreads depend on

 

’6 Banks’ dividends do not necessarily have to be zero. This is a sensible assumption for emerging
economies in general, and transition economies in particular.

22

the marginal cost of deposits and the additional cost of holding unremunerated reserves

with the central bank.

2.5. Government

The government is modeled similarly to Edwards and Végh (1997). It plays the
role of both monetary and ﬁscal authorities.

The government chooses the rate of devaluation e, and the reserve requirement
ratios for domestic currency and dollar deposits, 6, and 6 * , respectively. It receives
interest on its bond holdings, b,g , and seignorage revenue, 7r,h, , and makes lump sum

transfers, 1, , to households. Government’s net asset holdings a,g are

(2.28) 11,8 =b,g—h,-h,*(1+e,).
Government’s ﬂow constraint is given by

2.29 ' . . .. + ..
( ) a,g =ra,g+z,h,+(r,—7r,)h,(1+£,)+77(d,,z,)+K,77(d,,z,)—r,.

I have assumed the banking cost to be a private, but not a social, cost.17 Thus, the

govemment’s lifetime constraint is as follows, imposing the transversality condition

(2.30) ag + [[11, + 7r,h, + h,’ (1 + a, ) + "(11, , z, ) + K,17(d: , z: ) — r, ]exp(—rt)dt = 0.
0

 

'7 Edwards and Végh (1997) show how this assumption simpliﬁes the analysis by making the size of the
banking sector irrelevant.

23

 

 

 

equ
ch

stat

that

that

equa

2.6. Competitive Equilibrium

There are no intrinsic dynamics in this economy. The economy is in steady-state
equilibrium as long as the exogenous variables are constant over time. The adjustment to
changes in exogenous variables is instantaneous and the economy moves to a new steady—
state equilibrium.

The labor market equilibrium condition is l, = 1. Perfect capital mobility implies
that i, =1: + 8, , and perfect mobility in the goods market (the law of one price) implies

that it, = 7r, + 8,. Srnce I have assumed foreign mﬂatron to be zero, domestrc inﬂation rs

equal to the rate of devaluation.

I denote the economy’s net stock of bonds by k, . The economy’s ﬂow constraint

(equilibrium current account condition) is derived using Eqs. (2.7), (2.14), (2.21) and

(2.29)

(2.31) kt = rkt +y, —c, —p,n,.

The economy’s resource constraint is derived from Eqs. (2.8), (2.15), (2.23) and
(2.30), imposing the transversality condition
m
(2.32) k0 + ﬂy, —c, — p,n,)exp(—rt)dt = 0.
0
In equilibrium, the ﬁnancial dollarization ratios are given by Eqs. (2.4), (2.5),

(2.18) and (2.19)

1+
DD, :5, = 8‘

1+£,+i
r

24

 

 

 

CXI
SII’C
a SI

C0111

cost
aS‘u'
bank

doHaI

rate 0

regula

 

detern,

dOmeS,

 

CD,= Z,(I+8,) = 1

t _ U .
2: +Zt(1+51) “(E—)1 0[1+i,1‘ -(i, -£,)]

 

 

 

Pt 1 + i,’ — i,

The currency composition of deposits varies with the rate of devaluation of the
exchange rate and households’ preference for dollar deposits. When residents have a
strong preference for dollar deposits, a decrease in the rate of devaluation will have only
a small effect on deposit dollarization. Thus deposit dollarization depends on the initial
conditions in the economy and is likely to show persistence effects.

The currency composition of credit, on the other hand, depends on the relative
cost and elasticity of substitution between the foreign and domestic factors of production,
as well as the lending interest rate spreads. Thus credit dollarization varies with both
bank- and production-speciﬁc factors. Changes in deposit dollarization affect credit

dollarization through changes in the interest rate spreads.

3. Impact of Domestic and External Shocks on Financial Dollarization

In this section, I examine how ﬁnancial dollarization varies with changes in the
rate of devaluation of the exchange rate, preference of residents for dollar deposits,
regulation on dollar instrument holdings, and international terms of trade.

Timing is important. At the beginning of the period, households and banks
determine the level of domestic currency and dollar deposits. Similarly, ﬁrms and banks
determine the level of domestic currency and dollar loans. Banks “produce” the required

domestic currency deposits and credit. A shock hits then the economy. Households use

25

 

 

ra
ho.
cha

and

Shoe.

the ec

(”01156T

 

’hCI‘err

(3. //

 

 

all deposits to consume; ﬁrms use all credit to produce, and then sell the ﬁnal good on the
domestic or international market.

The economy is initially in steady-state equilibrium. When the shock hits, the
economy adjusts immediately and reaches the new steady-state equilibrium. I examine

these effects by comparative static analysis.

3.1. Impact of a Change in Households’ Preference for Dollar Deposits

Consider an exogenous increase in households’ preference for dollar deposits,

measured by A,. This variable captures all factors other than changes in the exchange

rate that make dollar deposits more “desirable”. For example, residents might get used to
holding dollar instead of domestic currency deposits.18 Another way to think about this
change is to compare economies with different persistence effects, everything else equal,
and see if they exhibit different patterns of ﬁnancial dollarization.

In addition to the effects on ﬁnancial dollarization, 1 study the propagation of
shock from the consumption to the production sides and examine the overall effects on
the economy.

Proposition 1: An increase in households ’ preference for dollar deposits, A,,
causes an increase in deposit dollarization, DD,. Dollarization of credit, CD,,

increases if the following (suﬂicient) condition holds

I
(3'1) p,(1+K,nz-)Sol‘”.

 

'8 Or, residents of transition economies might switch to holding Euro deposits, in preparation for the
accession to the European Union.

26

“Inv—

 

 

 

 

 

am

mi gh

ambi

 

Real output increases unambiguously; the eﬂect on the real wage, however, is

ambiguous. Consumption and thus welfare increases if the following condition holds
* - 1

Proof: See Appendix.

The algebraic derivation of the results is provided in the Appendix. In this
section, I present the interpretation of the effects and transmission mechanism for the
preference shock.

An increase in the preference for dollar deposits directly increases deposit
dollarization. As households use more dollar deposits to pay for consumption, holdings
of dollar deposits increase. The increase in preference for dollar deposits tends to reduce
domestic currency deposits; the indirect effect from a possible increase in consumption
might work in the opposite direction. Thus the effect on domestic currency deposits is
ambiguous. While more dollar deposits tend to increase consumption, potentially less
domestic currency deposits tend to reduce it. Condition (3.2) states that consumption
increases if the effective cost of domestic currency deposits, relative to that of dollar
deposits, is higher than the increase in the marginal cost of dollar deposits, relative to the
increase in the marginal cost of domestic currency deposits. Therefore, whether
households consume more or less depends on how costly it is for them to substitute
domestic currency with dollar deposits, and how costly it is for banks to produce more
dollar deposits relative to domestic currency deposits.

The shock is transmitted to the production side through the banking sector. Banks
jointly produce deposits and credit, and they match them by currency of denomination.

The cost of dollar credit decreases as more dollar deposits are produced. Holdings of

27

dollar loans increase. However, it is not clear what happens to holdings of domestic
currency loans, and thus to credit dollarization. Condition (3.1) states that credit
dollarization is more likely to increase, the lower the marginal cost of the foreign
intermediate good and the higher the elasticity of substitution between inputs in the
production function. The intuition for this result is as follows. As dollar loans cost less,
ﬁrms demand more of those, and then use the dollar loans to import more foreign input.
With more foreign intermediate good, production increases. The lower is the cost of an
additional unit of foreign good, the higher is ﬁrms’ demand for imports, and thus the
higher is the demand for dollar loans. As ﬁrms substitute labor with the foreign factor of
production, demand for labor falls. Labor is supplied inelastically, thus the real wage
tends to drop. The income effect from the increase in output, however, tends to increase
wages. The higher the substitution effect, the more likely wages are to drop. As
domestic currency loans are used to pay wages, the higher is the substitution effect, the
lower ﬁrms’ demand for domestic currency loans.

To summarize, an increase in deposit dollarization, due to a shift in households’
preference from domestic currency to dollar deposits, might lead to an increase in credit
dollarization as ﬁrms substitute domestic for foreign intermediate goods.19 Hence, credit
dollarization is associated with higher integration in the intermediate goods market.

If Condition (3.1) does not hold, the effect on credit dollarization is ambiguous.

Thus I show that deposit and credit dollarization do not necessarily move in tandem;

 

'9 Intuitively, the results would be similar if I assumed elastic labor supply. The key assumption that drives
the results is that the domestic factor is scarce relative to the foreign factor of production. Production can
increase only if imports of intermediate goods increase. This assumption is reasonable for most small open
emerging economies (major oil-supplier countries, for example Russia, are not considered).

28

 

(5

the

(32

cOnsr
Peﬁo.

bank- and ﬁrrn-speciﬁc factors, such as banking cost, restrictions on dollar instruments,
as well as the characteristics of the production process matter.

While output increases, the effect on the real wage is ambiguous.20 Welfare is
measured by households’ consumption. Thus the effect on welfare is ambiguous.
Whether the overall economy is better off with more ﬁnancial dollarization depends on
how costly dollar deposits are for households and banks, relative to domestic currency

deposits.

3.2. Impact of a Change in the Rate of Devaluation

Assume there is an exogenous change in the rate of devaluation, that is, 8,
increases. Then the following proposition holds:

Proposition 2: An increase in the rate of devaluation, 8,, causes an increase in

deposit dollarization, DD,. Dollarization of credit, CD,, increases if the following

(suﬂicient) condition holds
A
p,(1+K,nz.)_<_ol‘0.
Real output and real wage increase. Consumption and thus welfare increase if
the following condition holds

(3.3) {awn—(1+8. +3916. +ndr—(r6? +Kmd* )>(1_A,(1+.~, +-/:—)
t t

t
dt Kr’ld‘d‘ ‘dt’ldd-

 

2° The real wage can be interpreted as the return to any domestic factor of production. Rather than
considering labor, I could have looked at any non-traded intermediate good, and assumed that in each
period residents have a ﬁxed endowment of this good.

29

Proof: See Appendix.

The intuition for these results is as follows. An increase in the rate of devaluation
has direct effects on both the consumption and production sides of the economy. First,
the nominal interest rate increases, thereby raising the opportunity cost of holding
deposits and loans. Second, the value (in domestic currency) of dollar deposits and loans
increases. On the consumption side, there is the additional effect of households
demanding more dollar deposits as they become more desirable. In addition, there are
indirect effects ﬁom the consumption to the production side and vice versa, transmitted
through the banking cost. I assume the indirect effects to be small relative to the direct
effects (see Appendix).

Consider the production side. After the shock, dollar loans cost more, but their
value in domestic currency increases. I show in the Appendix that the valuation effect
dominates the cost effect. Hence, while actual holdings of dollar loans decrease, the level
of dollar credit in the economy increases. As the value of dollar loans increases and the
relative price of the intermediate good does not change, ﬁrms can import more and
produce more. As the demand for labor increases and labor is inelastically supplied, real
wage increases. Consequently, the demand for domestic currency loans increases.

Effects on credit dollarization are ambiguous as the level of both dollar and

domestic currency credit increases. However, a sufﬁcient condition for CD, to increase

is given by Condition (3.1). The lower the marginal cost of the foreign input and the
higher the elasticity of substitution between inputs in the production function, the more

likely credit dollarization is to increase. Thus, the lower the cost of an additional unit of

30

foreign good, the higher the demand of ﬁrms for imports and the higher the demand for
dollar loans.

Things are more ambiguous, however, on the consumption side. I have assumed
that deposit dollarization increases with the rate of devaluation. The return to dollar
deposits relative to domestic currency deposits increases. So does their value. In
addition, households want to hold more of their deposits as dollar deposits. There is a
negative effect as well, from the increase in the opportunity cost for both dollar and
domestic currency deposits as the nominal interest rate increases. The ﬁnal effect on the
level of domestic currency and dollar deposits is ambiguous.

The increase in deposit dollarization due to the increase in the value of dollar
deposits has a positive effect on consumption. The higher cost of both types of deposits,
however, has a negative effect. Consumption increases or decreases, depending on which
effect is larger. 1 show that consumption increases if Condition (3.3) holds. This
condition is similar to Condition (3.2). In addition, it includes the valuation effects for
deposits and required reserves.

Condition (3.3) states that consumption (welfare) increases if the effective cost of
domestic currency deposits, relative to that of dollar deposits, is higher than the increase
in the marginal cost of dollar deposits, relative to the increase in the marginal cost of
domestic currency deposits. The cost of required reserves for dollar deposits increases
with a devaluation of the exchange rate. The value of the dollar deposits, however,
decreases. Ultimately, the effect on consumption, and thus on welfare, depends on how

costly it is for households to substitute domestic currency with dollar deposits, and how

31

costly it is for banks to produce more dollar deposits, relative to domestic currency

deposits.

3.3. Impact of a Change in Restrictions on Dollar Instruments

Consider an exogenous change in the relative banking cost of dollar to domestic

currency deposits and loans, measured by K ,. This can be interpreted as an increase in

the regulation of or restrictions on dollar instruments. The effects on the economy are as
follows.

Proposition 3: An increase in restrictions on dollar deposits and loans, K ,, has,
by assumption, no effect on deposit dollarization, DD,. Credit dollarization, CD,,

decreases if the following (suﬂicient) condition holds

1
p,(1+K,nz+)Sol‘0.

Real output, real wage, and consumption decrease. Since consumption decreases,
welfare decreases as well.

Proof: See Appendix.

The change in regulations has a direct impact on both the production and
consumption sides of the economy. On the production side, the banking cost to produce
dollar loans increases. Thus the opportunity cost for ﬁrms to borrow in dollars increases,
and holdings of dollar loans decrease. As dollar loans are used to pay for the foreign
input, ﬁrms buy less foreign intermediate goods. Since they are already using all labor in
the economy, they cannot substitute the foreign factor for the domestic factor of

production. Therefore, production decreases, and real output drops. The income effect

32

dominates the substitution effect, and the real wage decreases. As ﬁrms need less
domestic currency loans to pay wages, these holdings also decrease. Consequently, when
dollar loans become more expensive relative to domestic currency loans the overall level
of credit in the economy decreases. This result follows from the assumptions that the two
types of loans jointly ﬁnance production and that the domestic factor is scarce relative to
the foreign factor of production.21

The effects on credit dollarization are ambiguous, since both dollar and domestic
currency credit decrease. However, Condition (3.1) provides a sufﬁcient condition for
dollar loans to decrease more, and thus for credit dollarization to drop. The lower the
marginal cost of the foreign input and the higher the elasticity of substitution between
inputs in the production function, the more likely credit dollarization is to decrease.

On the consumption side, as the cost of producing dollar deposits increases, their
return decreases. Thus dollar deposit holding drop. However, since deposit dollarization
does not change (there is no substitution of dollar deposits with domestic currency
deposits), consumption has to decrease, triggering a decrease in holdings of domestic

currency deposits.22

3.4. Impact of an External Supply Shock

Suppose there is a shock to terms of trade, that is, the relative price, p, , of the

foreign factor of production increases. Then the following proposition holds.

 

2' These are reasonable assumptions for emerging economies, especially for the case of transition
economies, where there is credit segmentation (see Bonin (2001)) and domestic resources are scarce.

22 In a more general case, deposit dollarization decreases as more restrictions on dollar deposits are
imposed. Intuitively, as banks have less dollar deposits, the marginal cost of credit increases even more
than without the effect from deposit dollarization. Thus credit dollarization is even more likely to drop.
Output still falls. However, the effect on consumption and thus welfare might become ambiguous.

33

Proposition 4: An increase in terms of trade, p,, has, by assumption, no effect
on deposit dollarization, DD,. Credit dollarization, CD,, increases if the following

(suﬂicient) condition holds

1
(3'4) p,(1+K,r]z+).<.(1—a)1“’.

The real wage decreases unambiguously; the effects on real output and
consumption are ambiguous. However, a suﬂicient condition for real output,

consumption and thus welfare to decrease is

1
(3-5) p,(1 + 191,2.) 2 (1 —o)"".

Proof: See Appendix.
On the production side, as the cost of the foreign input and thus the production
cost increases, demand of ﬁrms for labor falls. Since labor supply is inelastic, the real

23 This triggers a fall in the demand for domestic currency loans. Thus

wage falls.
holdings of domestic currency loans decrease.

The effects on real imports of intermediate goods and real output depend on the
marginal cost of the foreign input and the elasticity of substitution between factors of
production. The higher is the marginal cost of the foreign input, the higher is the effect
on output of an increase in the input price. The higher the elasticity of substitution is, the
higher the drop in volume of imports as their price increases.

There are two opposite effects on holdings of dollar loans. One is caused by the

increase in the price of imports, the other by a potential decrease in the volume of

 

23 Real wage falls even if labor supply were elastic, just not perfectly elastic.

34

I: \r'"? " .. ;:

 

 

 

 

 

$3
of

dol

cos
don
deer
reas
(Ion

COIlS

 

“inl

Cons

dolla

dona:

imports. Depending on which effect is larger, credit dollarization might increase or

decrease. A sufﬁcient condition for CD, to increase is given by Condition (3.4). It

states that the lower is the marginal cost of the foreign input and the lower is the elasticity
of substitution between inputs in the production ﬁrnction, the more likely credit
dollarization is to increase.“

The shock is propagated to the consumption side through the change in banking
costs. The lower level of domestic currency loans increases the marginal cost of
domestic currency deposits. Holdings by residents of domestic currency deposits
decrease. If Condition (3.5) holds, less dollar loans are produced and following the same
reasoning, dollar deposit holdings decrease. Thus consumption decreases. However, if
Condition (3.4) does not hold, and more dollar deposits are produced, the effect on
consumption is ambiguous. Additional dollar deposits tend to increase consumption,
while less domestic currency deposits tend to decrease it. Consequently, the effect on

consumption is ambiguous.

3.5. Determinants of Credit Dollarization

This section summarizes the effects of domestic and external shocks on credit
dollarization. Financial dollarization is connected to the real sector through the real

dollarization and trade dependence assumptions. However, the way it responds to shocks

 

2’ Consider the two extremes cases: 0—» 0 and 0—» l . In the ﬁrst case, the Leontief technology, an increase
in p, increases CD, as the price of the foreign input increases, and factors cannot be substituted one for the
other. In the second case, the Cobb-Douglas technology, the effect on CD, is ambiguous. As p, increases,
ﬁrms want to substitute foreign input for domestic labor. Domestic labor is supplied inelastically, however,
thus real imports and output decrease.

35

v—a

 

of :
inst
Hox
doll
due

cred
doll'

elast

dollri

inter
d011,
dollal

restrr

 

 

depends on ﬁrm- and bank-speciﬁc factors, such as banking cost, bank currency
matching and the characteristics of the production sector.

Assume ﬁrst that the marginal cost of the foreign input is small and the elasticity
of substitution is large. Then both an increase in households’ preference for dollar
instruments and an increase in the devaluation rate raise ﬁnancial dollarization.
However, the transmission mechanism for the two shocks differs. Higher preference for
dollar deposits and thus higher deposit dollarization triggers higher credit dollarization
due to bank currency matching. Higher rate of devaluation increases both deposit and
credit dollarization directly.

Restrictive measures on dollar instruments, on the other hand, reduce credit
dollarization (while leaving, by assumption, dollar deposits unchanged).25

Assume now that the marginal cost of the foreign factor is still small, but the
elasticity of substitution is also small. Then higher terms of trade increase credit
dollarization.

The model predicts that the more integrated is an economy in the international
intermediate goods market, the more likely it is to respond to an increase in deposit
dollarization with an increase in credit dollarization. Similarly, there is more credit
dollarization in economies with higher rates of devaluation of the exchange rate and less

restrictions on holdings of dollar instruments.

 

2’ Intuitively, a decrease in deposit dollarization would amplify the downturn in dollar loans.

36

 

 

do}

car

 

is L
the
the

COD

eff

redl
dire
doll
eco
Iron

and

In 1

Staf

 

4. Policies that Aim to Reduce Financial Dollarization

Governments in emerging markets might be concerned with the high level of
dollarization and attempt to reduce it. These restrictive policies can be ill advised as they
can end up having disinterrnediation effects.

Consider the increase in the regulation of dollar instruments, which increases the
cost of dollar deposits and loans. In section 3.3, I present the intuition for the resulting
drop in the total credit, real output, and consumption. To summarize, since dollar credit
is used to ﬁnance imports of foreign intermediate goods, any restriction on its level limits
the access to foreign goods. When the economy has limited domestic resources, which is
the case for most emerging markets, total credit and real output decrease. A sufﬁcient
condition for credit dollarization to decrease is Condition (3.1). Thus, while regulatory
measures might contain credit dollarization, they do have adverse disinterrnediation
effects.

Alternatively, consider a decrease in the rate of devaluation. The government can
reduce the rate of devaluation in an attempt to stabilize the economy. This policy might
directly target dollarization of the banking sector. Suppose ﬁrst that there is no
dollarization. Then the only channel through which the effects are transmitted to the
economy is the interest rate channel. A decrease in the rate of devaluation decreases the
nominal interest rate, and thus the cost of consumption and production. Consumption
and real output increase.

When the economy is partially dollarized, however, things can be very different.
In section 3.2 I show the effects of an increase in the rate of devaluation. When a

stabilization policy is implemented, the balance sheet channel, in addition to the interest

37

 

 

ran
less

dep

1A8
don
loar
stat“

Iota

 

 

 

enre
loar

rece

inte
the

1 sh

rate channel, propagates the effects within and across sectors.26 Dollar deposits are now
less desirable and households substitute them with domestic currency deposits. Thus
deposit dollarization decreases. If Condition (3.3) holds,_consumption increases.

The cost of dollar loans is lower now, but so is their value in domestic currency.
As less foreign factor of production is purchased, less ﬁnal good is produced and less
domestic currency loans are demanded. Holdings of both dollar and domestic currency
loans decrease. Credit dollarization decreases if Condition (3.1) holds. While the
stabilization policy might reduce ﬁnancial dollarization, it unambiguously reduces the

total credit and real output in the economy.

5. Conclusions

Recent studies highlight the high degree of banking sector dollarization in many
emerging economies. Macroeconomic uncertainty, which was originally believed to be
the main cause of dollarization, is no longer able to explain its trend and patterns. In
emerging market and transition economies, residents continue to use dollar deposits and
loans, even after successﬁrl stabilization policies are implemented. Furthermore, several
recent empirical studies identify an increasing trend in dollarization.

This chapter shows the connection between ﬁnancial dollarization and
international economic integration. As emerging economies become more integrated in
the international goods markets over time, ﬁnancial dollarization increases. Furthermore,

I show that ﬁnancial dollarization depends on macroeconomic variables, as well as bank-

 

“ As the debt of ﬁrms is partially denominated in dollars, a change in the rate of devaluation of the
exchange rate modiﬁes the value of debt.

38

 

 

of
(16V
sim
doll

hou

10811

81101

and

liabil

bOnd

 

and ﬁrm-speciﬁc factors. Bank regulations and restrictions on dollar instruments do
matter, as well as the characteristics of the production sector. To the extent that these
vary across countries, so do the levels of credit dollarization. Finally, I show that deposit
dollarization might cause credit dollarization as banks match the currency of
denomination of their deposits and loans.

If ﬁnancial dollarization is indeed a result of integration and globalization forces,
then policies that attempt to reduce it might be ill advised. I show how an increase in
regulation of dollar instruments and a decrease in the rate of devaluation end up hurting
the economy by causing ﬁnancial disinterrnediation.

The chapter can be extended in several ways by relaxing some of the assumptions
of the model. First, I assume that deposit dollarization depends only on the rate of
devaluation of the exchange rate and the “preference for dollars” of residents. This
simpliﬁcation keeps the analysis tractable. However, some effectsfrom credit to deposit
dollarization might be missing. The analysis could be extended to explicitly model
households’ choice between the two types of deposits.

Second, I assume a representative ﬁrm that uses domestic currency and dollar
loans to jointly ﬁnance production. Different ﬁrms might use different types of loans.
Therefore, a possible extension of the model is to introduce heterogeneous ﬁrms, and to
allow ﬁrms to use either domestic currency or dollar credit, or both.

Third, I assume that banks derive a beneﬁt from matching by currency deposits
and loans. Banks could instead match the currency of denomination of their assets and
liabilities. In that case, more dollar deposits might lead to more holdings of foreign

bonds, and the link between deposits and loans becomes weaker.

39

 

 

Oil

561

an

dol

 

 

Illnllf

Finally, the implications of the model should be tested empirically as more data
on credit dollarization become available. Does ﬁnancial dollarization depend on the real
sector and the characteristics of the banking sector? Further research on both theoretical
and empirical grounds is needed to shed lights on these issues. Until that research is

done, however, practitioners should be cautious when attempting to restrict dollarization.

40

 

A 3

Ar

Bai

Bar

801

Bur

Cab

Cali

Cah

Cats

can,

REFERENCES

Aghion, P., P. Bacchetta, and A. Banerjee. 2001. “A Corporate Balance Sheet Approach
to Currency Crises.” Studienzentrum Gerzensee Working Paper 01/05.

Aguiar, M. 2002. "Devaluation, Foreign Currency Exposure, and Investment: The Case
of Mexico." Mimeo. University of Chicago.

Arteta, C. O. 2002. "Exchange Rate Regimes and Financial Dollarization: Does
Flexibility Reduce Bank Currency Mismatches?" Board of Governors of the
Federal Reserve System, International Finance Discussion Paper 738.

Arteta, C. O. 2003. "Are Financially Dollarized Countries More Prone to Costly Crises?"
Board of Governors of the Federal Reserve System, International Finance
Discussion Paper 763.

Baliﬁo, T. J., A. Bennett, and E. Borenzstein et a1. 1999. “Monetary Policy in Dollarized
Economies.” International Monetary Fund Occasional Paper 171.

Barajas, A. and A. Morales. 2003. "Dollarization of Liabilities: Beyond the Usual
Suspects." International Monetary Fund Working Paper 03/ 1 1.

Bonin, J. 2001. "Financial Intermediation in Southeast Europe: Banking on the Balkans."
Mimeo. Vienna Institute for International Economic Studies.

Burnside, C., M. Eichenbaum, and S. Rebelo. 2000. "Hedging and Financial Fragility in
Fixed Exchange Rate Regimes." Mimeo. Northwestern University and National
Bureau of Economic Research.

Caballero, R. and A. Krishnamurthy. 2002. "Excessive Dollar Debt: Financial
Development and Underinsurance." Journal of Finance, 58:2, pp. 867-894.

Calvo, G. A. 2001. "Capital Markets and the Exchange Rate: With Special Reference to
the Dollarization Debate in Latin America." Journal of Money, Credit, and
Banking, 33:2, pp. 312-334.

