Financial Crises
Chapter

Chapter 2. Systemic Banking Crises

Author(s):
Stijn Claessens, Ayhan Kose, Luc Laeven, and Fabian Valencia
Published Date:
February 2014
Share
  • ShareShare
Show Summary Details
Author(s)
Luc Laeven and Fabián Valencia The authors thank Olivier Blanchard, Eduardo Borensztein, Martin Cihak, Stijn Claessens, Luis Cortavarria-Checkley, Giovanni Dell’Ariccia, David Hoelscher, Simon Johnson, Ashok Mody, Jonathan Ostry, and Bob Traa for comments and discussions, and Ming Ai, Chuling Chen, and Mattia Landoni for excellent research assistance. The views expressed in this chapter are those of the authors and do not necessarily represent those of the IMF or IMF policy.

Financial crises can be damaging and contagious, prompting calls for swift policy responses. The financial crises of the past have led affected economies into deep recessions and sharp current account reversals. Some crises turned out to be contagious, rapidly spreading to countries with no apparent vulnerabilities. Among the many causes of financial crises have been a combination of unsustainable macroeconomic policies (including large current account deficits and unsustainable public debt), excessive credit booms, large capital inflows, and balance sheet fragilities, combined with policy paralysis caused by a variety of political and economic constraints. In many financial crises, currency and maturity mismatches were a salient feature, while in others off–balance sheet operations of the banking sector were prominent.1

Choosing the best way to resolve a financial crisis and accelerate economic recovery is difficult. Little agreement has been found on what constitutes best practice or even good practice. Many approaches have been proposed and tried in attempts to resolve systemic crises more efficiently. This lack of agreement may occur, in part, because the objectives of the policy advice have varied. Some have focused on reducing the fiscal costs of financial crises; others on limiting the economic costs of lost output and on accelerating restructuring; still others on achieving long-term, structural reforms. Trade-offs are likely to arise between these objectives.2 Governments may, for example, through certain policies consciously incur large fiscal outlays in resolving a banking crisis, with the objective of accelerating recovery. Or structural reforms may only be politically feasible in the context of a severe crisis with large output losses and high fiscal costs.

This chapter introduces and describes a new data set on banking crises, with detailed information about the type of policy responses used to resolve crises in different countries. The emphasis is on policy responses to restore the banking system to health. The data set expands Caprio and others’ (2005) banking crisis database by including recent banking crises, information on currency and debt crises, and information on crisis containment and resolution measures. The database covers all systemically important banking crises for the period 1970 to 2007, and has detailed information on crisis management strategies for 42 systemic banking crises in 37 countries.

Governments have used a broad range of policies to deal with financial crises. Central to identifying sound policy approaches to financial crises is the recognition that policy responses that reallocate wealth toward banks and debtors and away from taxpayers face a critical trade-off. Such reallocations of wealth can help to restart productive investment, but they have large costs. These costs include taxpayers’ wealth that is spent on financial assistance, indirect costs from misallocations of capital, and distortions to incentives that may result from encouraging banks and firms to abuse government protections. Those distortions may worsen capital allocation and risk management after the resolution of the crisis.

Institutional weaknesses typically aggravate crises and complicate crisis resolution. Bankruptcy and restructuring frameworks are often deficient. Disclosure and accounting rules for financial institutions and corporations may be weak. Equity and creditor rights may be poorly defined or weakly enforced. And the judiciary system is often inefficient.

Many financial crises, especially those in countries with fixed exchange rates, turn out to be twin crises with currency depreciation exacerbating banking sector problems through the foreign currency exposures of borrowers or the banks themselves. In such cases, another complicating factor is the conflict between the desire to maintain currency pegs and the need to provide liquidity support to the banking system.

Existing empirical research has shown that providing assistance to banks and their borrowers can be counterproductive, resulting in increased losses to banks, which often abuse forbearance to take unproductive risks at government expense. The typical result of forbearance is a deeper hole in the net worth of banks, a crippling tax burden caused by financing bank bailouts, and an even more severe credit supply contraction and economic decline than would have occurred in the absence of forbearance.3

Cross-country analyses to date also show that accommodative policy measures (such as substantial liquidity support, explicit government guarantees on financial institutions’ liabilities, and forbearance from prudential regulation) tend to be fiscally costly and that these particular policies do not necessarily accelerate the economic recovery.4 Of course, the weakness in these findings is that a counterfactual to the crisis resolution cannot be observed, therefore, it is difficult to speculate how a crisis would unfold in absence of such policies. Better institutions are, however, uniformly positively associated with faster recoveries.

The remainder of the chapter is organized as follows: The first section presents new data on the timing of banking crises, currency crises, and sovereign debt crises. The second section presents variable definitions of the data collected on crisis management techniques for a subset of systemic banking crises. The third section presents descriptive statistics of data on containment and resolution policies, fiscal costs, and output losses. The fourth section discusses the ongoing (as of 2013) global liquidity crisis that originated within the U.S. subprime crisis of 2007–09.

Crisis Dates

Banking Crises

This analysis defines a systemic banking crisis as occurring when a country’s corporate and financial sectors experience a large number of defaults and financial institutions and corporations face great difficulties repaying contracts on time. As a result, nonperforming loans increase sharply and all or most of the aggregate banking system capital is exhausted. This situation may be accompanied by depressed asset prices (such as equity and real estate prices) on the heels of runups before the crisis, sharp increases in real interest rates, and a slowdown or reversal in capital flows. In some cases, the crisis is triggered by depositor runs on banks, though in most cases it is a general realization that systemically important financial institutions are in distress.

Using this broad definition of a systemic banking crisis that combines quantitative data with some subjective assessment of the situation, the starting years of systemic banking crises around the world since 1970 are identified. Unlike previous work (Caprio and Klingebiel, 1996; and Caprio and others, 2005), this chapter excludes banking system distress events that affected isolated banks but were not systemic. As a cross-check on the timing of each crisis, the analysis examines whether the crisis year coincides with deposit runs, the introduction of a deposit freeze or blanket guarantee, or extensive liquidity support or bank interventions,5 and thus confirms about two-thirds of the crisis dates. Alternatively, it must be apparent that the banking system has a large proportion of nonperforming loans and that most of its capital has been exhausted.6 This additional requirement applies to the remainder of crisis dates.

In sum, 124 systemic banking crises were identified between 1970 and 2007. This list is an updated, corrected, and expanded version of the Caprio and Klingebiel (1996) and Caprio and others (2005) banking crisis databases. Appendix Table 2A.1 lists the starting year of each banking crisis, as well as background information, including peak nonperforming loans, gross fiscal costs, output loss, and minimum real GDP growth rate. Peak nonperforming loans is the highest level of nonperforming loans as a percentage of total loans during the first five years of the crisis. Gross fiscal costs are computed over the five years following the start of the crisis using data from Hoelscher and Quintyn (2003); Honohan and Laeven (2005) IMF staff reports; and publications from national authorities and institutions. Output losses are computed by extrapolating trend real GDP based on the trend in real GDP growth up to the year preceding the crisis, and taking the sum of the differences between actual real GDP and trend real GDP expressed as a percentage of trend real GDP for the first four years of the crisis (including the crisis year).7 The minimum real GDP growth rate is the lowest real GDP growth rate during the first three years of the crisis.

