Content uploaded by Asli Demirguc-Kunt
Author content
All content in this area was uploaded by Asli Demirguc-Kunt
Content may be subject to copyright.
Monitoring Banking Sector Fragility
By Aykut Kibritçiog˘lu*
The Arab Bank
REVIEW
Vol. 5, No. 2
October 2003
Banking
Management
Introduction
The last two decades of the 20th century are mainly character-
ized by currency and banking crises. Particularly the 1994–95
Mexican, 1997–98 Asian and 1998 Russian financial crises
seem to have strongly stimulated academic research on the tim-
ing, duration, causes, effects, and cures of both currency and
banking crises. As also defined in the International Monetary
Fund’s World Economic Outlook (May 1998, p. 74), a currency
crisis may be said to occur when a speculative attack on the
exchange value of a national currency results in a devaluation
(or sharp depreciation) of the currency, or forces the authorities
to defend the currency by expending large volumes of interna-
tional reserves or by sharply raising interest rates. Accordingly,
many researchers construct a monthly or quarterly index,
which is called a foreign-exchange-market pressure (FEMP)
index, to identify and predict currency crisis episodes.
Typically, a FEMP index is calculated as the weighted average
of (1) percentage changes in nominal exchange rates, (2) neg-
ative of percentage changes in foreign exchange reserves, and
(3) international interest rate differential.1A crisis is then said
to arise when the index value exceeds an arbitrarily determined
threshold value. In this sense, the identification of currency
crisis episodes using the so-called FEMP index is an easy and
highly mechanical task, and it is highly useful in empirical
research.
A bank failure, on the other hand, refers to a situation in
which the excessively rising liquidity, credit, interest-rate, or
exchange-rate risk pushes the bank to suspend the internal con-
vertibility of its liabilities. If the bank failure problem under-
mines an entire banking system, the crisis turns out to be sys-
temic. Potential or actual difficulties in the domestic banking
sector sometimes may force the government to intervene into
the market to prevent their far reaching adverse effects, such as
that on the corporate sector and foreign exchange market. The
exact timing of government intervention and the extent of the
possible bailout costs obviously vary according to time and
space.
The last two decades have seen a dramatic increase of systemic
banking crises, as documented mainly by the comprehensive
studies of Caprio and Klingebiel (1996, 1999, 2002 and 2003)2
and Lindgren et al. (1996). The identified domestic crisis
episodes are recently reviewed and discussed by Frydl (1999),
Eichengreen and Arteta (2000) and Boyd et al. (2001), among
others. In the literature, it is usually argued that, contrary to the
case of currency crises, building a time series index to identify
banking crisis episodes is highly difficult, particularly because
of the lack of reliable sectoral data on banks’ financial activi-
ties. It is frequently stressed that the data on non-performing
loans in many countries are either not available or are system-
atically distorted (see Hawkins and Klau, 2000). Consequently,
existing methods that are widely used to pinpoint banking cri-
sis episodes are generally event-based. That is to say, they usu-
ally are based on the available ex-post figures, which are relat-
ed to banks’ losses and governments’ bailout costs. The years
attached to crises reviewed in the literature are those, which are
more or less generally accepted by finance experts familiar
with the countries (Caprio and Klingebiel, 2003). Furthermore,
in many studies, the crisis episodes are also identified with
the help of the country-specific banking information that is
available in the databases of some international financial
organizations (e.g., Bank for International Settlements, the
International Monetary Fund, and the World Bank) or that is
published in major daily newspapers or popular economy jour-
nals (e.g., Wall Street Journal, New York Times, and American
Banker).
51
Event-Based Approach
- It is relatively easy to find information on the date of
both government intervention and change in banking reg-
ulations.
- Pinpointing the dates of crises is possible only for the
annual data frequency. Therefore, it is not useful to apply
it to predict and discuss banking crises within the frame-
work of monthly data frequency.
- In general, the usage of crisis years is restricted to limit-
ed dependent variable models (logit, probit, etc.). Hence,
it usually requires a multi-country framework to increase
the number of crises considered.
- The date of government intervention, which is used to
pinpoint crisis dates, does not necessarily reflect the
actual beginning date of a particular crisis.
- It is not always easy to judge whether a crisis is systemic
or not, particularly if one uses information only on gov-
ernment intervention.
- For an individual researcher, it is not easy to collect
event-based information on banking sector difficulties
across the world.
Statistical Approach Employed in This Study
- The banking-sector-fragility (BSF) index is very useful
to monitor and interpret the developments in the sector
by using monthly banking data.
- The monthly BSF index can easily be employed within a
single-country framework.
- One can easily define criteria to differentiate between
systemic and non-systemic crises or fragility, based on
the fluctuations in the BSF index.
- Reliable and continuous monthly banking sector data for
every country in the world is not available.
- Some of the data may be biased because of the wrong
reporting practices (as in the case of non-performing
loans or interest rates) or country-specific legal regula-
tions.
- The data, and hence, the BSF index do not necessarily
reflects the exact date of government intervention.
Advantages
Disadvantages
Table 1
Comparison of Different Methods to Identify Episodes of Banking Crises and High Banking Fragility
The Arab Bank
REVIEW
Vol. 5, No. 2
October 2003
Banking
Management
The event-based mainstream approach, however, clearly has
some disadvantages against the statistical approach that may be
employed by constructing a banking sector fragility (BSF)
index using the available monthly time-series data. Table 1
summarizes and compares both the advantages and disadvan-
tages of these two methods. Apparently, the statistical approach
has some overwhelming advantages with respect to the widely
used event-based approach. Particularly, the measurement
of banking sector vulnerability by using "monthly" data is a
highly attractive feature of the time-series-based statistical
approach. A monthly BSF index may significantly contribute
to policy makers’ efforts towards early detection of approach-
ing banking sector difficulties. It also can be used as a reliable
method to identify crisis episodes, even though it cannot com-
pletely substitute the event-based approach. Departing from
this idea, this paper attempts to propose a weighted BSF index
to measure and monitor changes in the banking sector fragility
by using monthly data for selected countries. Despite the
above-mentioned arguments on the scarcity of some relevant
banking sector data, we aim to show that even the existing data
for national banking sectors definitely allows us to work with
"monthly statistics" instead of "events", if one intends to pin-
point banking crisis episodes in different countries.
The rest of the paper is organized as follows. In section 2, the
need for and possibility of constructing an index to measure
the tendency towards crisis in the banking system are demon-
strated in detail. The discussions there are primarily based on a
brief literature review. Conceptually, section 2 focuses on the
various financial risks that banks face. An empirically func-
tional BSF index is then explicitly formulated in section 3.
Section 4 covers both the statistical details of calculations and
visual presentations of the BSF indices for each of the selected
22 countries. Finally, in section 5, the results of the current
study are briefly evaluated with respect to previous, i.e. event-
based studies, such as Caprio and Klingebiel (2003) and
Lindgren et al. (1996).
Banks’ Net Worth, Economic Risks, and Potential
Fragility-Indicators
Banks are intermediaries, which aim to earn profits in f inan-
cial markets by acquiring funds, and investing these funds or
lending them to borrowers. Banks’ liabilities are the funds that
they acquire from savers in the form of deposits or as borrow-
ings, while their assets mainly include reserves, marketable
securities, and loans. The difference between the assets and lia-
bilities of a bank equals its net worth, which in fact shows the
bank’s remaining value, or equity capital, after it has met all of
its liabilities.3When the net worth of a bank turns into negative,
the bank becomes insolvent.
52
Explicitly, a bank is exposed to the risk that the values of its
assets and/or liabilities change in financial markets. That is, all
banks are potentially exposed to different types of economic
risks, such as (i) liquidity risk (i.e., massive bank runs), (ii)
credit risk (i.e., rising non-performing loans), and (iii)
exchange-rate risk (i.e., banks’ increasing unhedged foreign
currency liabilities).
Therefore, a bank’s net worth, and hence, a bank failure basi-
cally can be associated with excessive risk-taking of bank man-
agers. In fact, several empirical studies in the literature show
that massive bank runs and withdrawals, enormous lending
booms, and/or high increases in the foreign liabilities of the
banking sector are among the major leading indicators of
impending banking crises.4
1. Bank Runs and Liquidity Risk
No matter what the reason, savers’ massive run on deposits
may indeed trigger a new (or accelerate the ongoing) increase
in the fragility of the banking sector to crisis. However, it
should be noted that the presence of a so-called deposit insur-
ance system may prevent depositors from withdrawals, and
hence, this may significantly weaken the potential link between
bank runs and bank insolvency. Furthermore, Kaminsky and
Reinhart (1999) argue that "recent" banking problems world-
wide do arise from the assets side (i.e., increases in non-per-
forming loans) instead of the liability side (i.e., bank runs).
2. Lending Boom, Non-Performing Loans, and Credit Risk
A lending boom on the assets side of a bank’s balance sheet is
likely to be caused by the bank’s poor, or over-optimistic, eval-
uation regarding the investors’credit applications. Moreover, a
bank can credit risky projects (and thus, it may contribute to a
possible credit-boom process in the country), if the borrower is
an economic unit, which actually is somehow connected with
the bank. This is called insider, or connected, lending in the lit-
erature. Additionally, the existence of deposit insurance may
encourage bank managers to take excessive risk (moral hazard
problem) by loosening the credit taps further than expected.