Calvo, G. A. 2002. "On Dollarization." Economics of Transition, 10:2, pp. 393-403.

Catao L. and S. L. Rodriguez. 2000. “Banks and Monetary Shocks in Emerging Markets:
How Far Can We Go With the “Credit View”?” International Monetary Fund
Working Paper 00/68.

Catao, L. and M. Terrones. 2000. "Determinants of Dollarization: The Banking Side."
International Monetary Fund Working Paper 00/ 146.

41

 

 

 

 

Dc

Dc

Edv

 

Céspedes, L. F ., R. Chang, and A. Velasco. 2000. “Balance Sheets and Exchange Rate
Policy.” National Bureau of Economic Research Working Paper 7840.

Chang, R. and A. Velasco. 2001. “Monetary Policy in a Dollarized Economy Where
Balance Sheets Matter.” Journal of Development Economics, 66:2, pp. 445-464.

Chinn, M. D. and K. Kletzer. 2001. “International Capital Inﬂows, Domestic Financial
Intermediation and Financial Crises under Imperfect Information.” in Financial
Crises in Emerging Markets. R. Glick, R. Moreno, and M. M. Spiegel eds.
Cambridge University Press, pp. 196-237.

Devereux, M. B. 2001. “Financial Constraints and Exchange Rate Flexibility in
Emerging Market Economies.” Mimeo. University of British Columbia.

Devereux, M. B. 2002. “Monetary Policy Strategies for Emerging Market Countries: The
Case of the East European Accession Countries.” Mimeo. University of British
Columbia.

Diaz-Gimenez, J., E. C. Prescott, T. Fitzgerald, and F. Alvarez. 1992. “Banking in
Computable General Equilibrium Economics.” Journal of Economic Dynamics
and Control 16:3-4, pp. 533-559.

Edwards, S. and C. Végh. 1997. “Banks and Macroeconomic Disturbances under
Predeterrnined Exchange Rates.” Journal of Monetary Economics, 40:2, pp. 239-
278.

Fama, E. 1985. “What’s Different about Banks?” Journal of Monetary Economics, 15,
pp. 29-40.

Fisher, S. 1983. “A Framework for Monetary and Banking Analysis.” The Economic
Journal, 93, pp. l-l6.

Guidotti, P. E. and C. A. Rodriguez. 1992. “Dollarization in Latin America - Gresham’s
Law in Reverse?” International Monetary Fund Staﬂ Papers, 39:3.

Hausmann, R., U. Panizza, and E. Stein. 2001. "Why Do Countries Float the Way They
Float?" Journal of Development Economics, 66:2, pp. 387-414.

Honohan, P. and A. Shi. 2002. "Deposit Dollarization and the Financial Sector in
Emerging Economies." World Bank Working Paper 2748.

Ize, A. and E. Levy Yeyati. 2003. "Financial Dollarization." Journal of International
Economics, 59:2, pp. 323-347.

Jeanne, O. 1999a. “Foreign Currency Debt and Signaling.” Mimeo. International
Monetary Fund: Washington, D. C.

42

 

 

Jean
Jeanr

Kelol

Krugr

Martir
hIcFini

Mester.
Mester,

Pulley,

Santos,
Schneidt

Uribe, r

 

 

Jeanne, O. 1999b. "Foreign Currency Debt, Moral Hazard, and the Global Financial
Architecture." Mimeo. International Monetary Fund: Washington, DC.

Jeanne, O. 2001. "Why Do Emerging Markets Borrow in Foreign Cmrency?" Mimeo.
International Monetary Fund: Washington, DC.

Keloharju, M. and M. Niskanen. 2001. "Why Do Firms Raise Foreign Currency
Denominated Debt? Evidence from Finland." European Financial Management,
7:4, pp. 481-496.

Krugrnan, P. 1999. "Balance Sheets, the Transfer Problem, and Financial Crises."
International Tax and Public Finance, 6:4, pp. 459-472.

Martinez, L. and A. Werner. 2002. "The Exchange Rate Regime and the Currency
Composition of Foreign Debt: The Mexican Experience." Journal of Development
Economics, 69:2, pp. 315-334.

McKinnon, R. I. and H. Pill. 1999. “Exchange-Rate Regimes for Emerging Markets:
Moral Hazard and International Overborrowing.” Oxford Review of Economic
Policy, 15:3, pp. 19-38.

Mester, L. J. 1994. “How Efﬁcient are Third District Banks?” Federal Reserve Bank of
Philadelphia Business Review, 1994:1, pp. 3-18.

Mester, L. J ., L. I. Nakamura, and M. Renault. 2001. “Checking Accounts and Bank
Monitoring.” Federal Reserve Bank of Philadelphia Working Paper 01-3.

Pulley, L., A. Berger, and D. Humphrey. 1993. “Do Consumers Pay for One-stop
Banking?” Center for Financial Institutions Working Paper 94-01, Wharton
School of the University of Pennsylvania.

Santos, J. A. C. 1998. “Commercial Banks in the Securities Business: A Review.”
Journal of Financial Services Research 14:1, pp. 35-60.

Schneider, M. and A. Tomell. 2001. “Boom-Bust Cycles and the Balance Sheet Effect.”
Mimeo. University of California at Los Angeles.

Uribe, M. 1997. “Hysteresis in a Simple Model of Currency Composition.” Journal of
Monetary Economics, 40:1, pp. 185-202.

43

WWW F.

"W

 

dome

COIISII

(Al)

(Al),
(A2)

and hol

(2.3) an

(2'16) an

 

 

APPENDIX

A. First-Order Conditions and Comparative Static Analysis

Using the perfect capital mobility assumption and Fisher’s equation, the nominal

domestic interest rate, i, , can be expressed in terms of the rate of devaluation, 8, , and the

constant real foreign interest rate, r , (foreign inﬂation was assumed equal to zero)
(A.1) It =r+8,.

The ﬁrst-order condition for household’s optimization problem is given by Eqs.

(Al), (2.4) and (2.9)

 

 

1 I/A -d 6‘ .d't'
, ———=}. 1+ t r+8 —-r + t r—i ,

and holdings of domestic currency and dollar deposits are given by Conditions (2.2) and
(2.3) and Eq. (2.4)

I/A,
t=1 019
+8, +I/A,

 

3t

d* 1+8 = c.
t( ’) 1+8,+1/A,t

 

The ﬁrst-order conditions for frrm’s optimization problem are given by Eqs.
(2.16) and (2.17). Using Eq. (A. 1), they can be expressed in the following way

-1 1-1 [-1 i

11001: 6+0 (7)0—1 :wt[1+itI—(r+8t)]a

Use the labor market equilibrium condition to obtain

44

 

 

(A..

(21.4

and(

(2.23)
(A.5)
(A6)
(A.7)

(AS)

6x
Ogeno

(A9)

 

 

(A3) 1-—
(n, 0 +1)"‘1 =w,[1+z, 8,-r],
J. ,__1. _1_

(A-4) n, “(n, 5 +1)”‘1 =p,[1+i,1* —r].

Holdings of domestic currency and dollar loans are given by Conditions (2.11)
and (2.12), and using the labor market equilibrium condition

Zt =Wt’

up
21(1+8:)=P1"t~
The ﬁrst-order conditions for bank’s optimization problem are given by Eqs.

(2.23) - (2.26). Use then Eq. (Al) to obtain the following conditions

(11.5) (,1 —(r+e.)=nz(d..z.).
(A6) r,” —r=sz-(dftz;),
(A7) (r+8,)—i,d=(r+8,)6,+nd(d,,z,),

A.8 -* * * "
( ) r-ltd =70, +Kt'ld.(dt’zt).

Eqs. (A.2) - (A.8) can be used to express c,, 2,, and z; in terms of the

exogenous variables only

(A9) 1 1 A, A,
——=1+———— r+8 6 + ————c,z +
Ac, I+6,+A{( ”t ”d[1+£,+A t ’1}
—————r6 +K a. ———,z ,
1+£,+A{ ’ Md [1+8,+A, t»

45

-. m
_,,....a; _»_f..- . _

 

 

 

(A.1C

(11.11:

credit,
prefere
regulatr

to Eqs.

ordering

/
"\
1+8! +

(l/At)2

 

 

-\/

(I

1 1 1-0'

 

 

*— — C t _
201+8 01+K + t ,2 =2 "
(A.10) ’ ( ’) [ ”72 (1+8,+1/A, 1)] ‘p’
I/Ac
1+ t’ ,z ,
{ ”z[1+e,+1/A, ’1}
1'0 .11 _(1:3)
(A11) 1+zion+emn 0 =pf’”[1+K.rt*( 0’ 2‘11”".

 

2 1+8, +1/A,’ 1

The effects on consumption c, and the levels of domestic currency and dollar

credit, 2, and z: , respectively, are calculated for shocks to the following: ( 1) the
preference variable A,, (2) the rate of devaluation of the exchange rate 8,, (3) the
regulation variable K, , and (4) terms of trade p, . I apply the implicit function theorem

to Eqs. (A.9) - (All) and use Crarner’s rule to calculate the effects of shocks.

I calculate the determinant of the following system and I denote it by A (the

ordering of the variables is c,, z, , and z: )

 

-—L—[—I-L2(l+8,+I/A,)2+
1+8,+1/A1 1c, ‘1/A1'ldz _(1+gt)1(,,,d.z.
(1/A,)217,,,, +(1+a,)K,qd.d.]
I
I [z’éam ﬁx” 1-0 I L—a l
' I ((1.2‘- — -——-z.0 1+8 0
-o

t
— I-i-K 0 +02 K o o
[/Atztpta 'ldz] ( 1'12 t ('1, z )

I I [-20 _l_'_¢{
_ ‘—

—(1—a)p,a—1 (I—o)p, l[—zr ‘7 “+300

1+8, +l/A, 0 0

_ (I-a)’

(1+Km,- )" ’andy ——

—2
p1 " +(1+K1rt;)" szyl

 

46

 

 

”(1.0

not it

COHSUI.

\
s

.An mam

 

Below I assume that 17,1, and ”d’z' are not too large compared to ”dd: 17,, ,
”d'd' , and ”z’z' .27 That is, there are economies of scope within currencies, but they are
not too large. Then A can be shown to be positive.

The cost function is assumed strictly convex, thus the following conditions hold
2
”dd ”22 ”le > 0 ,

2
”dado 1720an ”drill- > 0 .

A.1. Proof of Proposition 1

Use the following vector, to calculate the effects of a change in A, on

. . . . *
consumptton, c, , domestrc currency credrt, z, , and dollar credrt, 2,

f I
(1+8,+1/A,)

 

2 (1+8.){[(r+8+)5t +’ld 461' —K.n,,+/

1 1-0 [-0
C t— —
(1 tI/A )2 (1+3t)[zt"(1+5r) ‘7 Kt’ld‘z‘ +ZtPta ’ldz]

 

C _ _
’ 2 (1-0)p1” 10+ Km,- )" 219%.;
ﬂ+q+Aﬂ

K /

 

 

 

The effect of an increase in the preference for dollar deposits on real consumption

is

 

27 An example of a cost function that satisﬁes all these assumptions is the translog function.

47

 

 

 

( ., ‘
t I
C A n t. t.
1m A .C

 

2. Ell-u. It...

 

act I (I—U)P:I—l at:
__ = 2 {{—(1+e, )[(r+8, )6, +27, -rc3. -Kmd*l
6A, (1+8,+1/A,) A

1-0 [-20
. 1 ..
(1+8t+1/At)+1/Atct'7dd—Cth'ld#d‘ lipta (1+’lz+zt’722)[;zt 0'

 

_(I-0' 2 [-0 1—0'

p, 0 (1+8,)0 +(1+K,q,.)"'21<,nz,z.]+1/A,c,(1+a,)z,p,0
1-20 1-0 2 {g

I t -
”:[32‘ 0 p. 0 (1+8.)“ +(1+Kt'lz*)a 219212.28]-

 

I-a

- I ‘—
c.(1+sz+)" ’Kfn’, , Ernst», 0 (1w. +Zttlzz)—1/At'ldz’ld*z:
d z

1 1-0 _1_
*__
6'th ‘a'zt 0 (144%)” Prim}-

If the following condition holds

I/A,c,

r+8 6+ +
( ’)’ ”‘1 1+8,+1/A,

 

 

at
>r6 +K .+ ‘ K . .,
”dd t ind 1+£t+1/At (”dd

6c . . . . A
then ——i > 0 . If thrs condrtron does not hold, then consumption decreases.
t

The effect of a change in A, on domestic currency loans is

 

o—l
921—: I (I ”)1” {(1+8,+1/A,)(1+8,)K,217 . .2(1+K,q,.)"‘2
5A1 (1+8,+I/A,)2 A ‘12
.1'_" 1~0 1

I t— T _
CtZtPta ’ldz+(1+8t+1/At)2;zt " (1+3t)”Pta [Kt’ld’z‘[1+(r+3t)5t+
[—20 _ 13,2 I-a

I I- — _
owl-[32+ 0 p1 " (1H,)“ +(1+Kzrtz*)" 2

 

+ I/A,
1+8, +1/A,

 

’Id

[—0

*
K.n,.,*](1+81+1/A1)221(1+8r)p1a mama. +Kr'ld“ +

Kindsd" ]} °

 

Ct
1+8, +1/A,

48

 

 

This et

(A12)

 

COllditic

ACCOFdir

0n Teal Vt

 

A sufﬁcient condition for z, to increase when A, increases is

 

 

 

 

.1 c
z 1+r6 +K . + t K . . <
t”d2[ t Ind 1+8t+1/At _t’7d d ]
st: [/24 C
2 K . — 1+ r+8 6 + + ’ ’ .
r (lid 2 I ( t) t ”d 1+31+1/Ar ’ldd]
The effect on dollar loans is
52: a-2 1 (1'0)Pta_l
—=(1+K,nz*) Kmd‘z’ 2 {[1+(r+8,)6,+
5A1 (1+8, +1/A, ) A
I/A La
11+ ’ 2 Ct’lddIPta (1+rtz+zntzz)(1+6t+1/A1)2-

(I+8,+1/A,)

I-O’

1/A,(1 + s, + 1/A, )p, 0‘ z,c,q,,,2}.
This effect is positive. Thus dollar loans always increase when A, increases.

The effect on credit dollarization can be determined using the following formula

.. [Zr 21 -
6A1 A [2, +z,(1+8,)]2 6A, 5A:

 

 

(A.12)

By substituting for the changes in dollar and domestic currency loans, a sufﬁcient
condition for CD, to increase is
l

p,(I+K,nz.)S(I—o)1“’.

. . . * * .
Imports of rnterrnedrate good 1ncrease as n, = z, (1 + 8, )/ p, and z, Increases.

According to Eq. (2.10), the real wage is equal to domestic currency loans, thus the effect

I o

1__ __

on real wage is ambiguous. Output increases, since y, = (n, a +1)"'1 and n, increases.

49

 

 

COHS

ICSI O

 

 

A.2. Proof of Proposition 2
The description of the effects is similar to the preference shock. The effects on
consumption, domestic currency, and dollar credit are presented ﬁrstly; the impact on the
rest of the variables is discussed afterwards.

I use the following vector, to calculate the effects of a change in 8, on c, , z, , and

( \
1
(1+ +1/A )2 {1/A,(1+8, +1/A,)[ra,’ +K,,,d. —(r+8,)6, -n,,]+
5: t

1/A,(1+8, +1/A, )26, —(1/A, )2 8,17,, —(1+g, )c,K,r;d.d. }

 

1 l [-0
I t— - T—
(1 1/A¢)2 [210(1+3r)UCth’ld‘z‘ _1/Atztpt a Ct'ldz ‘
+8, +

 

1 L10. -
(1+8, +I/A,)le:0(1+£,) 0 (1+K,nz+)]
a

1—0 1'20 —(l—a)2

1 (I—a)p.""[(1+e.+I/A1121217(1+e.)a p, 0 —
0

(1+8, + 1/A,)2
-2
K Ct(1+Ktﬂzt )0 Ktndtzt J

 

 

 

The effect of an increase in the rate of devaluation on real consumption is

50

6c, _

1 (I—U)Pa-l

 

68, _

 

2 ' {[1/A,(1 + e, + 1/A, )[(r + a, )6, + 17,, - r6: —
(1 + a, + I/A, ) A

Kt’ld*]_I/At(1+8t +1/At)251+(1/At)zct’ldd +(1.+31)Cth’ldtd'
_]—_0' 1 [-20 _(1__£)2 1:3
pta (1+'Iz+zt’lzz)[;zt ‘7 P, 0 (1+8,) 0 +(I+Kt’lz*)a—2

 

[-0 1 [—20 1‘0)2 1-0
2 at: ——
sz,2¢]—(1/A,) CtZth a '7‘; ['52: ‘7 P: 0 (1+8,) " +

 

1-0'

(1+K,nz+)U-2K,nz*z+ ] —(1+8,)Kt2n‘2t . p, ‘7 (1+1)z +z,nzz )c,
dz

1 1;:
(1+K.nz+)""’ 41411.1. l(1+., +1/A.1222‘o(1+e.) a 121""); - -
U

z z
[—0 [‘0'

_ I .— _’
Kt+(1+£t)Kt’7 “P11 0(I+’lz+Zt’Izz)(1+£t+I/At)2;zt 0' (1+5t) ‘7
d

1- I
1 *__U g 1—0
—1/A,ndznd:z+c,K,;z, " (“'31) Pt 1-

Consumption increases when 8, increases if the following condition holds

I at-
{[(r+8r)’(1+31+7)]61+’ld}—{r6t +Kt’ld‘}>
t

I t
I

If the above condition does not hold, then consumption decreases.

The effect of a change in 8, on domestic currency loans is

51

 

ofdevr

 

 

62 1 1—0 ”‘1 1 1
——t : ( )pt {[I—2(1+8,+1/A,)2 +(I/A,)2ﬂdd +
C

581 (1+ 8, +1/A, )3 A

 

I :3
0'

1-0 1
+ 2 1 .— -

1—0' 1-2,, _(I-a)2 I—a

_—

_ I * "‘ _
PtI 04,144,4th 'ldz[;2t " Pt 0 (“'51) a +(1+K,nzs)” 2

 

 

1 l
2 21 *r _ 1—
K,nz,z+]—(1+8,+1/A,)2K, 17d? Ez,0(1+8,)”p, “-

l—a 1’0

1 :——— “__ _
Kt’ldz’ldn-z’I/At(1+gt+1/At)2;Zr ‘7 zt(1+8t) 0 Pt] 0}-

This effect is positive, thus domestic currency loans always increase when the rate
of devaluation increases.

The effect on dollar loans is the following

a

(32 1_ 0—1
at = 1 3( a)!" {—[l—Iz—(I-i-E, +1/A,)2 +(1/Ar)2’7dd +
5, (1+8, +1/A, ) A 1 C:

[—0- [-20
I *——— _
(1+8,)K,nd,d‘](1+8,+I/A,)2;z, ‘7 (“'51) a P11 0(1+'lz+zt'lzz)+

 

 

[—0
P1“ (1+rtz +Zrnzz)(1+K,nz+)"‘2K,qd.z. (1+8, +1/A,)2[1+r6: +
1 1—0 [-20
Km."+1/A.6.1+(1/A.)’(1+e.+1/AD’122—z.‘ . 2.04.8.) 0 p1”.
0'

 

1 [—0
I t"' "‘ _
1/A,(1+8, +1/A,)2K,qd,qd,z. -0-z,0(1+8,) a p,’ a}.

The effect is negative; however, the effect on 2: (1 + 8, ) is positive. As both z,

and z:(1 + 8, ) increase, I calculate the effect on CD, using Eq. (A.12). Similar to the

preference shock, a sufﬁcient condition for CD, to increase is

52

 

 

incre

Thus

increz

2,,re

 

 

l
p,(1+K,nz.).<_(1-a)1-a.

Imports of intermediate good increase since n, = 2: (1+8, )/ p, and 2: (1+8, )
increases. According to Eq. (2.10), the real wage is equal to domestic currency loans.

1-1 1.

Thus the real wage increases. Output increases since y, =(n, " +1 )0" and n,

increases.

A.3. Proof of Proposition 3

Use the following vector to calculate the effects of a change in K, on c, , z, , and
'1' .
z! , respectively

{ (1+Et)’7d‘

I l
—z:0(1+8,)0nz.

 

 

—l —2
(-(I —a)p+" (1 + Km; 1" ’7;)
The effect of an increase in restrictions of dollar instruments on real consumption

is

53

68, 1 (I-o)p’0_1 :3
675 = 1+8 +1/A A {_(1+8t)”d*pt 0' (1+1? +Zt”ZZ)
t t t

1'20 _(_I'_0_)2 L19:
[321 0 Pt 0 (1+et)a +(1+Kt’72*)0-2Kt’lz*z‘]+

1—0' 1 [“7

 

 

”Amara/”'1; §2i7(1+81); +(I+Kt’lz*)a—2 112m: "
(1+nz +2,nzz )(1+8,)K,nd*z- }.
Thus consumption decreases, and so do the levels of domestic currency and dollar
deposits. However, deposit dollarization does not change, since it only depends on
households’ preference and rate of devaluation.

The effect of a change in K, on domestic currency credit is

 

 

—1
oz, 1 (1-0)p,“ 1 1 2 2
6K, 1+5, +1/A, A 1 6,2
[—0 1

I t— T 1- —
”dad‘];zt 0' (1+gt)apt 002’ _(I+81)Kt’7d'z’(1+Kt’Iz*)a 2’7 ‘

Z

I :1: — _
wrap,“ 112+11+e.)n,+[;z. «7 p, a 11%.)“ +(1+1<,.,,.)0 2

 

[—0 [—20 _ ﬂ)2 [—0

l— _
0' 10 1+,

1r- 1_
Ktnz'kz‘ II/Alztpt 0' ”d2 +I<tﬂd‘z.’7d‘I E2, 0 (1+Et)a pt 0'}.

This effect is negative, thus domestic currency loans decrease when K, increases. The

effect on dollar loans is

 

52' 1 (1—0) a" 1 1
1 pt 0—2 2 2
—= 1+K I. - + —— 1+8 +I/A +1/A
6K, 1+e.+I/A. A 1 In.) 11,1,c,( . .1 1 .1
1-0' 1-0'
2
’ldd +(1+31)K1’7d+d‘1P10 (1+llz +ztllzz)+(1/Ar)2’lz‘ztpta 'ldz +
1-0

Kt’ld‘z'pta (1+81)’1d‘](1 + ’72 + Zt’lzz ”-

54

The effect is negative. Since both 2, , and 2: increase, I calculate the effect on CD,

using Eq. (A.12). Similar to the two other shocks discussed before, a sufﬁcient condition
for CD, to decrease is
I

p,(1+K,172. )S(1—o)l“’.

Imports of intermediate good decrease as n, = 2: (1+8, )/ p, and 2; decreases.

According to Eq. (2.10), the real wage is equal to domestic currency loans, thus the real

,_1 _2'_

wage decreases. Output decreases, since y, = (n, a + 1)°"1 and n, decreases.

A.4. Proof of Proposition 4

Use the following vector to calculate the effects of a change in p, on c, , 2,, and

 

 

z
I
I 0 ‘
I l
zfa(I+e,)0(1+K,qz.)
—l -1
\(I—o)p,” (I+K,nz.)” —1)
The effect of a shock to terms of trade on real consumption is
_[ _
ac, , (I-a)pf I a ’

{1/A,z,’7(1+s, )0 ”d,(1+K,nz. + 2,“ 19:72.2. )—

 

ap, = 1+8, + I/A, A
L:

(1+8,)K, (1 up, 0 (1+1. +2.12. )111—a)p.”"(1+K.nz+)““’ —11}.
d z

55

A sufﬁcient condition for consumption to decrease is

I
p,,,+K,,,z.)2,,_,)1_a,

The effect of a change in p, on domestic currency loans is

 

 

0—1
:12): =1+5t1+1/At (1’03” {-[§é(1+8,+1/A, )2 +(1/A, 1211.111 +(1+e, )K,
1-0 1 ﬁg 1
11d.,,*12:7(1+81)0(1+1<1't,'+2: K.n,-,+)+(1+e.) a Kr’ndyzih
1_—_g
(1+8,)K, ”.2: p, ‘7 (1/A1)21’ldz[(1‘0)Pta—l(1+ Iona)“ —11}.

This effect is negative, thus domestic currency loans always decrease when terms of trade

increase.

The effect on dollar loans is

*

 

t t 2 2 .
= —— 1+8 +1/A + I/A + 1+3 K ,
1:3 1—0'
p,0 (1+1. +2.1121111—01p.”"(1+KID)“ 4141011122110. " (182111—01

I i
Pia-10+ Kt’lz" )a—l - 1] "(I/At Mitzi." (1 + 51)" Pia-thlld‘; 0‘07}-
If the following condition holds

1
Ptr(1+Kt’lz‘)5(-’—¢7)1—‘I ,

then 2, increases. Thrs 1s also a sufﬁcrent condrtron for CD, to 1ncrease, srnce 2, drops

in all cases. If the above condition does not hold, 2: decreases, and the effect on CD, is

ambiguous,

56

The real wage decreases unambiguously. Output and imports decrease if
2: decreases, that is, if the above condition does not hold. If the condition holds, the

effect on output and imports of intermediate goods is ambiguous.

57

Figure 1 Credit and Deposit Dollarization in Emerging Market and Transition Economies

(Credit Dollarization denoted by continuous line; Deposit Dollarization denoted by dotted
line. Data are from Arteta (2002) and (2003)).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

' Chile 1 Paraguay
100 -- -—--——- +— ~—~ - i 100
I 90- I 90+
80 - so +
70+ 701
I 60+ 90+
. 50+ 50+
40 I 40 ~
30+ 30+
rl 20* 20+
I 10 1 , 1o .
i 0 I I 7—T T T T I—T V ﬁT I l I if T I I I I I 0
Q 19781980198219941988198819901992199419981998 ( ,
____ L L
l—T“_—T'__#—T_—_— —TT F
I Peru
‘ 100 -_..rg _.._. .. 1 100 ,‘
90 I 901 F
80 + g 80 1 1
70+ 5 70+
60 I 80+
50 * g 50
40+ 1 40+
30 1 1 30+
20« \~, 20+
10+ I toj
O rrrrr . rrrrrrrrrrr 1 rrrrrrr ‘ O '—_r r r r 7 r r 7 T r r
l 1975 1978 1981 1984 1987 1990 1993 1996 1999! 1 198919901991 19921993 199419951996199719981999
r 1 A
I Lithuania I Russia
*' l
i 100 - —-- ~--—- ~— . 100
90 . m -
. 80+ 80+
70+ ' 70+
60- 60+ ,
50 + 50 + .
40+ 40+
30+ 30 ;
20 20 J I
10- 10 3
‘ o + . . . . 1 0 . . . , . . + A
I 1993 1994 1995 1996 1997 1998 1999 1993 1994 1995 1996 1997 1998 1999 2000
_ , L

 

 

 

 

Credit Dollarization (%): Dollar Credit/Total Credit (to the resident private sector issued
by resident banks)
Deposit Dollarization (%): Dollar Deposits/Total Deposits (of residents held in resident

banks)

58

CHAPTER 2
CREDIT DOLLARIZATION IN TRANSITION ECONOMIES:

IS IT FIRMS’ OR BANKS’ “FAULT”?