Appendix Table 2A.1Timing of Systemic Banking Crises
CountrySystemic banking crisis (starting date)Share of NPLs at peak (%)Gross fiscal cost (% of GDP)Output loss (% of GDP)Minimum real GDP growth rate (%)Comments
Albania199426.8−7.2Rapid growth in NPLs, reaching 26.8 percent of total loans in 1994, following the creation of a two-tier commercial banking system in 1992.
Algeria1990306.7−2.1In 1989, five government-owned banks were granted managerial and financial autonomy from the central government. In the transition to a market economy, NPLs (about 30 percent of total loans) created problems for some banks in 1990, and the central bank had to provide discount financing to these banks.
Argentina1980955.110.8−5.7In March 1980, a number of financial institutions were forced to rely heavily on central bank financial assistance when faced with deposit withdrawals. Failed institutions included the largest investment bank and the second largest private commercial bank. More than 70 institutions (accounting for 16 percent of commercial bank assets and 35 percent of finance company assets) were liquidated or subjected to intervention between 1980 and 1982.
Argentina198927610.7−7.0During the 1980s, a decline in the availability of external resources led to increased recourse to domestic financing. To fund its credit operations the central bank imposed reserve and investment requirements on deposits. They were replaced by frozen deposits at the central bank in August 1988. Central bank debt grew through the issuance of short-term paper (CEDEPS) to financial entities for purposes of monetary control. The central bank accelerated its placement of CEDEPS, which by midyear were being issued to finance interest payments on the central bank’s own debt. By mid-1989 the quasi-fiscal deficit of the central bank reached almost 30 percent of GDP, although most of it was reversed by year-end. On January 1, 1990, the government announced the bond conversion of time deposits and public sector debt coming due in 1990 (Bonex 89). The central bank kept liquidity tight and by end-February interest rates reached more than 1,000 percent a month for seven-day term deposits.
Argentina19951727.1−2.8After the Mexican devaluation, a small bond trader experienced a liquidity squeeze pushing it to closure by mid-January 1995. This development persuaded most banks to cut credit to bond traders, which, in turn, affected banks with large bond and open trading positions. Furthermore, provincial banks were having difficulties raising funds and people started moving funds to larger banks, particularly foreign banks that were perceived as more solvent, and by March 1995 capital flight intensified. Several measures were implemented to alleviate liquidity pressures. Eight banks were suspended and three banks collapsed. Out of the 205 banks in existence as of end-1994, 63 exited the market through mergers, absorptions, or liquidation by end-1997.
Argentina200120.19.642.7−10.9In March 2001, a bank run began caused by increasing doubts about the sustainability of the currency board, strong opposition from the public to the new fiscal austerity package sent to the Congress, the resignation of the president of the central bank, and the amendment to the convertibility law (change in parity from being pegged to the U.S. dollar, to being pegged to a basket composed of the U.S. dollar and the euro). During the second half of 2001, bank runs intensified. On December 3, 2001, as several banks were on the verge of collapsing, partial withdrawal restrictions (corralito) were imposed on transactional accounts while fixed-term deposits (CDs) were reprogrammed (corralon) to stop outflows from banks. On February 4, 2002, bank assets were asymmetrically pesified, adversely affecting the solvency of the banking system. In 2002, two voluntary swaps of deposits for government bonds were offered but received little interest by the public. In December 2002, the corralito was lifted. By August 2003, one bank had been closed, three banks nationalized, and many others had reduced their staff and branches.
Armenia19943.3Starting in August 1994, the central bank closed half of active banks. Large banks continued to suffer from high NPLs. The savings banks were financially weak.
Azerbaijan1995−13.0Twelve private banks were closed; three large state-owned banks were deemed insolvent; one large state-owned bank faced serious liquidity problems.
Bangladesh19872034.72.4In 1987, four banks accounting for 70 percent of credit had NPLs of 20 percent. From the late 1980s, the entire private and public banking system was technically insolvent.
Belarus1995−11.3Many banks were undercapitalized; forced mergers burdened some banks with poor loan portfolios.
Benin198880171.9−2.8All three commercial banks collapsed.
Bolivia1986300−2.6Five banks were liquidated. Banking system NPLs reached 30 percent in 1987; in mid-1988 reported arrears stood at 92 percent of commercial banks’ net worth.
Bolivia19946.2604.4Two banks with 11 percent of banking system assets were closed in 1994. In 1995, 4 of 15 domestic banks, accounting for 30 percent of banking system assets, experienced liquidity problems and suffered high NPLs.
Bosnia and Herzegovina1992−6.4Banking system suffered from high NPLs caused by the breakup of the former Yugoslavia and the civil war.
Brazil1990012.2−4.2Deposits were converted to bonds. Liquidity assistance was provided to public financial institutions.
Brazil19941613.202.1The Brazilian economy entered a new phase with the implementation of the “Plan Real” in July 1994. The plan triggered major structural changes, which aimed primarily at lowering inflation. With this process, remonetization of the economy took place and with it, liabilities and assets of banks expanded rapidly—loans to the private sector grew by 60 percent during the first year of the plan—despite higher reserve requirements. At the same time, a sharp deterioration in the trade account took place, to which the central bank responded by raising interest rates and imposing credit restrictions. The financial situation of banks weakened as bad loans increased noticeably and also because the banks lost their inflation revenues. The problems were particularly acute at public banks. For federal banks, the ratio of loans in arrears and in liquidation to total loans increased from 15.4 percent in June 1994 to 22.4 percent at end-1995, and to slightly more than 30 percent in October 1996. For government-owned banks, the ratio increased from 8 percent to almost 12 percent and more than 14 percent for the same dates. For privately held banks, the ratio increased from 5 percent in June 1994 to 9 percent in December 1995. The problems in the banking sector triggered a restructuring of government-owned banks and the resolution of private institutions. Most of the closures were small to medium banks, while large banks were resolved under a “good bank/bad bank” approach, which separating the performing assets, good bank, from the rest, bad bank.
Bulgaria199675141.3−8.0The 1996 banking crisis had its roots in bad loans made during 1991–95, but the deepening insolvency of the system was not reflected in sustained liquidity problems until the second half of 1994. Two ailing state banks required ongoing refinancing from the Bulgarian National Bank (BNB) and the State Savings Bank (SSB) until they were bailed out in mid-1995. The public began to lose confidence in banks after the collapse of pyramid schemes in some cities, and in response to reports about the ill health of other banks. In late 1995, withdrawals of deposits, especially from First Private Bank (the largest private bank), were reflected in substantial BNB refinancing and falling foreign reserves. By early 1996, the sector had a negative net worth equal to 13 percent of GDP. The banking system experienced a run in early 1996. The government then stopped providing bailouts, prompting the closure of 19 banks accounting for one-third of sector assets. Surviving banks were recapitalized by 1997.
Burkina Faso19901645.2−0.6In 1989, the system of sectoral credit ratios was abolished, and deposit and lending rates were partially liberalized. During 1990, the financial condition of the banking sector deteriorated sharply. NPLs increased to 23 percent of total credit, and commercial banks’ deposits in the money market declined sharply. Three major commercial banks urgently needed restructuring, and two other large banks continued to experience liquidity problems. In 1991, the government merged these three major commercial banks into one bank with minority government participation and rehabilitated the two other banks, while assuming their nonperforming assets.
Burundi19942566.3−8.0In 1995 one bank was liquidated.
Cameroon198765118.1−7.9Five commercial banks were closed and three banks were restructured.
Cameroon19953003.3Three banks were restructured and two were closed.
Cape Verde19933006.7In 1993, the former monobank was split into a central bank and a commercial bank with 90 percent of banking system deposits. The commercial bank had accumulated a large fraction of nonperforming assets and was recapitalized by the government in 1994 by converting its portfolio of NPLs into interest-bearing notes equivalent to 17.5 percent of GDP. All commercial banking interest rates were liberalized in 1994, with the exception of one benchmark interest rate on time deposits.
Central African Republic197602.5Four banks were liquidated.
Central African Republic1995401.1−8.1The two largest banks, accounting for 90 percent of assets, were restructured.
Chad198305.3All banking offices closed in 1979 and 1980 when N’Djamena was the scene of heavy fighting. The banking sector experienced solvency problems. With the collapse of world cotton prices in 1985, Cotontchad’s revenues dropped, and foreign exchange flowing into Chad declined. As a result, the Bank of Central African State’s (BEAC’s) exchange reserves dropped precipitously in 1986. Operations in the banking sector ground to a halt as Cotontchad fell into arrears on repayments of its short-term debt. In late 1986, the BEAC negotiated a rescheduling of some three-fourths of the short-term debt, allowing a ten-year maturity, including a five-year grace period with an interest rate of 6 percent. In 1983, the government imposed a five-year moratorium that froze all deposits and outstanding credits before 1980. The moratorium’s purpose was to prevent a run on banks and to staunch capital flight when banks restored operations in early 1983 under the new government.
Chad19923537.2−2.1The Chadian banking system came close to collapse in 1992, owing mainly to the vulnerable state of the economy and an expansionary credit policy. To avoid a major financial crisis, the monetary authorities embarked on a comprehensive rehabilitation program of the banking system, involving enhancement of central bank supervision through the Commission Bancaire de l’Afrique Centrale, and the liberalization of banking activity. In addition, they eased the liquidity crisis of the commercial banks in 1993 by consolidating into a long-term loan to the government the rediscounted commercial bank loans that had been extended mainly to public enterprises. Credit policy was tightened; the amount of direct advances to the treasury by the central bank was stabilized; and the Banque Internationale pour le Commerce et 1’Industrie du Tchad was liquidated. As a result, the net foreign assets position of the banking system was strengthened and the liquidity position of the banks was gradually restored.
Chile197603.5Entire mortgage system was insolvent.
Chile198135.642.992.4−13.6By the end of 1981, a six-year expansionary period ended abruptly. High international interest rates, a world recession, lower copper prices, and an abrupt cut of voluntary foreign credit to Latin America pushed Chile into a costly economic crisis. The problems were aggravated by unsound financial practices among banks, which included substantial connected lending ranging from 12 percent to 45 percent of the total loans portfolio. The financial system was affected in two waves. The first wave in 1981–82 included 11 liquidations (banks and finance companies), and all depositors were protected. The second wave was in 1983 and involved liquidations and rehabilitations. For the liquidations, domestic depositors were compensated only partially. Although foreign creditors were offered the same compensation, they threatened to cut trade credit lines and their claims were ultimately restructured under the external debt-restructuring plan.
China1998201836.87.6At the end of 1998, China’s four large state-owned commercial banks, accounting for 68 percent of banking system assets, were deemed insolvent. Banking system NPLs in 2002 and 2003 were 20 percent and 15 percent, respectively, of total loans. The restructuring costs are estimated to have reached about 1.8 trillion yuan based on estimates of capital injections and loans to AMCs to purchase assets, or 18 percent of 2002 GDP.
Colombia19824.1515.10.9During the early 1980s, an economic downturn affected the profitability of the banks. They came under pressure as the 1981 recession intensified. This, in turn, caused a sharp deterioration in asset quality through an increase in defaults. Colombia began experiencing capital outflows. Subsequent bank failures and nationalizations generated widespread decline in public confidence, which led to a massive government intervention. The central bank intervened in six banks accounting for 25 percent of banking system assets, and in eight finance companies.
Colombia1998146.333.5−4.2A capital account reversal during the first half of 1998 triggered by pressures in emerging markets led to a response from the central bank oriented toward defending the currency. As a result, interest rates increased in real terms, harming the quality of banks’ loan portfolios and putting downward pressure on asset prices and hence on the value of collateral, especially real estate. The already-weak large public banks faced severe asset quality deterioration, which spread to private banks and other financial entities.
Congo, Dem.198300.5Banking sector experienced solvency problems.
Republic of Congo, Dem.199181.0−13.5Four state-owned banks were insolvent; a fifth bank was recapitalized with private participation.
Republic of Congo, Dem.1994750−5.4Two state-owned banks were liquidated and two other state banks privatized. In 1997, 12 banks were having serious financial difficulties.
Republic of Congo,199263.2−5.5Two large banks were restructured and privatized. The remaining insolvent bank was liquidated. The situation was aggravated by the civil war.
Republic of Costa Rica198703.4In 1987, public banks accounting for 90 percent of total banking system loans were in financial distress because 32 percent of their loans were considered uncollectible. Implied losses were at least twice capital plus reserves. Pressure was put on the banks to negotiate a “Brady-type” settlement of foreign debt; settlement was reached in November 1989 at 16 cents on the dollar. Budgetary relief to the government enabled restructuring of state bank debts.
Costa Rica1994321.60.9One large state-owned commercial bank with 17 percent of deposits was closed in December 1994. The ratio of overdue loans (net of provisions) to net worth in state commercial banks exceeded 100 percent in June 1995. Implied losses were at least twice capital plus reserves.
Côte d’Ivoire198850250−1.1The recession of 1987 and problems with the cocoa and coffee markets (the country’s main exports) substantially increased the private sector’s NPLs. These problems were aggravated by a large amount of NPLs in the public enterprise sectors, the large accumulation of government payment arrears, the substantial decline in public and private deposits in the banking system, reduction in credit lines from abroad, and poor management in some banks. Four large banks were affected, accounting for 90 percent of banking system loans; three definitely and one possibly insolvent. Six government banks closed.
Croatia199810.56.90−0.9The introduction of a market-oriented legal framework in the early 1990s led to significant progress in establishing a modern banking system. The banking sector expanded vigorously until end-1997. Meanwhile, the incentives for sound bank behavior had not yet been fully established, and bad debt problems had been inherited from the old regime. These weaknesses were in part addressed with the Bank Rehabilitation Plan (Law of 1994) implemented in 1996–97. Four state-owned banks, accounting for 46 percent of total bank assets (as of 1995) entered rehabilitation, with an overall cost of 6.1 percent of GDP. However, a new wave of problems began in March 1998 with the failure of the fifth largest bank, Dubrovacka (5 percent of total assets). Problems at this bank triggered political turmoil, which, in turn, induced runs at other banks, perceived to be indirectly related to Dubrovacka. In July 1998, the sixth largest bank ran into problems and several small and medium institutions experienced liquidity difficulties in the fall of 1998 and early 1999.
Czech Republic1996186.8−0.8In 1994, a small bank (Banka Bohemia) failed because of fraud. Although all depositors were covered, partial deposit insurance coverage was introduced shortly after this first failure. The likelihood of facing material losses triggered runs at other small banks; by the end of 1995, two small banks failed (Ceska and AB Banka), which triggered a second phase of bank restructuring starting in 1996, aimed at 18 small banks (9 percent of industry assets).
Djibouti199122.6−6.7Two of six commercial banks ceased operations in 1991–92; other banks experienced difficulties.
Dominican Republic200392215.5−1.9In April 2003, the central bank took over Baninter (Banco Intercontinental), which declared bankruptcy in May and dissolved in July. Baninter’s liabilities exceeded its assets by 55 billion pesos (US$2.2 billion) and 15 percent of GDP. The central bank had been providing liquidity support to Baninter since September 2002. Two other banks, Bancredito and Banco Mercantil, were also given liquidity support from the central bank to deal with deposit withdrawals.
Ecuador198213.6−2.8A program was implemented that exchanged domestic for foreign debt to bail out the banking system.
Ecuador19984021.76.5−6.3Seven financial institutions, accounting for 25–30 percent of commercial banking assets, were closed in 1998–99. In March 1999, bank deposits were frozen for six months. By January 2000, 16 financial institutions accounting for 65 percent of system assets had either been closed (12) or taken over (4) by the government. All deposits were unfrozen by March 2000. In 2002, the blanket guarantee was lifted.
Egypt198038.12.2The government closed several large investment companies.
El Salvador19893701.0Nine state-owned commercial banks had NPLs averaging 37 percent.
Equatorial19830−2.3Two of the country’s largest banks were liquidated.
Guinea
Eritrea19932.3Most of the banking system was insolvent.
Estonia199271.9−21.6Banking problems surfaced in November 1992 when the state-owned North Estonian Bank, the Union Baltic Bank, and the Tartu Commercial Bank exhibited serious liquidity problems and delayed payments by three weeks. A second stress episode took place in early 1994, when the government reduced the level of its deposits from the Social Bank. The Social Bank, which controlled 10 percent of financial system assets, failed. Five banks’ licenses were revoked, and two major banks were merged and nationalized. Two other large banks were merged and converted to a loan recovery agency.
Finland19911312.859.1−6.2The three Nordic countries went through a financial liberalization process that led to a lending boom. However, they also suffered the adverse consequences of higher German interest rates. In Finland, the problems were exacerbated by the collapse of exports to the former Soviet Union. The first bank in trouble was Skopbank, which was taken over by the central bank in September 1991. Savings banks were badly affected; the government took control of three banks that together accounted for 31 percent of system deposits.
Georgia199133−44.9Largest banks virtually insolvent.
Ghana198235615.8−6.9During most of the 1980s Ghana suffered severe structural imbalances related to the cumulative impact of large budgetary deficits, rapid increases in domestic bank credit, a fixed exchange rate, and high inflation, which authorities aimed to control through price controls. The effects of these policies were exacerbated by a deterioration of capital equipment and inadequate price incentives in the agricultural and export sectors. As a result, real output in 1981 was 15 percent lower than its 1974 level. The situation deteriorated further toward the second half of the 1980s because of high fiscal deficits, financed primarily through domestic credit; directed credit policies (since 1981 banks had been obliged to lend at least 20 percent of their portfolios to the agricultural sector); a deterioration in cocoa exports; and a large depreciation of the currency (a 1,173 percent depreciation took place in 1983). Banks experienced liquidity pressures, but there were also deficiencies in banking supervision and regulation. As a result, 7 out of the 11 banks were insolvent and the problems were addressed by recapitalization and purchase of NPLs.
Guinea1985303.1Six banks—accounting for 99 percent of system deposits—deemed insolvent. Repayment of deposits amounted to 3 percent of 1986 GDP.
Guinea19934504.0Two banks were deemed insolvent; one other bank had serious financial difficulties.
Guinea-Bissau19954522.8−27.2At end-1996, the central bank had a negative capital position and Guinea-Bissau’s two commercial banks had substantial NPLs. In March-April 1997, the treasury recapitalized the central bank.
Guyana199305.1Before financial reforms started in 1989, directed credit programs had resulted in investments with low rates of return and large NPLs for the banks. State-owned banks were merged in May 1995 and a state-owned loan-recovery institution was subsequently established to recover the NPLs of the merged bank.
Haiti19949.3−11.6The central bank registered considerable losses because the majority of its assets, represented by credit to the government, were nonperforming.
Hungary19912310−11.9In the second half of 1993, eight banks (25 percent of financial system assets) were deemed insolvent.
India1993203.14.9Nonperforming assets reached 11 percent in 1993–94. Nonperforming assets of the 27 public banks were estimated to be 20 percent in 1995. At the end of 1998, NPLs were estimated to be 16 percent and at the end of 2001 they decreased to 12.4 percent.
Indonesia199732.556.867.9−13.1Through May 2002, Bank Indonesia had closed 70 banks and nationalized 13, of a total of 237. Official NPLs for the banking system were estimated to be 32.5 percent of total loans at the peak of crisis.
Israel19773001.0Almost the entire banking sector was affected, representing 60 percent of stock market capitalization. The stock exchange closed for 18 days, and bank share prices fell more than 40 percent.
Jamaica199628.943.930.1−1.2In 1994, a merchant banking group was closed. In 1996, Financial Sector Adjustment Company, a government resolution agency, provided assistance to five banks, five life insurance companies, two building societies, and nine merchant banks. The government recapitalized 21 troubled institutions using nontradable government-guaranteed bonds. By June 30, 2000, outstanding recap bonds were estimated to account for 44 percent of GDP.
Japan1997351417.6−2.0Banks suffered from sharp declines in stock market and real estate prices. In 1995, the official estimate of NPLs was 40 trillion yen (US$469 billion, or 10 percent of GDP). An unofficial estimate put NPLs at US$1 trillion, equivalent to 25 percent of GDP. Banks made provisions for some bad loans. At the end of 1998, banking system NPLs were estimated at 88 trillion yen (US$725 billion, or 18 percent of GDP). In 1999, Hakkaido Takushodu Bank was closed, the Long Term Credit Bank was nationalized, Yatsuda Trust was merged with Fuji Bank, and Mitsui Trust was merged with Chuo Trust. In 2002, NPLs were 35 percent of total loans; a total of 7 banks had been nationalized, 61 financial institutions closed, and 28 institutions merged. In 1996, rescue costs were estimated to be more than US$100 billion. In 1998, the government announced the Obuchi Plan, which provided 60 trillion yen (US$500 billion, or 12 percent of GDP) in public funds for loan losses, bank recapitalizations, and depositor protection. Fiscal cost rose to 14 percent of GDP.
Jordan19891066.6−10.7The third largest bank failed in August 1989. The central bank provided overdrafts equivalent to 10 percent of GDP to meet a run on deposits and allowed banks to settle foreign obligations.
Kenya198504.1Four banks and 24 nonbank financial institutions—accounting for 15 percent of financial system liabilities—faced liquidity and solvency problems.
Kenya199223.0−1.1Intervention occurred in two local banks.