These considerations imply that credit booms easily may be
linked to banking crises, at least at the theoretical level. How-
ever, Gourinchas et al. (2001) recently emphasized that, while
most banking crises may be preceded by a lending boom, most
lending booms are not followed by a banking crisis.5
3. Banks’ Unhedged Foreign Liabilities, Devaluation, and
Exchange-Rate Risk
Kaminsky and Reinhart (1999) present one of the broadest
frameworks to discuss the potential links between banking and
currency crises. Referring to crises since the early 1980s, they
briefly argue that problems in the banking sector typically pre-
cede a currency crisis but they are not necessarily the immedi-
ate cause of currency crises. In turn, however, the currency cri-
sis deepens the banking crisis, activating a vicious spiral.
Obviously, in the absence of regulations limiting banks’ open
foreign currency positions and if the domestic currency is
not expected to depreciate (or to be devaluated) in the near
future, banks are likely to be motivated to take excessive risk
by acquiring funds from international financial markets. If
domestic banks have large amounts of unhedged foreign cur-
rency debt, then a sudden devaluation may cause a sharp fall in
the net worth of banks thereby increasing the vulnerability of
the domestic banking sector. Therefore, banks may try to
reduce their foreign currency liabilities, if they foresee that the
domestic currency will be devaluated soon. Accordingly, they
also may attempt to reduce the high debt burden by increasing
the credit interest rates.6Hence, bank credits to the private sec-
tor may considerably decline in the aftermath of devaluation.
Furthermore, devaluation expectations and/or rises (falls) in
foreign (domestic) interest rates may trigger a massive bank
run, as also discussed by Calvo et al. (1994) and Obstfeld and
Rogoff (1995).
To examine the causes of banking crises empirically or to
develop a model to monitor and predict impending banking
sector problems, one first needs to be able to empirically iden-
tify the episodes and severity of previously occurred crises.
Our discussions so far show that there is a strong motivation to
design an empirically functional BSF index which is able to
reflect the changes in the excessive risk-taking behavior of
banks for monthly data frequency. Therefore, the next section
is devoted to the creation of this type of index.
Construction of a Monthly Banking Sector Fragility
Index
In this section, we begin to describe how a monthly BSF index
can be constructed, and how it can be used to decide whether a
national banking system is/was in crisis at a particular point in
time. The brief discussions in the previous sections indicate
that there are mainly three7leading sectoral indicators of bank-
ing crises, which may be used in construction of a BSF index:
(i) bank deposits, (ii) bank claims on (or credits to) the domes-
tic private sector, and (iii) foreign liabilities of banks. These
three variables are proxies, or indirect indicators, of changes
in the liquidity risk, credit risk and exchange rate risk in the
The Arab Bank
REVIEW
Vol. 5, No. 2
October 2003
Banking
Management
53
banking sector, respectively. In other words, the fluctuations in
these indicators are supposed to represent the changes in the
fragility of banking sector in any country. Therefore, consider-
ing the economic risks related to banks’ balance sheets, we pro-
pose the following general index (BSF3) to measure the fragili-
ty of banks to crisis by using monthly banking sector data:
(1)
where
(2)
(3)
and
(4)
In equation (1), the BSF3 index is defined as an average of
standardized8values of CPS, FL and DEP, where µand σstand
for the arithmetic average and standard deviation of these three
variables, respectively. In equations (2), (3) and (4), LCPS,
LFL and LDEP represent banking system’s total real claims
on the private sector, the banks’ real foreign liabilities, and the
total real deposits on banks, respectively. That is, CPS, FL and
DEP are simply the corresponding annual changes in each and
every one of these three variables. By using 12-month percent
changes in the monthly data instead of using monthly changes,
we avoid any seasonality, which may be incorporated into the
data. We also hope to be kept away from the risk of deriving
misleading interpretations, if we would consider simply
month-to-month changes. Indeed, banking crises should be
those types of far reaching financial difficulties that cannot be
signaled simply by "monthly" fluctuations in banking vari-
ables, such as the bank deposits, claims on private sectors, or
foreign liabilities. They must be caused by longer term and
powerful deteriorations in the banking sector.
The BSF3 index is proposed to measure the ups and downs in
the domestic banking sector.9Its mean for the sample period is
equal to zero, as implied by equation (1) above. As long as the
BSF3 does not deviate significantly from zero, historically
there is no reason to expect a severe banking sector problem in
the short run. Evidently every deep banking crisis is preceded
by a relatively significant increase in the BSF3 index, which
actually corresponds to a large extent to the excessive risk-tak-
ing behavior of banks, and hence, to an early period of increas-
ing possibility of crisis in the banking sector. This early warn-
ing phase of any approaching crisis is then followed by a rapid
decrease in the value of the BSF3, which in turn can be associ-
ated with substantial falls (i) in bank deposits (bank with-
drawals), (ii) in claims to private sector (as a response to sig-
nificant increases in non-performing loans), and/or (iii) in for-
eign liabilities (particularly in the face of an actual or potential
depreciation in the domestic currency). In this sense, it is obvi-
ous that a coincidence of these three events would enhance the
severity of the impending banking sector problem. The sudden
change in the pattern of risk-taking behavior of banks, or the
substantial fall in the BSF3 following an enormous increase,
may be triggered by a country-specific event, such as a politi-
cal scandal or individual failure of a major bank.
Every fall in the BSF index, on the other hand, does not neces-
sarily imply that a banking system is moving into a deep sys-
temic crisis. Therefore, we differentiate here between medium
and high fragility episodes by defining two arbitrary thresh-
olds. In this study, a national banking system is supposed to be
in a medium fragility period, if the value of the BSF3 index is
between 0 and –0.5:
0 > BSF3t> -0.5 (5)
If, however, the value of the BSF3 index is equal to or lower
than –0.5, we assume that the relevant banking sector is highly
fragile to systemic crisis:
–0.5 ≥BSF3t(6)
Accordingly, a banking system is only accepted to be fully
recovered from crisis when the BSF index reaches its sample
period average (i.e., zero) again.
The cyclical time pattern of a hypothetical banking crisis and
its five successive stages described above are summarized in
table 2 and illustrated in figure 1. In terms of the thresholds
defined in equations (5) and (6), we expect now that banking
crises which are identified in event-based studies mentioned
above occur in high fragility periods determined by our esti-
mations given below.
The Arab Bank
REVIEW
Vol. 5, No. 2
October 2003
Banking
Management
54
Figure 1
Time Path of the BSF Index and Five Phases of a
Hypothetical Banking Crisis
Note: For interpretation of both changes in the BSF index and phases
of crisis, see table 2 and the related part of the text. Clearly, it can
be accepted that this recovery period starts in some cases as the
BSF is increasing but is still below -0.5.
Before we proceed to the presentation of empirical results, we
shortly define an alternative index of banking sector fragility,
BSF2, to test the idea that bank runs do not play a major role
in modern banking crises:
(1’)
The BSF2 index above is simply calculated by omitting the role
of changes in real bank deposits on banks’ financial fragility,
and thus any deviation of the BSF2 from the BSF3 will help us
in understanding the relative importance of bank runs in bank-
ing crises.
Empirical Results
The BSF3 and BSF2 indices proposed above are calculated for
each of the selected 22 countries from which we know that they
experienced systemic, or at least significant, banking sector
problems within the last three decades. To ensure the interna-
tional comparability we use the International Monetary Fund’s
International Financial Statistics (IFS) database (CD-ROM
version, July 2003) as the common data source: LCPS is taken
from IFS’s line 22D, while LFL is taken from line 26C. Finally,
LDEP is considered as the sum of lines 24 and 25 in the IFS.
Notice that nominal series are deflated by using the correspon-
ding domestic consumer price index (CPI). If the CPI data
(IFS line 64), however, is not available for a particular country,
the wholesale, or producer, price index (IFS line 63) is used to
deflate the relevant nominal time series.
The Arab Bank
REVIEW
Vol. 5, No. 2
October 2003
Banking
Management
55
Table 2
Changes in the BSF Index and the Five Phases of a Hypothetical Baking Crisis
Banks’ Behavior Direction of the Banking Fragility Probability of Approaching
Change in the BSF Index Banking Crisis
Phase 1 excessively risk taking increases significantly falls * the probability starts
above zero (optimistic, or boom, phase) to increase *
Phase 2 generally risk avoiding suddenly begins to decrease starts to increase it increases furthermore
(probably panic arises)
Phase 3 risk avoiding falls below zero (but it’s increases significantly system is approaching
still above 0.5) (medium fragility) the borderline to crisis
Phase 4 risk avoiding falls below –0.5 continues to increase most probably, a crisis
(high fragility) occurs in this phase
Phase 5 gradually they start to increases towards zero it falls again (recovery period) crisis is over if the BSF is very
take risk again close or equal to zero again
* Although increases in the BSF index imply a fall in fragility in the short run, it actually must be interpreted as an alarming indi-
cator for impending crisis, if the increase in the index is significant and continues for a while. Hence, the probability of crisis
starts to increase in this initial phase, since banks’ take excessive risks during that period of time.
Value
of the
BSF
Index
Time
0
- 0.5
BSF Index
Phase 1 Phase 4P
h
a
s
e
2
P
h
a
s
e
3
P
h
a
s
e
5
The list of countries considered in this study, corresponding
sample periods imposed by the availability of reliable country
data, and country-specific standard deviations in CPS, FL,
DEP and inflation rates are all given in table 3. Note that the
standard deviation figures in this table show that, for each and
every one of the countries covered here, the FL variable has the
highest volatility among the three variables.