1. Introduction

Dollarization of bank deposits and loans is widespread in emerging market and
transition economies. Since levels of foreign currency deposits and credit vary widely
across countries and over time, the causes and consequences of ﬁnancial dollarization
deserve thorough analysis.1

Much of the current literature analyzes the risks imposed by ﬁnancial
dollarization and the policy options in highly-dollarized economies. It shows that
ﬁnancial dollarization complicates monetary policy and limits available policy
instruments. However, only a few studies evaluate the determinants of dollarization. In
particular, deposit dollarization has been theoretically and empirically studied in the
context of currency substitution. Credit dollarization and the respective contributions of
banks and ﬁrms to the dollarization process have received less attention. Is credit

dollarization the outcome of a unilateral decision by domestic banks? Or is it driven by

 

' Financial dollarization occurs when a signiﬁcant share of residents’ domestic ﬁnancial contracts
(including deposits and loans) are denominated in foreign currencies, not necessarily the US dollar. Dollar
deposits and loans denote all foreign currency deposits and loans. In addition, deposit dollarization and
credit dollarization refer to the phenomenon of denominating deposits and loans in foreign currencies.

59

domestic ﬁrms’ integration in the international goods market? This chapter attempts to
shed light on these issues.

Currently, only two empirical studies, Arteta (2002) and Barajas and Morales
(2003), analyze the determinants of credit dollarization. Arteta (2002) evaluates the
effect of the exchange rate regime on credit dollarization (and more generally on bank
currency matching of deposits and loans) for a sample of developing and transition
economies. He ﬁnds that credit dollarization, unlike deposit dollarization, does not vary
across exchange rate regimes.2 Thus, volatility in the exchange rate cannot fully explain
differences in dollarization patterns. Barajas and Morales (2003) use data on bank- and
ﬁrm-speciﬁc factors, in addition to macroeconomic variables, for fourteen Latin
American economies to explore possible biases introduced by domestic monetary policy
into the decisions by banks and ﬁrms to lend and borrow in dollars. Their results indicate
that both bank asset and ﬁrm liability allocation decisions are important determinants of
dollarization. They also ﬁnd limited evidence that the greater the volatility in the
exchange rate, the lower the level of credit dollarization, at least in the short run.

This chapter extends the existing work on credit dollarization, ﬁrst, by using a
comprehensive, newly-assembled dataset on transition economies, and, second, by
estimating and comparing the respective contribution of ﬁrms and banks to the ﬁnancial
dollarization phenomenon.

This work draws on two other literatures in economics and ﬁnancial management.

The ﬁrst literature concerns the determinants of currency composition of corporate debt,

 

2 He ﬁnds some limited evidence that credit dollarization decreases with a more ﬂexible exchange rate
regime. Overall, his paper provides evidence that a more ﬂexible exchange rate regime tends to amplify
bank currency mismatches as credit dollarization does not change (or it decreases), while deposit
dollarization increases.

60

regardless of the source (domestic or foreign investors). It highlights ﬁtms’ hedging
behavior as a potential determinant of credit dollarization. A portfolio of loans
denominated in more than one currency can provide an improved risk and cost position
(see Cotner (1991)). In addition, empirical papers provide evidence that ﬁrms borrow in
dollars for hedging reasons. Larger ﬁrms, and ﬁrms with exporting activities, borrow
more in foreign currencies (see Aguiar (2002), and Martinez and Werner (2002), for
evidence for ﬁrms from developing economies, Kelohatju and Niskanen (2001), Kedia
and Mozumdar (2002), and Giraldi and Hamaui (1991), for evidence for ﬁrms from
developed economies).

The second literature studies the determinants of deposit dollarization.
Macroeconomic factors, such as uncertainty and non-credible domestic policies, as well
as persistence effects,3 are found to play an important role for deposit dollarization (see
Honohan and Shi (2002), and Ize and Levy Yeyati (2003)). In addition, safety nets for
banks, such as insurance schemes and the presence of a lender of last resort, are also
believed to encourage deposit dollarization (see Broda and Levy Yeyati (2003)). I
control for these macroeconomic factors as they will most likely affect credit
dollarization.

This chapter analyzes the respective contributions of banks and ﬁrms to the
phenomenon of dollarization of credit in transition economies. Why do domestic banks
in these countries lend in foreign currencies? Why do domestic ﬁrms borrow

domestically in foreign currencies? Is credit dollarization mainly supply or demand-

 

3 Historically high levels of dollar deposits might persist over time, even after the initial causes of
dollarization (which are believed to be macroeconomic uncertainty and non-credible domestic policies) no
longer exist.

61

driven? Ultimately, I would like to contribute to a better assessment of the risks
associated with ﬁnancial dollarization, and suggest what policy measures (if any) should
be implemented to control or restrict it.

I use a newly-constructed dataset for twenty-two transition economies from
Central and Eastern Europe and Central Asia, for the period 1990 - 2001, and group
potential determinants into bank- and ﬁrm-speciﬁc (ﬁnancial and mainly real) factors.
Bank-speciﬁc factors include indicators of asset and liability management and currency
matching, proﬁtability, concentration, and risk management. As ﬁrm-speciﬁc factors, I
use measures of real dollarization4 and access to alternative ﬁnancing sources. In
addition, I include speciﬁc indicators of overall hedging opportunities, liberalization, and
deregulation of the foreign exchange market, uncertainty and lack of credibility of
domestic policies, and persistence effects, as well as measures of overall ﬁnancial and
economic development.

I estimate a reduced form equation, and use the pooled OLS estimator as a
benchmark. I then extend the analysis and use the ﬁrst difference and ﬁxed effects
estimators to control for unobserved heterogeneity. Main results seem to be robust across
speciﬁcations.

Empirical results provide evidence that credit dollarization in transition
economies is determined by banks’ optimization decisions (asset and liability

management variables). Banks match the currency of denomination of their assets and

 

‘ I deﬁne real dollarization as the denomination of the returns and costs of production for ﬁrms in foreign
currencies. That is, ﬁrms export the ﬁnal good and/or use imported intermediate goods in the production of
the ﬁnal good.

62

liabilities, that is, deposit dollarization drives credit dollarization, and net foreign assets
represent a substitute for foreign currency loans to domestic ﬁrms.

In addition, there is limited evidence of ﬁrms hedging the currency of their returns
of production and debt. A higher ratio of exports to GDP increases credit dollarization.
More generally, the more integrated is the economy in the international goods market
(higher trade to GDP ratio), the higher is credit dollarization. This result suggests that
integration of international goods and ﬁnancial markets, real and ﬁnancial dollarization
respectively, move in the same direction. It questions the previous (counterintuitive)
ﬁnding (see Arteta (2002) and Barajas and Morales (2003)) that the openness of the
economy has a negative effect on credit dollarization.

This chapter provides evidence that lack of credibility of monetary policy and
macroeconomic uncertainty (measured by the central bank’s net foreign reserves) raise
dollarization levels. In addition, there is some evidence of persistence effects in credit
dollarization. That is, historically high values of the devaluation/depreciation rate seem
to lead to higher credit dollarization over time.

To summarize, banks in transition economies hedge against exchange rate risk by
holding matched foreign exchange positions. However, as Ize and Parrado (2002) point
out, as long as ﬁnancial dollarization outpaces real dollarization, and this seems to be the
case in transition countries, there is a currency mismatch somewhere in the economy. I
show that banks pass the exchange rate risk to ﬁrms. This decreases banks’ exposure to
currency risk, but it increases their exposure to default risk, and ultimately the exposure
of the economy to ﬁnancial and currency crises Krugrnan (1999) is the ﬁrst author to

focus on ﬁrm currency mismatches, and their role as a main ingredient in twin crises).

63

My results have several policy implications. Since macroeconomic credibility
and past and current uncertainty have a positive effect on credit dollarization, reducing
them would also reduce dollarization. Restrictions on dollar lending by domestic banks
to residents, however, might lead to the “export” of deposits and ﬁnancial
disinterrnediation as banks want to have matched asset and liability positions, and could
substitute dollar loans with foreign assets instead of domestic currency loans.

The remainder of the chapter is organized as follows. Section 2 reviews the
theoretical background and presents the main hypotheses for the bank- and ﬁrm-related
effects on credit dollarization. Section 3 outlines the econometric model and the
estimation methods. Section 4 describes the data, while section 5 presents the empirical
results. I perform several robustness tests in section 6. Section 7 discusses policy

implications and directions for future work.

2. Conceptual Framework and Main Hypotheses

To assess the extent to which credit dollarization is mainly a ﬁnancial (bank-
determined) or a real (ﬁrm-determined) phenomenon, I separate potential determinants
into 1) factors that affect only the supply of foreign currency credit (relative to domestic
currency credit), 2) factors that affect only the demand for foreign currency credit
(relative to domestic currency credit), and 3) factors that affect both demand and supply

of credit, as well as additional controls.

64

2.1. Bank-speciﬁc Factors
2.1.1. Asset and Liability Management

I expect banks to match their foreign currency pesitions, either because they do
not want to be exposed to the exchange rate risk, or because they are required to do so
(open foreign exchange position limits are enforced), or both. Calvo (2001) and (2002)
proposes bank currency matching as a possible explanation for the high levels of dollar
loans in emerging economies, given that deposits are highly dollarized. Ize and Levy
Yeyati (2003) also argue that banks hold foreign exchange matched positions, not only
because of regulation requirements, but also because they are “generally adverse to
assuming risk directly” (pg. 8).

An increase in deposit dollarization should increase credit dollarization as banks
match the currency of denomination of their deposits and loans. If instead banks match
by currency the level of overall assets and liabilities, which seems to often be the case in
transition economies,5 then an increase in foreign assets and a decrease in foreign
liabilities6 will both decrease credit dollarization. More generally, an increase in net
foreign assets is expected to have a negative effect on credit dollarization as banks
substitute foreign currency loans to domestic ﬁrms with foreign assets, for a given level

of dollar deposits.

 

5 For several transition economies, there are limits imposed on the open foreign exchange positions, which
refer to banks’ balance-sheet foreign currency assets and liabilities in the corresponding foreign currencies
and off-balance-sheet commitments related to foreign exchange buying and selling transactions.

6 I can safely assume that foreign assets and liabilities are denominated entirely in foreign currencies.
There is a trend in the economies literature that tries to explain the inability of emerging economies to
borrow externally in their own currency, the so-called “original sin” (see Hausmann et a1. (2001), who has
identiﬁed this problem).

65

2.1.2. Proﬁtability and Risk-taking Behavior

There is little guidance from previous theoretical and empirical work as to what to
expect from the direction of effects of proﬁtability, risk-taking behavior, and
concentration on the currency of denomination of bank credit. Hence, my main goal is to
determine if there is any evidence that these bank and ﬁrm characteristics are
signiﬁcantly related to credit dollarization.

I implicitly assume that all domestic agents, banks and ﬁrms, expect domestic
currency to depreciate in the future. Hence, unless a ﬁrm’s return is perfectly indexed to
the exchange rate, the risk of default on foreign currency credit is higher than on
domestic currency credit. For banks, foreign currency credit is riskier than domestic
currency credit; the return is higher, provided that domestic currency indeed depreciates
before the credit is repaid; however, the default risk is also higher.7

I hypothesize that less proﬁtable banks are willing to incur higher risks. Hence,
they extend more dollar loans. Also, “riskier” banks, banks with higher asset riskiness,
will lend more in dollars.8

I also expect deposit insurance to have a positive effect on credit dollarization
when dollar deposits are covered by the insurance scheme,9 as it provides bailout
guarantees for banks. According to the moral hazard hypothesis, banks are more likely to

engage in excessive risk-taking (see Demirglic-Kunt and Huizinga (2003) and Broda and

 

7 Using a simple theoretical model, Arteta (2002) shows that the default risk for the dollar credit increases
with the currency risk.

8 I do not consider the effect of dollar lending on the proﬁtability and riskiness of banks. Since I am
interested in banks’ characteristics prior to lending, I use lagged measures in my estimations.

9 Data on deposit insurance in transition economies show that for all countries but the Czech Republic,
prior to 2001, the insurance scheme also covers the dollar deposits.

66

Levy Yeyati (2003)). The currency risk is shifted to the borrowers, and this ultimately

increases the default risk.

2.1.3. Concentration

The effect of bank concentration on credit dollarization is not straightforward.
Drawing on previous work, I can identify two alternative explanations, and ultimately the
expected effect is ambiguous.

Catao and Terrones (2000) theoretically show that the higher the concentration in
the banking industry, the lower the equilibrium dollarization levels. The higher banks’
market power, the more they will exercise it in lending. Naturally, banks’ market power
is higher for borrowers that do not have access to direct borrowing from abroad (denoted
below as “external borrowing”). These are borrowers in the non-traded goods sector, or
borrowers that do not have traded collateral. Thus, they demand domestic currency
credit. Therefore, concentration tends to increase lending in domestic currency and
reduce lending in dollars.

An alternative explanation is that high concentration of banks will lead to more
credit dollarization due to implicit bailout guarantees. The more concentrated the
banking sector is, the higher is the probability that a bank will be bailed out should a bank
run happen (as the risk of a systemic ﬁnancial collapse is higher). This encourages
excessive risk-taking by banks,10 and if indeed “riskier” banks lend more in dollars, it

should lead to higher dollarization in equilibrium.

 

'0 The argument is similar to the case of deposit insurance, whose simple existence might generate moral
hazard.

67

2.2. Firm-specific Factors

2.2.1. Liability Management

F irms that have access to alternative sources of external ﬁnance are expected to
rely less on credit from domestic banks. Hence, I expect ﬁrms’ credit from abroad to
have a negative effect on credit dollarization, since I believe that external loans are a
substitute for foreign currency loans from domestic banks.

On the other hand, it is possible that the access to external borrowing is
conditional on domestic borrowing in dollars. Foreign lenders (who presumably have
less information on domestic ﬁrms) might extend loans only to those that have already
received a dollar loan from domestic banks (which presumably have more information).
If this is the case, then external loans and foreign currency loans from domestic banks are
complements, and more credit dollarization will lead to more external credit. 11

, Barajas and Morales (2003) show that for the case of ﬁrms in Latin American

economies, external credit is a substitute for the domestic foreign currency credit.

2.2.2. Hedging Behavior

Conventional wisdom suggests that ﬁrms tend to match the currency composition
of their debt with that of costs and revenues. If this is the case, ﬁnancial dollarization is a

natural consequence of real dollarization.

 

” In this case, the level of credit dollarization would determine the level of external credit, and the external
credit would be endogenous in the speciﬁcation of credit dollarization. Aguiar (2002) shows a positive
relation between domestic dollar credit and external equity. According to his paper, ﬁrms that issue equity
abroad are also likely to borrow in dollars domestically. Testing the reverse causality is beyond the scope
of this chapter.

68

First, ﬁrms might use imported intermediate goods in the production process, and
thus want to hedge against the cost of interruption in the production process should the
exchange rate change from the moment the loan is contracted and until it is used to pay
for imports. If this is the case and foreign currency loans are indeed used to pay for
imported goods, then I should see a positive effect of intermediate good imports on credit
dollarization. Hence, the more dependent is the economy on foreign intermediate goods,
the higher is the credit dollarization.12

Second, ﬁrms with exporting activities, and thus returns denominated in foreign
currencies, are hedged against currency risk if they also borrow in foreign currency. The
higher are the exports, relative to domestic production, the more dollarized the economy
should be. Studies that use ﬁrm level data ﬁnd evidence that the productive structure
matters for the corporate debt currency composition.l3

Regardless of what ﬁrms hedge against, production interruptions or currency risk,
the trade openness of the economy, broadly deﬁned as total trade per GDP, should have a
positive effect on credit dollarization (see also Ize and lParrado (2002)). Therefore, I

14 This is the opposite effect

expect real dollarization to cause ﬁnancial dollarization.
from what previous empirical work using aggregate data ﬁnds. Both Arteta (2002) and

Barajas and Morales (2003) ﬁnd that openness has a negative impact on dollarization.

 

'2 Delgado et al. (2002) also mention the import ﬁnancing as a possible explanation for domestic lending in
foreign currency.

‘3 Aguiar (2002) and Martinez and Werner (2002) use data on Mexican ﬁrms, and show that exporting
ﬁrms borrow more in foreign currencies. Kedia and Mozumdar (2002) ﬁnd that US ﬁrms performing
foreign operations borrow more in foreign currencies. Keloharju and Niskanen (1997) also ﬁnd evidence
for the hedging explanation using data on Finnish ﬁrms. Giraldi and Hamaui (1991) show that the
currencies used to settle international trade determine the currency composition of debt for Italian ﬁrms.

1’ Ize and Levy Yeyati (2003) reach a similar conclusion using a MVP (minimum variance portfolio)
model. Real dollarization is however measured by the pass-through coefﬁcient of exchange rate changes
on prices.

69

Their result is counterintuitive (and contrary to what studies using ﬁrm level data ﬁnd),
and suggests either an omitted variable or an aggregation bias.

In addition to the share of imports, exports, and total trade, in GDP, I consider
changes in international prices, and their effects on dollar credit. I expect an increase in
import prices to increase ﬁrms’ borrowing in dollars, if ﬁrms use dollar credit for import
ﬁnancing. An increase in export prices should also increase credit dollarization, if
exporting ﬁrms ﬁnance (an increase in) production using dollar loans.

Another ﬁrm characteristic that should matter for the currency composition of
ﬁrm liabilities is ﬁrm size. Previous work ﬁnds evidence that large ﬁrms borrow more in
dollars, since they can better diversify and manage risks than small ﬁrms.15 Thus I
expect the average size of ﬁrms in the economy to have a positive effect on credit

dollarization.

2.2.3. Proﬁtability and Risk-taking Behavior

The effects of ﬁrm characteristics, other than size, on equilibrium dollarization
levels are not straightforward. I hypothesize that less proﬁtable ﬁrms and riskier ﬁrms
will choose to borrow in foreign currencies, everything else equal. I believe that dollar
loans are associated with a higher default risk. Therefore, ﬁrm proﬁtability should have a

negative effect and ﬁrm risk-taking behavior a positive effect on dollar credit.16

 

‘5 See Aguiar (2002), Martinez and Werner (2002), and Keloharju and Niskanen (1997). Martinez and
Werner (2002) show that ﬁrm size matters more under ﬁxed than under ﬂoating exchange rate regimes.
When switching to a ﬂoating exchange rate regime, a ﬁrm’s exporting activities become the most important
determinant. Alternatively, Aguiar (2002) hypothesizes that ﬁrm size matters for debt dollarization due to
a ﬁxed cost to borrowing in dollars, which makes it worthwhile only for large ﬁrms.

'6 Similarly to the bank case, it is difﬁcult to say a priori what the effect should be, due to a lack of
theoretical and empirical work on the link between proﬁtability and risk-taking behavior, on one hand, and
dollar borrowing, on the other hand. Jeanne (1999) evaluates proﬁtability in connection to dollar debt
However, he considers the reverse effect, from currency composition to proﬁts. He models dollar debt as a

70

2.3. Macroeconomic Factors and Other Controls

2.3.1. Overall Hedging Opportunities

In countries where the currency forward market exists and is functional, economic
agents can use it to hedge against currency risk. Therefore, I expect to see more
borrowing and lending in foreign currencies, given that agents do not have to worry about
unmatched foreign exchange positions.17

The exchange rate regime can play a similar role. The majority economists’ view
(see, for example, Burnside et a1. (2000), Mishkin (1996), and Obstfeld (1998)) is that a
ﬁxed exchange regime offers implicit guarantees (implicit exchange rate insurance) and
eliminates the need to hedge against the currency risk, provided that it is credible. This
implies that over time a more ﬂexible exchange rate should reduce credit dollarization.l8
It also suggests that across countries the levels of dollarization should be higher in

countries with ﬁxed exchange rate regimes, when controlling for any credibility issues.

2.3.2. Restrictions on the Foreign Excha_nge Operations

Regulations of the foreign exchange market should play an important role in

explaining the levels and patterns of dollarization in emerging market and transition

 

commitment device. According to his model, ﬁrms that borrow in dollars put in higher effort and end up
being more proﬁtable. Aguiar (2002), however, shows that in the wake of devaluation, the presence of
dollar debt reduces ﬁrms’ net wealth. Neither of these studies indicates if a priori more proﬁtable ﬁrms
will borrow in dollars or domestic currency, everything else equal.

'7 Burnside et al. (2000) argue that the forward market can be used to alleviate the bank currency
mismatches, and banks can borrow in dollars, and lend in domestic currency. If this is the case, the effect
on credit dollarization should be negative. A similar argument applies to exporting/importing ﬁrms: they
can borrow in domestic currencies, and hedge against the currency risk using derivatives.

'8 Martinez and Werner (2002) ﬁnd empirical support for this hypothesis, using micro data on Mexican
ﬁrms. Arteta (2002), on the other hand, ﬁnds evidence that credit dollarization does not vary with the
exchange rate regime (there is some limited evidence that it decreases for more ﬂexible regimes, but this

71

economies. The more restrictions placed on lending domestically in dollars, the lower is

9 More generally, I hypothesize that the overall

the level of credit dollarization.1
liberalization of trade and foreign exchange operations should have a positive effect on

dollarization.

2.3.3. Level of Financial Development

Previous theoretical work suggests that the level of development of the ﬁnancial
sector has a negative effect on credit dollarization. Caballero and Krishnamurthy (2002)
show that the less developed the domestic ﬁnancial market is, the less domestic economic
agents (with returns in domestic currency) value the insurance against currency risk
offered by domestic currency credit, and the more likely they are to borrow in dollars.
Another way to think about this effect is that more developed domestic banks internalize
the currency risk better and prefer to lend to ﬁrms in domestic currency. In the above
interpretations, ﬁnancial development should reduce the dollarization of credit,

everything else equal.

 

result is not robust across speciﬁcations). Furthermore, he shows that currency mismatches in the banking
sector increase with increased ﬂexibility of the exchange rate regime.

'9 Delgado et al. (2002) describe the patterns for restrictions on foreign currency lending in emerging
economies. The highly-dollarized economies tend to have no restrictions, while the prohibition of dollar
lending is very rare. The requirement that dollar lending has to be approved by ofﬁcials is most commonly
used as a controlling mechanism. Alternative restrictions involve access to dollar credit only by ﬁrms with
dollar returns and/or dollar payments.

72

2.3.4. Level of Economic Development

I expect overall economic development to reduce dollarization. Thus, GDP per
capita should have a negative effect on the dollarization of credit. The argument is

similar to that for overall ﬁnancial development.

2.3.5. Overall Uncertairnty, Lag}; of Credibility of Domestic Policies. a_nd

Persistence Effects

Higher uncertainty and lack of credibility of domestic policies should lead to
higher dollarization levels, for both deposits and loans. Jeanne (2001) argues that the
uncertainty with respect to domestic monetary policy increases the cost of
borrowing/lending in domestic currency (the so called “peso effect”) and the level of
borrowing/lending in dollars. Similarly, Delgado et al. (2002) argue that uncertainty
about potentially high future depreciation associated with a high real interest rate for
domestic currency lending make foreign currency lending more attractive, regardless of
the risks involved.

In general, higher inﬂation/devaluation might reduce the credibility of the
domestic currency, and thus cause a switch from domestic currency deposits and loans to
those denominated in foreign currency.20

An alternative measure of uncertainty and lack of credibility is the level of net

foreign reserves held by the monetary authority. A higher level of reserves might signal

 

20 Some authors (see Ize and Levy Yeyati (2003) and Ize and Parrado (2002)) argue that it is not the ﬁrst
moment, but the volatility of inﬂation and the real exchange rate that affect dollarization. I leave the
estimation of volatility effects on credit dollarization for future work.

73

higher credibility in domestic currency, and higher sustainability of the monetary regime.
This should have a negative effect on dollarization (see Ize and Parrado (2002)).
Historically high values of inﬂation and devaluation rates should positively affect
dollarization if there are any persistence/hysterises effects in the dollarization variables.
Higher past uncertainty with respect to the price and exchange rate level might have an
impact in the present, if economic agents get used to borrowing in foreign currencies
even after the uncertainty with respect to the domestic currency was curbed, or if the

stabilization is not credible.

3. Econometric Model and Estimation Methods

I combine the (relative) demand and supply equations, and estimate the following
reduced-form equation21
Credit_Dollarizati0n,~,= a'Bank_ Variables ;,+ ﬂ'Firm_ Variables,-,'+ y. Controlsu+ £,-,.

Credit_Dollarization denotes the level of dollar credit (from domestic banks to
domestic ﬁrms) in the economy, calculated as the ratio of foreign currency loans to total
loans, or using a broader measure, as the ratio of foreign currency loans to total bank
assets. Bank_ Variables stands for the set of variables related to the banking sector,
described further in Section 4.2, which identify the supply-side of credit dollarization.

F irm_ Variables denotes the set of variables related to the real sector, described in Section

4.3, which measure the demand-side of credit dollarization. Controls measures all other

 

2' The cost of foreign currency (relative to domestic currency) credit is eliminated from the reduced-form
equation. Data on foreign currency interest rates are very scarce (for some countries, they are entirely
missing). Furthermore, I estimate the effects of bank- and ﬁrm-speciﬁc variables on equilibrium credit
dollarization, rather than on supply and demand separately.

74

factors that affect both demand and supply of dollar credit, or which capture country-
speciﬁc effects. Finally, 8 is a disturbance term.

I use pooled OLS as the benchmark. I then extend the analysis and use the ﬁrst
difference and ﬁxed effects estimators22 to control for missing country-speciﬁc effects
and check the robustness of my results to alternative speciﬁcations. I use alternative
measures for dollarization and other variables to also check the robustness of results to
alternative measures employed.

There are several methodological issues that need to be considered when
estimating the above equation: 1) the identiﬁcation problem; 2) persistence of credit
dollarization; and 3) endogeneity.

In order to be able to assess the contribution of demand and supply to the
dollarization of credit, I need to have (at least) one demand-speciﬁc measure that does
not affect the supply of foreign currency credit, and (at least) one supply-speciﬁc measure
that does not affect the demand for foreign currency credit. Out of all ﬁrm variables, the
measures of integration in the international goods market, as well as the changes in the
international commodity prices, should have no effect on banks’ supply of credit. On the
supply side, bank asset and liability management variables (measures of currency
matching) should be exogenous to ﬁrms’ demand for credit. Therefore, I believe my
equation is well identiﬁed, and I can distinguish demand and supply effects on

dollarization.