Korea, Republic of19973531.250.1−6.9The devaluation of the Thai baht in July 1997, the subsequent regional contagion, and the crash of the Hong Kong stock market sent shock waves through the Korean financial system. Korea’s exchange rate remained broadly stable through October 1997. However, the high level of short-term debt and the low level of usable international reserves made the economy increasingly vulnerable to shifts in market sentiment. Although macroeconomic fundamentals continued to be favorable, the growing awareness of problems in the financial sector and in industrial groups led to increasing difficulties for the banks in rolling over their short-term borrowing. Through May 2002, five banks were forced to exit the market through purchase and assumption and 303 financial institutions were shut down (215 were credit unions); another four banks were nationalized.
Kuwait1982400−9.5Share dealings using postdated checks created a huge unregulated expansion of credit. The crash of the unofficial stock market finally came in 1982, when a dealer presented a postdated check for payment and it bounced. A house of cards collapsed. An official investigation revealed that total outstanding checks amounted to the equivalent of US$94 billion from about 6,000 investors. Kuwait’s financial sector was badly shaken by the crash, as was the entire economy. The crash prompted a recession that rippled through the society as individual families were disrupted by the investment risks taken by particular family members using family credit. The debts from the crash left all but one bank in Kuwait technically insolvent, which survived only because of support from the central bank. Only the National Bank of Kuwait, the largest commercial bank, survived the crisis intact. In the end, the government stepped in, devising a complicated set of policies embodied in the Difficult Credit Facilities Resettlement Program. The implementation of the program was still incomplete in 1990 when the Iraqi invasion changed the entire financial picture.
Kyrgyz Republic199585−5.8In 1995, more than half the commercial banks had a negative net worth. The public lost confidence in the banking system, and many people withdrew their funds, causing many of the banks to go out of business. The licenses of five small banks were withdrawn in 1994–95. Two banks were closed in 1999, following the Russian crisis.
Latvia1995203−2.1Between 1994 and 1999, 35 banks saw their license revoked, were closed, or ceased operations. In 1995, the negative net worth of the banking system was estimated at US$320 million, or 7 percent of 1995 GDP. Aggregate banking system losses in 1998 were estimated to be 100 million lats (US$172 million), about 3 percent of GDP.
Lebanon19904.2−13.4Four small and medium banks became insolvent and 11 had to resort to significant central bank lending. Bank of Lebanon claims on commercial banks reached 31 percent of reserve money in September 1990.
Liberia19910Seven of eleven banks were not operational; in mid-1995, their assets accounted for 64 percent of bank assets.
Lithuania199532.23.11.2In 1995, of 25 banks, 12 small banks were liquidated, 3 private banks (accounting for 29 percent of banking system deposits) failed, and 3 state-owned banks were deemed insolvent.
Macedonia, former Yugoslav Republic of19937032−7.5The government took over banks’ foreign debts and closed the second largest bank. The costs of banking system rehabilitation, obligations from assumption of external debt, liabilities for frozen foreign exchange, and contingent liabilities in banks together were estimated to be 32 percent of GDP.
Madagascar1988250−6.3After the formal abandonment in 1985 of the previous policy of bank specialization and the appointment in 1986 of separate boards of directors to replace the single board that was shared by all commercial banks, the rehabilitation of the banking system gained speed with the enactment in 1988 of a new banking law, which opened the system to private capital, and the decision in 1989 to write off most of the NPLs of the existing banks.
Malaysia19973016.450.0−7.4The persistent pace of credit expansion to the private sector at an annual rate of nearly 30 percent, particularly to the property sector and for the purchase of stocks and shares, exposed the financial system to potential risks from price declines in property and other assets that occurred in 1997. In the wake of market turbulence and contagion effects in the second half of 1997, concerns among market participants about the true condition and resilience of the financial system increasingly became a central issue, highlighted by the known fragilities among finance companies. The finance company sector was restructured, and the number of finance companies was reduced from 39 to 10 through mergers. Two finance companies, including the largest independent finance company, were taken over by the central bank. Two banks were deemed insolvent—accounting for 14 percent of financial system assets—and were merged with other banks. NPLs peaked between 25 percent and 35 percent of banking system assets and fell to 10.8 percent by March 2002.
Mali1987755.7−0.3Mali’s economic and financial prospects for 1986 and the medium term changed significantly because of the collapse in late 1985 of the world market price of cotton, Mali’s major export commodity. In 1987, although the government undertook some corrective measures, the economic and financial situation deteriorated rapidly. The expansion of credit was significantly higher than programmed, and as a result, NPLs at banks increased rapidly. Owing primarily to the overexposure of the largest commercial bank in its loans and guaranteed letters of credit, a liquidity crunch emerged in the banking system. The financial situation of the largest commercial bank deteriorated further in 1987, reflecting the heavy losses of the public enterprise sector that it had financed over the years, defaults by the private sector on unsecured loans, and inappropriate management. By mid-November 1997, the bank had become virtually illiquid and ceased functioning normally. Its NPLs amounted to some 70 percent of its outstanding credit.
Mauritania1984701502.0In 1984, five major banks had nonperforming assets ranging from 45 percent to 70 percent of their portfolios.
Mexico198151.3−3.5The government took over the troubled banking system.
Mexico199418.919.34.2−6.2Of 34 commercial banks in 1994, 9 were intervened in and 11 participated in the loan or purchase recapitalization program. The 9 banks that were intervened in accounted for 19 percent of financial system assets and were deemed insolvent. By 2000, 50 percent of bank assets were held by foreign banks.
Morocco198029.8−2.8The banking sector experienced solvency problems. A debt crisis occurred in 1980–83.
Mozambique198701.0The main commercial bank experienced solvency problems that became apparent after 1992.
Nepal19882904.3NPLs increased sharply during 1988–89 at the two largest commercial banks. Both banks were majority government owned and together accounted for more than 90 percent of bank assets and deposits. In 1989, loan recovery programs were put in place for these two commercial banks.
Nicaragua1990500−0.4During the 1980s, lending rates were subsidized and often set below deposit rates. Deposit rates were for the most part negative in real terms and contributed to a severe contraction of the banks’ deposit base. The central bank provided much of the funding for commercial banks, mainly by intermediating foreign loans and donations. Lack of bank supervision and prudential controls resulted in risky lending and contributed to the large percentage of NPLs in banks’ portfolios. In 1990, financial sector problems were acknowledged by a new government. In 1992, a financial reform package was announced to confront these problems. The state banking system was recapitalized and reorganized starting in 1992.
Nicaragua200012.713.600.8The largest bank in Nicaragua, Interbank, was found to have committed fraud and therefore was intervened in August 2000. Following the intervention, full protection for its depositors was announced. However, withdrawals continued until the bank was finally resolved in October 2000 through a purchase and assumption. Another institution ran into problems soon after the resolution of Interbank. Runs against other banks occurred in part because the authorities announced limited coverage of its depositors. However, in its resolution a few days later, all depositors were protected. Two additional institutions were resolved a few months later. All banks were resolved under purchase and assumptions and a blanket guarantee was passed by law after the first two failures.
Niger198350122.7−16.8In the mid-1980s, banking system NPLs reached 50 percent. Four banks were liquidated and three restructured in the late 1980s. In 2002, a new round of bank restructuring was launched. Four banks were experiencing serious difficulties.
Nigeria1991770.4−0.6In 1993, insolvent banks accounted for 20 percent of banking system assets and 22 percent of deposits. In 1995, almost half the banks reported being in financial distress.
Norway199116.42.702.8Financial deregulation undertaken during 1984–87 led to a credit boom (with real rates of credit growth of 20 percent year-over-year), accompanied by a boom in both residential and nonresidential real estate. In 1985, oil prices fell sharply, turning a 4.8 percent surplus in the current account into a 6.2 percent deficit in 1986 with ensuing pressures on the exchange rate. Meanwhile, rate increases by the Bundesbank following the reunification of Germany forced Norway to keep interest rates high throughout the economic recession, which started in 1988. Problems at small banks that began in 1988 were addressed via mergers and assistance from the guarantee fund, funded by banks. However, by 1990 the fund had been depleted and the financial condition at large banks began to deteriorate. The turmoil reached systemic proportions by October 1991, when the second and fourth largest banks had lost a considerable amount of equity.
Panama198837.8−13.4As a result of severe United States–led economic sanctions, including the freezing of Panamanian assets in U.S. banks, a nine-week bank holiday was declared beginning in March 1988. As a result of these developments, the financial position of most state-owned and private commercial banks was substantially weakened and 15 banks ceased operations.
Paraguay19958.112.900.4During the early 1990s, the banking system remained undercapitalized and NPLs rose sharply, coupled with insider lending practices. As early as 1989, an assessment by the superintendency revealed that about one-third of the banking system was insolvent. The crisis began in May 1995 when the third and fourth largest banks could not meet clearing obligations and were intervened in. The first line of response was liquidity support. However, as the crisis unfolded, an important amount of unrecorded deposits were discovered. A blanket guarantee covering intervened-in banks was announced, but pressures remained because at first, the guarantee covered only legitimate deposits, although later, all deposits were protected. Through a series of interventions, closures, and substantial liquidity support, the distress period lasted until 1999. In the end, between 1995 and 1999, 15 out of the 19 locally owned banks were either closed or absorbed by stronger institutions. By 1999, banks were predominantly foreign owned.
Peru198325.5−9.3Two large banks failed. The rest of the system suffered from high NPLs and financial disintermediation following the nationalization of the banking system in 1987.
Philippines198319360.1−7.3Problems occurred in two public banks accounting for 50 percent of banking system assets, six private banks accounting for 12 percent of banking system assets, 32 thrifts accounting for 53 percent of thrift banking assets, and 128 rural banks.
Philippines19972013.20−0.6Since January 1998 one commercial bank, 7 of 88 thrifts, and 40 of 750 rural banks have been placed under receivership. Banking system NPLs reached 12 percent by November 1998, and 20 percent in 1999.
Poland1992243.52.0In 1991, seven of nine treasury-owned commercial banks—accounting for 90 percent of credit—the Bank for Food Economy, and the cooperative banking sector experienced solvency problems.
Romania1990300.6−12.9In 1998, NPLs reached 25–30 percent in the six main state-owned banks. The Agricultural Bank was recapitalized on a flow basis. In 1998, the central bank injected US$210 million in Bancorex (0.6 percent of GDP), the largest state bank, and in 1999 another US$60 million.
Russian Federation19984060−5.3From mid-1997 to April 1998, the Central Bank of Russia (CBR) was relatively successful in defending the fixed exchange rate policy through a significant tightening of credit. However, the situation became increasingly untenable when significant political turmoil in Russia—starting with the president’s dismissal of the government of Prime Minister Chernomyrdin and prolonged by a stalemate over the formation of a new cabinet—cast increasing doubt on the political resolve to come to grips with Russia’s fiscal problems. From mid-July, when the Duma refused to pass key fiscal measures, the situation deteriorated rapidly, leading to a unilateral restructuring of ruble-denominated treasury bills and bonds on August 17, 1998. The ruble was allowed to float three days later despite previous announcements that it would not be devalued. A large devaluation in real effective terms (more than 300 percent in nominal terms), loss of access to international capital markets, and massive losses to the banking system ensued. However, well before the crisis, there was widespread recognition that the banking system had a series of weaknesses. In particular, bank reporting and bank supervision were weak, there was excessive exposure to foreign exchange rate risk, connected lending, and poor management. Two key measures implemented were a 90-day moratorium on foreign liabilities of banks and the transfer of a large fraction of deposits from insolvent banks to Sberbank. Nearly 720 banks, or half of those then operating, were deemed insolvent. These banks accounted for 4 percent of sector assets and 32 percent of retail deposits.
São Tomé and Príncipe19929000.7At the end of 1992, 90 percent of the monobank’s loans were nonperforming. In 1993, the commercial and development departments of the former monobank were liquidated, as was the only financial institution. At the same time, two new banks were licensed that took over many of the assets of their predecessors. The credit operations of one of the new banks were suspended in 1994.
Senegal1988501725.4−0.7In 1988, 50 percent of banking system loans were nonperforming. Six commercial banks and one development bank closed, accounting for 20–30 percent of financial system assets.
Sierra Leone19904532.6−9.6One bank’s license was suspended in 1994.
Slovak Republic19983500In 1998, NPLs reached 35 percent of total loans and a bank-restructuring program was put in place involving the major state-owned banks.
Slovenia199214.61.0−5.5Three banks—accounting for two-thirds of banking system assets—were restructured.
Spain19775.60.2In 1978–83, 24 institutions were rescued, 4 were liquidated, 4 were merged, and 20 small and medium banks were nationalized. These 52 banks (of 110), representing 20 percent of banking system deposits, were experiencing solvency problems.
Sri Lanka19893552.22.3State-owned banks comprising 70 percent of banking system assets were estimated to have NPLs of about 35 percent. The government recapitalized two large state-owned banks, Bank of Ceylon and the People’s Bank, in 1993 (representing two-thirds of banking system assets) to solve their solvency problem caused by high nonperforming assets.
Swaziland199521.62.7Meridien BIAO Swaziland was taken over by the central bank. The central bank also took over the Swaziland Development and Savings Bank, which faced severe portfolio problems.
Sweden1991133.630.6−1.2Nordbanken and Gota Bank, accounting for 22 percent of banking system assets, were insolvent. Sparbanken Foresta, accounting for 24 percent of banking system assets, was intervened. Overall, five of the six largest banks, with more than 70 percent of banking system assets, experienced difficulties.
Tanzania1987701003.8In 1987, the main financial institutions had arrears amounting to half their portfolios. In 1995, it was determined that the National Bank of Commerce, which accounted for 95 percent of banking system assets, had been insolvent since at least 1990.
Thailand19830.79.44.6Authorities intervened in 50 finance and security firms and 5 commercial banks (25 percent of financial system assets); 3 commercial banks were deemed insolvent (accounting for 14 percent of commercial bank assets). The fiscal costs for the 50 finance companies were estimated to be 0.5 percent of GDP; the fiscal cost for subsidized loans amounted to about 0.2 percent of GDP a year.
Thailand19973343.897.7−10.5Under the framework of a pegged exchange rate regime, Thailand had enjoyed a decade of robust growth performance, but by late 1996 pressures on the baht emerged. Pressure increased through the first half of 1997 amidst an unsustainable current account deficit; a significant appreciation of the real effective exchange rate; rising short-term foreign debt; a deteriorating fiscal balance; and increasingly visible financial sector weaknesses, including large exposure to the real estate sector, exchange rate risk, and liquidity risk. Finance companies had a disproportionately large exposure to the property sector and were the first institutions affected by the economic downturn. Following mounting exchange rate pressures and ineffective interventions to alleviate these pressures, the baht was floated on July 2, 1997. In light of weak supportive policies, the baht depreciated by 20 percent against the U.S. dollar in July. By May 2002, the Bank of Thailand had closed 59 (of 91) financial companies that in total accounted for 13 percent of financial system assets and 72 percent of finance company assets. It closed 1 (out of 15) domestic banks and nationalized 4 banks. A publicly owned AMC held 29.7 percent of financial system assets as of March 2002. NPLs peaked at 33 percent of total loans and were reduced to 10.3 percent of total loans in February 2002.
Togo199327.7−16.3The banking sector experienced solvency problems.
Tunisia1991302.2In 1991, most commercial banks were undercapitalized. During 1991–94, the banking system raised equity equivalent to 1.5 percent of GDP and made provisions equivalent to another 1.5 percent.
Turkey19822.503.4Three banks were merged with the state-owned Agriculture Bank and then liquidated; two large banks were restructured.
Turkey200027.6325.4−5.7Banks had high exposures to the government through large holdings of public securities and sizable maturity and exchange rate risk mismatches, making them highly vulnerable to market risk. In November 2000, interbank credits to some banks holding long-term government paper were cut, forcing them to liquidate the paper, which caused a sharp drop in the price of such securities, triggering a reversal in capital flows, a sharp increase in interest rates, and a decline in the value of the currency. Two banks closed and 19 banks were taken over by the Savings Deposit Insurance Fund.
Uganda199405.5Between 1994 and 1998, half of the banking system faced solvency problems. In 1998, two banks were closed and one was recapitalized and privatized. In 1999, two banks were closed. In 2002, one small bank was intervened in and two other banks were experiencing difficulties.
Ukraine199862.400−1.9Between 1995 and 1997, 32 of 195 banks were liquidated, and 25 others were undergoing financial rehabilitation. Bad loans accounted for 50–65 percent of assets even in some leading banks. In 1998, banks were further hit by the government’s decision to restructure government debt following the Russian debt crisis.
United Kingdom2007On September 14, 2007, Northern Rock, a mid-sized U.K. mortgage lender, received a liquidity support facility from the Bank of England, following funding problems related to global turmoil in credit markets caused by the U.S. subprime mortgage financial crisis. Starting on September 14, 2007, Northern Rock experienced a bank run, until a government blanket guarantee—covering only Northern Rock—was issued on September 17, 2007. On February 22, 2008, the bank was nationalized, following two unsuccessful bids to take it over. On April 21, 2008, the Bank of England announced it would accept a broad range of mortgage-backed securities and swap those for government paper for a period of one year to aid banks with liquidity problems. The scheme enabled banks to temporarily swap high quality but illiquid mortgage-backed assets and other securities for treasury bills for one year.
United States19884.13.74.1−0.2More than 1,400 savings and loan institutions and 1,300 banks failed. Cleaning up savings and loan institutions cost $180 billion, or 3.7 percent of GDP.
United States2007During the course of 2007, U.S. subprime mortgage markets melted down and global money markets were under pressure. The U.S. subprime mortgage crisis manifested itself first through liquidity issues in the banking system owing to a sharp decline in demand for asset-backed securities. Hard-to-value structured products and other instruments created during a boom of financial innovation had to be severely marked down because of the newly implemented fair value accounting. Credit losses and asset writedowns got worse with accelerating mortgage foreclosures, which increased in late 2006 and worsened further in 2007 and 2008. On August 16, 2007, Countrywide Financial ran into liquidity problems because of the decline in value of securitized mortgage obligations, triggering a deposit run on the bank. The Federal Reserve Bank intervened by lowering the discount rate by 0.5 percentage points and by accepting $17.2 billion in repurchase agreements for mortgage-backed securities to aid in liquidity. On January 11, 2008, Bank of America bought Countrywide for US$4 billion. Bear Stearns, the fifth largest investment bank at the time, required an emergency government bailout and was purchased by JP Morgan Chase with federal guarantees on its liabilities in March 2008. Profits at U.S. banks declined to $5.8 billion from $35.2 (an 83.5 percent decline) during the fourth quarter of 2007 compared with the previous year, due to provisions for loan losses. By June 2008, subprime-related and other credit losses or writedowns by global financial institutions hovered around $400 billion. The Fed introduced the Term Securities Lending facility to swap a broad range of mortgage-backed securities for treasury notes for a period of one month. On September 7, 2008, mortgage giants Fannie Mae and Freddie Mac were placed under conservatorship.
Uruguay198131.287.5−9.3Affected institutions accounted for 30 percent of financial system assets; insolvent banks accounted for 20 percent of financial system deposits.
Uruguay200236.32028.8−11.0The introduction of capital controls and deposit freezes in Argentina in December 2001 triggered liquidity problems at the two largest private banks, Banco Galicia Uruguay (BGU) and Banco Comercial (BC) (with combined assets of 20 percent of the system total), as a result of their high level of exposure to the Argentinean economy. In January 2002 alone, BGU lost 15 percent of deposits. BGU was intervened in in February and later suspended. A second wave of deposit withdrawals ensued in April 2002, following Uruguay’s downgrade from investment-grade status. By May, the runs expanded to the public banks (Republica and Hipotecario), accounting for 40 percent of the system’s assets, which were in a weak condition with NPLs of 39 percent as of 2001 (compared with 6 percent at private banks).
Venezuela199424159.6−2.3Insolvent banks accounted for 35 percent of financial system deposits. In 1994, the authorities intervened in 17 of 47 banks that held 50 percent of deposits and nationalized 9 banks and closed 7 others. The government intervened in another five banks in 1995.
Vietnam1997351019.74.8Two of the four large state-owned commercial banks—accounting for 51 percent of banking system loans—were deemed insolvent; the other two experienced significant solvency problems. Several joint stock banks were in severe financial distress. Banking system NPLs reached 18 percent in late 1998.
Yemen, Republic of19962.43.8Banks suffered from extensive NPLs and heavy foreign currency exposure, leaving many banks technically insolvent. The 1994 civil war drained Yemen’s economy leading to a financial crisis in 1996.
Zambia19951.40.5−2.8Meridian Bank, accounting for 13 percent of commercial bank assets, became insolvent.
Zimbabwe19952.40.1Two of five commercial banks had high NPLs.
Sources: IMF Staff reports; country authorities’ reports; and newspaper articles.Note: — = not available; AMC = asset management company; NPL = nonperforming loan.
Sources: IMF Staff reports; country authorities’ reports; and newspaper articles.Note: — = not available; AMC = asset management company; NPL = nonperforming loan.