As also mentioned above, we use the standardized values of the
three variables in construction of the BSF3 to avoid the possi-
bility that one of the three components dominates the BSF3
index. Therefore, one may think that there is nothing wrong
with the fact that the fluctuations in one of the variables are
significantly higher than those of the others. However, after
checking the three ratios calculated in the last three columns in
table 3, we conclude that for some countries the absolute value
of FL is so low that we do not necessarily need to consider FL
in the fragility index explicitly. Hence, we create a third ver-
sion of the fragility index, BSF2*, by excluding the FL from
the BSF3 index:
(1’’)
Note that the BSF2* index is calculated only for those coun-
tries, which have relatively high LCPS / LFL and quite
low LFL / (LCPS – LDEP) ratios (see table 3). These
countries are Chile, Kenya, Mexico, Trinidad and Tobago,
and Venezuela.
Table 3
Basic Characteristics of National Banking Systems According to the Components of the BSF Index
Standard Deviations in: Selected Banking Sector Ratios (period averages)
Country Sample Period 12-Month 12-Month 12-Month 12-Month Real Claims Real Total Real Foreign
Percent Percent Percent Percent on Private Deposits / Liabilities/(Real
Change in Change in Change in Change in Sector / Real Real Claims Claims on
Real Claims on Real Foreign Real Total the Consumer Foreign on Private Private Sector
Private Sector Liabilities Deposits Prices Index Liabilities Sector RealTotal Deposits)
(CPS) (FL) (DEP) (INF) (LCPS / LFL) (LDEP / LCPS) (LFL/ (LCPS-LDEP))
1 Argentina Jan. 1982 - Dec. 2002 17.7 35.8 25.2 2100.0 2.2 0.9 5.5
2 Bolivia Jan. 1965 - Dec. 2002 42.3 77.2 48.3 2025.5 9.3 0.8 0.5
3 Brazil June 1989 - Dec. 2002 23.7 85.6 20.6 1391.5 6.5 0.9 1.2
4 Chile Dec. 1979 - Dec. 2002 14.2 51.3 9.8 9.8 13.9 0.6 0.2
5 Indonesia Jan. 1981 - Dec. 2002 22.4 88.5 11.1 12.9 6.7 1.1 -1.4
6 Israel Jan. 1982 - Dec. 2002 6.1 9.7 6.5 105.4 2.8 1.1 -5.2
7 Japan Sep. 1968 - Dec. 2002 5.5 23.3 5.2 4.9 7.8 0.9 1.4
8 Jordan Jan. 1977 - Dec. 2002 11.2 44.7 10.1 7.0 2.2 1.2 -2.2
9 Kenya Jan. 1969 - Dec. 2002 13.1 45.3 11.9 10.1 16.3 1.3 -0.2
10 Malaysia May 1965 - Dec. 2002 7.9 30.8 6.7 3.5 8.5 0.9 1.3
11 Malta Jan. 1965 - Dec. 2002 14.8 98.5 6.6 3.8 1.4 1.4 -1.9
12 Mexico Jan. 1983 - Dec. 2002 30.9 53.8 38.0 40.9 74.1 1.3 0.0
13 Pakistan Jan. 1965 - Dec. 2002 8.5 33.4 9.3 6.0 7.4 1.2 -0.6
14 Peru Jan. 1965 - Dec. 2002 25.2 100.3 21.5 1345.5 5.3 1.4 -0.5
15 Philippines Dec. 1987 - Dec. 2002 15.5 34.6 7.8 3.7 2.9 1.3 -1.3
16 Poland Jan. 1991 - Dec. 2002 12.6 28.4 7.6 21.5 7.4 1.3 -0.4
17 South Korea Jan. 1968 - Dec. 2002 16.1 40.7 15.7 10.6 8.8 0.7 0.4
18 Sweden Jan. 1971 - Dec. 2000 7.9 15.2 5.2 4.0 1.6 1.0 -14.3
19 Thailand Jan. 1961 - Dec. 2002 11.0 33.8 8.7 7.8 5.7 0.9 2.2
20 Trinidad
and Tobago Dec. 1965 - Dec. 2002 11.4 42.3 9.4 5.3 16.3 1.3 -0.2
21 Turkey Jan. 1979 - Dec. 2002 20.1 95.4 15.4 24.0 3.7 1.0 -23.6
22 Venezuela Sep. 1968 - Dec. 2002 19.7 59.3 15.4 24.4 48.6 1.3 -0.1
Sample Average 16.3 51.3 14.4 325.8 11.8 1.1 -1.8
Source: IMF, International Financial Statistics, CD-ROM version, July 2003; author’s own calculations.
The Arab Bank
REVIEW
Vol. 5, No. 2
October 2003
Banking
Management
56
Figure 2
Banking Sector Fragility in Argentina
Figure 3
Banking Sector Fragility in Bolivia
Figure 4
Banking Sector Fragility in Brazil
Figure 5
Banking Sector Fragility in Chile
Figure 6
Banking Sector Fragility in Indonesia
Figure 7
Banking Sector Fragility in Israel
The Arab Bank
REVIEW
Vol. 5, No. 2
October 2003
Banking
Management
57
Feb. ’01
Dec. ’01
May. ’94
Oct. ’95
Mar. ’98
Oct. ’92
May ’90
May ’89
May ’86
Sep. ’85
Jul. ’82
Jul. ’83
-3.5
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Jan-82
Jan-83
Jan-84
Jan-85
Jan-86
Jan-87
Jan-88
Jan-89
Jan-90
Jan-91
Jan-92
Jan-93
Jan-94
Jan-95
Jan-96
Jan-97
Jan-98
Jan-99
Jan-00
Jan-01
Jan-02
High Fragility BSF3 BSF2
Nov. ’02
Aug. ’98
Dec. ’94
Jan. ’96
Jun. ’00
Nov. ’90
Jan. ’82
Nov. ’83
May ’90
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Dec-79
Dec-80
Dec-81
Dec-82
Dec-83
Dec-84
Dec-85
Dec-86
Dec-87
Dec-88
Dec-89
Dec-90
Dec-91
Dec-92
Dec-93
Dec-94
Dec-95
Dec-96
Dec-97
Dec-98
Dec-99
Dec-00
Dec-01
Dec-02
High Fragility BSF3 BSF2 BSF2*
Feb. ’02
Aug. ’73 Jul. ’96
Aug. ’88
Aug. ’86
Mar. ’93
Aug. ’85
Sep. ’82
Apr. ’80
Mar. ’78
Jul. ’69
-3.0
-1.5
0.0
1.5
3.0
4.5
6.0
7.5
9.0
10.5
12.0
Jan-65
Jan-67
Jan-69
Jan-71
Jan-73
Jan-75
Jan-77
Jan-79
Jan-81
Jan-83
Jan-85
Jan-87
Jan-89
Jan-91
Jan-93
Jan-95
Jan-97
Jan-99
Jan-01
High Fragility BSF3 BSF2
Feb. ’00
Aug. ’96
Sep. ’94
Jul. ’90
Apr. ’02
Jul. ’92
Jan. ’99 Mar. ’01
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Jun-89
Dec-89
Jun-90
Dec-90
Jun-91
Dec-91
Jun-92
Dec-92
Jun-93
Dec-93
Jun-94
Dec-94
Jun-95
Dec-95
Jun-96
Dec-96
Jun-97
Dec-97
Jun-98
Dec-98
Jun-99
Dec-99
Jun-00
Dec-00
Jun-01
Dec-01
Jun-02
Dec-02
High Fragility BSF3 BSF2
ggy
Apr. ’02
Jun. ’01
Mar. ’83
Mar. ’84
Oct. ’88
Aug. ’90
Sep. ’91
Jan. ’98
Jun. ’99
Mar. ’89
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Jan-81
Jan-82
Jan-83
Jan-84
Jan-85
Jan-86
Jan-87
Jan-88
Jan-89
Jan-90
Jan-91
Jan-92
Jan-93
Jan-94
Jan-95
Jan-96
Jan-97
Jan-98
Jan-99
Jan-00
Jan-01
Jan-02
High Fragility BSF3 BSF2
Nov. ’87
May. ’85
Dec. ’82
Jun. ’82
Dec. ’89
Dec. ’85
Aug. ’96
Nov. ’83
Apr. ’02
Nov. ’94
Oct. ’98
Nov. ’93
Jan. ’88 May ’87
Nov. ’85
Jun. ’86
Oct. ’83
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
Jan-82
Jan-83
Jan-84
Jan-85
Jan-86
Jan-87
Jan-88
Jan-89
Jan-90
Jan-91
Jan-92
Jan-93
Jan-94
Jan-95
Jan-96
Jan-97
Jan-98
Jan-99
Jan-00
Jan-01
Jan-02
High Fragility BSF3 BSF2
The Arab Bank
REVIEW
Vol. 5, No. 2
October 2003
Banking
Management
Figure 8
Banking Sector Fragility in Japan
Figure 9
Banking Sector Fragility in Jordan
Figure 10
Banking Sector Fragility in Kenya
Figure 11
Banking Sector Fragility in Malaysia
Figure 12
Banking Sector Fragility in Malta
Figure 13
Banking Sector Fragility in Mexico
58
Aug. ’99
Mar. ’01
May ’96
Mar. ’94
Jul. ’87
Nov. ’81
May ’80
Aug. ’77
Nov. ’71
Feb. ’70
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
Sep-68
Sep-70
Sep-72
Sep-74
Sep-76
Sep-78
Sep-80
Sep-82
Sep-84
Sep-86
Sep-88
Sep-90
Sep-92
Sep-94
Sep-96
Sep-98
Sep-00
Sep-02
High Fragility BSF3 BSF2