 

22 The ﬁxed effects estimator works better than the random effects estimator when using aggregate data
since the independent variables are likely to be correlated with the errors (the missing country-speciﬁc
effect).

75

The second issue is that there might be persistence in the measures of credit
dollarization, since I use measures of stock (and not ﬂow) of loans. However, bank loans
in transition economies have a shorter maturity than in most developed economies.23 I
start with the pooled OLS, which I use as benchmark. In addition, I use ﬁrst difference
and ﬁxed effects estimators to control for unobserved heterogeneity, and compare the
results obtained across speciﬁcations. In all speciﬁcations, the standard errors are
adjusted for heteroskedasticity and serial correlation to account for the persistence
problem.

A ﬁnal issue is the problem of endogeneity. Mainly, the measures of asset and
liability management for banks and liability management for ﬁrms might be endogenous.
While I can safely assume that deposit dollarization is supply-driven and hence
exogenous,24 the measure of banks’ net foreign assets is probably endogenous.
Furthermore, ﬁrms’ external borrowing is probably endogenous. I control for these

problems by using Instrumental Variable speciﬁcations, where lagged values of banks’

net foreign assets and ﬁrms’ external borrowing are used as IVs.

 

23 For example, the average credit period for bank loans (denominated in either domestic or foreign
currency) for a sample of large ﬁrms for the period 1990 - 2001 is 70 days. The median credit period is
even lower: 49.5 days.

2’ Even deposit dollarization might be endogenous, if ﬁrms are required to hold correspondent deposits
(they have to hold foreign currency deposits with domestic banks, in order to obtain foreign currency
credit). I show how the results change if I use households’ deposit dollarization instead of, or as an IV for,
overall deposit dollarization.

76

4. Data

4.1. Credit Dollarization Data

I have constructed a new unbalanced panel dataset for twenty-two transition
economies from Central and Eastern Europe and Central Asia, covering the period 1990 -
2001.25 Data on credit dollarization are annual, and were collected from various central
bank publications. For some countries, data starts in early 90s. However, for most
countries, data covers the second half of the 905 (see Table 1).

Two measures of credit dollarization are employed. The ﬁrst one is the ratio of
foreign currency credit to total credit, from domestic banks to domestic enterprises. The
second measure is the ratio of foreign currency credit to enterprises to total bank assets
for domestic banks. The ﬁrst measure captures the substitution between foreign currency
and domestic currency credit, while the second measure captures the substitution of
foreign currency credit with any other bank assets, either demestic or foreign,

denominated in either domestic or foreign currency.

4.2. Bank-specific Variables

I use the following data on the domestic banking sector in transition economies to
measure factors that affect the equilibrium level of dollarization through the relative
supply of foreign currency credit (a detailed description of the variables and data sources

are provided in the Data Appendix).

 

25 Currently, despite the recognized importance of credit dollarization, these data are not systematically
collected. Only data on deposit dollarization are included in the IMF International Financial Statistics.
Studies that use credit dollarization data are Arteta (2002) and Barajas and Morales (2003). Arteta has
credit data for forty developing and transition economies (for some transition economies, he has more data,
for others, I do), while Barajas and Morales (2003) have data for fourteen Latin American economies.

77

0 Deposit dollarization ratio. The ratio of foreign currency deposits to total deposits,
held by residents at domestic banks. Alternative measures employed are the ratio of
foreign currency deposits to total bank assets, and? the ratio of foreign currency
deposits to total bank credit.

0 Banks" net foreign assets. I use three alternative measures: the ratio of net foreign
assets to total bank deposits, the ratio of net foreign assets to total bank assets, and the
ratio of net foreign assets to total bank credit.

0 Interest rate spread. The spread between the average lending interest rate and
average deposit interest rate (over all currencies and maturities).

0 Banks ’ proﬁts. The ratio of banks’ pre-tax proﬁts to the total bank assets.

0 Banks’ risk-taking behavior. The ratio of banks’ total risk assets to total assets.26

0 Deposit insurance. I deﬁne a dummy variable that takes the value 1 if there is
explicit deposit insurance, zero if there is not, and use it as a measure of risk-taking
behavior of domestic banks.

0 Banks ’ concentration. The total number of banks per million people. As an
alternative measure, I use the number of foreign banks operating domestically per

million people.

4.3. Firm-speciﬁc Variables

The following variables measure ﬁrm-specific effects, through the demand for

credit, on equilibrium credit dollarization:

 

26 See the Data Appendix for the description of how banks’ total risk assets are measured.

78

Firms’ external borrowing. I use two alternative measures: the ratio of foreign loans
to non-bank private enterprises to total domestic credit, and the ratio of foreign loans
to non-bank private enterprises to GDP.

Trade. The ratio of trade (total imports plus total exports) to GDP.

Imports of intermediate goods. The ratio of intermediate goods imports to GDP. The
ratio of raw material imports to GDP and other components of imports are used as
alternative measures.

Exports. The ratio of total exports to GDP.

Index of total commodities. The index of fuel and non—fuel commodity prices,
aggregated over all countries.

Index of industrial inputs. The changes in prices of industrial inputs aggregated over
all countries.

Change in the oil price. The change in the average world oil price.

Average ﬁrm size. The average size of large ﬁrms”, calculated as the logarithm of
average total assets per company. An alternative measure I employ is the average
number of employees per company.

Firms’ proﬁt margin. The average proﬁt margin for large ﬁrms, calculated as the
ratio of net proﬁt after taxes to total sales for a given 12-month period.

Firms ' leverage ratio. The average leverage ratio for large ﬁrms, calculated as long-

tenn funds with ﬁxed interest to total capital.

 

27 All data on ﬁrms’ size, proﬁt, and leverage ratios are aggregated from Amadeus (a ﬁnancial database of
large companies in transition economies) and refer to (up) to ﬁfty largest companies in the economy (by
total assets). For some countries, data on less than ﬁfty ﬁrms were available.

79

4.4. Other Variables

Forward market. A dummy variable that takes the value 1 if a forward market for
foreign exchange exists and is functional, zero it does not exist or is underdeveloped.
Exchange rate regime: Dummy variables differentiating between ﬁxed, intermediate,
and ﬂoating “de facto” exchange rate regimes. I use three different speciﬁcations.
The ﬁrst is the new IMF classiﬁcation for “de facto” exchange rare regimes,
introduced in 1997. The second classiﬁcation is constructed by Bubula and Otker-
Robe (2002), and combines the new IMF classiﬁcation with additional IMF internal
data/country information. The third classiﬁcation is introduced by Reinhart and
Rogoff (2002) and classiﬁes the regimes taking into account information on parallel
exchange markets.

Index of foreign exchange and trade liberalization. This EBRD index takes values
from 1 (lowest) to 4.3 (highest liberalization). Alternatively, I use a dummy variable
that takes the value 1 if there are any repatriation or surrender requirements of export
proceeds, zero otherwise. Additionally, I use a dummy variable that takes the value 1
if there is any tax imposed on foreign exchange transactions, zero otherwise.
Financial development. The ratio of domestic credit to enterprises to GDP.
Alternative measures used are the ratio of total domestic deposits to GDP, total M2 to
GDP, as well as EBRD indices of banking sector and non-banking ﬁnancial
institutions reform, which take values from 1 (lowest) to 4.3 (highest deve10pment).
Economic development. The GDP per capita.

Persistence ejfects. The maximum historical monthly inﬂation and devaluation rates

(see the Data Appendix).

80

0 Net foreign reserves. The ratio of the central bank’s net foreign reserves to GDP.
Other explanatory variables include the annual inﬂation rate, the annual devaluation
rate, the EBRD index of initial conditions, and the shares of agriculture, industry, and

service in the GDP.

5. Empirical Analysis

Table 2 reports descriptive statistics of the dependent and explanatory variables
used in the estimations. A simple data analysis shows positive correlation of deposit and
credit dollarization (0.585, P-value of 0.000). Net foreign assets, on the other hand, do
not seem to be correlated with credit dollarization (-0.042, P-value of 0.607). Neither are
the measures of bank proﬁtability, concentration, and risk-taking behavior?“ 29

On the ﬁrm side, external borrowing to total domestic credit is not signiﬁcantly
correlated with credit dollarization (0.076, P-value of 0.350), while different measures of

imports and exports are negatively but insigniﬁcantly correlated with dollarization.

Firm’s characteristics (proﬁtability, size) are not correlated with dollarization.

5.1. Pooled OLS (POLS) Estimation

I use the pooled OLS (with errors adjusted for serial correlation and

heteroskedasticity) as a benchmark for estimating the following equation

 

28 The correlation between credit dollarization and the interest rate spread is 0.136 (with a P-value of
0.107), the correlation between credit dollarization and bank proﬁtability is 0.136 (P-value = 0.416), the
correlation between credit dollarization and bank riskiness is 0.034 (P-value = 0.678), and the correlation
between credit dollarization and the total number of banks is -0. 100 (P-value = 0.222).

’9 All measures of proﬁtability and risk-taking behavior for both banks and ﬁrms are used as one-period
lags, in order to avoid endogeneity problems.

81

Credit_Dollarization,-,= a'Bank_ Variables p+ B'Firm_ Variablesp+ y, Controlsa+ 85;.
Table 3a presents the results when credit dollarization is measured as the ratio of dollar
credit to total bank credit. Table 3b presents the results when scaling by total bank assets.
There are four different speciﬁcations, (1) - (4), presented in each table. Notice that the
sample size in Table 3b is by almost 25 percent smaller than in Table 3a.

In addition to the variables already described in Section 4, I use a dummy
variable, which takes the value one for Croatia, zero otherwise. According to Bonin
(2001), most loans in Croatia are indexed to the German mark/Euro. These loans are
equivalent to foreign currency loans for the purpose of this chapter, but are instead
recorded as domestic currency loans. This explains why in Croatia the average level of
deposit dollarization is 69 percent, while credit dollarization is only 27 percent.30

Regardless of the measure used for credit dollarization, currency matching in the
banking sector seems to be the main determinant of dollarization. Deposit dollarization
has a large positive effect, while net foreign assets have a negative effect on credit
dollarization. A one percentage point increase in deposit dollarization increases credit
dollarization by 0.6 - 0.7 percentage points (see Table 3a), while a one percentage point
increase in net foreign assets decreases credit dollarization by approximately 0.2
percentage points.31 It seems to be the case that domestic banks substitute foreign assets

for domestic dollar loans. This suggests that either the domestic deposits are exported or

 

3° Alternatively, I estimate the model without data on Croatia, and obtain very similar results for all my
speciﬁcations.

3 Barajas and Morales (2003) also ﬁnd a signiﬁcantly positive effect of deposit dollarization on credit
dollarization. However, the estimated magnitude is much larger: a one percentage point increase in deposit
dollarization increases credit dollarization by around 0.9 percentage points.

82

that the foreign currency loans are ﬁnanced with foreign funds, or both. Banks do seem
to match the currency of denomination of their assets and liabilities.

There is some evidence of a positive effect of banks proﬁtability and
concentration (measured by the interest rate spread and the inverse of total number of
banks respectively) on credit dollarization. Higher proﬁtability of banks tends to increase
the level of dollarization, which is the opposite effect from what I hypothesized. The
total number of domestic banks, which is the inverse of banking industry concentration,
seems to reduce credit dollarization. This might be consistent with the conjecture that
more banks reduce the probability of bailout should a bank run occur, and thus reduce the
excessive risk-taking by banks (assuming that dollar loans bear a higher total risk than
domestic currency loans). However, these variables might be capturing missing country
effects as the effects disappear when I use alternative speciﬁcations (see the results for
the FD estimator, in Section 5.2). In addition, I would like to emphasize that my
measures of proﬁtability and concentration, both for banks and ﬁrms, are most likely
poor and the effects should be interpreted with caution.

There is some evidence of a ﬁrm-determined component of credit dollarization.
However, the effect varies across different measures and speciﬁcations. External
borrowing by ﬁrms seem to increase dollarization when using the ﬁrst measure of credit
dollarization, while the effect becomes negative (but insigniﬁcant) with the second
measure. I further explore this effect in Section 6.

There is also limited evidence that real dollarization leads to ﬁnancial
dollarization (see Table 3b). The effect, however, is very small: a one percentage point

increase in total trade to GDP raises the ratio of dollar loans to total bank assets by only

83

0.07 percentage points, while a one percentage point increase in exports to GDP raises
dollarization by 0.16 percentage points.

With respect to macroeconomic determinants of dollarization, I ﬁnd evidence that
current and past uncertainty and lack of credibility of policies encourage ﬁnancial
dollarization. Net foreign reserves to GDP, which I interpret as a measure of current
credibility and policy sustainability has a strong negative effect. The maximum historical
inﬂation, which measures past uncertainty and its persistence over time, has a positive
effect with the broader measures of dollarization.

There is some marginal evidence that higher ﬂexibility of the exchange rate
reduces dollarization (especially when comparing ﬂexible with ﬁxed exchange rate
regimes), which is consistent with the common belief that ﬁxed exchange rate regimes
provide implicit insurance against changes in the exchange rate.32 Not surprisingly, the

dummy for Croatia is very signiﬁcant and has a large negative effect (see Table 3a).

5.2. First Difference (FD) Estimation

I expect there exist some persistence effects in my indicators of dollarization as
they are measured as stock rather than ﬂow variables. Also, I believe that the country
effects are very important, and there might be regulatory effects, in addition to initial
conditions/peculiarities of these economies, that I do not control for. When these missing

effects are correlated with the variables of interest, the estimated coefﬁcients are biased.

 

32 I use Bubula and Otker-Robe classiﬁcation in the benchmark speciﬁcations presented in this chapter
Similar results are obtained with the new IMF and Reinhart and Rogoff classiﬁcations.

84

I test for serial correlation in the errors, and ﬁnd that the autoregressive
coefﬁcient for the change in the initial errors (on) is around 0.15. If initial errors were
uncorrelated, this coefﬁcient would be around -0.5. Not only are initial errors correlated,
but also some serial correlation seems to be left even after taking ﬁrst differences.
Hence, FD (which assumes initial errors are integrated of order one) might provide more
reliable results than the ﬁxed effects estimator (which assumes initial errors are
uncorrelated). I use both estimators, and compare the obtained effects.

Tables 4a and 4b show the estimation results for the FD estimation, with standard
errors adjusted for serial correlation and heteroskedasticity. One drawback of the FD
estimator is that the sample size decreases, since the ﬁrst period observations are lost
when taking differences. I will focus on the results presented in Table 4a, when credit
dollarization is calculated as a ratio to total credit (I have less data on total bank assets,
thus the sample size diminishes even more when scaling by it).

Currency matching is still the main driving force of dollarization. Deposit
dollarization signiﬁcantly increases and net foreign assets signiﬁcantly decrease credit
dollarization. However, bank proﬁt and concentration variables, as well as ﬁrm
variables, do not seem to have a signiﬁcant effect on credit dollarization. External
borrowing by ﬁrms, as well as trade and its components (exports and imports), still have
a positive, but insigniﬁcant, effect.

The effect of the exchange rate regime dummies is even higher now, with higher
ﬂexibility of the exchange rate reducing credit dollarization. Similarly to the POLS
estimation, net foreign reserves reduce credit dollarization. The difference from the

POLS estimation is that the maximum historical devaluation rate increases dollarization,

85

while the historical inﬂation rate tends to decrease it. Finally, I can conclude that implicit
insurance against changes in the exchange rate offered by a ﬁxed exchange rate regime,
lack of credibility of macroeconomic policies, and high past devaluations increase

dollarization.

5.3. Fixed Effects (FE) Estimation

The third speciﬁcation for my model is the ﬁxed effects (within) estimation. This
also controls for country effects, similar to the FD, while better capturing the effects of
factors with less sample variation. Errors are adjusted for serial correlation and
heteroskedasticity.33

The FE estimator (see Tables 5a and 5b) shows a similar picture with the POLS
estimator. Currency matching in the banking sector, through deposits and foreign assets,

34 In addition, higher proﬁtability also leads to more loans in

drive credit dollarization.
dollars, while bank concentration does not seem to have any effect.

On the ﬁrm side results are very similar to the POLS estimation. First, external
borrowing by ﬁrms increases the amount of dollar loans they obtain domestically (see
Table 5a). External loans and domestic foreign currency loans seem to be complements,
not substitutes as I hypothesized. It might be the case that external lending is conditional
on domestic lending, in which case my explanatory variable is endogenous. Second, total

trade and exports appear to have a positive and sometimes signiﬁcant effect on

dollarization of credit. Thus, there is some limited evidence that real dollarization does

 

3’ Similar results are obtained when using a FE estimator with errors assumed to follow an autoregressive
process of order one (AR(1)).

Deposit dollarization has a positive effect, but is insigniﬁcant with the second measure of dollarization
(see Table 5b).

86

indeed cause ﬁnancial dollarization. However, the estimated effect is very small (see
Table 5b).

There is mixed evidence for the exchange rate regime effects, while the lack of
credibility of macroeconomic policies (proxied by net foreign reserves) still leads to
higher dollarization. No persistence effects are found, but a large positive effect for the

trend variable.

6. Robustness Tests

6.1. Endogeneity Issues

Summing up, all the evidence presented so far suggests the same conclusion:
currency matching in the banking sector drives the dollarization of credit. I have also
identiﬁed two effects coming from the ﬁrm side, mainly from real dollarization and
ﬁrms’ alternative sources of dollar loans. However, these effects are not that robust
across alternative speciﬁcations. The question that rises is whether the results are driven
by the potential endogeneity of asset and liability management variables.

To address this issue I proceed in two ways. First, I estimate the POLS using a
one-period lag for banks’ net foreign assets and ﬁrms’ external borrowing instead of their

35 While the external borrowing becomes insigniﬁcant, net

contemporaneous values.
foreign assets have still a signiﬁcant negative effect (and of similar magnitude as in

Section 5.1). It is possible then that the access to external credit is conditional on

borrowing in dollars from domestic banks; in this case, the causality is from credit

 

35 I do not report the results in the chapter due to space constraints. However, they are available upon
request.

87

dollarization to external borrowing. If this is the case, my model is well speciﬁed if I
exclude ﬁrms’ external borrowing from it. No results change when I do so.

Second, I estimate the model using Instrlnnental Variables (IV). I believe that
past values of ﬁrms’ external borrowing and bank holdings of external assets (external
liabilities) are good measures of ﬁrms’ access to external ﬁnance and banks’ current
holdings of external assets (external liabilities). Statistically, lagged values are indeed
good IV s as they are very signiﬁcant in the reduced-form ﬁrst stage estimation.

The results for the 2SLS estimations are reported in Tables 6a and 6b. The
standard errors are again adjusted for both serial correlation and heteroskedasticity. Net
foreign assets are signiﬁcantly negative again, and the effect is slightly larger with IV
than with OLS. Firms’ external borrowing, on the other hand, is never signiﬁcant, and

36 The Hausman test does not reject the

even in magnitude the effect becomes smaller.
null hypothesis of exogeneity in external borrowing and foreign assets by ﬁrms and
banks, respectively. Therefore, there is no evidence that the OLS and the IV estimations
are signiﬁcantly different. In addition, there is no evidence of serial correlation in the
standard errors, which seems to suggest that the ZSLS in ﬁrst moments is the correct way
to proceed.

One might argue that deposit dollarization is endogenous as well, if ﬁrms are
required to hold dollar deposits in order to receive dollar loans. I do not have any

speciﬁc information if there is indeed such a requirement in transition economies. Even

if there is not, unless banks make an immediate payment abroad or ﬁrms immediately sell

 

3’ Barajas and Morales (2003) ﬁnd evidence of a negative effect on ﬁrms’ external borrowing on domestic
dollar borrowing. It might be the case that ﬁrms in Latin American economies have higher access to
external ﬁnancial markets than ﬁrms in transition economies, and they can indeed substitute dollar loans
from abroad with dollar loans from domestic banks.

88

the dollars they obtain, dollar loans will at least temporarily increase the level of dollar
deposits.

I deﬁne household deposit dollarization as the ratio of household dollar deposits
to total household deposits, and use it as an IV for the total deposit dollarization ratio.
Household deposit dollarization can be safely assumed to be uncorrelated to the errors in
the main model, and thus exogenous. Furthermore, it has a strong positive effect on
deposit dollarization in the reduced-form ﬁrst stage regression. The effect of deposit
dollarization on credit dollarization is still positive, but smaller and insigniﬁcant.
However, the Hausman test fails to reject the null hypothesis of exogeneity in deposit
dollarization. If instead I use a FE estimator with N8, deposit dollarization is highly
signiﬁcant and positive, and the effect is even larger than without IV s.

In conclusion, IV estimates suggest that the results are not driven by potential
endogeneity problems. Bank currency matching still remains the main driving force of
credit dollarization, with the effect of net foreign assets becoming even larger, and a
possibly smaller but still positive effect of deposit dollarization. As expected, ﬁrms’
access to external borrowing is not signiﬁcant at all, suggesting that the causality is from

domestic to external borrowing, and not the other way around.

6.2. Alternative Specifications

To assess the robustness of these ﬁndings, I conduct several other tests. First, I
use alternative measures for bank and ﬁrm variables. Second, I discuss the effects on

dollarization of ﬁrm characteristics such as size, proﬁtability, and risk-taking behavior,

89

which were not included in the main speciﬁcations. Third, I undertake additional
sensitivity tests.

Besides the interest rate spread and the number of total banks, I use several other
measures of bank proﬁtability, concentration, and in addition riskiness, such as bank
proﬁts, number of foreign banks, bank risk-taking behavior, and deposit insurance (see
the Data Appendix, for the description of these variables). While bank proﬁts still have a
positive but insigniﬁcant effect, suggesting that bank proﬁtability might be positively
linked to credit dollarization, all the other variables have mixed and insigniﬁcant effects.

On the ﬁrm side, I try alternative measures for external supply shocks, such as
changes in import or export prices. Using the POLS estimator, changes in the oil price
seems to positively affect dollarization in non oil producing countries, while the opposite
is true for oil producing countries. However, the effect vanishes when I control for
country-speciﬁc effects. Other prices do not seem to have a signiﬁcant effect. I also test
to see if dollar borrowing ﬁnances speciﬁc components of imports, or if ﬁrms in speciﬁc
sectors are more prone to borrowing in dollars. For example, one might think that ﬁrms
in the natural resource export sector might borrow more in dollars. The ratio of natural
resource exports to GDP, while signiﬁcantly positive when using the FE estimator, is
insigniﬁcant when using all other estimators. To conclude, I cannot ﬁnd any effect ﬁom
the real side of the economy other than that coming from the overall level of imports and
exports.

I do not include the ﬁrm characteristics variables (size, proﬁtability, and
riskiness) in the main estimations, due to missing data on these variables. However, I

separately estimate the model controlling for these effects, and show they are all very

90

insigniﬁcant. As I have already indicated, my measures of bank and ﬁrm proﬁtability
and concentration are very poor. In addition, I might miss effects because of using
aggregated data. Thus the results should be interpreted .with caution. These ﬁndings
rather suggest that better data have to be collected before these effects could be seriously
studied.

Additional sensitivity tests I perform are the following:

0 Using standardized coefﬁcients for the FD estimator, to compare the magnitude of
bank- and ﬁrm-speciﬁc effects.37

0 Using the dollarization and net foreign assets variables, scaled by total credit
instead of total deposits, to check if scaling by different variables changes the
results.

0 Using alternative regulatory measures: indicators of restrictions on domestic
lending in foreign currencies,38 forward market, repatriation or surrender
requirements of the export proceedings, and tax on the foreign exchange
transactions.

0 Using alternative measures for the overall level of development ‘of ﬁnancial
markets: domestic deposits to GDP, M2 to GDP, the index of banking reform, and
the index of non-banking reform.

0 Using year dummies for the time effects, instead of the time trend.

 

37 Changes in deposit dollarization and net foreign reserves still have the largest relative effect on credit
dollarization: a one percent standard deviation increase in deposit dollarization raises credit dollarization by
0.446 standard deviations, and a one percent increase in net foreign reserves raises it by 0.378 standard
deviations (P-value for both effects is 0.000). Among the ﬁrm variables, trade has the largest relative effect
(0.120, but P-value=0.186). The exchange rate regime dummies and net foreign reserves have a signiﬁcant
negative impact: -0.228 (P-value=0.010), —0.241 (P-value=0.023), and -0.l82 (P-value=0.014) respectively.
38 Unfortunately, I have very limited data on foreign currency lending restrictions. Hence including this
variable in the model reduces my sample size by half.

91

0 Using the alternative classiﬁcations for the exchange rate regimes: the new IMF
and the Reinhart and Rogoff (2002) classiﬁcations. In addition, I specify the
exchange regimes as freely falling versus non-freely falling, to better capture the
uncertainty associated with large movements of the exchange rate.

0 Using a dummy variable for the oil producing countries.

0 Using other controls for the overall level of development of the economy: GDP
per capita, the index of initial conditions, the shares of agriculture, industry, and
service in GDP.

None of these sensitivity tests changes the main results.

7. Conclusions and Future Extensions

Domestic banks in emerging markets and transition economies often grant dollar
credit to residents instead of credit denominated in domestic currency. Is this credit
dollarization phenomenon coming mostly ﬁom the ﬁnancial side (banks), or from the real
side (ﬁrms)? In the presence of ﬁnancial dollarization without real dollarization, there
exist balance sheet mismatches somewhere in the economy and the economy is highly
exposed to ﬁnancial and currency crises.

This chapter analyzes the respective contributions of banks and ﬁrms to credit
dollarization, and suggest what policy measures (if any) should be implemented to
control or restrict it. I group potential determinants into bank- and ﬁrm-speciﬁc
(ﬁnancial and mainly real) factors, and use a newly-assembled database for twenty-two

transition economics.

92

I show that bank currency matching is the main driving force of credit
dollarization. I ﬁnd some evidence that real dollarization also causes ﬁnancial
dollarization, but this effect is relatively small. There are. currency mismatches in these
economies, and they seem to be concentrated in the real sector. Banks, who are supposed
to do a better job at internalizing the exchange rate risk, just pass it over to ﬁrms. This
decreases bank exposure to currency risk, but it increases their exposure to default risk,
and ultimately the exposure of the economy to ﬁnancial and currency crises.

Not surprisingly, my ﬁndings suggest that one way to reduce dollarization is to
reduce macroeconomic uncertainty and lack of monetary policy credibility. Otherwise,
as long as domestic deposits are highly dollarized, and banks want to match the
currencies of their assets and liabilities, restrictions placed on dollar loans will most
likely lead to “exports” of deposits and domestic disinterrnediation.