Currency Crises

Building on the approach in Frankel and Rose (1996), a “currency crisis” is defined as a nominal depreciation of the currency of at least 30 percent that is also at least a 10 percent increase in the rate of depreciation compared with the previous year. To measure exchange rate depreciation, the percentage change of the end-of-period official nominal bilateral dollar exchange rate from the IMF’s World Economic Outlook database is used. For countries that meet the criteria for several continuous years, the first year of each five-year window is used to identify the crisis. This definition yields 208 currency crises during the period 1970–2007. (This list also includes large devaluations by countries that had fixed exchange rate regimes.)

Sovereign Debt Crises

Episodes of sovereign debt default and restructuring are identified and dated by relying on information from Beim and Calomiris (2001); World Bank (2002); Sturzenegger and Zettelmeyer (2006); and IMF staff reports. The information compiled includes year of sovereign defaults to private lending and year of debt rescheduling. Using this approach, 63 episodes of sovereign debt defaults and restructurings are identified since 1970.

Appendix Table 2A.2 lists the complete set of starting years of systemic banking crises, currency crises, and sovereign debt crises.

Appendix Table 2A.2Timing of Financial Crises
CountrySystemic banking crisis (starting date)Currency crisis (year)Debt crisis (default date)Debt restructuring (year)
Albania1994199719901992
Algeria19901988, 1994
Angola1991, 199619881992
Argentina1980, 1989, 1995, 20011975, 1981, 1987, 20021982, 20011993, 2005
Armenia19941994
Azerbaijan19951994
Bangladesh19871976
Belarus19951994, 1999
Benin19881994
Bolivia1986, 19941973, 198119801992
Bosnia and Herzegovina1992
Botswana1984
Brazil1990, 19941976, 1982, 1987, 1992, 199919831994
Bulgaria1996199619901994
Burkina Faso19901994
Burundi1994
Cambodia1971, 1992
Cameroon1987, 1995199419891992
Cape Verde1993
Central African Republic1976, 19951994
Chad1983, 19921994
Chile1976, 19811972, 198219831990
China1998
Colombia1982, 19981985
Comoros1994
Congo, Dem. Republic of1983, 1991, 19941976, 1983, 1989, 1994, 199919761989
Congo, Rep. of1992199419861992
Costa Rica1987, 19941981, 199119811990
Côte d’Ivoire198819941984, 20011997, n.a.
Croatia1998
Czech Republic1996
Djibouti1991
Dominica2002n.a.
Dominican Republic20031985, 1990, 20031982, 20031994, 2005
Ecuador1982, 19981982, 19991982, 19991995, 2000
Egypt19801979, 199019841992
El Salvador19891986
Equatorial Guinea19831980, 1994
Eritrea1993
Estonia19921992
Ethiopia1993
Fiji1998
Finland19911993
Gabon19941986, 20021994
Gambia, The1985, 200319861988
Georgia19911992, 1999
Ghana19821978, 1983, 1993, 2000
Greece1983
Grenada20042005
Guatemala1986
Guinea1985, 19931982, 200519851992
Guinea-Bissau19951980, 1994
Guyana1993198719821992
Haiti19941992, 2003
Honduras199019811992
Hungary1991
Iceland1975, 1981, 1989
India1993
Indonesia19971979, 199819992002
Iran1985, 1993, 200019921994
Israel19771975, 1980, 1985
Italy1981
Jamaica19961978, 1983, 199119781990
Japan1997
Jordan1989198919891993
Kazakhstan1999
Kenya1985, 19921993
Korea, Republic of19971998
Kuwait1982
Kyrgyz Republic19951997
Lao P. D. R.1972, 1978, 1986, 1997
Latvia19951992
Lebanon19901984, 1990
Lesotho1985
Liberia19911980n.a.
Libya2002
Lithuania19951992
Macedonia, former Yugoslav Rep. of1993
Madagascar19881984, 1994, 200419811992
Malawi199419821988
Malaysia19971998
Maldives1975
Mali19871994
Mauritania19841993
Mexico1981, 19941977, 1982, 199519821990
Moldova199920022002
Mongolia1990, 1997
Morocco1980198119831990
Mozambique1987198719841991
Myanmar1975, 1990, 1996, 2001, 2007
Namibia1984
Nepal19881984, 1992
New Caledonia1981
New Zealand1975, 1984
Nicaragua1990, 20001979, 1985, 199019801995
Niger1983199419831991
Nigeria19911983, 1989, 199719831992
Norway1991
Pakistan1972
Panama198819831996
Papua New Guinea1995
Paraguay19951984, 1989, 200219821992
Peru19831976, 1981, 198819781996
Philippines1983, 19971983, 199819831992
Poland199219811994
Portugal1983
Romania1990199619821987
Russian Federation1998199819982000
Rwanda1991
São Tomé and Príncipe19921987, 1992, 1997
Senegal1988199419811996
Serbia, Republic of2000
Sierra Leone19901983, 1989, 199819771995
Slovak Republic1998
Slovenia1992
South Africa198419851993
Spain19771983
Sri Lanka19891978
Sudan1981, 1988, 199419791985
Suriname1990, 1995, 2001
Swaziland19951985
Sweden19911993
Syrian Arab Republic1988
Tajikistan1999
Tanzania19871985, 199019841992
Thailand1983, 19971998
Togo1993199419791997
Trinidad and Tobago198619891989
Tunisia1991
Turkey1982, 20001978, 1984, 1991, 1996, 200119781982
Turkmenistan1993
Uganda19941980, 198819811993
Ukraine1998199819981999
United Kingdom2007
United States1988, 2007
Uruguay1981, 20021972, 1983, 1990, 20021983, 20021991, 2003
Uzbekistan1994, 2000
Venezuela19941984, 1989, 1994, 200219821990
Vietnam19971972, 1981, 198719851997
Yemen, Republic of19961985, 1995
Yugoslavia, SFR19831988
Zambia19951983, 1989, 199619831994
Zimbabwe19951983, 1991, 1998, 2003
Source: Authors’ calculations.Note: No entry implies no crisis; SFR = Socialist Federal Republic.
Source: Authors’ calculations.Note: No entry implies no crisis; SFR = Socialist Federal Republic.

Frequency of Crises and Occurrence of Twin Crises

Table 2.1 reports the frequency of different types of crises (banking, currency, and sovereign debt), as well as the occurrence of twin (banking and currency) crises or triple (banking, currency, and debt) crises. A twin crisis in year t is defined as a banking crisis in year t, combined with a currency crisis during the period t − 1 through t + 1; and a triple crisis in year t is defined as a banking crisis in year t, combined with a currency crisis during the period t − 1 through t + 1 and a sovereign debt crisis during the period t − 1 through t + 1.8

Table 2.1Frequency of Financial Crises
YearBanking crises (number)Currency crises (number)Sovereign debt crises (number)Twin crises1 (number)Triple crises2 (number)
1970
19711
19725
19731
1974
19755
1976241
1977211
197853
197932
19803433
19813961
198255911
1983712921
19841104
19852103
1986143
1987661
1988751
198948311
19907102
19911061
19928911
1993781
199411252
19951342
1996461
1997764
1998710233
199982
200024
200113211
2002154
200314111
200411
20051
2006
20072
Total12420863268
Source: Authors’ calculations.

Twin crisis indicates banking crisis in year t and currency crisis during t − 1 through t + 1.

Triple crisis indicates banking crisis in year t and currency crisis during t − 1 through t + 1 and debt crisis during t − 1 through t + 1.

Source: Authors’ calculations.

Twin crisis indicates banking crisis in year t and currency crisis during t − 1 through t + 1.

Triple crisis indicates banking crisis in year t and currency crisis during t − 1 through t + 1 and debt crisis during t − 1 through t + 1.

Banking crises were found to be most frequent during the early 1990s, with the largest number of systemic banking crises, 13, starting in 1995. Currency crises were also common during the first half of the 1990s, with a peak in 1994 of 25 episodes. Sovereign debt crises were also relatively common during the early 1980s, with a peak of 9 debt crises in 1983. In total, 124 banking crises, 208 currency crises, and 63 sovereign debt crises occurred during the period 1970–2007. Several countries experienced multiple crises. Of these 124 banking crises, 26 are considered twin crises and 8 can be classified as triple crises, using the definition above.

Crisis Containment and Resolution

In reviewing crisis policy responses it is useful to differentiate between the containment and resolution phases of systemic restructuring (Honohan and Laeven, 2005; and Hoelscher and Quintyn, 2003, for further details). During the containment phase, the financial crisis is still unfolding. Governments tend to implement policies aimed at restoring public confidence to minimize the repercussions on the real sector of the loss of confidence by depositors and other investors in the financial system. The resolution phase involves the actual financial, and to a lesser extent operational, restructuring of financial institutions and corporations. Although policy responses to crises naturally divide into immediate reactions during the containment phase of the crisis, and long-term responses toward resolution of the crisis, immediate responses often remain part of the long-term policy response. Poorly chosen containment policies undermine the potential for successful long-term resolution. It is thus useful to recognize the context within which policy responses to financial crises occur.

For a subset of 42 well-documented systemic banking crisis episodes (in 37 countries), detailed data have been collected on crisis containment and resolution policies using a variety of sources, including IMF staff reports, World Bank documents, and working papers from central bank staff and academics. This section explains in detail the type of data collected, and defines the variables in the process, organized by the following categories: initial conditions, containment policies, resolution policies, macroeconomic policies, and outcome variables.

Overview and Initial Conditions

This section starts with information on the initial conditions of the crisis, including whether banking distress coincided with exchange rate pressures and sovereign debt-repayment problems, initial macroeconomic conditions, the state of the banking system, and institutional development of the country.

  • CRISIS DATE is the starting date of the banking crisis, including year and month, when available. The timing of the banking crisis follows the approach described in the “Crisis Dates” section of this chapter.