Mar. ’95
May ’02
May ’01
Jun. ’97
Mar. ’92
Sep. ’89
May ’88
Feb. ’86
Aug. ’80
Mar. ’79
Jan. ’78
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Jan-77
Jan-78
Jan-79
Jan-80
Jan-81
Jan-82
Jan-83
Jan-84
Jan-85
Jan-86
Jan-87
Jan-88
Jan-89
Jan-90
Jan-91
Jan-92
Jan-93
Jan-94
Jan-95
Jan-96
Jan-97
Jan-98
Jan-99
Jan-00
Jan-01
Jan-02
High Fragility BSF3 BSF2
ggyy
Sep. ’00
Jun. ’72
Aug. ’77 Sep. ’86
Jan. ’96
Nov. ’90
Aug. ’82
Feb. ’92
Mar. ’94
May ’75
Jul. ’74
Dec. ’70
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Jan-69
Jan-71
Jan-73
Jan-75
Jan-77
Jan-79
Jan-81
Jan-83
Jan-85
Jan-87
Jan-89
Jan-91
Jan-93
Jan-95
Jan-97
Jan-99
Jan-01
High Fragility BSF3 BSF2 BSF2*
Jan. ’99
Feb. ’97
Jan. ’95
Sep. ’92
Sep. ’87
Dec. ’83
Nov. ’82
Oct. ’80
Jan. ’75
Dec. ’73
Apr. ’70
Dec. ’68
Jan. ’66
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
May-65
May-67
May-69
May-71
May-73
May-75
May-77
May-79
May-81
May-83
May-85
May-87
May-89
May-91
May-93
May-95
May-97
May-99
May-01
High Fragility BSF3 BSF2
ggy
May. ’99
Jan. ’78
Oct. ’72
Jan. ’95
Aug. ’67 Mar. ’69
Jan. ’72
Jan. ’77
Mar. ’02
Jan. ’96
Mar. ’88
Dec. ’73
Mar. ’68
Aug. ’65
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
Jan-65
Jan-67
Jan-69
Jan-71
Jan-73
Jan-75
Jan-77
Jan-79
Jan-81
Jan-83
Jan-85
Jan-87
Jan-89
Jan-91
Jan-93
Jan-95
Jan-97
Jan-99
Jan-01
High Fragility BSF3 BSF2
Aug. ’98
Dec. ’94
Jan. ’01
Feb. ’96
Apr. ’90
Nov. ’86
Mar. ’85
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Jan-83
Jan-84
Jan-85
Jan-86
Jan-87
Jan-88
Jan-89
Jan-90
Jan-91
Jan-92
Jan-93
Jan-94
Jan-95
Jan-96
Jan-97
Jan-98
Jan-99
Jan-00
Jan-01
Jan-02
High Fragility BSF3 BSF2 BSF2*
The Arab Bank
REVIEW
Vol. 5, No. 2
October 2003
Banking
Management
Figure 14
Banking Sector Fragility in Pakistan
Figure 15
Banking Sector Fragility in Peru
Figure 16
Banking Sector Fragility in Philippines
Figure 17
Banking Sector Fragility in Poland
Figure 18
Banking Sector Fragility in South Korea
Figure 19
Banking Sector Fragility in Sweden
59
Aug. ’99
Sep. ’71
Jul. ’69
Aug. ’02
Nov. ’97
Mar. ’83
Oct. ’96
Oct. ’93
Dec. ’90
Mar. ’84
Aug. ’88
Jul. ’86
Dec. ’80
Dec. ’76
Jul. ’74
Dec. ’72
Jun. ’66
Jun. ’65
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Jan-65
Apr-66
Jul-67
Oct-68
Jan-70
Apr-7
Jul-72
Oct-73
Jan-75
Apr-76
Jul-77
Oct-78
Jan-80
Apr-81
Jul-82
Oct-83
Jan-85
Apr-8
Jul-87
Oct-88
Jan-90
Apr-91
Jul-92
Oct-93
Jan-95
Apr-96
Jul-97
Oct-98
Jan-00
Apr-01
Jul-02
High Fragility BSF3 BSF2
Aug. ’83
Sep. ’98
May ’82
Feb. ’00
May ’67
Jun. ’68
Jun. ’74
Jun. ’77
Feb. ’75
Jan. ’89
Aug. ’91
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Jan-65
Apr-66
Jul-67
Oct-68
Jan-70
Apr-71
Jul-72
Oct-73
Jan-75
Apr-76
Jul-77
Oct-78
Jan-80
Apr-81
Jul-82
Oct-83
Jan-85
Apr-86
Jul-87
Oct-88
Jan-90
Apr-91
Jul-92
Oct-93
Jan-95
Apr-96
Jul-97
Oct-98
Jan-00
Apr-01
Jul-02
High Fragility BSF3 BSF2
Oct. ’01
Oct. ’00
Jul. ’88
Jan. ’99
Nov. ’96
Nov. ’93
Aug. ’94
Sep. ’91
Dec. ’90
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Dec-87
Jun-88
Dec-88
Jun-89
Dec-89
Jun-90
Dec-90
Jun-91
Dec-91
Jun-92
Dec-92
Jun-93
Dec-93
Jun-94
Dec-94
Jun-95
Dec-95
Jun-96
Dec-96
Jun-97
Dec-97
Jun-98
Dec-98
Jun-99
Dec-99
Jun-00
Dec-00
Jun-01
Dec-01
Jun-02
Dec-02
High Fragility BSF3 BSF2
ggy
Jun. ’02
Jun. ’01
Jun. ’00
Dec. ’98
Oct. ’97
May. ’96
Oct. ’94
Nov. ’92
Nov. ’91
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
Jan-91
Jul-91
Jan-92
Jul-92
Jan-93
Jul-93
Jan-94
Jul-94
Jan-95
Jul-95
Jan-96
Jul-96
Jan-97
Jul-97
Jan-98
Jul-98
Jan-99
Jul-99
Jan-00
Jul-00
Jan-01
Jul-01
Jan-02
Jul-02
High Fragility BSF3 BSF2
Sep. ’82
Jan. ’70
Apr. ’72
Mar. ’75
Jan. ’76
Sep. ’85
Oct. ’84
Jan. ’88
Dec. ’91
Jan. ’94
Nov. ’97
Nov. ’98
May ’73
Jan. ’81
Jan. ’79
Mar. ’97
Apr. ’74
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Jan-68
Apr-69
Jul-70
Oct-71
Jan-73
Apr-74
Jul-75
Oct-76
Jan-78
Apr-79
Jul-80
Oct-81
Jan-83
Apr-84
Jul-85
Oct-86
Jan-88
Apr-89
Jul-90
Oct-91
Jan-93
Apr-94
Jul-95
Oct-96
Jan-98
Apr-99
Jul-00
Oct-01
High Fragility BSF3 BSF2
Sep. ’00
Nov. ’76
Jul. ’88
Apr. ’90
Feb. ’86
Jan. ’72
May ’75
Dec. ’79
Oct. ’83
Feb. ’94
Mar. ’92
Nov. ’98
Sep. ’99
Jun. ’74
Mar. ’81
Feb. ’83
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
Jan-71
Apr-72
Jul-73
Oct-74
Jan-76
Apr-77
Jul-78
Oct-79
Jan-81
Apr-82
Jul-83
Oct-84
Jan-86
Apr-87
Jul-88
Oct-89
Jan-91
Apr-92
Jul-93
Oct-94
Jan-96
Apr-97
Jul-98
Oct-99
Jan-01
Apr-02
High Fragility BSF3 BSF2
The Arab Bank
REVIEW
Vol. 5, No. 2
October 2003
Banking
Management
Figure 20
Banking Sector Fragility in Thailand
Figure 21
Banking Sector Fragility in Trinidad & Tobago
Figure 22
Banking Sector Fragility in Turkey
Figure 23
Banking Sector Fragility in Venezuela
Source: IMF, International Financial Statistics, CD-ROM version, July 2003;
author’s own calculations.
Note: BSF3 and BSF2 are two alternative indices of banking sector fragility
that are calculated as defined in equations (1) and (1’), respectively. For
interpretation of changes in indices, please see table 2. Note that gray
vertical bands in figures designate the periods of high fragility (see
equation (6)) with respect to the BSF2 index. For Chile, Kenya, Mexico,
Trinidad and Tobago, and Venezuela, we consider BSF2* instead of
BSF2, as justified in section 4 above.
The national BSF3 and BSF2 indices - and if calculated, the
BSF2* index - are graphically shown in figures 2 to 23. The
episodes of medium and high banking-sector fragility are cal-
culated according to the above-described criteria by consider-
ing the BSF2 (or BSF2*) index. The country-specific high
fragility periods determined according to equation (6) are
marked by gray vertical bands in figures.
Concluding Comparison and Final Remarks
In recent years, there has been a high interest in research on the
timing, duration, causes, effects, and cures of banking crises.
In this paper, we proposed a monthly, weighted banking-sector
fragility (BSF) index that may easily be used to measure and
monitor the changes in the banking sector fragility to crisis.
Apparently, this type of index is capable of providing more
information about the ups and downs in the banking sector
with respect to certain crisis-years in event-based studies.
Table 4 compares our findings shown in figures 2-23 with the
results of major studies in the event-based tradition.