This chapter is one of the ﬁrst attempts to empirically study credit dollarization.
There are several possible extensions that deserve further consideration. First, I would
like to extend the analysis to include other developing economies, especially Latin
American countries where dollarization reaches very high levels. That would provide not
only more data, but also the generality of the conclusions. Transition economies are a
good starting point as they provide a lot of variation in data; this can also be a downside,
since it is sometimes hard to separate one effect ﬁ'om the others.

Second, using disaggregated data, such as individual loan data, would provide

better insights into the effect of ﬁrm and bank characteristics on ﬁnancial dollarization.

93

Third, more empirical work in general is necessary to better identify both the
determinants and the effects of ﬁnancial dollarization, and to evaluate policy implications

and the exchange rate exposure of various economic agents and the overall economy.

94

REFERENCES

Aguiar, M. 2002. "Devaluation, Foreign Currency Exposure, and Investment: The Case
of Mexico." Mimeo. University of Chicago.

Arteta, C. O. 2002. "Exchange Rate Regimes and Financial Dollarization: Does
Flexibility Reduce Bank Currency Mismatches?" Board of Governors of the
Federal Reserve System, International Finance Discussion Paper 738.

Barajas, A. and A. Morales. 2003. "Dollarization of Liabilities: Beyond the Usual
Suspects." International Monetary Fund Working Paper 03/11.

Bonin, J. 2001. "Financial Intermediation in Southeast Europe: Banking on the Balkans."
Mimeo. Vienna Institute for International Economic Studies.

Broda C. and E. Levy Yeyati. 2003. "Endogenous Deposit Dollarization." Federal
Reserve Bank of New York Staff Report 160.

Bubula, A. and I. Otker-Robe. 2002. "The Evolution of Exchange Rate Regimes since
1990: Evidence from De Facto Policies." International Monetary Fund Working
Paper 02/155.

Burnside, C., M. Eichenbaum, and S. Rebelo. 2000. "Hedging and Financial Fragility in
Fixed Exchange Rate Regimes." Mimeo. Northwestern University and National
Bureau of Economic Research.

Caballero, R. and A. Krishnamurthy. 2002. "Excessive Dollar Debt: Financial
Development and Underinsurance." Journal of Finance, 58:2, pp. 867-894.

Calvo, G. A. 2001. "Capital Markets and the Exchange Rate: With Special Reference to
the Dollarization Debate in Latin America." Journal of Money, Credit, and
Banking, 33:2, pp. 312-334.

Calvo, G. A. 2002. "On Dollarization." Economics of Transition, 10:2, pp. 393-403.

Catao, L. and M. Terrones. 2000. "Determinants of Dollarization: The Banking Side."
International Monetary Fund Working Paper 00/146.

Cotner, J. 1991. "Currency Risk in Long-term Borrowing." Journal of Multinational
Financial Management, 1:2, pp. 49-64.

Delgado, F. L., D. S. Kanda, G. M. Casselle, and R. A. Morales. 2002. "Domestic
Lending in Foreign Currency." in Building Strong Banks through Surveillance
and Resolution. C. Enoch, D. Marston, and M. Taylor eds. International Monetary
Fund: Washington DC, pp. 40-61.

95

Demirgiic-Kunt, A. and H. Huizinga. 2003. " Market Discipline and Deposit Insurance."
Journal of Monetary Economics, forthcoming.

EBRD. 1998-2002. "Transition Report." European Bank for Reconstruction and
Development: London.

Giraldi, C. and R. Hamaui. 1991. "Foreign Exchange Market Liberalization and the
International Diversiﬁcation of Borrowing by Italian Firms." Journal of
Economics and Business, 43:4, pp. 309-327.

Hausmann, R., U. Panizza, and E. Stein. 2001. "Why Do Countries Float the Way They
Float?" Journal of Development Economics, 66:2, pp. 387-414.

Honohan, P. and A. Shi. 2002. "Deposit Dollarization and the Financial Sector in
Emerging Economies." World Bank Working Paper 2748.

IMF. 1990-2002. "Annual Report on Exchange Arrangements and Exchange
Restrictions." International Monetary Fund: Washington, DC.

Ize, A. and E. Levy Yeyati. 2003. "Financial Dollarization." Journal of International
Economics, 59:2, pp. 323-347.

Ize, A. and E. Parrado. 2002. "Dollarization, Monetary Policy, and the Pass-through."
International Monetary Fund Working Paper 02/188.

Jeanne, O. 1999. "Foreign Currency Debt, Moral Hazard, and the Global Financial
Architecture." Mimeo. International Monetary Fund: Washington, DC.

Jeanne, O. 2001. "Why Do Emerging Markets Borrow in Foreign Currency?" Mimeo.
International Monetary Fund: Washington, DC.

Kedia, S. and A. Mozumdar. 2002. "Foreign Currency Denominated Debt: An Empirical
Examination." The Journal of Business: forthcoming.

Keloharju, M. and M. Niskanen. 2001. "Why Do Firms Raise Foreign Currency
Denominated Debt? Evidence from Finland." European Financial Management,
7:4, pp. 481-496.

Krugrnan, P. 1999. "Balance Sheet, the Transfer Problem, and Financial Crises."
International Tax and Public Finance, 6:4, pp. 459-472.

Martinez, L. and A. Werner. 2002. "The Exchange Rate Regime and the Currency
Composition of Foreign Debt: The Mexican Experience." Journal of Development
Economics, 69:2, pp. 315-334.

Mishkin, F. 1996. "Understanding Financial Crises: A Developing Country Perspective."

in Annual World Bank Conference on Development Economics 1996. M. Bruno
and B. Pleskovich eds. Washington DC, pp. 29-62.

96

 

Obs

Reir

 

 

Obstfeld, M. 1998. "The Global Capital Market: Benefactor or Menace?" Journal of
Economic Perspectives, 12:4, pp. 9-30.

Reinhart, C. M. and K. S. Rogoff. 2002. "The Modern History of Exchange Rate
Arrangements: A Reinterpretation." National Bureau of Economic Research
Working Paper 8963.

97

APPENDIX

Data Definitions and Sources

Abbreviations: AREAER: IMF Annual Report on Exchange Arrangements and
Exchange Restrictions (various issues). CB: Central Banks publications (various issues).
IFS: International Monetary Fund International Financial Statistics. IMF_CP: IMF
Commodity Prices. WB_DI: World Bank Deposit Insurance Database. WDI: World
Bank World Development Indicators. EBRD: European Bank for Reconstruction and
Development Transition Reports (various issues). BIS: Bank of International Settlements
International Banking Statistics. UN: United Nations Trade Statistics.

I use two datasets which include information on individual ﬁrms and banks in
transition economies. Amadeus provides data on ﬁrms’ balance sheet and proﬁt and loss
account, while Bankscope provides data on banks’ balance sheet and income statement.

I also use Bonin (2001) for data on total bank assets for several transition
economics, as well as AREAER, Bubula and Otker-Robe (2002), and Reinhart and

Rogoff (2002) for alternative classiﬁcations of the exchange rate regimes.

 

Variable Definition Source

 

Credit dollarization Foreign currency denominated credit to total CB
credit (total bank assets), from domestic banks to
domestic enterprises, calculated as a percent.
Data for a few countries include credit to
individuals, ﬁnancial enterprises and/or

 

government.
Deposit Foreign currency denominated deposits to total CB
dollarization deposits (total bank assets, or total credit) held by
residents at domestic banks, calculated as a
percent.

 

 

 

 

Net foreign assets Bank foreign assets minus bank foreign liabilities IFS, CB

 

 

98

 

 

 

lntc

 

Ban

 

Ban
beh;

Dept

Tota

M
Exte

 

TEE

Impo
inten

 

 

(line 21 . . .ZF. .. - line 26C. . .ZF), as a percent of
total bank deposits (total bank credit, or total bank
assets).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Interest rate spread Average lending interest rate minus average EBRD, IFS,
deposit interest rate (over all currencies and CB
maturities). Data for most countries is from
EBRD.

Bank proﬁts Banks’ pre-tax proﬁts as a percent of total assets Bankscope
(line 6780*100/line 2025).

Bank risk-taking Banks’ total risk assets as a percent of total assets Bankscope

behavior (line 7480*100/1ine 2025). Total risk assets
include: total loans (net), other listed securities,
other securities, equity investments, investment
securities, bonds, non-listed securities, other
investments, deferred tax receivable, other non
earning assets, and intangible assets.

Deposit insurance Dummy variable=l if there exists explicit deposit WB_DI
insurance, =0 if there does not.

Total number of Total banks per million people. EBRD

banks

Number of foreign Foreign banks per million people. EBRD

banks

External borrowing Foreign claims on non-bank private enterprises as BIS (Table
a percent of total domestic credit (GDP). 9A, col. G)

Trade Total trade (imports plus exports of goods and WDI
services) as a percent of GDP.

Imports of Imports of intermediate goods UN

intermediate goods (SITC2+SITC3+. . .+SITC7: crude materials,
mineral fuels, oil and fats, chemicals, basic
manufactures, and machinery and transport
equipment) as a percent of GDP.

Imports of raw Imports of agricultural materials, ﬁrels, and ores WDI

materials and metals, as a percent of GDP.

Exports Total exports of goods and services as a percent WDI, IFS
of GDP.

Index of total The index of fuel and non-fuel commodity prices IMF_CP

commodities for all countries (1995:100).

Index of industrial The changes in prices of industrial inputs for all IMF_CP

inputs countries (1995:100).

Change in the oil Change in the average world crude spot price, IFS

price USD/barrel @ode 00176AAZZF).

Average ﬁrm size Logarithm of average total assets per company, Amadeus

(total assets) average for (up) to 50 largest companies (by total
assets).

Average ﬁrm size Number of employees, average for (up) to 50 Amadeus

(no employees)

 

largest companies (by total assets).

 

 

99

 

 

Variable

Definition

Source

 

F inn proﬁt margin

Net proﬁt after taxes as a percent of total sales for
a given 12-month period, average for (up) to 50
largest companies (by total assets).

Amadeus

 

Firm leverage ratio

Long-term funds with ﬁxed interest as a percent
of total capital, average for (up) to 50 largest
companies (by total assets).

Amadeus

 

Forward market

Dummy variable=l if a forward market for
foreign exchange exists and is functional, =0 if it
does not exist or is underdevelgied.

AREAER

 

Fixed exchange
rate regime

Intermediate
exchange rate
regime

Floating exchange
rate regime

Dummy variables differentiating between ﬁxed,
intermediate, and ﬂoating exchange rate regimes.
These are all measures of the “de facto” regimes,
which might be very different from the “de jure”
ones. I use three different speciﬁcations. The
ﬁrst is the new IMF classiﬁcation for “de facto”
exchange rare regimes, introduced in 1997. The
second classiﬁcation is constructed by Bubula and
Otker-Robe (2002) and combines the new IMF
classiﬁcation with additional IMF internal
data/country information. The third classiﬁcation
is introduced by Reinhart and Rogoff (2002) and
classiﬁes the regimes taking into account
information on parallel exchange markets.

AREAER,
Bubula and
Otker-Robe
(2002),
Reinhart and
Rogoff
(2002)

 

 

 

 

Index of forex
liberalization

 

Index of trade and foreign exchange rate system
liberalization, =1 if widespread import and/or
export controls or very limited legitimate access
to foreign exchange, =2 if some liberalization of
import and/or export controls, almost full current
account convertibility in principle but with a
foreign exchange regime that is not ﬁrlly
transparent (possibly with multiple exchange
rates), =3 if removal of almost all quantitative and
administrative import and export restrictions,
almost full current account convertibility, =4 if
removal of all quantitative and administrative
import and export restrictions (apart form
agriculture) and all signiﬁcant export tariffs,
insigniﬁcant direct involvement in exports and
imports by ministries and state-owned trading
companies, no major non-uniformity of customs
duties for non-agricultural goods and services, full
current account convertibility, =4.3 if standards
and performance norms of advanced industrial
economies, removal of most tariff barriers, WTO
membership.

 

EBRD

 

100

 

 

Variable

Deﬁnition

Source

 

Foreign currency
lending

Dummy variable=1 if domestic lending in foreign
currency is free, =0 if it is prohibited or there are
any restrictions imposed.

AREAER

 

Repatr/Surrend

Dummy variable=l if there are any repatriation or
surrender requirements of export proceeds, =0
otherwise.

AREAER

 

Exchange tax

Dummy variable=l if there is any tax on foreign
exchange transactions, =0 otherwise.

AREAER

 

Domestic credit

Domestic credit to enterprises as a percent of
GDP.

CB

 

Domestic deposits

Total domestic deposits as a percent of GDP.

CB

 

M2

Money and quasi-money as percent of GDP ((line
34. . .ZF. . .+line 35. . .ZF. . .)*100/GDP).

IFS

 

Index of banking
reform

Index of banking reform and interest rate
liberalization, =1 if little progress beyond
establishment of a two tier system, =2 if
signiﬁcant liberalization of interest rates and
credit allocation, =3 if substantial progress in
establishment of bank solvency and of a
ﬁ‘amework for prudential supervision and
regulation, ﬁlll interest rate liberalization with
little preferential access to cheap reﬁnancing,
signiﬁcant lending to private enterprises and
signiﬁcant presence of private banks, =4 if
signiﬁcant movement of banking laws and
regulations towards BIS standards, well
functioning banking competition and effective
prudential supervision, signiﬁcant term lending to
private enterprises, substantial ﬁnancial
deepening, =4.3 if standards and performance
norms of advanced industrial economies, full
convergence of banking laws and regulations with
BIS standards, provision of full set of competitive
banking services.

EBRD

 

 

Index of non-
banking reform

 

Index of securities markets and non-bank

ﬁnancial institutions development, =1 if little
progress, =2 if formation of securities exchanges,
market makers and brokers, some trading in
government paper and/or securities, rudimentary
legal and regulatory ﬁamework for the issuance
and trading of securities, =3 if substantial issuance
of securities by private enterprises, establishment
of independent share registries, secure clearance
and settlement procedures, and some protection of
minority shareholders, emergence of non-bank

 

EBRD

 

101

 

 

ﬁnancial institutions (e. g. investment funds,
private insurance and pension funds, leasing
companies) and associated regulatory framework,
=4 if securities laws and regulations approaching
IOSCO standards, substantial market liquidity and
capitalization, well-ﬁlnctioning non-bank
ﬁnancial institutions and effective regulation,
=4.3 if standards and performance norms of
advanced industrial economies, full convergence
of securities laws and regulations with IOSCO
standards, fully developed non-bank
intermediation.

 

 

 

 

 

 

 

 

 

 

 

GDP GDP at market prices (current USD) per capita. WDI
Net foreign Net reserves of the central bank as a percent of IFS
reserves GDP (line 1L..DZF. . .*100/GDP).
Inﬂation rate Percentage change in the index of consumer EBRD
prices.
Devaluation rate Percentage change of the average exchange rate IFS
(expressed as domestic currency per USD, line
RF..ZF. . .).
Maximum Maximum historical inﬂation rate, calculated as IFS
historical inﬂation the maximum of all past monthly inﬂation rates
(for which data are available) and current monthly
inﬂation rate. The annual data is the historical
maximum of monthly inﬂation rates at the end of
the year.
Maximum Maximum historical devaluation rate, calculated IFS
historical as the maximum of all past monthly devaluation
devaluation rates (for which data are available) and current
monthly devaluation rate. The annual data is the
historical maximum of monthly devaluation rates
at the end of the year.
Index of initial Index derived from factor analysis and represents EBRD
conditions a weighted average of measures for the level of
development, trade dependence on CMEA,
macroeconomic disequilibria, distance to the EU,
natural resources endowments, market memory,
and state capacity. Higher values denote more
favorable starting positions. Data are from the
EBRD Transition Report, 2000, pg. 21.
Agriculture The percent of value-added in agriculture in total EBRD
GDP.
Industry The percent of value-added in industry in GDP. EBRD
Services The percent of value-added in the service sector in EBRD

 

total GDP (GDP-share of agriculture-share of
industry).

 

 

102

 

Albania
Armenia
Azerbaijan
Bulgaria
Croatia

Czech Republic
Estonia
Georgia
Hungary
Kazakhstan
Kyrgyz Republic
Latvia
Lithuania
Macedonia
Moldova
Poland
Romania
Russia

Slovak Republic
Slovenia
Ukraine

Yugoslavia

1998 - 2001

2000 - 2001

2000 - 2001

1990 - 2001

1993 - 2001

1993 - 2001

1993 - 2001

1997 — 2001

1992 - 2001

1996 - 2001

1999 - 2001

1994 - 2001

1993 - 2001

1995 - 2001

1999 - 2001

1997 - 2001

1990 - 2001

1996 - 2001

I992 - 2001

1991 - 2001

1992 - 2001

1998 - 2001

Table 1 Country and Dollarization Data Coverage

103

Mean
Standard deviation
Number of observations

Mean
Standard deviation
Number of observations

Mean
Standard deviation
Number of observations

Mean
Standard deviation
Number of observations

Mean
Standard deviation
Number of observations

Mean
Standard deviation
Number of observations

Table 2 Summary Statistics

Dependent Variables
Credit Dollarization ' Credit Dollarization
(% of total credit) (% of total bank assets)
38.192 11.878
22.438 8.173
156 110
Explanatory Variables

Deposit Dollarization
(% of total deposits)

36.689
18.431
172

Net foreign assets
(% of total deposits)

7.426
36.826
167

Interest rate spread
(one period lagged)

21.475
46.855
191

Deposit insurance
(dummy)

0.275
0.447
272

External borrowing
(% of total credit)

33.755
49.074
152

Bank-specific Variables

Deposit Dollarization
(% of total bank assets)

17.347
9.697
120

Net foreign assets
(% of total bank assets)

5.429
10.760
118

Bank profits
(% of total assets)

2.532
4.927
193

Total number of banks
(per million people)

7.648
6.086
213

F inn-speciﬁc Variables

External borrowing
(% of GDP)

3.905
5.334
202

104

Deposit Dollarization
(% of total credit)

57.490
58.868
156

Net foreign assets
(% of total credit)

21.371
73.084
154

Bank risk-taking
(%, risk assets/total assets)

57.705
16.090
197

Number of foreign banks
(per million people)

1.550
1.321
191

Imports of lnterm goods
(% of GDP)

33.004
14.208
145

Mean
Standard deviation
Number of observations

Mean
Stande deviation
Number of observations

Mean
Standard deviation
Number of observations

Mean
Standard deviation
Number of observations

Mean
Standard deviation
Number of observations

Mean
Standard deviation
Number of observations

Mean
Standard deviation
Number of observations

Table 2 (cont’d)

Explanatory Variables
F irm-speciﬁc Variables
Exports Index of total commodities Index of indust inputs
(% of GDP) (1995=100) (1995=100)
41.133 95.538 85.349
17.497 13.549 9.770
249 198 286
Average ﬁrm size Average firm size Firm profit margin
(log of total assets) (no of employees)
5.346 3840.958 3.408
1.385 4487.276 5.132
95 85 89
Firm leverage ratio
48.737
45.560
80
Other Variables
Forward market
(dummy)
0.277
0.448
238
Exchange rate regime- new IMF classification
Fixed Intermediate Floating
0.304 0.103 0.591
0.461 0.304 0.492
223 223 223
Exchange rate regime- Bubu1a&Otker-Robe classification
Fixed Intermediate Floating
0.307 0.213 0.478
0.462 0.410 0.500
234 234 234
Exchange rate regime- Reinhart&Rogoff classification
Fixed Intermediate Floating
0.213 0.355 0.431
0.410 0.479 0.496
225 225 225

105

Mean
Standard deviation
Number of observations

Mean
Standard deviation
Number of observations

Mean
Standard deviation
Number of observations

Mean
Standard deviation
Number of observations

Mean
Standard deviation
Number of observations

Table 2 (cont’d)

Explanatory Variables
Other Variables

Index of forex liberalization
(from 1=lowest to 4.3=highest)

3.332
1.092
241

Domestic credit
(% of GDP)

19.849
15.816
156

Index of banking reform
(from 1=lowest to 4.3=highest)

2.237
0.825
241

Inflation rate
(annual rate)

94.892
424.972
154

Devaluation rate
(annual rate)

45.000

144.757
153

106

Repatr/Surrend Net foreign resv
(dummy) (% of GDP)
0.746 10.576

0.435 6.700

217 214

M2 GDP

(% of GDP) (per capita)
28.824 2372.244

19.635 1907.543

206 252

Index of non-banking reform
(from 1=lowest to 4.3=highest)

1.832
0.717
241

Maximum historical inﬂation
(monthly rate)

2292.606
5586.61 1
196

Maximum historical devaluation
(monthly rate)

333.520
1042.129
207

Table 3a Pooled OLS

Dependent Variable: Credit Dollarization (% of Total Credit)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ll) (2) (3) (4)
Deposit dollarization 0611*" 0715*“ 0610*" 0711*“
(% of total deposits) (0.186) (0.193) (0.204) (0.1 Q)
Net foreign assets 0261*" -0.201*** 0227*" 0203*“
(% of total deposits) (0.055) (0.050) 40.064) (0.050)
Interest rate spread 0.053 0068*" 0.056 0067*"
(one period lagged) @134) (0.022) (0.060) (0.022)
Number of total banks -0.742* -0.721 -0.698
(per capita) (0.413) (0.495) (0.408)
External borrowing 0.677 * 0630* 0.469 0.679"
(% of GDP) (0.366) (0.352) (0.408) (0.325)
Trade 0.019 0.023
(% of GDP) (0.081) (0.085)
Imports of intermediate goods 0.199
(% of GDP) (0.223)
Exports -0.010
(% of GDP) (0.172)
Index of forex liberalization 5.221 0.420 2.195 0.376
(3.932) (4.266) (5.266) (4.037)
Intermediate exchange rate regime -4.471 -6.844 -2.999 -7.318
(5.994) (5.531) 45.322) (5.402)
Floating exchange rate regime 4.659 -10.035* -5.910 -10.363*
@921) (5.327) (6.004) (5.266)
Net foreign reserves -0.827*** 0958*" -1.149*** 0918*"
(% of GDP) (0.242) (0.273) (0.339) (0.248)
Domestic credit -0.416* -0.188 -0.399 -0.182
(% of GDP) (0.233) (0.191) (0.234) (0.183)
Inﬂation rate 0003 -0.007"'* -0.0010 -0.007**
(0.003) (0.003) (0.016) (0.002)
Maximum historical inflation -0.000
(0.000)
Croatia -34.496*** -35.443*** -29.315"”" 85.872"
(8.943) (9.387) (10.594) (8.883)
Trend 2.026 2.125” 2.119 2.129“
(1.175) (1.183) (1.254) (1.176)
_cons -4.155 14.805 11.669 16.851
(14.062) (15.530) (15.639) (15.890)
Obs. 136 139 115 139
Adjusted R-squared 0.691 0.690 0.713 0.689

 

Standard errors in parentheses (adjusted for heteroskedasticity and serial correlation)

*** Significant at 1%; ** signiﬁcant at 5%; * signiﬁcant at 10%

107

Table 3b Pooled OLS

Dependent Variable: Credit Dollarization (% of Total Bank Assets)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(1) (2) (3) (4)
Deposit dollarization 0.242 0.294" 0.372" 0.288"
(% of total bank assets) (0.166) (0.131) (0.138) (0.125)
Net foreign assets -0.355"* 0360*” -0.501*** 0347"
(% of total bank assets) (0.094) (0.084) (0.053) (0.081)
Interest rate spread -0.066 0.023 0.038 0.020
(one period lagged) (0.062) (0.059) (0.069) (0.061)
Number of total banks 0444" -0.332* 0498"
(per capita) (0.188) (0.160) (0.188)
External borrowing 0.053 -0.037 -0.057 -0.081
(% of GDP) (0.099) (0.0845) (0.128) (0.084)
Trade 0.034 0.070"
(% of GDP) (0.033) (0.031)
Imports of intermediate goods 0.093
(% of GDP) (0.098)
Exports 0.162"
(% of GDP) (0.059)
Index of forex liberalization -0.881 -1.098 -2.525 -0.426
(1.724) (1.234) (1.757) (1.237)
Intermediate exchange rate regime 0.325 0.784 0148 0.518
(2.236) (1.849) (1.895) (1.694)
Floating exchange rate regime -2.249 -1 .705 -1.622 -l.979*
(2.129) (1.253) (1.449) (1.065)
Net foreign reserves 0344'” -0.350** -0.310** 0413*”
(% of GDP) (0.154) (0.126) (0.132) (0.123)
Domestic credit -0.044 -0.080 -0.013 -0.081
(% of GDP) (0.062) (0.055) (0.044) (0.051)
Inﬂation rate 0.018 -0.006 -0.007 -0.006
(0.014) (0.013) (0.017) (0.013)
Maximum historical inﬂation 0004*" 0003*" 0003"" 0003*"
(0.001) (0.001) (0.001) (0.001)
Croatia -4.153 -3.043 -7.l64 -2.175
(5.106) (4.083) (4.065) (4.086)
Trend 0.5742 0.327 0.317 0.395“
(0.378) (0.213) (0.333) (0.208)
_cons 8.098 11.005* 17.845" 8.981
(7.422) (6.022) (6.5 13) (6.312)
Obs. 99 99 94 99
Adjusted R-squared 0.678 0.731 0.765 0.744

 

Standard errors in parentheses (adjusted for heteroskedasticity and serial correlation)

*** Signiﬁcant at 1%; ** signiﬁth at 5%; * signiﬁcant at 10%

108

Table 4a First Difference Estimator

Dependent Variable: Credit Dollarization (% of Total Credit)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(1) (2) (3) (4)
Deposit dollarization 0638*" 0596*" 0.411" 0582*"
(A. of total deposits) (0.133) (0.127) (0.190) (0.133)
Net foreign assets 0175*" 0182*" -0.173* 0186*"
(% of total deposits) (0.043) (0.044) (0.082) (0.043)
Interest rate spread 0.021* -0.005 0.051 0006
(one period lagged) (0.012) (0.008) (0.032) (0.008)
Number of total banks 0.468
(per capita) (0.807)
External borrowing 0.271 0.276 0.292 0.252
(We of GDP) (0.172) (0.175) (0.244) (0.179)
Trade 0.108 0.103
(% of GDP) (0.070) (0.067)
Imports of intermediate goods 0.133
(% of GDP) (0.138)
Exports 0.191
(% of GDP) (0144)
Index of forex liberalization -1.067 0.969 -l.349 0.859
(2.400) (2.269) (4.573) (2.127)
Intermediate exchange rate regime -5.708** -4.058 -0.438 -3.802
(2.491) (2.746) (2.993) (2.638)
Floating exchange rate regime -6.069** -4.451* -3.011 4.358"
(2.322) (2.240) (4.316) (2.085)
Net foreign reserves -0.517* -0.584* -0.234 -0.593*
(% of GDP) (0.267) (0.327) (0.295) (0.318)
Domestic credit -0.239 -0.146 -0.205 -0.166
(% of GDP) (0.178) (0.169) (0.201) (0.164)
Inﬂation rate 0002" 0.003 0025*" 0.002
(0.001) (0.002) (0.007) (0.002)
Maximum historical inﬂation -0.010***
(0.003)
Maximum historical devaluation 0.321" 0.523“ 0.306M
(0.154) (0.208) (0.144)
Trend -0.673 -0.605 -0.102 -0.608
40.460) (0.448) (0.463) (0.440)
_cons 8.762 7.91 1 2.526 7.833
(5.222) (5.036) (5.323) (4.963)
Obs. 1 15 l 15 93 115
Adjusted R-squared 0.413 0.423 0.326 0.419