  • CURRENCY CRISIS indicates whether a currency crisis occurred during the period t − 1 through t + 1, where t denotes the starting year of the banking crisis. If the currency experiences a nominal depreciation of at least 30 percent that is also at least a 10 percentage point increase in the rate of depreciation in both years t − 2 and t − 1, with t the starting year of the banking crisis, year t − 1 is treated as the year of the currency crisis for the purposes of creating this variable. YEAR OF CURRENCY CRISIS is also listed.

  • SOVEREIGN DEBT CRISIS indicates whether a sovereign debt crisis occurred during the period t − 1 through t + 1. YEAR OF SOVEREIGN DEBT CRISIS is also listed.

  • This is followed by a brief description of the crisis, denoted as BRIEF DESCRIPTION OF CRISIS.

Information on the following macroeconomic variables was collected. Each of these variables are computed at time t − 1 using data from the IMF’s International Financial Statistics (IFS) and World Economic Outlook.

  • FISCAL BALANCE/GDP is the ratio of the general government balance to GDP for the precrisis year t − 1.9

  • PUBLIC DEBT/GDP is the ratio of the general government gross debt to GDP for the precrisis year t − 1.

  • INFLATION is the percentage increase in the consumer price index during the precrisis year t − 1.

  • NET FOREIGN ASSETS (CENTRAL BANK) is the net foreign assets of the central bank in millions of U.S. dollars for the precrisis year t − 1.

  • NET FOREIGN ASSETS/M2 is the ratio of net foreign assets (central bank) to broad money (M2) for the precrisis year t − 1.

  • DEPOSITS/GDP is the ratio of total deposits at deposit-taking institutions to GDP for the precrisis year t − 1.

  • GDP GROWTH is real growth in GDP during the precrisis year t − 1.

  • CURRENT ACCOUNT/GDP is the ratio of the current account to GDP for the precrisis year t − 1.

The following information was collected on the state of the banking system.

  • PEAK NPL is the peak ratio of nonperforming loans to total loans (percent) during the years t through t + 5. This is an estimate using data from Honohan and Laeven (2005) and IMF staff reports. In all cases, the country’s definition of nonperforming loans was used.

  • GOVERNMENT OWNED is the percentage of banking system assets that are government owned in year t − 1. Data are from La Porta, Lopez-De-Silanes, and Shleifer (2002) and refer to the year 1980 or 1995, whichever is closer to the starting date of the crisis, t. When more recent data are available from IMF staff reports, such data is used instead.

  • SIGNIFICANT BANK RUNS indicates whether the country’s banking system experienced a depositors’ run, defined as a one-month percentage drop in total outstanding deposits in excess of 5 percent during the period t through t + 1. This variable is constructed using data from the IMF’s IFS.

  • CREDIT BOOM indicates whether the country experienced a credit boom leading up to the crisis, defined as three-year precrisis average growth in private credit to GDP in excess of 10 percent per year, computed for the period t − 4 through t − 1. This variable is constructed using data from IFS.

As a proxy for institutional development, data were collected on the degree of protection of creditor rights in the country.

  • CREDITOR RIGHTS is an index of protection of creditors’ rights from Djankov, McLiesh, and Shleifer (2007). The index ranges from 0 to 4, and higher scores denote better protection of creditor rights. The score from year t is used.

Crisis Containment Policies

Initially, the government’s policy options are limited to those policies that do not rely on the formation of new institutions or complex new mechanisms. Immediate policy responses include (1) suspension of convertibility of deposits, which prevents depositors from seeking repayment from banks; (2) regulatory capital forbearance,10 which allows banks to avoid the cost of regulatory compliance (for example, by allowing banks to overstate their equity capital to avoid the costs of contractions in loan supply); (3) emergency liquidity support to banks; or (4) a government guarantee of depositors. Each of these immediate policy actions are motivated by adverse changes in the condition of banks.

Banks suffering severe losses tend not only to see rising costs but also to experience liability rationing, either because they must contract deposits to satisfy their regulatory equity capital requirement, or because depositors at risk of loss prefer to place funds in more stable intermediaries. Banks, in turn, will transmit those difficulties to their borrowers in the form of a contraction of credit supply (Valencia, forthcoming). Credit will become more costly and financial distress of borrowers and banks more likely.

The appropriate policy response will depend on whether the trigger for the crisis is a loss of depositor confidence (causing a deposit run), regulatory recognition of bank insolvency, or the spillover effects of financial asset market disturbances outside the banking system, including exchange rate and wider macroeconomic pressures.

Deposit withdrawals can be addressed by emergency liquidity loans, usually from the central bank when market sources are insufficient, by an extension of government guarantees of depositors and other bank creditors, or by a temporary suspension of depositor rights in what is often called a bank holiday. Each of these techniques is designed to buy time, and the first two help to restore depositor confidence. The success of each technique will depend on the credibility and creditworthiness of the government.

Preventing the looting of an insolvent or near-insolvent bank requires a different set of containment tools. Administrative intervention, including the temporary assumption of management powers by a regulatory official, may be used. Alternatively, closure could be used, which could include the subsidized compulsory sale of a bank’s good assets to a sound bank, together with the assumption by that bank of all or most of the failed entity’s banking liabilities; or could simply consist of an assisted merger. The availability of the necessary legal powers is critical, given the incentive for bank insiders to hang on, as well as the customary cognitive gaps causing insiders to deny the failure of their bank.

Most complex of all are the cases in which disruption of banking is part of wider financial and macroeconomic turbulence. In this case, the bankers may be innocent victims of external circumstances, and special care is needed to ensure that regulations do not become part of the problem. Regulatory forbearance on capital and liquid reserve requirements may prove to be appropriate in these conditions.

Adopting the correct approach to an emerging financial crisis calls for a clear understanding of the underlying cause of the crisis, as well as a quick judgment of the likely effectiveness of the alternative tools that are available. The actions taken at this time will have a possibly irreversible impact on the ultimate allocation of losses in the system. In addition, the longer-term implications for future moral hazard need to be taken into account.

All too often, central banks favor stability over cost in the heat of the containment phase: if so, they may extend loans too liberally to an illiquid bank that is almost certain to prove insolvent anyway. Also, closure of a nonviable bank is often delayed for too long, even when there are clear signs of insolvency (Lindgren, 2005). Because bank closures face many obstacles, there is a tendency to rely instead on blanket government guarantees, which, if the government’s fiscal and political position makes them credible, can work but at the cost of placing the burden on the budget, typically squeezing future provision of needed public services.

Information was collected on the following crisis containment policies.

First, information was collected on whether the authorities impose deposit freezes, bank holidays, or blanket guarantees to stop or prevent bank runs.

  • DEPOSIT FREEZE indicates whether the authorities imposed a freeze on deposits. If a freeze on deposits was implemented, information is collected on the duration of the deposit freeze (in months), and the type of deposits affected.

  • BANK HOLIDAY indicates whether the authorities installed a bank holiday. If a bank holiday was introduced, information is collected on its duration (in days).

  • BLANKET GUARANTEE indicates whether the authorities introduced a blanket guarantee on deposits (and possibly other liabilities). If a blanket guarantee was introduced, information is collected on the date of introduction and the date of removal, and the duration of the guarantee is calculated (in months). Information is also collected on whether a previous explicit deposit insurance arrangement was in place at the time of the introduction of the blanket guarantee, the name of the administering agency of the blanket guarantee, and the coverage of the guarantee (deposits only or also other liabilities).

  • TIMING OF FIRST BANK INTERVENTION indicates the date (month and year) that the authorities intervened for the first time in a bank.

  • TIMING OF FIRST LIQUIDITY ASSISTANCE indicates the date (month and year) that the first loan under liquidity assistance was granted to a financial institution.

Next, information is collected on the timing and scope of emergency liquidity support to financial institutions.

  • LIQUIDITY SUPPORT indicates whether emergency liquidity support, measured as claims from monetary authorities on commercial banks (IFS line 12E) to total deposits, was at least 5 percent and at least doubled with respect to the previous year during the period t through t + 3.

    Information is also collected on whether liquidity support was different across banks, or whether emergency lending was remunerated. If liquidity support was remunerated, information is collected on whether interest was at market rates.

    Information is also collected on the peak of liquidity support (as a percentage of deposits), computed as the maximum value (percent) of the ratio of claims from monetary authorities on commercial banks (IFS line 12E) to total deposits during the period t through t + 3.

  • LOWERING OF RESERVE REQUIREMENTS denotes whether authorities lowered reserve requirements in response to the crisis.

Crisis Resolution Policies

Once emergency measures have been put in place to contain the crisis, the government faces the long-term challenge of crisis resolution, which means the resumption of a normally functioning credit system and legal system, and the rebuilding of banks’ and borrowers’ balance sheets.

At this point, the crisis has left banks and nonfinancial firms insolvent and many are in government ownership or under court or regulatory administration. Economic growth is unlikely to resume on a secure basis until productive assets and banking franchises are back in the hands of solvent private entities.

The financial and organizational restructuring of financial and nonfinancial firms during the crisis resolution phase is a large task, typically entailing much detailed implementation work in the bankruptcy courts, as well as the use of informal or ad hoc work-out procedures. There are also important trade-offs, such as that between speed and durability of the subsequent economic recovery on the one hand, and the fiscal costs on the other.

Crisis resolution involves inherently complicated coordination problems between debtors and creditors. The fate of an individual corporation or financial institution and the best course of action for its owners and managers will depend on the actions of many others and the general economic outlook. Because of these coordination problems, as well as a lack of capital and the importance of the financial system to economic growth, governments often take the lead in systemic restructuring, especially of the banking system. In the process, governments can incur large fiscal costs, presumably with the objective of accelerating recovery from the crisis.

The most-asked question arising at this time is whether an overindebted corporate entity should be somehow subsidized or forgiven some of its debt, or whether its assets should be transferred to a new corporate structure and new management. This question applies to undercapitalized banks and to overindebted nonbank corporations alike. The feasibility of making such decisions on a case-by-case basis becomes difficult during a systemic crisis resulting in thousands of insolvencies, so a systematic approach needs to be established. General principles have proved elusive and, as well as depending on the scale of the crisis and the quality of existing legal and other governance institutions, the best answer is likely to depend on the source of the crisis.

If the problem results from an economy-wide crash, the best prospect for future performance of banks and their borrowing customers may be with their existing owners and managers, given the information and other intangible forms of firm- or relationship-specific capital they possess. However, if bank insolvency has been the result of incompetent, reckless, or corrupt banking, or the use of government-controlled banks as quasi-fiscal vehicles or for political purposes, the relevant stock of information and relationship capital is unlikely to be of much social value. Therefore, separating the good assets from their current managers and owners offers better prospects in such circumstances and establishes a better precedent for avoiding moral hazard. Information capital is also likely to be relatively unimportant for real estate ventures, which have been central to many recent banking crises.

The main policy approaches used in the resolution phase of recent crises include (1) conditional government-subsidized, but decentralized, work-outs of distressed loans; (2) debt forgiveness; (3) the establishment of a government-owned asset management company to buy and resolve distressed loans; (4) government-assisted sales of financial institutions to new owners, typically foreign; and (5) government-assisted recapitalization of financial institutions through injection of funds. The latter three deal with bank insolvency, and are thus the focus here.

In an attempt to let the market determine which firms are capable of surviving if given some modest assistance, some official schemes have offered loan subsidies to distressed borrowers conditional on the borrowers’ shareholders injecting some new capital. There have also been schemes offering injection of government capital funds for insolvent banks whose shareholders were willing to provide matching funds.

To the extent that they are discretionary, debt-relief schemes for bank borrowers carry the risk of moral hazard because debtors stop trying to repay in the hope of being added to the list of scheme beneficiaries.

Generalized forms of debt relief, such as are effectively provided by inflation and currency depreciation, can be regarded as relationship-friendly in the sense introduced above. Inflation is also a solution that reduces the budgetary burden. After all, if the crisis is big enough, the government’s choices may be limited by what it can afford. Its capacity to subsidize borrowers or inject capital into banks is constrained by its ability over time to raise taxes or cut expenditure. These reasons are behind the inflationary solutions or currency devaluations that have been a feature of crises resolutions in the past. It amounts to generalized debt relief and a transfer of the costs of the crisis to money holders and other nominal creditors. In this case, the banks as well as the nonbank debtors receive relief without a climate of debtor delinquency being created. Of course, these are questions of monetary and macroeconomic policy as much as banking policy and must be considered in light of the need to preserve an environment of macroeconomic stability into the future.

In contrast, the carving out of an insolvent bank’s bad loan portfolio, and its organizational restructuring under new management and ownership, represents the opposite pole, appropriate if large parts of the bank’s information capital was dysfunctional. The bad loan portfolio may be sold back into the market, or disposed of by a government-owned asset management company (AMC). The effectiveness of government-run AMCs has been mixed: better if the assets to be disposed of have been primarily real estate, worse if loans to large, politically connected firms dominated (Klingebiel, 2000).

Government itself often retains control and ownership of troubled banks for much of the duration of the resolution phase. Regardless of whether control of the bank passes into public hands, it should eventually emerge, and it must be adequately capitalized. Depending on how earlier loss allocation decisions were made, the sums of money involved in the recapitalization of the bank so that it can safely be sold into private hands may be huge. Many governments have felt constrained by fiscal and monetary policy considerations from properly doing the financial restructuring. Putting the bank on a sound financial footing should be the priority. Without this, banks will be undercapitalized, whatever the accounts state, and will have an incentive to resume reckless behavior.

Countries typically apply a combination of resolution strategies, including both government-managed programs and market-based mechanisms (Calomiris, Klingebiel, and Laeven, 2005). Both depend for their success on efficient and effective legal, regulatory, supervisory, and political institutions. Furthermore, a lack of attention to incentive problems when designing specific rules governing financial assistance can aggravate moral hazard problems, especially in environments in which these institutions are weak, unnecessarily raising the costs of resolution. Accordingly, policymakers in economies with weak institutions should not expect to achieve the same level of success in financial restructuring as is achieved in more advanced economies, and they should design resolution mechanisms suited to their institutional capacity.

Information on the following crisis resolution policies was collected.

  • FORBEARANCE indicates whether there is regulatory forbearance during the years t through t + 3. This variable is based on a qualitative assessment of information contained in IMF staff reports. As part of this assessment, information was also collected on whether banks were permitted to continue functioning despite being technically insolvent, and whether prudential regulations (such as for loan classification and loan loss provisioning) were suspended or not fully applied during the first three years of the crisis.

With regard to actual bank restructuring, information was collected on nationalizations, closures, mergers, sales, and recapitalizations.

  • LARGE-SCALE GOVERNMENT INTERVENTION indicates whether there was large-scale government intervention in banks, such as nationalizations, closures, mergers, sales, and recapitalizations of large banks, during the years t through t + 3.