60
ggy
Jan. ’99
Mar. ’64
Jul. ’66
Feb. ’71
Dec. ’72
Jan. ’79
Jun. ’80
Mar. ’84
Mar. ’86
Nov. ’88
Jul. ’92
Mar. ’94
Sep. ’00
Jul. ’97
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
Jan-61
Apr-62
Jul-63
Oct-64
Jan-66
Apr-67
Jul-68
Oct-69
Jan-71
Apr-72
Jul-73
Oct-74
Jan-76
Apr-77
Jul-78
Oct-79
Jan-81
Apr-82
Jul-83
Oct-84
Jan-86
Apr-87
Jul-88
Oct-89
Jan-91
Apr-92
Jul-93
Oct-94
Jan-96
Apr-97
Jul-98
Oct-99
Jan-01
Apr-02
High Fragility BSF3 BSF2
ggy g
Feb. ’70
Aug. ’67
Jul. ’68
Apr. ’74
Dec. ’72
Sep. ’76
Nov. ’80
Nov. ’82
Sep. ’87
Oct. ’91
Aug. ’97
May ’00
Dec. ’92
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Dec-65
Mar-67
Jun-68
Sep-69
Dec-70
Mar-72
Jun-73
Sep-74
Dec-75
Mar-77
Jun-78
Sep-79
Dec-80
Mar-82
Jun-83
Sep-84
Dec-85
Mar-87
Jun-88
Sep-89
Dec-90
Mar-92
Jun-93
Sep-94
Dec-95
Mar-97
Jun-98
Sep-99
Dec-00
Mar-02
High Fragility BSF3 BSF2 BSF2*
Feb. ’02
Aug. ’97
Feb. ’01
Sep. ’88
Feb. ’87
Nov. ’83
Aug. ’82
May ’80
Jun. ’79
Dec. ’82
Jan. ’86
Nov. ’90
Nov. ’91
Oct. ’93
Apr. ’94
Oct. ’94
Jul. ’99
Oct. ’00
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Jan-79
Jan-80
Jan-81
Jan-82
Jan-83
Jan-84
Jan-85
Jan-86
Jan-87
Jan-88
Jan-89
Jan-90
Jan-91
Jan-92
Jan-93
Jan-94
Jan-95
Jan-96
Jan-97
Jan-98
Jan-99
Jan-00
Jan-01
Jan-02
High Fragility BSF3 BSF2
ggy
Feb. ’01
Jun. ’99
Oct. ’97
Nov. ’93
Jul. ’91
Jan. ’87
May ’89
Jan. ’85
Dec. ’75
Feb. ’80
Feb. ’84
Jun. ’96
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Sep-68
Dec-69
Mar-71
Jun-72
Sep-73
Dec-74
Mar-76
Jun-77
Sep-78
Dec-79
Mar-81
Jun-82
Sep-83
Dec-84
Mar-86
Jun-87
Sep-88
Dec-89
Mar-91
Jun-92
Sep-93
Dec-94
Mar-96
Jun-97
Sep-98
Dec-99
Mar-01
Jun-02
High Fragility BSF3 BSF2 BSF2*
The Arab Bank
REVIEW
Vol. 5, No. 2
October 2003
Banking
Management
61
Table 4
Episodes of Major Banking Crises and High Fragility in Selected Countries
Caprio and Lindgren, Hardy and Demirgüç- Kaminsky and Reinhart Martinez Glick and Bordo and Current Study (the BSF2 , or BSF2* , index)
Klingebiel Garcia Pazarba?? Kunt and (1996 and 1999) Peria Hutchison Eichengreen Beginning Date of Episode of High Fragility
(1996, 1999 and Saal, o∂lu Detragiache Beginning Peak of (2000) (2000) (2002) of the Highest (if applicable)
2002 and (1996) (1998) (1997 of the Crisis Distress Fragility
2003) and 1998) the Crisis
Argentina 1980-1982 1980-1982 Mar. 1980 July 1982 1980-1982 1980 d.n.a. Jul. 1983 Jul. 1983 - Mar. 1985
May 1985 June 1989 Oct. 1985 May 1986 Apr. 1986 - Nov. 1986
1989-1990 1989-1990 1989-1990 1989 Jun. 1989 May 1990 Oct. 1988 - Jan. 1992
1995 Dec. 1994 Mar. 1995 1995-1997 1995 Jun. 1994 Oct. 1995 medium fragility
2001-present Mar. 2001 Dec. 2001 Sep. 2001 - Feb. 2002
Bolivia Apr. 1978 Apr. 1980 Aug. 1979 - Oct. 1981
Oct. 1982 Aug. 1985 Mar. 1983 - Sep. 1985
1986-1988 1986-1987** Oct. 1987 June 1988 1986-1987 Sep. 1986 Aug. 1988 May 1988 - Dec. 1989
1994-? 1994-pres.** 1994-1997 April 1993 Feb. 2002 Feb. 2000 - Jun. 2002
Brazil Nov. 1985 Nov. 1985 d.n.a. d.n.a.
1990 1990 1990 d.n.a. July 1990 Nov. 1989 - Apr. 1991
1994-1999 1994-pres.** Dec. 1994 Mar. 1996 1994-1997 1994 Apr. 1995 Aug. 1996 medium fragility
Feb. 1999 Feb. 2000 Jan. 2000 - Mar. 2000
Chile 1976 1976 1976 d.n.a. d.n.a.
1981-1986 1981-1987 1981-1987 Sep. 1981 Mar. 1983 1981-1987 1981-1983 1981 Feb. 1982 Nov. 1983 Jan. 1983 - Apr. 1987
June 1990 Nov. 1990 Aug. 1990 - July 1991
Sep. 1998 June 2000 Feb. 2000 - Jan. 2001
Mar. 2002 Nov. 2002 Jul. 2002 - Dec. 2002
Indonesia 1994* 1994-pres.** 1992* 1992-1994 Nov. 1992 present 1994 1992 Sep. 1990 Sep. 1991 medium fragility
1997-present 1997 1997 1997-1998 Feb. 1998 June 1999 Sep. 1998 - Dec. 2002
Israel 1977-1983 1977 d.n.a. d.n.a. d.n.a.
1983-1984** 1983-1984 Oct. 1983 June 1984 Jan. 1983 Dec. 1985 Nov. 1983 - Jun. 1986
Dec. 1987 Dec. 1989 Jan. 1988 - Aug. 1991
Japan Dec. 1971 Aug. 1977 Dec. 1976 - July 1978
1991-present 1992-pres.** 1992* 1992-1994 1992-pres 1992-1997 1992 Aug. 1987 Mar. 1994 July 1991 - July 1995
June 1996 Aug. 1999 June 1997 - Dec. 2002
Jordan Mar. 1977 Jan. 1978 Nov. 1977 - Mar. 1978
1989-1990* 1989 1989-1990 1989-1990 Jun-88 Sep. 1989 Dec. 1988 - June 1991
Apr. 1995 Jun. 1997 Sep. 1996 - Aug. 1998
Kenya Aug. 1974 May 1975 Jan. 1975 - Jan. 1976
Sep. 1977 Aug. 1982 Oct. 1980 - July 1984
1985-1989 1985-1989 1985-1989 Oct. 1986 Nov. 1990 Mar. 1988 - Oct. 1988
1992 see the next row
1993-1995 1993* 1993* 1993 1993-1995 1992-1997 Mar. 1992 Mar. 1994 Mar. 1990 - June 1994
1996-?* Feb. 1996 Sep. 2000 Jun. 2000 - May 2001
Malaysia 1985-1988* 1985-1988 1985 1985-1988 July 1985 Aug. 1986 1985-1988 1985-1988 1985 Jan. 1984 Sep. 1987 Oct. 1986 - Mar. 1989
1997-present Sep. 1997 present 1997 1998 Mar. 1997 Jan. 1999 Apr. 1998 - Dec. 2002
Malta no crisis no crisis Nov. 1972 Dec. 1973 Feb. 1973 - Sep. 1975
between between Feb. 1977 Jan. 1978 Nov. 1977 - Jan. 1980
1990-1995 1992-1997 Jun. 1999 Mar. 2002 Jul. 2001 - Jun. 2002
Mexico 1981-1991 1982 1982 1982 Sep. 1982 June 1984 1981-1991 1981 Apr. 1985 Nov. 1986 Dec. 1985 - Dec. 1988
1994-1997 1994-pres.** 1994 1994-1995 Oct. 1992 Mar. 1996 1995-1997 1994 Jan. 1995 Feb. 1996 Aug. 1995 - Dec. 1996
Sep. 1998 Jan. 2001 Aug. 2000 - Apr. 2001
Pakistan Jan. 1973 July 1974 June 1973 - June 1975
1980-pres.** Aug. 1986 Aug. 1988 Apr. 1988 - Nov. 1988
Sep. 1999 Aug. 2002 Sep. 2001 - Dec. 2002
Peru June 1967 June 1968 June 1968 - Nov. 1968
July 1974 June 1977 June 1977 - July 1979
1983-1990 1983-1990** 1983* 1983-1990 Mar. 1983 Apr. 1983 1983-1990 1983 June 1982 Jan. 1989 July 1987 - Jan. 1991Philippines
1981-1987 1981-1987 1981 1981-1987 Jan. 1981 June 1985 1981-1987 1981 d.n.a.. d.n.a.. d.n.a.