 

Standard errors in parentheses (adjusted for heteroskedasticity and serial correlation)

*** Significant at 1%; ** signiﬁcant at 5%; * significant at 10%

109

Table 4b First Difference Estimator

Dependent Variable: Credit Dollarization (% of Total Bank Assets)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11) (2) (3) (4)
Deposit dollarization 0.131 0.125 0.113 0.126
(% of total bank assets) (0.117) (0.114) (0.102) (0.112)
Net foreign assets 0228*" -0.244"‘** -0.338*** 0239*“
(% of total bank assets) (0.063) (0.052) (0.051) QIOSOL
Interest rate spread 0.045 0.046 0.047 0.048
(one period lagged) (0.041) (0.043) (0.039) (0.043)
Number of total banks 0.148
(per capita) (0.361)
External borrowing -0.053 -0.027 -0.063 -0.024
(% of GDP) (0.083) (0.109) (0.055) (0.103)
Trade 0008 -0.009
(% of GDP) (0.030) (.028)
Imports of intermediate goods -0.061
(% of GDPL (0.050)
Exports -0.032
(% of GDP) (0.070)
Index of forex liberalization -0.118 -0.027 -1.968 -0.044
(1.279) (1.522) (2.044) (1.547)
Intermediate exchange rate regime 0.373 1.790 1.426 1.916
(1.165) (1.567) (1.337) (1.724)
Floating exchange rate regime 0.355 2.033 1.189 2.081
(1.552) (2.128) (1.904) (2.225)
Net foreign reserves -0.050 -.024 -0.022 -0.033
(% of GDP) (0.080) (0.102) (0.070) (0.082)
Domestic credit 0.097 0.103 0.191" 0.103
(% of GDP) (0.070) (0.075) (0.067) (0.074)
Inﬂation rate -0.011 -0.012 -0.010 -0.013
(0.010) (0.010) (0.007) (0.010)
Maximum historical inﬂation -0.003
(0.002)
Maximum historical devaluation 0.179" 0.180 0188*
(0.083) (0.104) (0.092)
Trend 0.1 10 0.142 0.274 0.155
(0.224) (0.203) (0.195) (0.215)
_cons -0.323 -0.751 -1.893 -0.927
(2.529) (2.226) (2.248) (2.400)
Obs. 83 83 77 83
Adjusted R-squared 0.380 0.388 0.498 0.385

 

Standard errors in parentheses (adjusted for heteroskedasticity and serial correlation)

*** Signiﬁcant at 1%; ** signiﬁcant at 5%; * signiﬁcant at 10%

110

Table 5a Fixed Effects Estimator

Dependent Variable: Credit Dollarization (% of Total Credit)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(1) (2) . (3) (4)
Deposit dollarization 0616*“ 0603*" 0664"" 0565*"
(% of total deposits) (0.188) (0.182) (0.247) (0.175)
Net foreign assets 0176" -0.183“ -0.167* -0.160**
(% of total deposits) (0.086) (0.076) (0.093) (0.074)
Interest rate spread 0.045" 0045*" 0.067 0.044"
(one period lagged) (0.021) (0.016) (0.084) (0.017)
Number of total banks -0.149
(per capita) (1.366)
External borrowing 0.513" 0.513“ 0.444 0.497"
(% of GDP) (0.228) (0.249) (0.268) (0.214)
Trade 0.1 10 0.108
(% of GDP) (0.141) (0.150)
Imports of intermediate goods 0.370
(% of GDP) (0292)
Exports 0.292
(% of GDP) (0.267)
Index of forex liberalization 0.247 1.203 -4.712 -1.456
(3.1 11) (2.928) (6.010) (2.990)
Intermediate exchange rate regime 2.076 3.296 5.173 1.411
(3.949) (4.535) (4.965) (4.075)
Floating exchange rate regime -5.730* -4.560 -4.803 -7.469*
(3.378) (3.825) (6.196) (3.949)
Net foreign reserves -1.285*** -1.352*** 4.293" -1.422“*
(% of GDP) (0.463) (0.506) (0.547) (0.485)
Domestic credit -0.363 -0.343 -0.336 -0.232
CA1 of GDP) (0.311) (0.322) (0.365) (0.287)
Inﬂation rate 0003* -0.002 -0.005 0004*"
(.002) (0.003) (0.022) (0.002)
Maximum historical inﬂation -0.003
(0.006)
Maximum historical devaluation 0.093
(0.181)
Trend 2.433” 2.335“ 2.312“ 2.607"
(1.213) (1.292) (1.294) (1.174)
_cons 9.235 -16.643 19.852 9.194
(26.212) (43.889) (32.271) (20.867)
Obs. 136 136 115 140
Adjusted R-squared 0.874 0.874 0.880 0.877

 

Standard errors in parentheses (adjusted for heteroskedasticity and serial correlation)

*** Significant at 1%; ** signiﬁcant at 5%; * signiﬁcant at 10%

111

Table 5b Fixed Effects Estimator

Dependent Variable: Credit Dollarization (% of Total Bank Assets)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(1) (2) (3), (4)
Deposit dollarization 0.136 0.108 0.137 0.1262
(% of total bank assets) (0.209) (0.224) (0.208) (0.196)
Net foreign assets 0358*” 0317*" -0.406*** 0363*"
(% of total bank assets) (0.127) (0.105) (0.090) (0.119)
Interest rate spread 0.081 0.090 0.113“ 0.071
(one period lagged) (0.062) (0.069) (0.065) (0.061)
Number of total banks 0.072
(per capita) @532)
External borrowing -0.022 -0.018 -0.032 -0.049
(% of GDP) (0.067) (0.090) (0.117) (0.064)
Trade 0.088" 0.078“
(% of GDP) (0.044) (0.044)
Imports of intermediate goods 0.043
(% of GDP) (0.083)
Exports 0.290"
Q/o of GDP) (0.083)
Index of forex liberalization -0.131 -0.142 -4.518 0.025
(2.773) (2.452) (3.107) (2.683)
Intermediate exchange rate regime 0.617 -2.584 3.183* 0.767
(2.330) (2.675) (1.599) (2.255)
Floating exchange rate regime -0.566 -3.932 1.330 -0.622
(2.421) (2.881) (2.260) (2.315)
Net foreign reserves -0.306* -0.208 -0.1 17 -0.320"'*
(% of GDP) (0.164) (0.210) (0.124) (0.157)
Domestic credit -0.083 -0.115 0.005 -0.056
(% of GDP) (0.118) (0.117) (0.088) (0.109)
Inﬂation rate -0.021 -0.021 -0.027 -0.019
(0.014) (0.015) (0.016) (0.014)
Maximum historical inﬂation 0.006 0.007
(0.005) (0.005)
Maximum historical devaluation 0.065
(0.215)
Trend 0.644M 0713*" 0.816” 0.672"
(0.260) (0.270) (0.324) (0.283)
_cons -1.350 -2.913 19.406 -3.782
(6.429) (16.970) (11.681) ((5.273)
Obs. 99 99 94 99
Adjusted R-squared 0.839 0.833 0.868 0.846

 

Standard errors in parentheses (adjusted for heteroskedasticity and serial correlation)

*** Signiﬁcant at 1%; ** signiﬁcant at 5%; * signiﬁcant at 10%

112

Table 6a IV Regressions

Dependent Variable: Credit Dollarization (% of Total Credit)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(1) (2) (3) (4)
Deposit dollarization 0564*“ 0673*" 0.560" 0668“"
(% of total deposits) (0.198) (0.205) (0.219) (0.209)
Net foreign assets 0324*" -0.245*** -0.255*** 0246*"
(% of total deposits) (0.079) (0.062) @073) (0.063)
Interest rate spread 0.032 0.107“ 0.089 0.099“
(one period lagged) (0.046) (0.048) (0.059) (0.047)
Number of total banks -1.l87*** -1.167** -l.l31*"
(per capita) (0.400) (0.414) (0.398)
External borrowing 0.439 0.419 0.054 0.466
(Va of GDP) (0.476) (0.433) (0.45L (0.398)
Trade 0.029 0.044
(% of GDP) (0.085) (0.092)
Imports of intermediate goods 0.376“
(% of GDP) (0.194)
Exports 0.0213
(% of GDP) (0.185)
Index of forex liberalization 5.683 -0.531 0.382 -0.372
(4.172) (4.547) (4.608) (4.362)
Intermediate exchange rate regime -4.067 -8.038 -2.600 -8.768
(5.919) (5.278) (4.788) (5.160)
Floating exchange rate regime -4.498 40.3163“ -5.675 40.906"
(6.985) (5.015) (5.557) (4.982)
Net foreign reserves 0813*" 0836*" -1.207*** 0809*"
(% of GDP) (0.268) (0.292) (0.306) (0.284)
Domestic credit -0.483* -0.255 -0.426* -0.242
(% of GDP) (0.256) (0.191) (0.233) (0.182)
Inﬂation rate 0.007 -0.021 -0.010 -.002
(0.013) (0.013) (9.016) (0.013)
Maximum historical inﬂation -0.000
(0.000)
Croatia -32.391*** -30.240**"‘ -23.396"‘ -30.669***
(9.504) (10.022) (11.232) (9.518)
Trend 2.558“ 2.170 2.558" 2.213
(1.426) (1.467) (1.376) (1.466)
_cons -7.446 21.701 16.402 23.436
(14.833) (16.489) (15.249) (16.877)
Obs. 129 132 109 132
Adjusted R-squared 0.676 0.695 0.733 0.694

 

Standard errors in parentheses (adjusted for heteroskedasticity and serial correlation)

*** Signiﬁcant at 1%; ** signiﬁcant at 5%; * signiﬁcant at 10%

Instrumental variables in ZSLS: one-period lagged net foreign assets and external

borrowing

113

Table 6b IV Regressions

Dependent Variable: Credit Dollarization (% of Total Bank Assets)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(1) (2) (3) (4)
Deposit dollarization 0.242 0.301“ 0355*" 0.297"
(% of total bank assets) (0.160) (0.114) (0.117) (0.111)
Net foreign assets -0.406*** 0431*" -0.557"‘** -0.418***
(% of total bank assets) (0.126) (0.116) (0.104) (0.117)
Interest rate spread -0.038 0.072 0.104" 0.069
(one period lagged) (0.059) (0.054) (0.054) @055)
Number of total banks 0564*" -0.470*** 0604*“
(per capita) (0.124) (0.095) (0.126)
External borrowing 0.084 -0.040 -0.064 -0.075
(% of GDP) (097) (0.069) (0.125) (0.075)
Trade 0.021 0.063M
(% of GDP) (0.029) (0.025)
Imports of intermediate goods 0.106
(% of GDP) (0.079)
Exports 0.138"
(% of GDP) (0.049)
Index of forex liberalization -3.046 -3.311** -3.977* -2.788*
(2.539) (1.353) (1.989) (1.348)
Intermediate exchange rate regime 0.811 1.074 0.632 0.723
(2.522) (1.633) (1.565) (1.557)
Floating exchange rate regime -2.461 -1 .823 -1.225 -2.123
(2.370) (1.380) (1.304) (1.257)
Net foreign reserves -0.316 -0.290* -0.233 0338"
ﬁt of GDP) (0.189) (0.147) (0.154) (0.145)
Domestic credit 0.002 -0.050 -0.035 -0.051
(% of GDP) (0.090) (0.070) (0.059) (0.066)
Inﬂation rate 0.011 -0.018 -0.023* -0.018
(0.012) (0.011) (0.012) (0.011)
Maximum historical inﬂation 0004*" 0003*" 0.003“ 0.003"
@001) (0.0017) (0.001) (0.001)
Croatia -5.630 4.166 -6.907** -3.460
(4.906) (3.240) (2.865) (3.294)
Trend 0.573 0.275 0.272 0.341
(0.364) (0.237) (0.292) (0.231)
_cons 16.722 20.801*** 24.022*** 19.466
(9.985) (4.512) (6.462) (4.580)
Obs. 89 89 85 89
Adjusted R-squared 0.736 0.821 0.821 0.827

 

Standard errors in parentheses (adjusted for heteroskedasticity and serial correlation)

*** Signiﬁcant at 1%; ** signiﬁcant at 5%; * signiﬁcant at 10%

Instrumental variables in ZSLS: one-period lagged net foreign assets and external

borrowing

114

CHAPTER 3

REGULATION, FINANCIAL DEVELOPMENT, AND CAPITAL FLOWS

1. Introduction

Banks play an important role in any economy by channeling savings ﬁom
consumers to producers. In emerging market and transition economies, bank credit is one
of the main sources of external ﬁnance for domestic agents, due to underdeveloped
capital markets. However, there is almost always higher demand than supply of funds,
and thus insufﬁcient bank credit.

Riess et al. (2002) show that the level of intermediation for ten transition
economies, accession countries to the European Union (EU), is low when compared to
developed economies and even to other countries with similar income per capita. These
economies, however, have excess funds and seem unable to allocate them efﬁciently to
domestic agents through the banking sector. Rather, since the mid-19908, the banks have
been net external creditors, as the increase in bank deposits has led to an increase in net
assets with foreign banks. Thus, banks seem unwilling to take more risks by lending

domestically and prefer to deposit the excess funds abroad.

115

Transition economies have different degrees of dollarization of bank deposits and
loans.1 When bank deposits are highly-dollarized and there are restrictions on domestic
lending in dollars, banks might hold foreign assets rather than lend domestically (in
domestic currency) as they want to keep matched foreign exchange positions. Could then
restrictions on domestic dollar lending explain the low level of domestic intermediation?
These regulatory effects have not been previously studied.

This chapter analyzes the effects of the liberalization of foreign exchange and
capital account transactions on capital ﬂows and the domestic level of ﬁnancial
development. Mainly, it disentangles the impact of controls on domestic foreign
exchange operations (borrowing and lending in dollars) and on cross-border transfers, on
banks’ choice between domestic credit and foreign assets.

Dollarization of bank deposits and loans has been long seen as a source of
economic instability, macroeconomic volatility, and potentially increased exposure to
ﬁnancial and currency crises. Consequently, regulators impose restrictions on dollar
deposits and loans from domestic banks to residents in an attempt to discourage the use
of foreign currencies. Restricting lending in dollars might cause banks to substitute
dollar credit by cross-border deposits or loans, instead of substituting it by domestic
currency credit. Consequently, capital outﬂows might increase and the overall level of
credit in the economy might fall.

Honohan and Shi (2002) show, using data on ﬁfty-eight emerging economics for
the years 1990 - 2000, that around half of the domestic dollar deposits are exported

through the banking sector. However, they do not take into account any regulation

 

' Dollarization of deposits and loans occurs when a signiﬁcant share of deposits and loans from domestic
banks to domestic agents are denominated in foreign currencies. I use the term “dollar” to denote any
foreign currency in general.

116

effects. Moreover, restricting dollar deposits might reduce the total level of bank
liabilities, as households choose to hold dollars (the “money under the mattress”
phenomenon) or hold dollar deposits abroad. Garcia-Herrero (1997) argues that allowing
for domestic dollar deposits helped stabilize the economy in Argentina and Paraguay
during their banking crises, while the large offshore dollar deposits of residents
destabilized the economy in Venezuela.

Ize and Levy Yeyati (2003) develop a general equilibrium minimum variance
portfolio model and determine the benchmark level of dollarization as a function of
inﬂation and the real exchange rate volatility. They consider then monetary policy
induced deviations from benchmark dollarization. They theoretically show that capital
ﬂight and disinterrnediation can occur when a tax wedge is introduced, that is when
ﬁnancial intermediation in domestic and foreign currency is not taxed at the same rate.

To date, there are no empirical studies that explicitly consider the effects of
foreign exchange and capital regulation on banks’ choice between domestic credit and
foreign assets. Moreover, there are no empirical studies that evaluate the effects of
controls on foreign exchange operations separately from controls on current and capital
account transactions.

This chapter draws on three literatures. First, it relates to the literature on the
effects of capital account regulation on ﬁnancial development and capital ﬂows. There is
a tendency in this literature to aggregate the exchange and capital controls (denoted
below by capital controls), and hence not to differentiate between restrictions on

domestic operations (between residents) and international operations (between residents

117

and non residents). However, this distinction is necessary if one were to study banks’
portfolio allocation.

Previous studies have found mixed results for the effects of capital controls on
ﬁnancial development.2 Chinn and Ito (2002) ﬁnd a link between capital controls and
ﬁnancial development. However, this link depends on the measures used and the level of
economic development. In a simple correlation analysis on a cross-country sample of
forty-one countries, IMF (1999) ﬁnds that exchange and capital controls are associated
with less domestic credit provided by banks.

Several studies evaluate the effectiveness of capital controls, thus the effects of
controls on the volume and composition of capital inﬂows and outﬂows. Johnston and
Ryan (1994) ﬁnd mixed results for the effects of capital controls on capital outﬂows,
while Montiel and Reinhart (1999) show that capital controls affect the structure but not
the volume of capital inﬂows.

Second, this chapter relates to the literature on the effects of banking regulation
on banks’ portfolio allocation. Several studies evaluate the effects of banking regulation
(such as the capital adequacy requirements) on banks’ choices between alternative
domestic assets (see, for example, Furﬁne (2001)). There are no studies, however, that
evaluate the effects of controls on domestic banks’ borrowing and lending in dollars on

their portfolio allocation.3 A good survey of prudential supervision of foreign exchange

 

2 Regulatory measures on current and capital account transactions and their effects on the economy have
been studied since the 19805. There are studies on the effects of capital controls on economic growth and
macroeconomic volatility (see, for example, Prasad et a1. (2003)), on trade (see Stockrnan and Hernandez
(1988) and IMF (1999)), on the international activities of banks (see Buch (2003)), as well as on the
stability of the ﬁnancial sector, inﬂation, real interest rate, consumption-smoothing behavior, and welfare.

3 The only empirical study I could ﬁnd on restrictions on the domestic use of domestic and foreign
currencies is IMF (2003), which considers the restrictions’ effects on nominal effective exchange rate, and
ﬁnds no statistically signiﬁcant effect.

118

operations in nineteen countries in 1996 and the methods used to manage foreign
exchange risk in the banking sector is Abrams and Beato (1998).4

Third, there are several empirical studies that evaluate determinants of banks’
international activities. Most work in this area focuses on foreign direct investment
(FDI). Furthermore, papers that analyze determinants of banks’ foreign assets refer
mainly to banks from developed countries (mostly US, Germany, and Japan), and
consider mainly factors from the host country. Regulation ﬁom the host country matters
for banks’ holdings of foreign assets, and so do demand conditions in the local markets,
FDI in non banks, and relative returns (see Goldberg and Johnson (1990), Moshirian and
Van der Laan (1998), Miller and Parkhe (1998), and Buch (2000), (2003)).

This chapter’s contribution to the literature is twofold. First, it separates,
theoretically and empirically, the effects of controls on foreign exchange operations and
capital ﬂows, on domestic credit and banks’ foreign assets. It shows. how these effects
can vary, depending on the type of activity, domestic or international, that is restricted.
Second, it examines the determinants of foreign assets for banks in transition economies,
and contributes to the literature on determinants of banks’ international activities. I ﬁnd
that for banks in transition economies, conditions from the home countries, as well as
controls on foreign exchange and capital ﬂows are important determinants of their
decision to go abroad.

I derive a simple international portfolio choice model for a domestic bank (similar
to Buch (2003) and Freixas and Rochet (1997, Chapter 8)). Each period the bank

chooses its level of assets and liabilities to maximize expected proﬁts while minimizing

 

" There is extensive literature on other aspects of banking regulation, such as supervision, regulatory
monitoring, limits on activities that banks can engage in, loan classiﬁcation, regulation of information
disclosure, and their effects on bank stability and development, proﬁtability, and foreign bank entrance.

119

overall portfolio risk. The assets are composed of domestic currency and dollar loans to
residents, and foreign assets. Liabilities consist of domestic currency and dollar deposits
from residents. There are several types of regulation costs the bank incurs: a cost per unit
of dollar deposits taken, a cost per unit of dollar loans granted, and a cost per unit of
foreign assets. The bank chooses between domestic credit and foreign assets based on
expected relative returns and regulatory costs.

The empirical analysis uses a panel dataset for twenty-two transition economies.
Measures of controls on foreign exchange operations include restrictions on domestic
banks’ lending in dollars domestically, restrictions on residents’ accounts in dollars held
domestically, and restrictions on banks’ holdings of local securities denominated in
dollars. The measures of controls on cross-border capital outﬂows include restrictions on
residents’ accounts in dollars held abroad and restrictions on ﬁnancial credit between
residents and nonresidents.

The main contribution of this chapter is to show that foreign exchange controls
and capital ﬂow controls have opposite effects on the level of domestic credit. I show
that liberalization of foreign exchange operations increases the level of domestic credit,
possibly because of keeping domestic capital in the economy, and discouraging capital
ﬂight, while the liberalization of capital ﬂows reduces it. Thus the mixed effects of
capital account controls on ﬁnancial development found in previous studies might be due
to the aggregation of both types of restrictions into one indicator. I also ﬁnd that banks
increase their holdings of foreign assets when capital outﬂows are liberalized.
Furthermore, I ﬁnd strong evidence that the liberalization of domestic lending in dollars

increases banks’ foreign assets. This last effect might be due to banks holding deposits

120

with correspondent foreign banks, which are just the dollar loans they granted to
domestic ﬁrms, and the ﬁrms use them to make payments abroad through the banking
sector.

The remainder of the chapter is organized as follows. Section 2 presents the
theoretical framework. Section 3 outlines the econometric model and the estimation
methods. Section 4 describes the data and the measures of regulation employed. I
present the empirical results in Section 5, while in Section 6 I perform several robustness

tests and discuss possible biases. Section 7 concludes.

2. A Simple International Portfolio Choice Model

I use a simple partial equilibrium, international portfolio choice model for a
domestic bank, similar to Buch (2003) and Freixas and Rochet (1997, Chapter 8). The
model is highly stylized, and it is intended to highlight the different impact of different
restrictions on domestic and international operations on banks’ portfolio choices.

Each period the domestic bank chooses its level of assets and liabilities. The bank
assets are composed of domestic currency and dollar loans to residents, and foreign

assets. Liabilities consist of deposits of residents in domestic currency or dollars
(1) W+D+D*=L+L*+Rt,
where W denotes initial wealth, D (D. ) denotes deposits of residents in domestic

currency (dollars), L (L. ) denotes loans granted to residents, denominated in domestic

currency (dollars), and R,' is foreign assets. I assume that foreign liabilities are either

zero, or supply-deterrnined.

121

At the end of the period, the bank maximizes the following objective ﬁrnction:

(2) MaxE(17)-%702(17).

where E(I7) denotes expected proﬁts, 02(17) is the portfolio risk, and y is the
coefﬁcient of risk aversion.

Banks are assumed to be small; they take the returns on assets and liabilities as
given. The supply of deposits and the demand for loans in both domestic currency and

dollars are assumed inﬁnitely elastic. The gross interest rates paid on domestic currency

~ ~

and dollar deposits are rD and r011 ,5 and the gross returns for domestic currency and

~ ~

dollar loans are r 1, and r 1,8 , respectively. The foreign asset is assumed to be risk-free,

with a return of r*. In addition, there are three types of regulation costs that the bank
incurs: a cost per unit of dollar deposits taken (d D ) , a cost per unit of dollar loans

granted (d L ) , and a cost per unit of foreign assets (a) . Thus, domestic operations in

dollars are taxed at rates d D and d L , while international operations (the majority of

which are in dollars, for the case of emerging market and transition economies) are taxed

at rate a .

Bank’s proﬁts are

~

(3) II=rLL+(r[,* —dL)Lt +(rt -a)R* -rDD-("D* +dD)D*,

Combine Eqs. (1) and (3) to obtain

 

5 I use the symbol ~ to denote random variables, and the symbol - for the mean of a random variable.

Hence, E( r D ) = ; D , and similarly for the other interest costs/returns.

122

(4) 17=(rL —r* +a)L+(rL* ‘dL —r* +a)L‘ +(rt —a)W—(r1) —r. +a)D—
(rut +dD -r* +a)D‘,
Assume further that the variances of interest costs/retums on deposits/loans

denominated in the same currency are equal, and that the variance for domestic currency

returns is higher than that for dollar returns6
(5) 02(rD)=o’z(rL)=a202(rDe)=a202(rLe)=a202,

where a > 1 is a parameter.
Also, assume that the correlation of returns denominated in the same currency is
twice as large as that of returns in different currencies due to currency matching
(6) corr(rD,rDe ) = corr(rD,rLs ) = corr(rL ,rDt- ) = corr(rL,rL4 ) = ,u ,
(7) corr(rD ,rL ) = c0rr(rDs,rLt ) = 2,11 ,
where ,u is a parameter and 02 denotes the variance of interest costs/retums on domestic

currency deposits/loans.

Using Eqs. (2) and (4) — (7), bank’s maximization problem becomes

Mpx ,{(?L —r’ +a)L+(;1,* —dL r’ +a)L’ +(r’ —-a)W—-
D,D ,L,L
(8) (PD-r‘+a)D—(;D*+dD—r*+a)D.—éy0’2
2 2 4
[a’L’+L’ +a’D’+D’ +2,u(—2a2LD+aLL —
aLD’ —aDL’ +aDD’ —2D’L’ )1},

 

6 The assumption of higher uncertainty for domestic currency returns implicitly incorporates the uncertainty
and lack of credibility of domestic policies. Notice, also, that the exchange rate volatility is not explicitly
modeled, as the focus of the model is on regulation costs.

123

The optimal portfolio allocation of a domestic bank is given by the following set

of ﬁrst-order conditions7

(9) ;L_r*+0—y02[a21,+,u(aL*—2a2D-aD*)]=0,
(10) ;L._dL_,-'+a_ygz[L‘+,u(aL—aD—ZD*)]=0,
(11) ;D_r’+a+y02[azD+'u(—2a2L—aL*+aD.)]=0,
(12)

ipa +dD —r’ +a+yo’[D’ +rt(—aL—2L’ +aD)]=O.