  • INSTITUTIONS CLOSED indicates the share (percentage) of bank assets liquidated or closed during the years t through t + 3. Information is also collected on the number of banks in year t and the number of banks in year t + 3.

  • BANK CLOSURES indicates whether banks were closed during the period t through t + 3. Information was also collected on the number of banks closed or liquidated during the period t through t + 3.

    Information is collected separately on whether financial institutions other than banks were closed (OTHER FI CLOSURES), and on whether shareholders of closed institutions were made whole (SHAREHOLDER PROTECTION).

    Information is also collected on whether banks were nationalized (NATIONALIZATIONS), merged (MERGERS), or sold to foreigners (SALES TO FOREIGNERS) during the period t through t + 5. For mergers, information was also collected on whether private shareholders or owners of banks injected capital, and for sales to foreigners information was collected on the number of banks sold to foreigners during the period t through t + 5.

    Next, information is collected on whether a bank-restructuring agency (BANK RESTRUCTURING AGENCY) was set up to deal with bank restructuring, and whether an AMC (ASSET MANAGEMENT COMPANY) was set up to take over and manage distressed assets. If an AMC was set up, information is collected on whether it was centralized or decentralized, the entity in charge, its funding, and the type of assets transferred.

As part of crisis resolution, systemically important (or government-owned) banks are often recapitalized by the government.

  • RECAPITALIZATION denotes whether banks were recapitalized by the government during the period t through t + 3.

    Banks can be recapitalized using a variety of measures. Information is collected on whether recapitalization occurred in the form of cash, government bonds, subordinated debt, preferred shares, purchase of bad loans, credit lines, assumption of bank liabilities, ordinary shares, or other means.

When available, information is also collected on the targeted recapitalization level of banks (expressed as a percentage of assets) and an estimate was made of the gross recapitalization cost (as a percentage of GDP) to the government during the period t through t + 5. The latter variable is denoted as RECAP COST (GROSS).

Next, information was collected on the recovery of recapitalization costs.

  • RECOVERY denotes whether the government was able to recover part of the recapitalization cost.

  • RECOVERY PROCEEDS denotes the recovery proceeds (as a percentage of GDP) during the period t through t + 5.

  • RECAP COST (NET) denotes the net recapitalization cost to the government, expressed as a percentage of GDP, computed as the difference between the gross recapitalization cost and recovery proceeds.

For deposit insurance and depositor compensation, the following information was collected from Demirgüç-Kunt, Kane, and Laeven (2008) and IMF staff reports.

  • DEPOSIT INSURANCE indicates whether an explicit deposit insurance scheme is in place at the start of the banking crisis. Note that deposit insurance arrangements put in place after the first year of the crisis are ignored.

  • FORMATION reports the year that the deposit insurance scheme was introduced.

  • COVERAGE LIMIT denotes the coverage limit (in local currency) of insured deposits at the start of the banking crisis. This variable is set to zero if there is no explicit deposit insurance.

  • COVERAGE RATIO is the ratio of the coverage limit to per capita GDP at the start of the banking crisis. This variable is set to zero if there is no explicit deposit insurance.

  • WERE LOSSES IMPOSED ON DEPOSITORS? denotes whether losses were imposed on depositors of failed banks, and if so, whether these losses were severe (implying large discounts and a substantial number of people affected) was reported.

Macroeconomic Policies

Governments also tend to change macroeconomic policy to manage banking crises and reduce the negative impact on the real sector. Therefore, in addition to crisis containment and resolution policies, information on monetary policy and fiscal stance during the first three years of the crisis is collected. Although these measures are somewhat crude, they serve the purpose of providing some understanding of the policy stance.

  • MONETARY POLICY INDEX is an index of monetary policy stance during the years t through t + 3. The index indicates whether monetary policy is expansive (+1), if the average percentage change in reserve money during the years t through t + 3 is between 1 and 5 percent higher than during the years t − 4 through t − 1; contractive (−1), if the average percentage change in reserve money during the years t through t + 3 is between 1 and 5 percent lower than during the years t − 4 through t − 1; or neither (0).

    The average change in reserve money (in percent) during the years t through t + 3, where t denotes the starting year of the banking crisis, was also reported.

  • FISCAL POLICY INDEX is an index of the fiscal policy stance during the years t through t + 3, where t denotes the starting year of the crisis. The index indicates whether fiscal policy is expansive (+1), if the average fiscal balance during the years t through t + 3 is less than −1.5 percent of GDP; contractionary (−1), if the average fiscal balance during the years t through t + 3 is greater than 1.5 percent of GDP; or neither (0).

    The average fiscal balance (as a percentage of GDP) during the years t through t + 3 is also reported.

Finally, whether an IMF program was put in place around the time of the banking crisis (IMF PROGRAM) is reported, including the year the program was put in place.

Outcome Variables

Information on fiscal costs and output losses, the outcome variables, was collected.

  • FISCAL COST (NET) denotes the net fiscal cost, expressed as a percentage of GDP, during the period t through t + 5. The gross fiscal costs are also reported, as are the recovery proceeds during the period t through t + 5, which is the difference between the two. Fiscal cost estimates are from Hoelscher and Quintyn (2003), Honohan and Laeven (2005), IMF staff reports, and publications from national authorities and institutions.

  • OUTPUT LOSS is computed by extrapolating trend real GDP, based on the trend in real GDP growth up to the year preceding the crisis, and taking the sum of the differences between actual real GDP and trend real GDP expressed as a percentage of trend real GDP for the period t through t + 3. A minimum of three precrisis real GDP growth observations were required to compute the trend real GDP numbers.11

Descriptive Statistics

Appendix Table 2A.3 summarizes the data collected on crisis containment and resolution policies for a subset of 42 systemic banking crises. The crisis countries comprise Argentina (four crises), Bolivia, Brazil (two crises), Bulgaria, Chile, Colombia (two crises), Côte d’Ivoire, Croatia, the Czech Republic, the Dominican Republic, Ecuador, Estonia, Finland, Ghana, Indonesia, Jamaica, Japan, the Republic of Korea, Latvia, Lithuania, Malaysia, Mexico, Nicaragua, Norway, Paraguay, the Philippines, the Russian Federation, Sri Lanka, Sweden, Thailand, Turkey, Ukraine, the United Kingdom, the United States, Uruguay, Venezuela, and Vietnam. Note that the financial crises in the United Kingdom and the United States were ongoing at the time of writing of this chapter, so the analysis of crisis containment and resolution policies for these two countries is preliminary and incomplete.