Jan. 1991 Sep. 1991 Mar. 1991 - Mar. 1992
1998-present 1997* July 1997 present 1997 Dec. 1996 Jan. 1999 Aug. 1998 - Dec. 2002
Table 4
Episodes of Major Banking Crises and High Fragility in Selected Countries (continued)
Caprio and Lindgren, Hardy and Demirgüç- Kaminsky and Reinhart Martinez Glick and Bordo and Current Study (the BSF2 , or BSF2* , index)
Klingebiel Garcia Pazarba?? Kunt and (1996 and 1999) Peria Hutchison Eichengreen Beginning Date of Episode of High Fragility
(1996, 1999 and Saal, o∂lu Detragiache Beginning Peak of (2000) (2000) (2002) of the Highest (if applicable)
2002 and (1996) (1998) (1997 of the Crisis Distress Fragility
2003) and 1998) the Crisis
Poland 1990s d.n.a.. Nov. 1991 Jul. 1991 - Jun. 1995
South June 1973 Apr. 1974 July 1973 - Sep. 1974
Korea Feb. 1979 Jan. 1981 Mar. 1980 - June 1980
mid-1980s** Oct. 1985 Jan. 1988 Mar. 1987 - Apr. 1989
1997-present 1997 1997 1997-1998 Apr. 1997 Nov. 1998 Mar. 1998 - Feb. 1999
Sweden Mar. 1983 Oct. 1983 Oct. 1983 - June 1984
1991 1990-1993 1992 1990-1993 Nov. 1991 Sep. 1992 1990-1993 Aug. 1988 Feb. 1994 Apr. 1991 - Feb. 1995
Thailand Mar. 1979 Mar. 1979 Feb. 1979 June 1980 Dec. 1979 - Mar. 1981
1983-1987 1983-1987 1983 1983-1987 Oct. 1983 June 1985 1983-1987 1983 Apr. 1984 Mar. 1986 Nov. 1985 - Nov. 1987
1997-present 1997 May 1996 present 1997 Apr. 1994 Sep. 2000 Feb. 1998 - Dec. 2002
Trinidad Jan. 1973 Apr. 1974 Nov. 1973 - Oct. 1974
& Tobago 1982-1993* Early 1982- 1982-1993 Dec. 1982 Sep. 1987 June 1983 - Feb. 1991
1993**
Nov. 1991 Dec. 1992 Apr. 1992 - Nov. 1994
Turkey July 1979 May 1980 Jan. 1979 - Nov. 1980
1982-1985 1982 1982 1982-1985 1982 Sep. 1982 Nov. 1983 medium fragility
Mar. 1987 Sep. 1988 Apr. 1988 - Oct. 1989
1991 1991 Jan. 1991 Mar. 1991 1991 Dec. 1990 Nov. 1991 Nov. 1991 - Mar. 1992
1994* 1994** 1994-1995 1994-1995 Nov. 1993 Oct. 1994 Apr. 1994 - Apr. 1995
Sep. 1997 July 1999 Mar. 1999 - Mar. 2000
2000-present Nov. 2000 Feb. 2002 June 2001 - Dec. 2002
Venezuela Late 1970s
& 1980s* 1978-1986 1980 Jan. 1976 Feb. 1980 Oct. 1979 - Aug. 1980
Feb. 1987 May 1989 Aug. 1987 - Mar. 1990
1994-1995 1994 1993-1995 Oct. 1993 Aug. 1994 1994-1997 1993 Aug. 1991 June 1996 Jan. 1993 - Feb. 1997
Nov. 1997 June 1999 Sep. 1998 - Jan. 2000
Mar. 2001 Dec. 2002? Feb. 2002 - Dec. 2002
* Borderline, or non-systemic, banking crisis.
** Significant, or extensive, unsoundness short of a crisis. d.n.a.: Data not available.
The Arab Bank
REVIEW
Vol. 5, No. 2
October 2003
Banking
Management
62
It should be noted that, in this table, we interpret the first
month, when a country-specific fragility index started to
decline before entering an episode of high fragility, as the first
sign, or beginning date, of an approaching banking sector dis-
tress. Table 4 also shows both the specific months of highest
fragility and episodes of high fragility, which are marked by
gray vertical bands in figures from 2 to 23. Considering the
comparative information presented both in figures 2-23 and in
table 4, we briefly conclude that:
(a) Both the definition of banking crisis and the identification
of crisis episodes are essential, if one attempts to predict
and explain banking crises empirically. Depending on the
timing and duration of a crisis that is to be explained, the
result of the analysis is expected to vary remarkably. The
crisis episodes in most of the subsequent studies are pri-
marily adapted from the information given in Caprio and
Klingebiel (1996 or 2003) and/or Lindgren et al. (1996).
However, there are some important differences between the
crisis episodes given in these two studies and between the
episodes considered in those studies that are also men-
tioned in table 4. The Mexican banking crisis in the 1980s,
for example, is to be said by different researchers to occur
in 1982, between 1982 and 1984, or in the 1981-1991 peri-
od. It is obvious that the result of an empirical analysis of
the Mexican crisis will strongly depend on which year or
years we initially assumed as crisis years.
(b) Many studies in the literature (see table 4) do not differen-
tiate between systemic and non-systemic (borderline)
crises. However, the analysis of a banking crisis must also
be affected by the initial assumption on the extent of the
crisis. The BSF index proposed here not only captures cri-
sis times in terms of the defined high fragility periods in
this study, but it also roughly describes the whole develop-
ment process of a banking sector problem, even if it is only
a significant unsoundness short of a crisis.
The Arab Bank
REVIEW
Vol. 5, No. 2
October 2003
Banking
Management
(c) Overall, the high fragility episodes determined according
to the BSF calculations in this study overlap with the crisis
episodes mentioned in those studies that are considered in
table 4. Moreover, medium fragility episodes dated here are
to a large extent in accordance with borderline-crisis
episodes mentioned in the literature. Clearly, the informa-
tion content of changes in a monthly BSF index is signifi-
cantly higher than that of the simple years of crisis that are
identified based on country-specif ic information or rele-
vant events. A monthly BSF index explicitly detects the ups
and downs even within a single year, and hence, it elimi-
nates the risk of labeling an entire year as crisis year even
if the crisis has arisen, let’s say, only on the last two months
of that year.
63
Table 5
Country-Specific Literature on Banking Sector Performance and Fragility
Study Country Coverage Period Major Issues
Drees and Finland, Norway, and Sweden 1980s & 1990s Macroeconomic determinants of banking crises; role of
Pazarbas
,ıog˘lu (1995) financial liberalization in financial crises
García-Herrero (1997) Argentina, Paraguay, 1990s Causes of banking crises stemming from both macroeconomic
and Venezuela and bank-specific factors; macro economic effects of banking
crises
Baliño and Ubide (1999) South Korea 1993-1997 The sources of the 1997 Korean financial crisis, and the measures
taken to deal with it
Englund (1999) Sweden 1980s & 1990s Discussion of the 1985 deregulation and other causes of the
banking crisis in early 1990s; the relation between the European
Exchange-Rate Mechanism (ERM) crisis in 1992 and Swedish
banking crisis
Lindgren et al. (1999) Indonesia, South Korea, 1997-1999 Policy responses of Indonesia, Korea, and Thailand to the 1997
Thailand, Malaysia, and Asian crisis and comparison of their actions with those of
the Philippines Malaysia and the Philippines, which were buffeted by the crisis
Kanaya and Woo (2000) Japan 1990s Causes of banking crisis; reasons for the unnecessary prolongation
of the recovery process
Kane and Rice (2000) African countries 1980-1999 Effects of banks’ unbooked losses on banking stress; government
corruption and duration of banking crises
Levine (2000) Chile 1980-1999 Possible effects of banking sector concentration on financial
development, economic growth and banking sector fragility
Chang and Velasco (2001) East Asian and Latin 1997-1998 Macro- and microeconomic roots of international illiquidity in
American countries countries considered
Chopra et al. (2001) South Korea 1997-1998 Korean stabilization and reform program implemented in response
to the currency and banking crisis in 1997-98; recovery from deep
recession; the lessons learned
Duvan (2001) Turkey 1999-2001 Rapid growth of non-performing loans; debt restructuring
between creditor banks and borrowing corporate sector companies
Enoch et al.. (2001) Indonesia 1988-1999 Chronological evaluation of developments in the weak of the 1997
banking crisis, and the effects of government policies on the
recovery process
Gruben and Welch (2001) Brazil late 1990s Brazil’s January 1999 currency crisis; links between banking and
currency crises
Hardy and Bonaccorsi Pakistan 1980s & 1990s The effects of the 1988-1992 financial sector reform Bonaccorsi di
di Patti (2001) on the profitability and efficiency of the Pakistani banking system
Koo and Kiser (2001) South Korea 1997-1998 The chronology and causes of currency and banking crises in South
Korea; recovery from a twin crisis
Nakaso (2001) Japan 1990s The chronology of events and the policy responses by the authori-
ties; identification of factors that explain why it has taken so long
to bring the crisis under control; lessons learnt from the crisis
Worrell, Cherebin and Caribbean countries 1990s Review of financial sector performances and quantitative analysis
Polius-Mounsey (2001) (incl. Trinidad and Tobago) of bank soundness in the Caribbean
Barajas and Steiner (2002) Argentina, Bolivia, Brazil, 1960-2000 Causes of the slowdown in bank credit to the private sector in the
Chile, Colombia, Mexico, 1990s
Peru, and Venezuela
Pangestu and Habir (2002) Indonesia 1990s Effects of currency and interest-rate shocks on the vulnerability
of the Indonesian banking system, measures taken to deal with it,
and the lessons learned
Ertug˘rul and Yeldan (2002) Turkey 2000-2001 Recent disinflation attempt in Turkey and its negative effects on the
baking sector fragility
The Arab Bank
REVIEW
Vol. 5, No. 2
October 2003
Banking
Management
(d) For an individual researcher, the interpretation, or justifi-
cation, of variations in large numbers of national BSF
indices is not an easy undertaking because it requires some
degree of country expertise as well as additional microeco-
nomic information related to the relevant sector. Thus, a
certain group of studies in the financial crises literature
(see table 5) can be considered as a benchmark to examine
the chronological and institutional background of changes
in the BSF index. The country-specific chronological
explanations in those studies, which are listed in table 5,
strongly support both the results of and the main motiva-
tion behind the current study. To be more precise, let us
consider the case of the Swedish banking crisis in the early
1990s. The total length of this crisis spans from one year to
four years, depending on what study is considered as a ref-
erence for crisis episodes (see table 4). Even the timing of
the same crisis varies across the different studies.10 Figure
19 however implies that, after 1985, real bank deposits,
banks’ foreign liabilities, and credits to the private sector
all started to increase simultaneously, and that it reached a
peak in July 1988. The BSF approach used here suggests
that the following falls in the BSF2 (or BSF3) after this
date can be interpreted as a serious increase in the Swedish
banks’ vulnerability to crisis. The period from autumn of
1988 to autumn of 1990 is a period where all of the three
components of the index were clearly decreasing. Now,
according to Englund’s (1999) comprehensive analysis of
the Swedish banking crisis, these occurrences can be justi-
fied and understood as follows:
"Newly deregulated credit markets after 1985 stimulated a
competitive process between financial institutions where
expansion was given priority. Combined with an expansive
macro policy, this contributed to an asset price boom. The
subsequent crisis resulted from a highly leveraged private
sector being simultaneously hit by three major exogenous
events: a shift in monetary policy with an increase in pre-
tax interest rates, a tax reform that increased after tax inter-
est rates, and the ERM crisis. Combined with some overin-
vestment in commercial property, high real interest rates
contributed to breaking the boom in real estate prices, trig-
gering a downward price spiral resulting in bankruptcies
and massive credit losses. The government rescued the
banking system by issuing a general guarantee of bank
obligations. The total direct cost to the taxpayer of the sal-
vage has been estimated at around 2 per cent of GDP."