Eqs. (9) — (12) show that the equilibrium levels of banks’ assets and liabilities

depend on banks’ degree of risk aversion (y), the parameters and variances and

o I _ *
covarlances between risky assets (a, p, and 02), expected excess returns (rD —r ,

PD: —r*, M —r* , and F]; —r*), and regulatory costs (dD, d), and a). Next, I use
the above equations to evaluate the effects of changes in the regulatory measures of
foreign exchange and capital account operations on banks’ portfolio allocation decisions.

Consider ﬁrst a change in the regulatory cost of dollar loans (d L ) on the total

level of domestic credit (L + L* ) and the level of foreign assets R* . It can be shown that

 

 

(13) £=_[(1+4#)+2#2(1-2ﬂ)<0
adt 702(1+4ﬂ)(1-4#2)
6L u
_= >09
(14) adL ay02(1+4u)

 

7 For the system to have a solution, I have to assume that ,u < 0.5 .

124

a(L+L’) =_[(1-4#2)#-(1+4u)a-2a#2(1-2#)] < 0.
(15) 6‘11. a1a’(1+4tt)(1—4u’)

Hence, the liberalization of foreign exchange lending (a decrease in d L ) increases

the level of dollar loans and decreases domestic currency loans.8 The level of credit in
the economy increases. The direction of change in foreign assets depends on the variance

and covariance parameters

 

an’ era +13) 1

(16) — = —2 — 2 .
adL 5dr. ya (1—2u)

Holdings of foreign assets are more likely to decrease when the correlation

between interest costs and returns for deposits and loans denominated in the same

currency (,u) is small. Then, an increase in dollar loans as a result of foreign exchange

liberalization will trigger a smaller change in dollar deposits. Thus, there will be more
substitution of foreign assets by dollar loans.9

Second, consider a change in the regulatory cost of cross-border ﬂows (a). The

effects on banks’ assets are the following

 

 

6L 6D 1
(17) —=-—-= 2 2 (HM-274a).
6a 6a ya a (1+4u)
at“ 813’ 1
(18) —a =- = , [0(1+2,U)-2#)]>0,
a 6a yo a(1+4p)

 

8 I abstract from any changes in the default risk, and the effects the regulatory measures might have on the
variance of returns on loans.

9 If I instead assume that the cost of dollar instruments is the same, d L = d D , then the liberalization of
foreign exchange operations will still increase domestic credit (by rising the level of dollar loans), while
leaving the level of foreign assets unchanged.

125

8(L+L’)_ 1

a _ 2 2 [(I+a2)(1+2,u)—2,ua)]>0.
(19) a }’0' a ([+4]!)

 

The liberalization of capital ﬂows (a decrease in a) will lead to a decrease in
dollar loans and the overall level of domestic credit. The total level of deposits increases;

hence, foreign assets have to increase, in order for banks to maintain a balanced portfolio

9:
6a

(20) < 0.

To summarize, the two sets of restrictions have opposite effects on domestic
credit. The liberalization of domestic foreign exchange operations increases domestic
credit, while the liberalization of international operations (capital outﬂows) decreases it.
Foreign asset holdings increase when capital outﬂows are liberalized, while the effect of

foreign exchange liberalization depends on model parameters.

3. Empirical Model and Estimation Methods

In order to assess the effects of foreign exchange and capital account regulation

on ﬁnancial development and capital ﬂows, I estimate the following two equations

Domestic Credit ,-,= a, ’ F orex Regulation ;,+ ,8] . Cross-border Flows Regulation g,
+ y; Controls ,,+ 8,1,

(21)

Foreign Assets ,-,= a 2 ’ F orex Regulation ,,+ ,82' Cross-border F lows Regulation g,

(22) + yz Controls 11+ (it.

The two equations are speciﬁed under the assumption that direct intervention on

the foreign exchange market and capital account has a short-run effect (similar to Montiel

126

and Reinhart (1999)) as banks make a one-time adjustment of their portfolio to the policy
change. However, I also test for persistence effects.

Eq. (21) captures the effect of regulatory measures on the level of domestic credit,
while Eq. (22) measures the effects of regulation on bank foreign assets. Forex
Regulation measures the effect of the liberalization of domestic foreign exchange
operations (mainly of dollar deposits and loans from domestic banks). Cross-border
Flows Regulation denotes the measures of liberalization of capital ﬂows, referring mainly
to bank lending and/or holding deposits abroad. Controls denote all other factors that
might affect banks’ decision to lend domestically vs. hold foreign assets, including
relative returns, as they are further described in this section and Section 4. Finally, 8 and
C are disturbance terms.

I estimate each equation separately, using pooled OLS (POLS) as the benchmark.
In order to control for unobserved heterogeneity, I use then the ﬁxed effects (FE), random
effects (RE), and ﬁrst difference (FD) estimators, and compare the results. Since
domestic credit and foreign assets are measured as stock variables and show a high level
of persistence, I expect the ﬁrst difference estimator to provide the most reliable results.

There are several issues that have to be addressed when measuring regulatory
effects: 1) if and how regulation affects foreign banks that operate domestically; 2) which
banks can perform foreign exchange operations; 3) how to separate the regulation effects
from all other possible effects (the identiﬁcation problem); and 4) the possible
endogeneity of regulatory measures. I discuss each of these issues separately.

First, our measures of ﬁnancial development and bank foreign assets refer to all

domestic depository institutions, and thus include foreign banks that operate locally.

127

Since the presence of foreign banks in many transition economies is large10 and rapidly
growing over time, one question that arises when studying regulation effects is if and
how domestic regulatory measures affect foreign banks. Mainly, do regulatory measures
affect both domestic and foreign banks in the same way? According to IIB (2002)11 and
the banking laws in different countries, foreign banks operating domestically are mostly
under local supervision and their activities are regulated by domestic legislation. Thus,
the effects of regulations should affect banks in the same way regardless of their country
of origin.

Second, not all banks are allowed to perform operations in foreign currencies, and
thus take dollar deposits and extend dollar loans. In some countries, a permit or a license
is required for a bank to engage in foreign exchange or international operations. Abrams
and Beato (1998) survey the prudential regulation and management of foreign exchange
rate risk, and show that seven out of the nineteen countries considered have licensing
requirements for banks in order to perform general foreign exchange operations. Ideally,
one would use measures of domestic credit and foreign assets for only the banks that can
engage in foreign exchange and international operations, in order to capture the foreign

exchange regulation effects. However, due to data constraints, 1 use aggregate data for

 

'0 There is a large diversity of patterns for foreign banks’ presence in transition economies. While the
Croatian banking sector is almost entirely foreign-controlled, foreign banks play only a marginal role in
other countries, e.g. Slovenia and Russia. For example, according to BIS (2002), the share of foreign
banks’ local claims in domestic bank credit in 2001 is 68 percent in the Czech Republic, 52 percent in
Poland, 40 percent in Hungary, and only 2 percent in Russia.

'1 According to IIB (2002), foreign banks in the Czech Republic, Estonia, Latvia, and Poland are under the
host country supervision standards. Foreign banks’ subsidiaries and even foreign banks’ branches comply
with the regime applied to domestic banks. Romania, on the other hand, relies on the global supervision by
home country for branches of non-domestic banks. However, in practice, these foreign branches are still
subject to the rules imposed by the National Bank of Romania.

128

the banking sector, and thus implicitly assume that foreign exchange and capital
restrictions affect all banks.‘2

Third, in order to identify the effect of regulation, Incontrol for other factors that
might affect the level of domestic credit and foreign assets, and might be correlated with
the foreign exchange and capital restrictions: (1) demand for credit, (2) banking crises,
(3) the “follow the customer” strategy, and (4) the overall environment in which banks
operate.

Several papers (see, for example, Riess et al. (2002), Hajkova et al. (2002),
Varhegyi (2002), and Hristov and Zaimov (2003)) attribute the low level of credit in
transition economies to banks’ inability to identify proﬁtable investment projects or the
high competition for “good borrowers” from foreign investors/lenders, as well as existent
shortcomings in risk-appraisal and the legal environment. I include several proxies for
the riskiness of domestic lending, ﬁlms’ external borrowing, and enterprise reform and
competition policy.

Banking crises are usually followed by the tightening of the institutional
framework, and thus changes in regulation. I also include an indicator of banking crisis,
to control for the drop in lending due to severe banking distress or bank runs, when all or

most of bank capital has been exhausted. Also, banks can substitute domestic credit by

 

’2 I expect that the banks that are licensed to perform foreign exchange and international operations are the
large banks. Since I use aggregate and not bank-level data, I capture the effects coming from the large
banks anyway. Also, there is evidence that most banks in transition economies are licensed for foreign
exchange operations. In Bulgaria, in 2002, all banks have a foreign exchange license. In Kazakhstan, in
2002, only one out of 42 banks is restricted to operations in domestic currency only. In Macedonia, in
2001, 18 out of 22 banks have a license for international operations. In Russia, in 2000, 764 banks out of
1311 can conduct operations in foreign exchange, while in Ukraine in 1993 and 1999, 115 out of 228, and
153 out of 203, respectively, are licensed for foreign exchange operations. In Yugoslavia, all banks can
perform foreign exchange operations.

129

domestic securities, in addition to foreign assets. Hence, I include a measure of the
relative cost of domestic credit to domestic securities.

Foreign assets of banks from transition economies are mostly composed of
deposits with foreign banks and foreign securities. Thus I can assume a perfectly elastic
supply, which is consistent with the fact that banks in developing economies are small
relative to the other participants on the international ﬁnancial markets. There is also the
possibility that banks’ international activities are determined by the “follow the
customer” strategy. Thus, I include a measure of the outstanding stock of FDI abroad as

a potential determinant of foreign assets.

4. Data

1 use aggregate annual data on twenty-two transition economies from Central and
Eastern Europe and Central Asia. Data for some countries start as early as 1990.

However, for most countries the dataset covers the second half of the 90s (see Table l).

4.1. Regulatory Measures

In 1997, the IMF introduced three measures for controls on exchange, current,
and capital account operations, based on data ﬁom the AREAER (Annual Reports on
Exchange Arrangements and Exchange Restrictions): the index of controls on current
payments and transfers, the index of capital controls, and the index of exchange and
capital controls, which is a combination of the ﬁrst two. These indices include 142
individual types of exchange and capital controls. However, these measures do not

differentiate among restrictions on operations in dollars between residents and operations

130

in dollars between residents and nonresidents. Hence, they cannot be used to estimate the
regulatory effects on banks’ portfolio allocations.

Currently, the most commonly used restrictions on capital account operations are
controls on transactions by commercial banks. For the period 1998 - 2000, 85 percent of
the IMF members had some form of limits on bank cross-border borrowing or lending,
bank holdings of external accounts, and dollar borrowing and lending domestically. To
date, few studies examine the effects of these foreign exchange and capital restrictions on
banks’ portfolio allocation.

Restrictions on dollar lending in emerging economies can take different forms
(see Delgado et al. (2002) and the AREAER). The highly-dollarized economies tend to
have no restrictions (for example, Armenia, Latvia, or the Kyrgyz Republic); the
prohibition of dollar lending, as the other extreme, is very rare (Macedonia, in 1996 -
2000, and Slovenia, in 1991). The requirement that dollar lending has to be approved by
ofﬁcials is most commonly used as a controlling mechanism. Alternative restrictions
involve access to dollar credit limited to ﬁrms with dollar returns and/or dollar payments,
the use of dollar credit to ﬁnance critical imports only, and credit ceilings on individual
banks’ outstanding dollar debt.

I use data from the AREAER, as well as from the central banks’ publications and
banking and foreign exchange legislation, to deﬁne an index of foreign exchange lending
(IXFXLEND), which takes the value 0 if lending in dollars is prohibited, 1 if there are
any restrictions on banks’ lending and ﬁrms’ borrowing in dollars, or if there are
approval requirements, and 2 if there are no restrictions, or if only registration of the loan

is required. Alternatively, I deﬁne a dummy variable (DFXLEND) that takes the value 1

131

if there are no restrictions on domestic banks’ dollar lending, zero otherwise (see Tables
2a and 2b for the description and summary statistics of the variables).

Residents can hold dollar accounts domestically, subject to different restrictions,
such as: approval required for depositing into them, only individuals/enterprises can
maintain them, and only one account per person can be opened. I deﬁne an index of
foreign exchange accounts held domestically (IXRDACC), which takes the value 0 if
dollar accounts are prohibited, 1 if there are any restrictions on the opening of dollar
accounts, or if there are approval requirements, 2 if there are no restrictions, or only
registration of the account is required. Alternatively, I deﬁne a dummy variable
(DRDACC) that takes the value 1 if there are no restrictions on residents’ dollar
accounts, zero otherwise.

Banks might be substituting dollar loans by dollar securities issued domestically.
I use as a control an index of the liberalization of banks’ holdings of local securities
denominated in dollars (IXFXSEC), which takes the value 0 if these holdings are
prohibited, 1 if there are any restrictions or approval requirements, and 2 if there are no
restrictions. Alternatively, I use a dummy variable (DFXSEC), which takes the value 1 if
there are no controls, and zero if there are controls on holdings of domestic dollar
securities.

Similarly, I deﬁne measures of restrictions on capital outﬂows: the index of
foreign exchange accounts held abroad (IXRAACC), which takes the value 0 if these
accounts are prohibited, 1 if there are any restrictions or approval requirements, and 2 if
they are liberalized (or only registration is required) and a dummy variable (DRAACC),

which is 1 if there are no restrictions, and zero if there are restrictions. In addition, I

132

construct a dummy variable for capital outﬂows liberalization (DKOUTFLOW), which
takes the value 1 if either residents’ deposits abroad are restricted, or ﬁnancial credit from
residents to nonresidents is limited, zero if neither is restricted.

Table 3a presents the correlations among the regulatory measures employed in the
analysis. Based on the theoretical model, I expect the liberalization of dollar lending
(IXFXLEND or DFXLEND) to have a positive effect on the overall level of domestic
credit, while the effect on foreign assets is ambiguous. In addition, the liberalization of
capital outﬂows (IXRAACC, DRAACC, or DKOUTFLOW) should decrease domestic
credit and raise foreign assets.

There are several caveats of the regulation measures that have to be at least
mentioned. First, these measures capture the intensity of the controls up to a very limited
point. Second, they are all “de jure” measures, and do not take into account the
enforcement of controls. The regulation effects on credit and foreign assets might be
biased upward for countries with weak enforcement or low circumvention cost. I discuss
possible biases in Section 6. Third, the prohibition of domestic accounts in foreign
exchange seems to be one of the most controversial and poor measure from all the 142
measures used by the IMF, according to the IMF specialists (see HVIF (2003)). Last, it is
possible that countries do not report changes in regulations right away, and thus there
might be a lag in reporting. Also, there is an improvement in countries reporting changes

in regulation and a greater coverage of regulation towards the end of the 90s.

133

4.2. Financial Intermediation and Capital Flows

I use three alternative measures for domestic credit: the ratio of domestic credit to
enterprises to GDP (LTGDP), the ratio of domestic credit to the private sector (including
both enterprises and households) to GDP (DCPR),13 and the ratio of domestic credit to
enterprises to total bank assets (LTTBA). Similarly, I use the ratios of banks’ foreign
assets to GDP and banks’ foreign assets to total bank assets, denoted by FORASSGDP

and FORASSTBA, respectively.”

4.3. Controls

Several controls are used, in addition to the relative returns for bank assets, to
account for changes in domestic credit and foreign assets that are not caused by
restrictions on foreign exchange operations or capital ﬂows. These controls are presented

below. Also, the correlations among the measures employed are presented in Table 3b.

Relative Returns

The relative cost of domestic lending to foreign assets is measured by the external
interest rate spread (ESPREAD), deﬁned as the spread between the average domestic
lending rate and the LIBOR rate. I also use a domestic spread (DSPREAD), measured as

the average lending rate minus the domestic risk-free rate (the three-month Treasury-bill

 

'3 The ﬁrst measure includes credit to state enterprises, and is an imperfect measure of ﬁnancial
intermediation and depth in an economy. The second measure includes only credit to private enterprises.
The drawback of this measure is that an increase might simply capture the privatization of large state
enterprises, rather than the issuance of new loans. Thus, I use both measures and compare the results.

” I use the gross measure for foreign assets, rather than net foreign assets. Buch (2003) argues that gross
measures are more appropriate when studying the effect of regulatory measures, as effects might cancel out
with the net measures.

134

rate, or if this is not available, the money market rate, discount rate, or reﬁnancing rate),
which captures the cost of domestic lending relative to other domestic investment

opportunities.

Demand for Domestic Credit

Changes in the riskiness of domestic lending might cause banks to reallocate their
resources among different investment opportunities. I use the EBRD index of enterprise
reform (IXENTPR), which takes values from 1 (little progress) to 4.3 (substantial
progress; standards and performance of enterprises typical of advanced industrial
countries), to measure the overall development of enterprises. Alternative measures are:
the EBRD index of competition policy (IXCOMPET), similarly deﬁned,15 a dummy
variable DBANKRUPTL, which takes the value 1 if a bankruptcy law was adopted, zero
otherwise, and a measure of the country risk (MOODYS), which is‘an index based on
Moody’s long-terrn foreign currency sovereign rating and takes values from 1 (lowest
country risk) to 21 (highest country risk).

To measure an increase in ﬁrms’ demand for ﬁnance, I use INVESTGDP, which
is the ratio of reinvestment to GDP. I also use the ratio of external credit (credit from
foreign investors) to GDP (NONBGDP), as a measure of ﬁrms’ access to alternative

sources of ﬁnance.

 

‘5 The indices of enterprise reform and competition policy build on the judgment of the EBRD’s Office of
the Chief Economist. The index of enterprise reform includes information on the existence of soft budget
constrains, credit and subsidy policy, enforcement of bankruptcy legislation, competition and corporate
governance, and enterprise investment. The index of competition policy refers to competition legislation,
enforcement and institutions, actions to reduce abuse of market power, and elimination of entry restrictions.

135

 

Banking Crises

Following Caprio and Klingebiel (2003), I include a dummy variable DBCRISIS,
which takes the value 1 in the years when there is a banking crisis, either systemic or
borderline, zero otherwise. A systemic banking crisis is deﬁned as a situation when an
economy faces large-scale ﬁnancial and corporate distress within a short period, and
much or all bank capital is exhausted. A borderline crisis is a non-systemic crisis, such as

bank runs ofindividual banks.l6

“Follow the Customer” Hypothesis
One reason banks might go abroad is to follow their customers. I include the ratio
of total outstanding stock of FDI abroad to GDP (FDIOUTSGDP) to control for this

effect.

Overall Environment in which Banks Operate

Due to the speciﬁcs of the transition economics, the EBRD index of initial
conditions (IXIC) is used to control for the state of the economy at the beginning of the
transition process. This includes the initial level of development, trade dependence on
CMEA (Council for Mutual Economic Assistance), macroeconomic disequilibria,
distance to the EU, natural resource endowments, and state capacity. The index takes
values ﬁom 1 to 4.3, a higher value denoting a more favorable starting position.

Following Barth et al. (2002) and Demirgiic-Kunt and Huizinga (1999), I use the

following controls for the macroeconomic conditions that might inﬂuence banks’

 

'6 Data on banking crises in Kazakhstan and Moldova have been added from Tang et al. (1998). According
to their deﬁnition, a banking crisis is a severe banking distress (large share of non performing loans) and/or
bank runs.

136

activities: the rate of devaluation of the exchange rate (DEVRATE) and alternatively the
inﬂation rate (ICPI), the ratio of banks assets owned by the government to total bank

assets (STATBANK),l7 and the grth rate of GDP per capita (GGDPUSD).

5. Empirical Analysis

Table 3b reports the correlations among the variables employed in the main
speciﬁcation. Notice ﬁrst the high correlation (0.530) between the rate of devaluation of
the exchange rate (DEVRATE) and the interest rate spread (ESPREAD). The spread is
calculated using nominal interest rates, thus it includes the depreciation of the exchange
rate. However, I separately include the rate of devaluation, to capture additional
valuation effects (changes in stock variables due to changes in the exchange rates).

Second, two other independent variables, the index of enterprise reform
(IXENTPR) and the ratio of state-owned bank assets (STATBANK) are highly correlated
with other explanatory variables. As one would expect, the index of enterprise reform is
highly correlated with the index of initial conditions (0.456) and trend (0.520). It is also
correlated with the interest rate spread (-0.458), the rate of devaluation (-O.344), and the
state ownership of banks (-0.355). The state ownership of banks decreases over time as
more banks are privatized (the correlation of STATBANK and TREND is -0.429).

First, I estimate Eqs. (21) and (22) using pooled OLS (POLS). However, since
there are country effects that are probably missing from the speciﬁcation, I also estimate

the two equations using FE, RE, and FD. I choose to present the results for all estimators

 

‘7 This variable is also used by Riess et al. (2002), and described as an indicator of low proﬁtability and
high concentration of the banking sector. Buch (2003) uses it to measure limits on foreign bank entrance.

137

for comparison reasons, and since each has advantages and drawbacks (see Wooldridge
(2002)).

Since I use stock measures, there is persistence in the dependent variables.18
Furthermore, I ﬁnd that the autoregressive coefﬁcients for the changes in the errors, A8,-
and A6", are 0.240, 0.216, and -0.086 for the credit measures, and 0.000 and -0.105 for
the bank asset measures. This suggests there is serial correlation in the initial errors.
Hence, the ﬁrst difference (FD) estimator provides the most reliable results. However, it
also reduces the sample size, and it eliminates the effect of factors that do not vary over
time.19 The ﬁxed effects (FE) estimator uses all the observations in the sample. It tends
to provide large economic effects, but smaller statistical effects. The random effects
(RE) estimator tends to work better than the FE estimator when there is not too much
variation in the explanatory variables. However, it assumes that the independent
variables are not correlated with the errors (the missing country-speciﬁc effect), and this

tends to be false with aggregate data.20

5.1. Effects on Financial Intermediation

Tables 4a — 4c present the effects of the regulatory measures on domestic
ﬁnancial intermediation, using all four speciﬁcations (POLS, FE, RE, and FD). In Table

48, ﬁnancial intermediation is measured as the ratio of domestic credit to enterprises to

 

'8 The autoregressive coefﬁcients for LTGDP, DCPR, LTTBA, FORASSGDP, and FORASSTBA are 0.86,
0.84, 0.79, 0.99, and 0.90, respectively.

‘9 There is more variation across countries than over time in the regulation measures. There are only 6
changes in foreign exchange restrictions over time, and 12 in capital outﬂows restrictions.

2° I perform the Hausman tests for the appropriateness of the RE estimator. For all three measures of
domestic ﬁnancial intermediation, I reject the null hypothesis that the random effects and the regressors are
uncorrelated, while for the two measures of foreign assets, I cannot reject the null hypothesis. Thus, the RE
estimation seems to be appropriate for measuring the regulatory effects on foreign assets, but not on
domestic credit.

138

GDP. Table 4b presents the results for the domestic credit to private sector to GDP.
These two measures capture mostly the magnitude effect (increase or decrease in credit).
I also use the ratio of domestic credit to enterprises to total bank assets, to capture the
substitution of credit with other types of assets effect. These results are presented in
Table 4c.

The liberalization of foreign exchange operations (dollar lending domestically)
has a positive and (mostly) signiﬁcant effect on the level of domestic credit. It increases
the total level of credit by around 2.5 — 6 percentage points and the level of credit to the
private sector by 1 — 5 percentage points (see Tables 4a and 4b). There is a substitution
effect, of other bank assets with domestic credit when dollar lending is liberalized, but the
effect is not statistically signiﬁcant (see Table 4c).

The liberalization of capital ﬂows has a negative, but mostly insigniﬁcant effect
on domestic lending. Notice the very large effect with the POLS estimator in Tables 4a
and 4c, which probably captures missing country effects, since it disappears when I
control for these effects.21

The other variables have the following effects. The spread has a positive and
signiﬁcant effect, as higher return for domestic lending relative to foreign assets will
induce banks to lend more. A 10 percentage points increase in the spread will increase
domestic lending by around 0.1 - 0.4 percentage points, thus the economic effect is
relatively small. The rate of devaluation of the exchange rate has mixed effects, which

vary with the measure and the speciﬁcation used. It has a negative and signiﬁcant effect

 

2' I obtain very similar results if I use IXRAACC instead of DKOUTFLOW. In Table 4a, the effect
becomes statistically signiﬁcant in the RE estimation.

139

on the ratio of credit to total bank assets, which might capture either the substitution
effect, or a valuation effect.

The index of initial conditions matters for the magnitude of domestic credit (see
Tables 4a and 4b). As one would expect, the level of credit in transition economies still
depends on economic conditions when they embarked on the transition path.22 The
banking crisis indicator has a counterintuitive positive effect in the POLS estimation,
while with the FE, RE, and FD estimators it has a negative and sometimes signiﬁcant
effect. The index of enterprise reform, which I use to proxy for the demand for credit, is
positive and highly signiﬁcant with the POLS, and mostly negative with the other
estimators. Since the FE, RE, and FD effects are most likely more consistent than the
POLS effects, then banks seem to lend less domestically as the enterprises become more
competitive. It might be the case that the index is a poor measure of demand conditions,
or that it captures some other effects (see the high correlations with other explanatory
variables, and possibly with missing variables). I present alternative proxies for the

demand conditions in Section 6.

5.2. Effects on Capital Outﬂows

Tables 5a and 5b present the effects of regulatory measures on capital outﬂows. I
present the results for bank foreign assets measured as a ratio to GDP, in Table 5a, and as
a ratio to total bank assets, in Table 5b. Similarly to the domestic credit case, the ﬁrst
measure captures the magnitude, while the second measure captures substitution of

foreign assets by other assets on the bank’s balance sheet.

 

22 Around 37 percent of the variance in LTGDP and DCPR is explained by IXIC alone.

140

The effect of domestic liberalization of dollar lending is surprisingly positive and
very signiﬁcant. Allowing banks to lend in dollars increases the level of domestic credit,
as I show in Section 5.1, but it also increases banks’ holdings of foreign assets. The ratio
of foreign assets to GDP increases by 4 — 5 percentage points (see Table 58), while the
ratio of foreign assets to total bank assets increases by 4 — 8 percentage points (see Table
5b). A possible explanation for this ﬁnding is that after the liberalization of foreign
exchange operations, banks grant dollar loans to domestic ﬁrms that use them to make
payments abroad. Firms make these payments through the domestic banks, which
increase their deposits with foreign correspondent banks, in anticipation of the future
payments. The fact that a high percent of foreign assets are deposits with foreign banks
seems to conﬁrm this hypothesis. Thus, both domestic credit and foreign asset increases
when dollar lending is liberalized.