Appendix Table 2A.3Crisis Containment and Resolution Policies for Selected Banking Crises
ArgentinaArgentinaArgentinaArgentinaBoliviaBrazilBrazilBulgaria
Banking crisis date (year and month)Mar. 1980Dec. 1989Jan. 1995Dec. 2001Nov. 1994Feb. 1990Dec. 1994Jan. 1996
Currency crisis (Y/N) t − 1 through t + 1YYNYNYYY
Year of currency crisis198119882002198919931996
Sovereign debt crisis (Y/N) t − 1 through t + 1NNNYNNNN
Year of sovereign debt crisis2001
Initial conditions
Fiscal balance/GDP at t − 1 (percent)−2.65−4.420.03−3.61−3.000.000.27−5.63
Public sector debt/GDP at t − 1 (percent)10.2089.8033.7050.8076.0022.2023.00106.40
Inflation at t − 1 (percent)139.74387.813.85−0.738.521,972.912,477.1532.66
Net foreign assets/M2 at t − 1 (percent)34.21−16.9925.9024.167.890.0122.699.66
Deposits/GDP at t − 1 (percent)22.2421.2514.9628.2234.87133.25101.4359.81
GDP growth at t − 1 (percent)7.10−1.966.25−0.794.673.204.93−1.60
Current account/GDP at t − 1 (percent)0.55−1.23−2.83−3.15−3.990.21−0.12−0.20
Peak NPLs (percentage of total loans)9.0027.0017.0020.106.2016.0075.00
Government-owned bank (percentage of assets) at t − 171.9460.5041.0030.000.0031.7031.7085.68
Significant bank runs (Y/N)YYYYYYYY
Largest one-month percentage drop in deposits (>5%), t through t + 113.7626.658.366.847.0114.639.259.44
Credit boom (Y/N)YNYNYN
Annual growth in private credit to GDP, t − 4 through t − 1 (percent)23.62−1.7018.906.1022.505.80
Creditor rights in year t11112111
Containment phase
Deposit freeze (Y/N)NYNYNYNN
Introduction of deposit freeze198920011990
Duration of deposit freeze (in months)1201229
Coverage of deposit freeze (time deposits only? Y/N)YNN
Bank holiday (Y/N)NYNYNNNN
Introduction of bank holiday19902001
Duration of bank holiday (days)45
Blanket guarantee (Y/N)NNNNNNNN
Date of introduction
Date of removal
Duration of guarantee (months)
Previous explicit deposit insurance arrangement (Y/N)YYYYNNNY
Timing of first bank interventionMar. 1980Feb. 1989Jan. 1995Apr. 2002Nov. 1994Jul. 1994Early 1996
Timing of first liquidity assistanceFeb. 1990
Significant liquidity support or emergency lending (Y/N)YYNYYYYY
Support different across banks? (Y/N)NYYY
Collateral requiredYYYN
Remunerated (Y/N)YYY
If remunerated, interest at market rates (Y/N)Y
Peak support (percent of deposits)15.60300.004.1524.3013.905.0023.2022.90
Lowering of reserve requirements (Y/N)YNYNYNNN
Resolution phase
Forbearance(Y/N)YNNYYNYY
Banks not intervened in despite being technically insolventNNYYYY
Prudential regulations suspended or not fully appliedYNYYYY
Large-scale government intervention (Y/N)YNNYYNYY
Institutions closed (percent of banks assets)160.62011.000small24.00
Number of banks in t214177205841722924645
Number of banks in t + 3203165143731424523834
Bank closures (Y/N)YYYNYNYY
Number of bank closures during t through t + 32128524116
Other financial institution closures (Y/N)YYNNNNY
Shareholder protection (shareholders made whole? Y/N)NNNNNN
Nationalizations (Y/N)YNNYNNNY
Mergers (Y/N)YYYNYYN
Did bank shareholders inject new capital? (Y/N)YYYY
Sales to foreigners (Y/N)YNYNYNYY
Number of banks sold to foreigners during t to t + 51004034
Bank-restructuring agency (Y/N)NNNYYNN
Asset management company (Y/N)NNNNYNNY
Centralized (Y); Decentralized (N)YN
Recapitalization (Y/N)NNYYYNYY
Recap measures
CashY
Government bondsYY
Subordinated debtYY
Preferred shares
Purchase of bad loansY
Credit line
Assumption of bank liabilities
Ordinary shares
Recap level (percent)8.004.00
Recap cost (gross) (percentage of GDP)0.289.580.954.982.31
Recovery (Y/N)NNYNN
Recovery proceeds during t to t + 50.95
Recap cost (net) (percentage of GDP)0.289.580.004.982.31
Deposit insurance (Y/N)YYYYNNNY
Formation19791979197919791996
Coverage limit (in local currency) at tFullFull30,00030,0005,000
Coverage ratio (coverage limit to GDP per capita) at t4.044.192.37
Were losses imposed on depositors? (Y/N)NYNYYNNN
If yes, severe = 1 and moderate = 2112
Macro policies
Monetary policy index110−101−11
Average change in reserve money during t through t + 3 (percent)324.162,046.8536.2818.801,673.69939.63245.13
Fiscal policy index11111−111
Average fiscal balance during t through t + 3 (percent)−6.82−3.86−2.24−7.23−3.020.27−5.14−3.09
IMF program (Y/N)YYYYNYNY
IMF program put in place (year)198319901995200019891996
Outcome variables
Fiscal cost net (percentage of GDP)55.106.002.009.582.650.0010.2013.90
Gross (percent)55.106.002.009.586.030.0013.2014.00
Recovery during t through t + 5 (percent)00003.370.003.000.10
Output loss
Output loss during t through t + 3 (percent)10.8110.707.1342.650.0012.230.001.30
ChileColombiaColombiaCôte d’IvoireCroatiaCzech RepublicDominican RepublicEcuador
Banking crisis date (year and month)Nov. 1981Jul. 1982Jun. 19981988Mar. 19981996Apr. 2003Aug. 1998
Currency crisis (Y/N) t − 1 through t + 1YNNNNNYY
Year of currency crisis198220031999
Sovereign debt crisis (Y/N) t − 1 through t + 1NNNNNNYY
Year of sovereign debt crisis20031999
Initial conditions
Fiscal balance/GDP at t − 1 (percent)4.99−2.26−3.95−7.19−2.01−1.29−1.37−3.02
Public sector debt/GDP at t − 1 (percent)30.1926.7012.4726.8061.75
Inflation at t − 1 (percent)31.2426.3317.687.485.01107.8610.5130.67
Net foreign assets/M2 at t − 1 (percent)42.1745.9531.12−35.0528.6732.51−1.038.35
Deposits/GDP at t − 1 (percent)26.6224.7836.1420.5741.4262.2434.8023.25
GDP growth at t − 1 (percent)7.942.283.43−0.506.806.364.434.05
Current account/GDP at t − 1 (percent)−6.35−4.06−5.39−14.93−12.61−0.09−3.69−3.02
Peak NPLs (percentage of total loans)35.604.1014.0050.0010.5018.009.0040.00
Government-owned bank (percentage of assets) at t − 119.7257.6753.6220.601.0452.0015.509.00
Significant bank runs (Y/N)YNNNYYNY
Largest one-month percentage drop in deposits (>5%), t through t + 18.486.115.6711.09
Credit boom (Y/N)YNNNNNN
Annual growth in private credit to GDP, t − 4 through t − 1 (percent)34.105.407.000.007.607.709.40
Creditor rights in year t20003320
Containment phase
Deposit freeze (Y/N)NNNNNNNY
Introduction of deposit freeze1999
Duration of deposit freeze (in months)6
Coverage of deposit freeze (time deposits only? Y/N)N
Bank holiday (Y/N)NNNNNNNY
Introduction of bank holiday1999
Duration of bank holiday (in days)5
Blanket guarantee (Y/N)NNNNNYNY
Date of introductionMid-1996Dec. 1998
Date of removalJan. 1998Jan. 2002
Duration of guarantee (months)1837
Previous explicit deposit insurance arrangement (Y/N)NNYNYYNY
Timing of first bank interventionNov. 1981Jul. 19821988Apr. 1998Dec. 1995Apr. 2003Apr. 1998
Timing of first liquidity assistance
Significant liquidity support or emergency lending (Y/N)YYYYNNYY
Support different across banks? (Y/N)NYN
Collateral requiredNN
Remunerated (Y/N)YN
If remunerated, interest at market rates (Y/N)NN
Peak support (percentage of deposits)124.0014.909.2059.001.702.3061.6015.30
Lowering of reserve requirements (Y/N)YYNNYNNN
Resolution phase
Forbearance (Y/N)YNYYYNYY
Banks not intervened in despite being technically insolventNNNYNNNY
Prudential regulations suspended or not fully appliedYNYYYNYY
Large-scale government intervention (Y/N)YYYYYYYY
Institutions closed (percentage of banks assets)20.0009.907.061.500.0050.20
Number of banks in t61392060551440
Number of banks in t + 345271443451122
Bank closures (Y/N)YNYYYYNY
Number of bank closures during t through t + 3812611414
Other financial institution closures (Y/N)YNYYNNNY
Shareholder protection (shareholders made whole? Y/N)NNNNNN
Nationalizations (Y/N)NYYNYNNY
Mergers (Y/N)YNYNYYNY
Did bank shareholders inject new capital? (Y/N)YYYNYNN
Sales to foreigners (Y/N)YNNYYYN
Number of banks sold to foreigners during t through t + 51552
Bank-restructuring agency (Y/N)NNYYYNY
Asset management company (Y/N)NNYYYYYY
Centralized (Y); Decentralized (N)YYYYYY
Recapitalization (Y/N)YYYYYYNY
Recap measures
CashYYYYY
Government bondsYYY
Subordinated debtY
Preferred shares
Purchase of bad loansY
Credit lineYY
Assumption of bank liabilities
Ordinary shares
Recap level (percent)10.009.00
Recap cost (gross) (percentage of GDP)34.331.874.263.200.981.90
Recovery (Y/N)YNYNNNY
Recovery proceeds during t through t + 5 (percent)27.871.560.30
Recap cost (net) (percentage of GDP)6.461.872.703.200.981.60
Deposit insurance(Y/N)NNYNYYNY
Formation1988199719941998
Coverage limit (in local currency) at t10,000,00050,000100,0007,416
Coverage ratio (coverage limit to GDP per capita) at t3.291.80.753.81
Were losses imposed on depositors? (Y/N)YNNYNNNY
If yes, severe = 1 and moderate = 2211
Macro policies
Monetary policy index−100−1−1−111
Average change in reserve money during t through t + 3 (percent)9.9721.0011.97−6.5723.197.9945.95
Fiscal policy index−11111110
Average fiscal balance during t through t + 3 (percent)0.81−3.93−4.28−12.69−5.19−3.35−6.45−0.66
IMF program (Y/N)YNNYNNYY
IMF program put in place (year)1983198520042000
Outcome variables
Fiscal cost net (percentage of GDP)16.805.002.5425.006.905.8020.8016.26
Gross (percent)42.905.006.2825.006.906.8022.0021.70
Recovery during t through t + 5 (percent)26.1003.740.000.001.001.205.44
Output loss
Output loss during t through t + 3 (percent)92.3515.1133.520.000.0015.516.49
EstoniaFinlandGhanaIndonesiaJamaicaJapanKoreaLatvia
Banking crisis date (year and month)Nov. 1992Sep. 19911982Nov. 1997Dec. 1996Nov. 1997Aug. 1997Apr. 1995
Currency crisis (Y/N) t − 1 through t + 1YNYYYNYN
Year of currency crisis19911983199819951998
Sovereign debt crisis (Y/N) t − 1 through t + 1NNNNNNNN
Year of sovereign debt crisis
Initial conditions
Fiscal balance/GDP at t − 1 (percent)5.255.56−0.12−1.131.99−5.130.24−3.86
Public sector debt/GDP at t − 1 (percent)14.0426.4090.89100.488.8014.89
Inflation at t − 1 (percent)4.8816.796.0425.550.604.9326.27
Net foreign assets/M2 at t − 1 (percent)57.6312.73−0.0621.5819.071.6215.6236.32
Deposits/GDP at t − 1 (percent)72.3352.286.2044.7440.73252.4136.5521.15
GDP growth at t − 1 (percent)−7.910.08−6.917.821.012.757.002.20
Current account/GDP at t − 1 (percent)59.70−4.91−0.32−2.91−4.371.42−4.14−3.61
Peak NPLs (percentage of total loans)7.0013.0035.0032.5028.9035.0035.0020.00
Government-owned bank (percentage of assets) at t − 125.7013.4060.0042.300.000.0023.419.90
Significant bank runs (Y/N)YNYYNNYY
Largest one-month percentage drop in deposits (>5%), t through t + 119.9411.7422.6012.005.81
Credit boom (Y/N)NNNNNN
Annual growth in private credit to GDP t − 4 through t – 1 (percent)8.00−19.904.50−3.100.101.10
Creditor rights in year t3132333
Containment phase
Deposit freeze (Y/N)NNNNNNNN
Introduction of deposit freeze
Duration of deposit freeze (in months)
Coverage of deposit freeze (time deposits only? Y/N)
Bank holiday (Y/N)NNNNNNNN
Introduction of bank holiday
Duration of bank holiday (in days)
Blanket guarantee (Y/N)NYNYYYYN
Date of introductionFeb. 1993Jan. 1998Feb. 1997Nov. 1997Nov. 1997
Date of removalDec. 1998Jul. 2005Mar. 1998Apr. 2005Dec. 2000
Duration of guarantee (in months)7078118937
Previous explicit deposit insurance arrangement (Y/N)NYNNNYYN
Timing of first bank interventionSep. 1991Jun. 2005Nov. 1997Dec. 1994Apr. 1997Oct. 1997May 1995
Timing of first liquidity assistance
Significant liquidity support or emergency lending (Y/N)YYNYYNYN
Support different across banks? (Y/N)YN
Collateral requiredNNN
Remunerated (Y/N)NYY
If remunerated, interest at market rates (Y/N)NYY
Peak support (percentage of deposits)31.645.500.0053.8012.400.4028.903.01
Lowering of reserve requirements (Y/N)YNYNNYNY
Resolution phase
Forbearance (Y/N)YYYYNYYN
Banks not intervened in despite being technically insolventNNYNNNYN
Prudential regulations suspended or not fully appliedYYYYNYNN
Large-scale government intervention (Y/N)YYYYYYYY
Institutions closed (percentage of banks’ assets)15.000013.504.1509.0040.00
Number of banks in t2151911238365956
Number of banks in t + 31834711165203142
Bank closures (Y/N)YNNYYNYY
Number of bank closures during t through t + 3110066102214
Other financial institution closures (Y/N)YNNNYYY
Shareholder protection (shareholders made whole? Y/N)NYNNNN
Nationalizations (Y/N)YYNYYYYN
Mergers (Y/N)YYNYYYYN
Did bank shareholders inject new capital? (Y/N)NNNNYY
Sales to foreigners (Y/N)NNYYYYN
Number of banks sold to foreigners during t through t + 5218
Bank-restructuring agency (Y/N)NYNYYYYN
Asset management company (Y/N)YYYYYYYN
Centralized (Y); Decentralized (N)NYYYYYY
Recapitalization (Y/N)YYYYYYYN
Recap measures
CashYYY
Government bondsYYY
Subordinated debtYYYY
Preferred sharesYYYY
Purchase of bad loansYY
Credit line
Assumption of bank liabilitiesYY
Ordinary sharesYY
Recap level (percent)6.004.00
Recap cost (gross) (percentage of GDP)1.268.636.0037.3013.906.6119.31
Recovery (Y/N)YYNNYYY
Recovery proceeds during t to t + 50.271.724.950.093.50
Recap cost (net) (percentage of GDP)0.996.916.0037.308.956.5215.81
Deposit insurance (Y/N)NYNNNYYN
Formation196919711996
Coverage limit (in local currency) at tFullFull20,000,000
Coverage ratio (coverage limit to GDP per capita) at t2.18
Were losses imposed on depositors? (Y/N)YNNNNNNY
If yes, severe = 1 and moderate = 211
Macro policies
Monetary policy index0−10100−10
Average change in reserve money during t through t + 3 (percent)1.7546.9347.6619.358.884.05
Fiscal policy index01−111110
Average fiscal balance during t through t + 3 (percent)−0.66−5.07−0.04−2.47−5.71−6.17−1.66−1.36
IMF program (Y/N)YNNYNNYY
IMF program put in place (year)1993199819981993
Outcome variables
Fiscal cost net (percentage of GDP)1.6311.086.0052.3038.9513.9123.203.00
Gross (percent)1.9012.806.0056.8043.9014.0031.203.00
Recovery during t through t + 5 (percent)0.271.7204.604.950.098.000.00
Output loss
Output loss during t through t + 3 (percent)59.0815.7967.9530.0817.5650.10
Banking crisis date (year and month)Dec. 1995Jul. 1997Dec. 1994Aug. 2000Oct. 1991May 1995Jul. 1997Aug. 1998
Currency crisis (Y/N) t − 1 through t + 1NYYNNNYY
Year of currency crisis1998199519981998
Sovereign debt crisis (Y/N) t − 1 through t + 1NNNNNNNY
Year of sovereign debt crisis1998
Initial conditions
Fiscal balance/GDP at t − 1 (percent)−4.221.98−2.46−3.302.542.73−0.18−16.96
Public sector debt/GDP at t − 1 (percent)8.0035.1627.34191.3128.9215.8052.49
Inflation at t − 1 (percent)45.103.348.019.284.3618.317.1411.05
Net foreign assets/M2 at t − 1 (percent)39.6323.2018.12−14.1010.3438.8619.039.47
Deposits/GDP at t − 1 (percent)17.43119.5126.8237.0254.4427.6848.6114.59
GDP growth at t − 1 (percent)−9.7710.001.957.001.933.735.851.40
Current account/GDP at t − 1 (percent)−3.86−4.36−5.80−24.902.50−2.02−0.180.00
Peak NPLs (percentage of total loans)32.2030.0018.9012.7016.368.1020.0040.00
Government-owned bank (percentage of assets) at t − 148.009.9328.160.0043.6848.0227.2332.98
Significant bank runs (Y/N)YYYNNYNY
Largest one-month percentage drop in deposits (>5%), t through t + 16.266.0314.007.6821.00
Credit boom (Y/N)NYNYYN
Annual growth in private credit to GDP, t − 4 through t − 1 (percent)7.1022.502.9017.6017.709.50
Creditor rights in year t13042111
Containment phase
Deposit freeze (Y/N)NNNNNNNN
Introduction of deposit freeze
Duration of deposit freeze (in months)
Coverage of deposit freeze (time deposits only? Y/N)
Bank holiday (Y/N)NNNNNNNN
Introduction of bank holiday
Duration of bank holiday (days)
Blanket guarantee (Y/N)NYYYNNNN
Date of introductionJan. 1998Dec. 1993Jan. 2001Jul. 1995
Date of removalAug. 2005Jan. 2003Jul. 2002Jun. 1996
Duration of guarantee (months)911091411
Previous explicit deposit insurance arrangement (Y/N)NNYNYNYN
Timing of first bank interventionDec. 1995NoneNov. 1994Aug. 2000Fall 1988May 1995
Timing of first liquidity assistanceAug. 2000Sep. 1997Sep. 1998
Significant liquidity support or emergency lending (Y/N)NYYYYYNY
Support different across banks? (Y/N)NYNY
Collateral requiredNY
Remunerated (Y/N)YY
If remunerated, interest at market rates (Y/N)YN
Peak support (percentage of deposits)4.6012.2067.609.606.2020.802.5031.50
Lowering of reserve requirements (Y/N)YNNNNYNY
Resolution phase
Forbearance (Y/N)YYYNYYNY
Banks not intervened in despite being technically insolventYNNNNY
Prudential regulations suspended or not fully appliedYYYYYY
Large-scale government intervention (Y/N)YYYYYYNY
Institutions closed (percentage of banks assets)15.000001.0023.001.004.00
Number of banks in t28475212164341,0031,476
Number of banks in t + 31443376153229251,318
Bank closures (Y/N)YNNNYYYY
Number of bank closures during t through t + 3140002926399
Other financial institution closures (Y/N)NNYNY
Shareholder protection (shareholders made whole? Y/N)NNNNN
Nationalizations (Y/N)YYYNYNNY
Mergers (Y/N)NYYNYNNY
Did bank shareholders inject new capital? (Y/N)YYY
Sales to foreigners (Y/N)NNYNYNN
Number of banks sold to foreigners during t through t + 54
Bank-restructuring agency (Y/N)NYYNYNNY
Asset management company (Y/N)YYNYNNNY
Centralized (Y); Decentralized (N)YYNY
Recapitalization (Y/N)YYYNYYYN
Recap measures
CashY
Government bondsY
Subordinated debtYYY
Preferred sharesY
Purchase of bad loansYYY
Credit line
Assumption of bank liabilities
Ordinary sharesY
Recap level (percent)9.009.008.00
Recap cost (gross) (percentage of GDP)1.7016.403.802.611.220.20
Recovery (Y/N)YYYYNN
Recovery proceeds during t through t + 50.2011.301.302.00
Recap cost (net) (percentage of GDP)1.505.102.500.611.220.20
Deposit insurance (Y/N)NNYNYNYN
Formation198619611963
Coverage limit (in local currency) at tFullFull10,000
Coverage ratio (coverage limit to GDP per capita) at t3.22
Were losses imposed on depositors? (Y/N)YNNNNNNY
If yes, severe = 1 and moderate = 222
Macro policies
Monetary policy index0−1100−1−11
Average change in reserve money during t through t + 3 (percent)−2.7822.0317.635.5724.137.0347.21
Fiscal policy index10110−110
Average fiscal balance during t through t + 3 (percent)−5.05−1.28−4.77−3.80−0.65−0.02−2.75−1.32
IMF program (Y/N)NNYNNNYY
IMF program put in place (year)199519981999
Outcome variables
Fiscal cost net (percentage of GDP)2.905.1018.0012.570.6010.0013.206.00
Gross (percent)3.1016.4019.3013.612.7012.9013.206.00
Recovery during t through t + 5 (percent)0.2011.301.301.042.102.9000
Output loss
Output loss during t through t + 3 (percent)50.044.250.000.000.000.000.00
Sri LankaSwedenThailandTurkeyUkraineUnited KingdomUnited StatesUruguayVenezuelaVietnam
Banking crisis date (year and month)1989Sep. 1991Jul. 1997Nov. 20001998Aug. 2007Aug. 2007Jan. 2002Jan. 1994Fall 1997
Currency crisis (Y/N) t − 1 through t + 1NYNYYNNYYN
Year of currency crisis19922001199820021993
Sovereign debt crisis (Y/N) t − 1 through t + 1NNNNNNNYNN
Year of sovereign debt crisis2002
Initial conditions
Fiscal balance/GDP at t − 1 (percent)−8.593.392.40−14.97−5.56−2.56−2.61−0.22−2.92−2.36
Public sector debt/GDP at t − 1 (percent)108.7214.1551.3129.8843.0460.1039.05
Inflation at t − 1 (percent)15.1010.944.7768.7910.122.782.573.5945.944.59
Net foreign assets/M2 at t − 1 (percent)5.804.7925.1317.84−1.681.400.9827.1555.2924.66
Deposits/GDP at t − 1 (percent)22.0140.6276.9137.286.81139.6672.0175.008.33
GDP growth at t − 1 (percent)2.301.015.90−3.37−2.992.912.87−3.380.289.34
Current account/GDP at t − 1 (percent)−0.23−2.57−7.89−0.55−2.66−3.62−6.15−2.87−3.33−9.86
Peak NPLs (percentage of total loans)35.0013.0033.0027.6062.404.8036.3024.0035.00
Government-owned bank (percentage of assets) at t − 171.3923.2017.0935.0012.230.000.0040.909.8092.00
Significant bank runs (Y/N)YYNNNNNYYN
Largest one-month percentage drop in deposits (>5%), t through t + 17.515.569.1214.06
Credit boom (Y/N)NYNYNNYN
Annual growth in private credit to GDP, t – 41.6010.606.1015.006.065.2213.100.50
through t – 1 (percent)
Creditor rights in year t2232341231
Containment phase
Deposit freeze (Y/N)NNNNNNNYNN
Introduction of deposit freeze2002
Duration of deposit freeze (in months)36
Coverage of deposit freeze (time deposits only? Y/N)Y
Bank holiday (Y/N)NNNNNNNYNN
Introduction of bank holiday2002
Duration of bank holiday (days)5
Blanket guarantee (Y/N)NYYYNNNNNN
Date of introductionSep. 1992Aug. 1997Dec. 2000
Date of removalJul. 1996Jan. 2005Jul. 2004
Duration of guarantee (months)468943
Previous explicit deposit insurance arrangement (Y/N)YNNYYYYYYN
Timing of first bank interventionNoneSep. 1991Mar. 1997Nov. 19971995Sep. 2007Mar. 2008Feb. 2002Jan. 1994Fall 1998
Timing of first liquidity assistance
Significant liquidity support or emergency lending (Y/N)NYYYYNNYYN
Support different across banks? (Y/N)YNNY
Collateral requiredNYYY
Remunerated (Y/N)YYY
If remunerated, interest at market rates (Y/N)YYY
Peak support (percentage of deposits)3.109.4025.9022.1616.302.0631.0031.205.20
Lowering of reserve requirements (Y/N)NNNNNNNYY
Resolution phase
Forbearance (Y/N)YYNYNNNYY
Banks not intervened in despite being technically insolventYNYNY
Prudential regulations suspended or not fully appliedYYYYY
Large-scale government intervention (Y/N)YYYYYNNYYY
Institutions closed (percentage of banks’ assets)002.008.002.000018.8323.002.00
Number of banks in t1184180230315183
Number of banks in t + 310340541782139
Bank closures (Y/N)NNYYYNNYYY
Number of bank closures during t through t + 311248N5125
Other financial institution closures (Y/N)YYNYNY
Shareholder protection (shareholders made whole? (Y/N)YNNNNNN
Nationalizations (Y/N)NYYYNYNYYN
Mergers (Y/N)NYYYNYNNY
Did bank shareholders inject new capital? (Y/N)NYYY
Sales to foreigners (Y/N)NYYNNNYY
Number of banks sold to foreigners during t through t + 53225
Bank-restructuring agency (Y/N)NYYNYNNNYN
Asset management company (Y/N)NYYYNYYNY
Centralized (Y/N)YYYNYY
Recapitalization (Y/N)YYYYNYYYYY
Recap measures
CashYY
Government bondsYYYYY
Subordinated debtY
Preferred sharesY
Purchase of bad loans
Credit line
Assumption of bank liabilitiesY
Ordinary sharesYY
Recap level (percent)8.008.508.0010.00
Recap cost (gross) (percentage of GDP)3.601.8518.8024.500.206.185.595.00
Recovery (Y/N)NYNYNN
Recovery proceeds during t through t + 5 (percent0.361.16
Recap cost (net) (percentage of GDP)3.601.4918.8024.500.205.025.595.00
Deposit insurance (Y/N)YNNYYYYYYN
Formation1987198319982001193320021985
Coverage limit (in local currency) at t100,000Full1,20035,000100,000100,000250,000
Coverage ratio (coverage limit to GDP per capita) at t7.180.591.92.261.40.96
Were losses imposed on depositors? (Y/N)NNYNYNNNYN
If yes, severe = 1 and moderate = 2212
Macro policies
Monetary policy index010−1−1−111
Average change in reserve money during t through t + 3 (percent)15.3921.6312.9533.9933.2617.3779.3224.17
Fiscal policy index11111−111
Average fiscal balance during t through t + 3 (percent)−7.67−7.33−2.51−10.55−2.000.04−1.64−2.74
IMF program (Y/N)NNYYYYYN
IMF program put in place (year)19982000199519961996
Outcome variables
Fiscal cost net (percentage of GDP)5.000.2034.8030.700.0010.8312.5010.00
Gross (percent)5.003.6043.8032.000.0020.0015.0010.00
Recovery during t through t + 5 (percent)03.409.001.3009.172.500
Output loss
Output loss during t through t + 3 (percent)2.2030.6097.665.350.0028.799.6219.72
Sources: IMF staff reports; country authorities’ reports; newspaper articles; and authors’ calculations.Note: t denotes the starting year of the crisis; NPL = nonperforming loan; M2 = broad money.
Sources: IMF staff reports; country authorities’ reports; newspaper articles; and authors’ calculations.Note: t denotes the starting year of the crisis; NPL = nonperforming loan; M2 = broad money.