Englund’s explanations, which are only partially quoted
here, perfectly illustrate the macroeconomic background
behind the time path of the Swedish BSF curve shown in
figure 19 in this study. Not to expand the extent of the cur-
rent study unnecessarily, we prefer to restrict our country-
specific remarks here to the case of Sweden. But interested
readers easily may examine the reliability of the BSF index
proposed here by considering the explanations in country-
specific studies, such as those that are listed in table 5,
among others.
(e) As shown in figures 1-23, for many countries, the BSF3
and BSF2 curves appear to have followed a very similar
pattern, roughly implying that bank runs in many countries
may not have an important role in triggering banking
crises. Nevertheless, for particular countries, such as
Mexico, the developments in bank deposits must be close-
ly watched to detect possible banking sector problems.
To sum up, all of the discussions above show that the BSF
index proposed in this study is highly helpful in monitoring
and identifying the banking sector difficulties by using month-
ly data. Since the BSF index is reflecting the changes in the
sectoral climate more precisely and timely, it significantly
reduces the possibility that the crisis or high fragility episodes
are misidentified, contrary to the case of event-based identifi-
cation strategies. The BSF index presents the chance to be able
to work with higher frequency data on banking crises. Its infor-
mation content is significantly high. Therefore, in the future,
studies that aim to empirically investigate the causes, timing
and effects of banking crises can easily depart from the time-
series-based statistical approach developed here.
Notes
1To construct an index of vulnerability to currency crisis, some
researchers employ the difference between domestic and foreign
interest rates, or percentage changes in domestic interest rates, while
many others avoid using it because many developing countries do not
have reliable interest-rate data.
2Caprio and Klingebiel frequently update their well-known table of
episodes of systemic and borderline banking crises and publish it also
on the web, i.e. http://www.worldbank.org. Note that some of the
crises episodes mentioned in later versions of the table differ from
those episodes which are given in earlier versions.
3That is, the bank’s net worth includes the capital contributed by the
bank’s shareholders and accumulated prof its from doing business as
intermediary in financial markets.
4See Kaminsky and Reinhart (1996, 1999), Demirgüç-Kunt and
Detragiache (1998, 1999, 2000), Kaminsky (1999), Hardy and
Pazarba_ıo_lu (1998, 1999), the IMF’s World Economic Outlook
(May 1998, Ch. 4), Hutchison (1999), Goldstein et al. (2000),
Martinez Peria (2000), Bordo and Schwartz (2000), Gourinchas et al.
(2001), Hutchison and Neuberger (2002), and Bordo and
Eichengreen (2002).
64
The Arab Bank
REVIEW
Vol. 5, No. 2
October 2003
Banking
Management
5That means, credit overexpansions may reflect fundamental improve-
ments in investing opportunities that are beneficial to output growth
in the long run.
6In this case, this may lead to output recession, if bank credit is the
primary funding source of activities in the real economy as it is often
observed in many developing economies (see Disyatat, 2001).
7We neglect here (real) interest rates (or real interest rate differential)
as the fourth component of the BSF index because many developing
countries do not have internationally comparable and continuous
time series on market-determined interest rates. One may also argue
that interest-rate-risk (i.e., difficulties in maturity transformation)
actually is indirectly considered in calculations here by the mean of
deposits, claims, or foreign liabilities.
8By using the standardized values of CPS, FL and DEP, we equalize
the variance of the three components, and thus avoid the possibility
that any one of three components dominates the BSF3 index.
9Apparently, Hawkins and Klau’s (2000) index of banking system vul-
nerability is the most similar one to the BSF index proposed here.
The authors use the following five proxies to measure the banking
system vulnerability, by departing from the suggestion that banking
crises are typically preceded by overvalued exchange rates, inade-
quate international reserves, recessions, high real interest rates, and
excessive credit growth: (i) the rate of growth of domestic bank cred-
it, (ii) the growth of borrowing from international banks, (iii) the
external borrowing by banks as a percentage to domestic credit, (iv)
the level of real interest rates, and (v) "stand-alone" credit ratings of
the leading banks. Their index, however, differs from the BSF3 index
proposed here in many aspects. It is calculated for 24 emerging mar-
ket economies but the sample period is limited to 1996-1998 for
quarterly data frequency. The BSF3 index, on the other hand, aims to
cover a broader period of time for another set of 22 countries, includ-
ing also a few developed market economies, and the data frequency
is decided to be monthly. Furthermore, the Hawkins-Klau index is
based on a weighted scoring methodology, contrary to the calculation
methodology of the BSF3 index, which is actually similar to that of
the FEMP index that is used to measure the pressure in foreign cur-
rency market.
10Note that Caprio and Klingebiel (2003) accept the Swedish banking
crisis as occurred in 1991, although they considered the whole 1991-
1994 period as the crisis episode in an earlier version of their useful
table of banking crises.
References
- Baliño, T. J. T., and A. Ubide (1999). The Korean Financial Crisis of
1997 – A Strategy of Financial Sector Reform. Washington, DC:
International Monetary Fund, Working Paper No. 99/28.
- Barajas, A., and R. Steiner (2002). Credit Stagnation in Latin
America. Washington, DC: International Monetary Fund, Working
Paper No. 02/53.
- Bordo, M., and A. J. Schwartz (2000). Measuring Real Economic
Effects of Bailouts: Historical Perspectives on How Countries in
Financial Distress Have Fared With and Without Bailouts.
Cambridge, MA: National Bureau of Economic Research Working
Paper Series, No. 7701.
- Bordo, M., and B. Eichengreen (2002). Crises Now and Then: What
Lessons from the Last Era of Financial Globalization? Cambridge,
MA: National Bureau of Economic Research Working Paper Series,
No. 8716. S
,
- Boyd, J. H., P. Gomis, S. Kwak, and B. D. Smith (2001). A User’s
Guide to Banking Crises. Internet:
www.eco.utexas.edu/facstaff/Smith/bankingcrisestextnew.pdf, manuscript.
- Caprio, G., and D. Klingebiel (1996). Bank Insolvencies: Cross-
Country Experiences. Washington, DC: The World Bank, Policy
Research Working Paper No. 1620.
- Caprio, G., and D. Klingebiel (1999). Episodes of Systemic and
Borderline Financial Crises. Washington, DC: The World Bank,
manuscript.
- Caprio, G., and D. Klingebiel (2002). Episodes of Systemic and
Borderline Banking Crises. In: D. Klingebiel and L. Laeven, eds.,
Managing the Real and Fiscal Effects of Banking Crises.
Washington, DC: The World Bank, Discussion Paper No. 428,
p. 31-49.
- Caprio, G., and D. Klingebiel (2003). Episodes of Systemic and
Borderline Financial Crises. Washington,DC: The World Bank, man-
uscript.
- Calvo, G. A., L. Leiderman, and C. M. Reinhart (1994). The Capital
Inflows Problem: Concepts and Issues. Contemporary Economic
Policy, 12 (3): 54-66.
- Chang, R., and A. Velasco (2001). The 1997-98 Liquidity Crisis:
Asia Versus Latin America. Central Bank of Chile, Working Papers
No. 120.