The liberalization of capital outﬂows reduces banks’ cost of holding foreign
assets. Hence, holdings of foreign assets increase by around 2 — 4 percentage points (see
Tables 5a and 5b).

The interest rate spread decreases foreign asset holdings, as a higher return for
domestic loans relative to foreign assets cause banks to lend more domestically.
However, the economic effect is pretty small: a 10 percentage point increase in the spread
decreases the foreign assets ratio (to GDP and total bank assets, respectively) by 0.2 — 0.5
percentage points.

The index of initial conditions is again signiﬁcant and has a positive effect; more
developed economies seem to have banks with higher holdings of foreign assets.23

Surprisingly, the indicator of banking crisis has no effect on holdings of foreign assets,

 

23 The index alone explains some 35 percent of total variation in the foreign asset ratios.

141

regardless of the measure employed. The share of state-owned assets and the index of
enterprise reform have mixed effects. The trend has a large positive effect on the

measure of magnitude, but no effect on the substitution measure.

6. Alternative Measures and Possible Biases

In this section, I discuss the results obtained using alternative speciﬁcations and
measures for the regulatory indicators and the controls, in order to assess the robustness
of my ﬁndings. I also discuss the issue of endogeneity of the regulatory measures and
possible biases in the estimated effects.24

When using a dummy (DFXLEND) instead of the index of foreign exchange
liberalization (D(FXLEND), results do not change.25 To test for persistence in the
regulatory measures (the measures have a gradual effect), I estimate a regression with all
variables but the regulatory dummies in ﬁrst differences, and ﬁnd no evidence of
persistence in either the domestic credit or the foreign assets. It seems to be the case that
regulations of foreign exchange operations and capital ﬂows have a one-time effect on
banks’ assets.

The index of dollar lending IXFXLEND might be capturing the effect of dollar
deposits liberalization. If this is the case, the positive effect on foreign assets can be
simply explained by banks’ matching of assets and liabilities by currency. I control for
the liberalization of dollar deposits, the level of dollar deposits and the level of dollar

liabilities using the following measures: the index of liberalization of domestic dollar

 

2’ These results are not presented here, but are available upon request.

25 I try using an alternative measure of restrictions on foreign exchange operations, a dummy variable,
which takes the value 1 if there is a tax imposed on foreign exchange operations (and thus, a higher cost of
dollar instruments), and zero otherwise. This has no signiﬁcant effects on banks’ assets.

142

accounts (IXRDACC), the ratio of dollar deposits to total bank assets, and the ratio of
dollar bank liabilities to total bank assets, respectively. The positive effect of
IXFXLEND on domestic credit and foreign assets does notchange. In addition, to check
if the dollar lending regulation effect is not coming ﬁ'om the trade liberalization in
general, I also include the ratios of exports, imports, and total trade (imports plus exports)
to GDP. Although these measures have a positive and signiﬁcant effect on foreign assets,
the regulatory effects of interest remain the same.

I estimate the model replacing the index of enterprise reform by other controls for
the demand for credit. These controls are: the proxy for the country risk (MOODYS), the
measure of the existence of a bankruptcy law (DBANKRUPTL), the index of
competitiveness (IXCOMPET), ﬁrms’ external borrowing (NONBGDP), and the
reinvestment ratio (INVESTGDP). I ﬁnd that a higher country risk will reduce the
domestic lending, while having no signiﬁcant effect on foreign assets.26 Higher demand
for credit by ﬁrms, measured by a higher reinvestment rate, increases the level of
domestic credit and has no effect on foreign assets. However, this variable is most likely
endogenous. Domestic ﬁrms’ credit ﬁ'om foreigners seems to be a complement, and not
a substitute for domestic credit. It also increases the level of foreign assets, since it
probably captures some of the liberalization of capital account Operations. However, this
variable is also most likely endogenous. The bankruptcy law and the index of
competitiveness have no signiﬁcant effect on banks’ assets. In all these alternative
speciﬁcations, the regulatory effects do not change.

I perform the following additional sensitivity tests:

 

2’ The sample size drops by 30 observations when using MOODYS.

143

0 Using the outstanding stock of FDI abroad (FDIOUTSGDP) as a determinant of
foreign assets, to capture the “follow the customer” effect. This variable has
mixed effects, which might be due to endogeneity issues (as it measures the total
stock of F DI, including banks and non banks).
0 Using the inﬂation rate (ICPI) instead of the rate of devaluation of the exchange
rate (DEVRATE).
0 Using the growth rate of GDP (GGDPUSD) to control for the overall environment
in which banks operate.
0 Using year dummies for the time effects, instead of the time trend.
0 Controlling for the substitution of domestic credit with other domestic assets. 1
use the domestic spread instead/in addition to the external spread. I also use the
index of liberalization of the domestic issue of dollar securities (IXFXSEC), as
well as other regulatory measures that might cause substitution between
alternative domestic assets, such as the adoption of a capital adequacy ratio
requirement.
0 Using dummy variables for the countries that have signed an association or
partnership agreement with the EU.
None of these sensitivity tests changes the main results.

Several potential caveats of the regulatory measures deserve some discussion.
First, as I have already mentioned in Section 4, these are “de jure” measures, and they
might be upward biased for countries with weak enforcement or low circumvention cost.
The real regulatory effect for these countries might be smaller than the estimated effect.

Second, these measures might be upward biased towards the end of the sample. Third,

144

there might be a lag in reporting. However, with one-period leads instead of the current
values, the regulation measures are not signiﬁcant anymore. Fourth, I use an unbalanced
panel and there are missing observations for most countries. As long as the reason some
data are missing is not related to the error term, the results will be unbiased. I expect
countries to report changes in regulations, and missing data might indicate that for a
given period there were no changes in regulation. If this is the case, my estimates do not

suffer from a missing data bias.

7. Conclusions

How does liberalization of foreign exchange and capital account transactions
inﬂuence capital ﬂows and the domestic level of ﬁnancial intermediation? Using data on
twenty-two transition economies, I show that the liberalization of foreign exchange
operations increases the level of domestic credit, possibly because of keeping domestic
capital in the economy, and discouraging capital ﬂight, while the liberalization of capital
outﬂows reduces it. I also ﬁnd that banks increase their holdings of foreign assets when
capital outﬂows are liberalized. Surprisingly, the liberalization of domestic lending in
dollars also increases banks’ foreign assets.

Is a dollarized ﬁnancial system better than none? This chapter suggests there is a
disinterrnediation cost associated with restrictions on domestic dollar lending. However,
the question deserves further examination. Policy makers might have to weigh the costs
of ﬁnancial dollarization, such as higher complexity of monetary policy, limited

availability of policy instruments, and possibly higher exposure to ﬁnancial and currency

145

crises, against the beneﬁts of higher ﬁnancial intermediation, and decide if they prefer a

dollarized ﬁnancial system to none.

146

REFERENCES

Abrams, R. K. and P. Beato. 1998. "The Prudential Regulation and Management of
Foreign Exchange Risk." International Monetary Fund Working Paper 98/37.

AREAER. 1990-2002. "Annual Report on Exchange Arrangements and Exchange
Restrictions." International Monetary Fund: Washington, DC.

Barth, J. R., L. G. Dopico, D. E. Nolle, and J. A. Wilcox. 2002. "An International
Comparison and Assessment of the Structure of Bank Supervision." Corporate
Finance Review, 6:6, pp. 9-34.

BIS. 2002. "Quarterly Review: International Banking and Financial Market
Development." March 2002, Bank of International Settlements: Basel,
Switzerland.

Buch, C. M. 2000. "Why Do Banks Go Abroad? - Evidence from German Data. " Journal
of Financial Markets, Instruments and Institutions 9:1, pp. 9-30.

Buch, C. M. 2003. "Information or Regulation: What Drives the International Activities
of Commercial Banks?" Journal of Money, Credit, and Banking, (forthcoming).

Caprio, G. and D. Klingebiel. 2003. "Episodes of Systemic and Borderline Financial
Crises." Mimeo. World Bank: Washington, D. C.

Chinn, M. D. and H. Ito. 2002. "Capital Account Liberalization, Institutions and
Financial Development: Cross Country Evidence." Mimeo. University of
California at Santa Cruz.

Delgado, F. L., D. S. Kanda, G. M. Casselle, and R. A. Morales. 2002. "Domestic
Lending in Foreign Currency." in Building Strong Banks through Surveillance
and Resolution. C. Enoch, D. Marston, and M. Taylor eds. International Monetary
Fund: Washington DC, pp. 40-61.

Demirgiic-Kunt, A. and H. Huizinga. 1999. “Market Discipline and Financial Safety Net
Design.” World Bank Policy Research Working Paper 2183.

EBRD. 1998-2002. "Transition Report." European Bank for Reconstruction and
Development: London.

Freixas, X. and J.-C. Rochet. 1997. Microeconomics of Banking. The MIT Press:
Cambridge, MA.

147

Furﬁne, C. 2001. "Bank Portfolio Allocation: The Impact of Capital Requirements,
Regulatory Monitoring, and Economic Conditions." Journal of Financial Services
Research 20:1, pp. 33-56.

Garcia-Herrero, A. 1997. "Banking Crises in Latin America in the 1990s: Lessons from
Argentina, Paraguay, and Venezuela." International Monetary Fund Working
Paper 97/140.

Goldberg, L. G. and D. Johnson. 1990. " The Determinants of US Banking Activity
Abroad." Journal of International Money and Finance, 9:2, pp. 123-137.

Héjkova, D., J. Hanousek, and L. Nemecek. 2002. "The Czech Republic's Banking
Sector: Emerging from Turbulent Times." European Investment Bank Papers, 7:1,
pp. 55-73.

Honohan, P. and A. Shi. 2002. "Deposit Dollarization and the Financial Sector in
Emerging Economies." World Bank Working Paper 2748.

Hristov, K. and M. Zaimov. 2003. "Shadowing the Euro: Bulgaria's Monetary Policy
Five Years On." Bank of International Settlements Papers, 17, pp. 61-79.

IIB. 2002. "Global Survey of Regulatory and Market Developments in Banking,
Securities and Insurance." Institute of International Bankers: New York, NY.

IMF. 1990-2002. "Annual Report on Exchange Arrangements and Exchange
Restrictions." International Monetary Fund: Washington, DC.

IMF. 1999. " Exchange Rate Arrangements and Currency Convertibility: Developments
and Issues." World Economic and Financial Surveys, International Monetary
Fund: Washington, DC.

IMF. 2003. " Exchange Arrangements and Foreign Exchange Markets: Developments
and Issues." World Economic and Financial Surveys, International Monetary
Fund: Washington, DC.

Ize, A. and E. Levy Yeyati. 2003. "Financial Dollarization." Journal of International
Economics, 59:2, pp. 323-347.

Johnston, R. B. and C. Ryan. 1994. "The Impact of Controls on Capital Movements on
the Private Capital Accounts of Countries’ Balance of Payments: Empirical
Estimates and Policy Implications." International Monetary Fund Working Paper
94/78.

Miller, S. R. and A. Parkhe. 1998. " Patterns in the Expansion of the US. Banks' Foreign
Operations." Journal of International Business Studies, 29:2, pp. 359-390.

148

Montiel, P. and C. M. Reinhart. 1999. "Do Capital Controls and Macroeconomic Policies
Inﬂuence the Volume and Composition of Capital Flows? Evidence from the
1990s." Journal of International Money and Finance, 18:4, pp. 619-635.

Moshirian, F. and A. Van der Laan. 1998. “Trade in Financial Services and the
Determinants of Banks’ Foreign Assets.” Journal of Multinational Financial
Management 8, pp. 23-38.

Prasad, E., K. Rogoff, S. Wei, and M. A. Kose. 2003. "Effects of Financial Globalization
on Developing Countries: Some Empirical Evidence." Mimeo. International
Monetary Fund: Washington, DC.

Riess, A., R. Wagenvoort, and P. Zajc. 2002. "Practice Makes Perfect: A Review of
Banking in Central and Eastern Europe." European Investment Bank Papers, 7:1,
pp. 31-53.

Stockman, A. S. and A. Hernandez. 1988. “Exchange Controls, Capital Controls, and
International Financial Markets.” The American Economic Review, 78:3, pp. 362-
374.

Tang, H., E. Zoli, and I. Klytchnikova. 1998. "Banking Crises in Transition Countries:
Fiscal Costs and Related Issues." World Bank Policy Research Paper 2484.

Vélrhegyi, E. 2002. "Hungary's Banking Sector: Achievements and Challenges."
European Investment Bank Papers, 7:1, pp. 75-89.

Zajc, P. 2002. "Slovenian Banks a Decade Later." European Investment Bank Papers,
7:1, pp. 91-106.

Wooldridge, J. M. 2002. Econometric Analysis of Cross Section and Panel Data. The
MIT Press: Cambridge, MA.

149

Table 1 Country and Regulatory Measures Coverage

Albania
Armenia
Azerbaijan
Bulgaria
Croatia

Czech Republic
Estonia
Georgia
Hungary
Kazakhstan
Kyrgyz Republic
Latvia
Lithuania
Macedonia
Moldova
Poland
Romania
Russia

Slovak Republic
Slovenia
Ukraine

Yugoslavia

1989 - 1990
1996

1997 - 2001
1994 - 2001
1998 - 2001
1996 - 2001
1992 - 2001
1997 - 1998
1996, 2001
1996 - 2001
1998 - 2001
2001

1993 - 2001
1996 - 2001
1996 - 2001
1998 - 2001
1996 - 2001
1996 - 2000

1996 - 2001

1991 — 1992, 1996 - 2001

1997 - 2001

2001

150

Table 2a Coding and Data Sources

 

 

 

Code Description Data Sources

Regulatory Measures

IXFXLEND Index of liberalization of foreign exchange lending AREAER

DFXLEND Dummy for the liberalization of foreign exchange AREAER
lending

IXRDACC Index of liberalization of residents' foreign exchange AREAER
accounts held domestically

DRDACC Dummy for the liberalization of residents' foreign AREAER
exchange accounts held domestically

IXFXSEC Index of liberalization of banks’ holdings of local AREAER
securities denominated in dollars

DFXSEC Dummy for the liberalization of banks’ holdings of AREAER
local securities denominated in dollars

IXRAACC Index of liberalization of residents' foreign exchange AREAER
accounts held abroad

DRAACC Dummy for the liberalization of residents' foreign AREAER
exchange accounts held abroad

DKOUTFLOW Dummy for the liberalization of capital outﬂows AREAER

Financial Intermediation and Capital Flows

LTGDP Domestic credit to enterprises to GDP, % Central Banks’

publications
DCPR Domestic credit to the private sector to GDP, % EBRD
LTTBA Domestic credit to enterprises to total bank assets, % Central Banks’
publications
FORASSGDP Banks’ foreign assets to GDP, % IFS
FORASSTBA Banks’ foreign assets to total bank assets, % IFS

151

Table 2a Coding and Data Sources (cont’d)

 

 

Code Description Data Sources

Controls

ESPREAD Spread between average domestic lending rate and EBRD,
LIBOR rate Economagic

DSPREAD Spread between average lending rate and domestic EBRD, IFS
risk-free rate

IXENTPR Index of enterprise reform EBRD

IXCOMPET Index of competition policy EBRD

DBANKRUPTL Dummy if there is a bankruptcy law EBRD

MOODYS Index of Moody’s long-term foreign currency Moody's
sovereign rating

INVESTGDP Reinvestment to GDP, % EBRD

NONBGDP External borrowing by ﬁrms to GDP, % BIS

DBCRISIS Dummy if there is a banking crisis Caprio&

Klingebiel

FDIOUTSGDP Total outstanding stock of F DI abroad to GDP, % UNCTAD

IXIC Index of initial conditions EBRD

DEVRATE Rate of devaluation of the exchange rate IFS

ICPI Rate of inﬂation IFS

STATBANK Banks assets owned by government to total bank EBRD
assets, %

GGDPUSD Growth rate of GDP per capita IFS

 

152

 

 

 

 

 

 

 

 

 

am a _N on em m: .3.. a;

memo case so; was 32 sec .Em

o _ a N _ SE52

83 $3 83 2: SI :82
acéhDOvnn— “Hug— UHmE— UU<§X~ ﬁlms—NEVA“ 03:25»
a: _ SN ma 2 _ e: I .2... .2
82m 33% Spam 3% $38 Eases—z
43m owed 89¢ 83 33 83552
$2: was 292 News e52 son ._..m
852 E: 3 x 892 men: 5:82
:3: em: 3?: maven cmwé =32
<mkmm§0h mn—UmmAJ—Om ”ma—On <mhﬁs~ awn—9H1— 03:25,

 

 

 

mozmcﬁm Ewan—am an «Emu.

153

8855...:— E mos—aim can gem—atoms: we non—5.2

 

 

88.8.:0 88.8.20 28.8.80 80.8.80 88.8.80 88.8% 98.8.8: 28.8.8:
48.8 88 88 Ed 88 $3 22 83 30.5.5005
88.8.:0 28.8.80 88.8.80 88.8.20 28.8.20 28.8.8: 28.8.8:
888 :2 82 N88 88 82 82 0030
28.8.:0 88.8.80 88.8.20 28.8.20 88.8.8: 28.8.8:
:3 88 Name 88 E8 :2 005.5:
88.8 .80 8 8.8 .80 88.8 .80 88.8 .S 88.8 .3
38 82 RS 88.8 88.8 00895
8:8 .80 88.8 .80 88.8 .8 88.8 .E
88.8 88.8 83 82 00890:
88.8 .80 88.8 .2 0 E 3 .2 c
888 9:8 83 00<SE
88.8 r: s 88.8 .2 0
8.8 888 00¢.me—
88.8.2:
88.8 020.085
003: 00325 czar—02o 08550: 00<$E 004500: :53me nzmdnua .888...

85882 Dean—aom clan-:32 uneﬁa 8:23—9:00 an 0.5—«._.

 

154

$85.53.. 5 83?»..— EE 3.5an3.. ._e .395. Z

 

 

88.8.30 88.8.30 88.8.80 88.8.80 88.8.80 88.180 88.8.80 8:82:
8.2 84.? 8:8 824 824 88.8 88 83 ESE.
88.8.80 88.8.80 88.8.80 88.8.8: 88.8.80 88.8.80 843.2:
an? 88 van? was. 84.8 Ed :8 «.2250
88.8.8: 88.8.8: 88.8.8: 88.8.8: 88.8.8: 88.8.80
8.3 88 88 888 8a? 888 0.725.545
88.8.80 88.8.8: 88.8.80 88.8.80 88.8.0:
82.8. 28.8. :88 Rod 88.? Ems—0m:
88.8 .80 88.8 .8: 82.8 .8: 83.8 .8:
82 82.9 48.8- 8:8 8:.ng
89.8 .8: 82.8 .8: :88 .2 0
~88. 8:.9 8:? 345—me
88.8 .80 88.8 .2 a
$8- :88- 000
88.8.8:
83 BQESOVE
€0.sz Mz<mpﬁm mam—M00: 0.59585 355mm 00: BQESOMG Aims—02x— 8:88.:

833...; 05—05920:— ﬁa—Z waeﬁa «noun—oteo an «Bun.

 

155

8.: 2. 2282...».a a 2.8m 2. 2582...“... a... .8. 2. 2.32285 3..
5h 15. mac.— ..8 Hera—9:3 _.._.—8 can 533308—8232— ..8 008305 8855...:— 5 nae-:0 0.30.35
.ma 1:2 H...— 28 tones—8rd _...-.30 .e .52 meh 28 0932—8.”— 183:3.

 

 

 

 

 

 

 

 

 

 

 

 

0.3 5.8 2.8 88.8 2228.2
8 8 8 8 .30
82.0 88.0 8430 80...:
Sn. 8.. 0...: 88.8- 2.81
88.0 800 890 88.0
. .2.- 8.- 8m. 83 .25:-
880 88.0 88.0 88.0
8...- 88... new? .820 2.37.5:
88.0 88.0 :80 88.0
48. 08.. e8. .58- 0.72.0025
8.5 888% 88.0 88.0
.8.- em.... .9480 :8: man—023
88.0 88.0 88.0 88.0
.88.- a8.- 8. Rod- Ham-gm:
88.0 88.0 98.9 88.0
:88. .8. m8. .388 3.2080
88.0 83.0
_......8N.m :23 000
88.0 88.0 88.0 98.0
88.0 83- $ . .m- .88.. .. BOA-:03:—
830 93.0 88.0 80.0
23 .83. :83 an...” 020-0900
E— m: a... 30..

£20 2. 80 280 22.2.8: 8.8.29 8888..

we n._.—_.._.

156

.8. 2. 335:»... . :8 2. 335:“: 2. .8. 2. 33588 .3
an 8:. mac.— ..8 5.2.—9:8 _...-.8 8:. 3.939.603.8303 .8 6325.... 8855...:— S «.3.-8 _......aaam
.5. ..:.. m... 2.. 323...... :35 A... :5. mac. .2.. 3.2.8... 3.2.8...

 

 

 

 

 

 

 

 

 

 

 

 

8.8 8.8 .8... 88... 3.2.3...
8 .8 .8 .8 .25
8.8.0 8:0 8.2.0 88.8..
888 .8... :8. m. 9. .8...- 8:3:
9.8.. 8...... 98.0 9.8..
8.0- 8...- ...8. 88.- 97.5....
88.0 98.0 88.0 85.0
.88- 88.- ..8887 .88.... «37......
88.0 98.. 88.. 88.0
.8. .8.- .8. 8..- 87.5.58..-
880 8883 88.0 88.0
88..- :3.- ..80 88.. 88..qu
88.. 8.3 8.3 8:...
.8. N8. 88. 8.... ”.835...—
880 88.. 9 8.. 88.0
:88. .8. .8. 8.... 3.5.8...
98.. 9.8...
***026 ..:.:..mN~.m DE—
880 88.0 88.0 98.0
88... 88..- 88...- .80 Boa-.59.:
88... 88.0 88.0 8.8.0
_....88 8.8... .88... 80. 87.qu5..
P... H... m... 8.0..

8.... 2. 8. 2.33m .23.... 2.. 3 23.0 332:... 322......» .3333:

5. 038,—.

157

$8 a 595:3“ .. 2% a £85.53 2. 2.x; a 38:35 .3
3h can mach .3.. 5539-58 3.3m can manta—33.838.— ..8 Bung—H13 8855...:— 5 «._ch Ewe—35
.3 can a .3.— eouaavvrm Enos .E can mac.— ..ou 3.3.—caug— Banana

 

 

 

 

 

 

 

 

 

 

 

 

@33- iam .55 a. $3 :85 2.853 33529 32.5%:

2. as;

3.3 $3 83 $3 2.23.:
8 E K K .25
$3.3 9.1.5 8.2.5 $23
33- 803 338$ an: 28:
$3 $23 85 33.3
Re. :3;- 5- :3.- nzmmh
§ _.3 62.3 28.3 886 3
Mann. EB- ._.-5.3- E: ”_._.—.sz
$8.3 A83 53 £23
”8. m8. :5. 33.. uz<mh<em
$3. 3 23.3 3&3 59m 3
ex.- 52- 82- £2 Ems—Um:
88.3 98.3 3.8.3 $8.3
E... :33.- ..53- .33.- ”_._.—.352
3.83 $8.3 $8.3 8.8.3
**mmo. *mvo. *Nvo. _.._..Vmo. gag—mm
$3.: $53
52- £3. Dun
A213 8:3 €83 €83
3M:- 3? 3m. 23%.:- Bod-535
ASN.3 98.3 28.3 92.3
3v. 3.? a3. mmw.m azmdpu:
E. .5. mm mac.—

158

 

$3 a 385:3... .. 2x3. 3 335:2... r. 2x; a 2.85:5 .1.

Sr.— E—a was: .2... acts—9:3 3.3 E: 566868.858.— ..8 683533 8855...:— 5 29:0 FEE—Sm

.5— an: E ._8 3353.2 .1230 A; can m-HOA .3.. 3.3.—awry— 183:3

 

 

 

 

 

 

 

 

 

 

 

 

33 SE. was Rm... vegan-m
ﬂ 3 3 3 .25
$8.: 833 9:23 638
5:- £35- .55- 83- 28..
9:3 :83 82.3 53
. wON. ***-Van. 3.30:. ***hww. ﬁlmy—H.
82.3 33.3 62.3 98.3
9&- ea. 8”. SE E57352
A23 $8.3 $8.3 a8. 3
:o. 38. .8. NS.- Mania-m
$33 $3 £83 93.3
8m.- SN: 85. SE mas—um:
A83 $8.3 $8.3 489W
coo. 8o. 8o. :o. 559m:
A83 8:: 2 :3 38.3
twee.- voo- 8o.- Nsc 333mm
38.3 ~33
***omw.— ***Vawé DE
:33 En: 3.3.3 €33
3.1%th .1526 **mmc.m wmhﬁ BCQu—HDCMG
83.3 $3.3 $2.3 $2.3
_.._....ﬂamwéﬂ 3.2.. g de _.._.... :56 www. GZHENH
P..— mm mm 39.

Enw a. $3 333 nuts-a .938 3:35» 2.2.595

am «Bah.

159

$8 3 33552 .. ”exam 3 385:»: .3 2x; 2. 23555 ._.:

a..— 65 man: 28 nous—0.23 3.28 can 533333—8232— ..8 $325.3 $855.83 5 2.2.8 _...-65$
.8— .23 mm .8 2.253. .395 .E 2: m5.— 5 83.52: 82...:

 

 

 

 

 

 

 

 

 

 

 

 

 

33 e2; :3 33 2:23.”—
3 K K K .25
233 68.3 22.3 @333
mamm- 33 83 _- ......Emdm 2:5:
283 $3 233 22.3
5.. cs. 5:. NK. 272-5....
$8.8 28.3 233 83.3
23 «K. .5; 1.2322- ”32.7252
223 28.3 K83 23.3
.93. mg. as. ~23.- Mama-«hm
2:3 $8.3 28.3 6K3
3a.- 8: 82 $2.- m-mEomn
28.3 25.3 2:: 2:3
to. coo. 25.- _.._.-so. 5555
28.3 28.3 28.3 2.8.3
***Vmor _.. :uor GNOV ***vmor gamma
KR: 223
82- .23 25
K8: 23.3 amnmﬂ 28.3
52.2295.“q aawﬁ Nhod ***ONNdﬁ BOéHDOv—Q
23.3 28.3 83.3 22.3
3.....ng *0mm.v **mom.m mm”? ﬁlm—g—
nu mm mm 32

2883.. “._.—am :35. .5 XL 3023‘ aye-8m .3.—am ”035...; .5653:

am 2an

160