- Chopra, A., K. Kang, M. Karasulu, H. Liang, H. Ma, and A.
Richards (2001). From Crisis to Recovery in Korea: Strategy,
Achievements, and Lessons. Washington, DC: International
Monetary Fund, Working Paper No. 01/154.
- Demirgüç-Kunt, A., and E. Detragiache (1998). The Determinants of
Banking Crises: Evidence from Developing and Developed
Countries. IMF Staff Papers 45 (1): 81-109.. Formerly printed as:
International Monetary Fund, Working Paper No. 97/106, 1997,
Washington, DC.
- Demirgüç-Kunt, A., and E. Detragiache (1999). Financial
Liberalization and Financial Fragility. In: Pleskovic, B., and J. E.
Stiglitz, eds., Annual World Bank Conference on Development
Economics 1998. Washington, D.C.: World Bank. Formerly printed
as: International Monetary Fund, Working Paper No. 98/83, 1998,
Washington, DC.
- Demirgüç-Kunt, A., and E. Detragiache (2000). Monitoring Banking
Sector Fragility: A Multivariate Logit Approach. The World Bank
Economic Review, 14 (2): 287–307.
- Dress, B., and C. Pazarba_ıo_lu (1995). The Nordic Banking Crises:
Pitfalls in Financial Liberalization? Washington, DC: International
Monetary Fund, Working Papers, No. 95/61.
(http://www.imf.org/external/pubs/cat/longres.cfm?sk=1138.0)
Also published as: IMF Occasional Paper, No. 161, 1998.
- Disyatat P. (2001). Currency Crises and the Real Economy: The
Role of Banks. Washington, DC: International Monetary Fund,
Working Paper No. 01/49.
- Duvan, B. (2001). Türk Bankacılık Sisteminde Tahsili Gecikmi_
Alacakların Tasfiyesi ve S
,irket Borçlarının Yapılandırılması
(= Resolution of Non-Performing Loans in the Turkish Banking
System, and Restructuring of Corporate Sector Debts). Ankara:
State Planning Organization.
- Edison, Hali J. (2000). Do Indicators of Financial Crises Work? An
Evaluation of An Early Warning System. Board of Governors of the
Federal Reserve System, International Finance Discussion Papers,
No. 675.
- Eichengreen, B., and C. O. Arteta (2000). Banking Crises in
Emerging Markets: Presumptions and Evidence. Berkeley, CA: 65
The Arab Bank
REVIEW
Vol. 5, No. 2
October 2003
Banking
Management
University of California at Berkeley, Center for International and
Development Economics Research Working Paper C00-115.
Forthcoming in: Blejer, M., and M. Skreb, eds., Exchange Rate and
Financial Vulnerability in Emerging Markets, The MIT Press.
- Eichengreen, B., and A. Rose (1998). Staying Afloat When the Wind
Shifts: External Factors and Emerging-Market Banking Crises.
Cambridge, MA: National Bureau of Economic Research, Working
Paper Series No. 6370.
- Englund, P. (1999). The Swedish Banking Crisis: Roots and
Consequences. Oxford Review of Economic Policy, 15 (3): 80-97.
- Enoch, C., B. Baldwin, O. Frécaut, and A. Kovanen (2001).
Indonesia: Anatomy of a Banking Crisis – Two Years of Living
Dangerously, 1997-99. Washington, DC: International Monetary
Fund, Working Paper No. 01/52.
- Ertug
˘rul, A., and E. Yeldan (2002). On the Structural Weaknesses of
the Post-1999 Turkish Dis-Inflation Program. Ankara: Bilkent
University, manuscript, internet: http://www.bilkent.edu.tr/~yel-
dane/Er&y2002.pdf.
- Frydl, E. (1999). The Length and Cost of Banking Crises.
Washington, DC: International Monetary Fund, Working Paper No.
99/30.
- García-Herrero A. (1997). Banking Crises in Latin America in the
1990s: Lessons from Argentina, Paraguay, and Venezuela.
Washington, DC: International Monetary Fund, Working Paper No.
97/140.
- Goldstein, M., G. L. Kaminsky, and C. M. Reinhart (2000).
Assessing Financial Vulnerability: An Early Warning System for
Emerging Markets. Washington, DC: Institute for International
Economics.
- Gourinchas, P.-O., R. Valdes, and O. Lenderretche (2001). Lending
Booms: Latin America and the World. Cambridge, MA: National
Bureau of Economic Research, Working Paper Series, No. 8249.
- Gruben, W. C., and J. H. Welch (2001). Banking and Currency Crisis
Recovery: Brazil’s Turnaround of 1999. Federal Reserve Bank of
Dallas, Economic and Financial Review, Fourth Quarter, 12-23.
- Hardy, D., and E. Bonaccorsi di Patti (2001). Bank Reform and
Bank Efficiency in Pakistan. Washington, DC: International
Monetary Fund, Working Paper Series No. 01/138.
- Hardy, D., and C. Pazarba_ıo_lu (1998). Leading Indicators of
Banking Crises: Was Asia Different? Washington, DC:
International Monetary Fund, Working Paper Series No. 98/91.
- Hardy, D., and C. Pazarbas
, ıog
˘lu (1999). Determinants and Leading
Indicators of Banking Crises: Further Evidence. Washington, DC:
International Monetary Fund Staff Papers, 46 (3): 247-258.
- Hawkins, J., and M. Klau (2000). Measuring Potential
Vulnerabilities in Emerging Market Economies. Basel: Bank for
International Settlements, Working Papers, No. 91.
- Hutchison, M. M. (1999). Determinants of European Banking
Distress: Evaluating the Risks from EMU and Financial Integration.
Prepared for the workshop "Structural Change and European
Economic Integration" held at Liseleje, Denmark, September 10-11,
1999, manuscript.
- Hutchison, M. M., and I. Neuberger (2002). How Bad Are Twins?
Output Costs of Currency and Banking Crises. Santa Cruz, CA:
University of California, manuscript.
- International Monetary Fund (1998). World Economic Outlook (Ch.
IV: Financial Crises: Characteristics and Indicators of
Vulnerability). No. 5: 74-97.
- International Monetary Fund (2002). International Financial
Statistics – Database and Browser; CD-ROM. Washington, DC: IMF.
- Kaminsky, G. (1999). Currency and Banking Crises: The Early
Warnings of Distress. Presented at the IMF Institute course on An
Early Warning System for Financial Crises, May 4, 1999, manu-
script.
- Kaminsky, G. and C. Reinhart (1996). Banking and Balance-of-
Payments Crises: Models and Evidence. Washington, DC: Board of
Governors of the Federal Reserve System, manuscript.
- Kaminsky, G. and C. Reinhart (1999). The Twin Crises: The Causes
of Banking and Balance-of-Payments Problems. American
Economic Review, 89 (3): 473–500. Formerly published as: Board of
Governors of the Federal Reserve System, International Finance
Discussion Papers No. 544, 1996, Washington, DC.
- Kanaya, A., and D. Woo (2000). The Japanese Banking Crisis of the
1990s. Washington, DC: International Monetary Fund, Working
Paper No. 00/07.
- Kane, E. J., and T. Rice (2000). Bank Runs and Banking Policies:
Lessons for African Policymakers. Chestnut Hill, MA: Boston
College, manuscript, internet:
www2.bc.edu/~kaneeb/BankRuns.pdf.
- Koo, J., and S. L. Kiser (2001). Recovery from a Financial Crisis:
The Case of South Korea. Federal Reserve Bank of Dallas,
Economic and Financial Review, Fourth Quarter, 24-36.
- Levine, R. (2000). Bank Concentration: Chile and International
Comparisons. Central Bank of Chile, Working Papers No. 62.
- Lindgren, C.-J., G. Garcia, and M. I. Saal (1996). Bank Soundness
and Macroeconomic Policy. Washington, D.C.: IMF.
- Lindgren, C. J., T. J. T. Baliño, C. Enoch, A.-M. Gulde, M. Quintyn,
and L. Teo (1999). Financial Sector Crisis and Restructuring:
Lessons from Asia. Washington, DC: International Monetary Fund,
Occasional Paper No. 188.
- Martinez Peria, M. S. (2000). The Impact of Banking Crises on
Money Demand and Price Stability. Washington, DC: The World
Bank, Policy Research Working Paper No. 2305.
- Nakaso, H. (2001). The Financial Crisis in Japan During the 1990s:
How the Bank of Japan Responded and the Lessons Learnt. Basel:
Bank for International Settlements, BIS Papers No. 6.
- Obstfeld, M. and K. Rogoff (1995). The Mirage of Fixed Exchange
Rates. Journal of Economic Perspectives, 9 (4): 73-96.
- Pangestu, M., and M. Habir (2002). The Boom, Bust, and
Restructuring of Indonesian Banks. Washington, DC: International
Monetary Fund, Working Paper No. 02/66.
- Sachs, J., A. Tornell, and A. Velasco (1996). Financial Crises in
Emerging Markets: The Lessons from 1995. Brookings Papers on
Economic Activity, 0 (1): 147–98.
- Worrell, D., D. Cherebin, and T. Polius-Mounsey (2001). Financial
System Soundness in the Caribbean: An Initial Assessment.
Washington, DC: International Monetary Fund, Working Paper No.
01/123.
* Aykut Kibritçiog
˘lu is Associate Professor at the Department
of Economics, Faculty of Political Sciences, Ankara
University.
66