The selection of crisis episodes was determined by the availability of detailed information on applicable policies. A variety of sources were relied upon, including IMF staff reports and working papers, World Bank documents, and central bank and academic publications. The exact sources of the data may be found in the electronic version of the database.12

Initial Conditions

Table 2.2 reports summary statistics for the initial conditions variables. The data show that the selected banking crises tend to coincide with currency crises, whereas they rarely coincide with sovereign debt crises. In 55 percent of the cases, the banking crisis coincides with a currency crisis, but in only 11 percent of cases does the banking crisis coincide with a debt crisis.

Table 2.2Descriptive Statistics of Initial Conditions of Selected Banking Crises
VariableNumber of crisesMeanStandard deviationMinimumMaximum
Start year of banking crisis4219956.10019802007
Currency crisis (Y/N)420.5480.5040.0001.000
Sovereign debt crisis (Y/N)420.1190.3280.0001.000
Fiscal balance/GDP (percent)42−0.0210.045−0.1700.056
Debt/GDP (percent)330.4640.3950.0801.913
Inflation (percent)411.3714.862−0.00724.772
Net foreign assets/M2 (percent)420.1740.189−0.3510.576
Deposits/GDP (percent)420.4910.4540.0622.524
GDP growth (percent)420.0240.045−0.0980.100
Current account/GDP (percent)41−0.0390.049−0.2490.025
Peak NPLs (percent of total loans)400.2520.1550.0400.750
Government-owned banks (percent of total assets)420.3090.2450.0000.920
Bank runs (Y/N)420.6190.4910.0001.000
Largest one-month drop in deposits to GDP (percent)260.1120.0580.0560.267
Credit boom (Y/N)330.3030.4670.0001.000
Annual growth in private credit to GDP before crisis (percent)330.0830.098−0.1990.341
Creditor rights (index)411.7801.1290.0004.000
Source: Authors’ calculations.Note: M2 = broad money; NPL = nonperforming loan.
Source: Authors’ calculations.Note: M2 = broad money; NPL = nonperforming loan.

Macroeconomic conditions are often weak before a banking crisis. Fiscal balances tend to be negative (−2.1 percent on average), current accounts tend to be in deficit (−3.9 percent), and inflation often runs high (137 percent on average) at the onset of the crisis. However, the role of macroeconomic fundamentals has evolved across generations of crises. Whereas crises such as those in Russia in 1998, Argentina in 2001, and most crises of the 1980s were precipitated by large macroeconomic imbalances, particularly unsustainable fiscal policies, the East Asian crises of the late 1990s had more to do with the maturity composition of debt and foreign exchange risk exposures, rather than the levels of public debt and the fiscal deficit.

Nonperforming loans tend to be high at the onset of a banking crisis, running as high as 75 percent of total loans and averaging about 25 percent of loans. It is not always clear, though, to what extent the sharp rise of nonperforming loans is caused by the crisis itself or whether it reflected the effects of the tightening of prudential requirements during the aftermath of the crisis. In Chile, for instance, nonperforming loans peaked at 36 percent of total loans only in 1986, several years after the start of the crisis. However, part of the unsound banking practices that led to the Chilean banking crisis was the existence of substantial connected loans, which ranged from 12 to 45 percent of banks’ total loan portfolios (Sanhueza, 2001).

Government ownership of banks is common in crisis countries, with the government owning about 31 percent of banking assets on average. In many cases, government ownership may have become a vulnerability—problems at state-owned banks have been major contributors to the cost and unfolding of crises, with many exhibiting low asset quality before the onset of a crisis. In Uruguay, for instance, state-owned banks Republica and Hipotecario—accounting for 40 percent of the system’s assets—exhibited nonperforming loans rates of 39 percent of total loans as of 2001, compared with 5.6 percent at private banks in Uruguay (IMF, 2003). In Turkey, duty losses at state-owned banks were estimated at 12 percent of GNP as early as in 1999, and state-owned bank Bapindo in Indonesia had experienced important losses as early as 1994, three years before the onset of its crisis (Enoch and others, 2001).

Bank runs are a common feature of banking crises, with 62 percent of crises experiencing momentary sharp reductions in total deposits. The largest one-month drop in the ratio of deposits to GDP for countries experiencing bank runs averaged about 11.2 percent, and was as high as 26.7 percent in one case. Severe runs are often system wide, but it is also common to observe a flight to quality within the system from unsound banks to sound banks, which implies no or moderate systemic outflows. During the Indonesian crisis in 1997, for instance, private national banks lost 35 trillion rupiah in deposits between October and December 1997, whereas state-owned banks and foreign and joint-venture banks gained 12 and 2 trillion rupiah, respectively (Batunanggar, 2002). A similar situation occurred in Paraguay following the interventions in the third and fourth largest banks and the uncovering of unrecorded deposits. Depositors migrated from these banks to those perceived as more solid.

Banking crises are also often preceded by credit booms, with rapid precrisis credit growth in about 30 percent of crises. Average annual growth in the ratio of private credit to GDP before the crisis was about 8.3 percent across crisis countries, and was as high as 34.1 percent in Chile. Credit booms were often preceded by financial liberalization processes, such as the one that led to the crisis in the Nordic countries in the 1990s (Drees and Pazarbasioglu, 1998).

Crisis-affected countries often suffer from weak legal institutions, rendering a speedy resolution of distressed assets hard to accomplish. Scores on the creditor rights index in the selected crisis countries averaged about 1.8, ranging from a low of 0 to a high of 4 (the maximum possible score).

In summary, initial conditions are important because they may shape the market’s and policymakers’ responses during the containment phase. If macroeconomic conditions are weak, policymakers have limited buffers to cushion the impact of the crisis, and the burden falls on the shoulders of containment and resolution policies. Moreover, sudden changes in market expectations may gather strength rapidly depending on how weak initial conditions in the country are, in particular, the macroeconomic setting, the institutional environment, and the banking sector.

Take, for instance, Turkey in 2000. The trigger of the crisis was the collapse of interbank loans from large banks to a few small banks on November 20, in particular to DemirBank, which depended greatly on overnight funding. Turkey was widely known to have macroeconomic vulnerabilities, with inflation hovering around 80 percent per year during the 1990s, high fiscal deficits, large public debt, high current account deficits, and a weak financial system. Banks had high exposures to the government through large holdings of public securities, and also had sizable maturity and exchange rate risk mismatches, making them highly vulnerable to market risk. When credit lines to DemirBank were cut, several small banks were forced to sell their government securities. This sell-off caused a sharp drop in the price of government securities and triggered panic among foreign investors, a reversal in capital flows, sharp increases in interest rates, and declines in the value of the Turkish lira. Within a few weeks of these developments, the Turkish government announced a blanket guarantee.

An opposite example is Argentina in 1995, where the contagion from the Tequila crisis was weathered successfully with a substantial consolidation of the banking sector and small fiscal costs, in large part due to robust macroeconomic performance during the preceding years.

Crisis Containment

Table 2.3 reports summary statistics for the crisis containment and resolution policies of the 42 selected banking crisis episodes.

Table 2.3Descriptive Statistics of Crisis Resolution Policies of Selected Banking Crisis Episodes
VariableNumber of crisesMeanStandard deviationMinimumMaximum
Deposit freeze (Y/N)420.1190.32801
Duration of deposit freeze (months)540.60046.0306120
Coverage of deposit freeze: time deposits only? (Y/N)50.4000.54801
Bank holiday (Y/N)420.0950.29701
Duration of bank holiday (days)44.7500.50045
Blanket guarantee (Y/N)420.2860.45701
Duration of guarantee (months)1453.07133.99211109
Previous explicit deposit insurance arrangement (Y/N)420.5240.50501
Liquidity support or emergency lending (Y/N)420.7140.45701
Liquidity support different across banks? (Y/N)180.5000.51401
Collateral required for liquidity provision150.4670.51601
Collateral provided is remunerated (Y/N)130.8460.37601
If remunerated, interest at market rates (Y/N)110.6360.50501
Peak liquidity support (fraction of deposits)410.2770.49703
Lowering of reserve requirements (Y/N)410.3660.48801
Forbearance (Y/N)420.6670.47701
Banks not intervened in despite being technically insolvent370.3510.48401
Prudential regulations suspended or not fully applied370.7300.45001
Large-scale government intervention in banks (Y/N)420.8570.35401
Fraction of financial institutions closed390.0830.11700.500
Bank closures (Y/N)420.6670.47701
Other financial institution closures (Y/N)340.5000.50801
Were shareholders made whole? (Y/N)300.0670.25401
Nationalizations (Y/N)420.5710.50101
Mergers (Y/N)410.6100.49401
Did private bank shareholders inject fresh capital? (Y/N)240.6670.48201
Sales to foreigners (Y/N)370.5140.50701
Bank-restructuring agency (Y/N)400.4750.50601
Asset management company (Y/N)420.5950.49701
Centralized asset management company (Y/N)250.8400.37401
Recapitalization of banks (Y/N)420.7620.43101
Recap level (percent)130.0780.0200.0400.100
Recap cost to government (gross) (fraction of GDP)320.0780.0960.0020.373
Recovery of recap expense (Y/N)310.5160.50801
Recovery proceeds (fraction of GDP)310.0190.05300.279
Recap cost to government (net) (fraction of GDP)320.0600.07900.373
Deposit insurance (Y/N)420.5240.50501
Coverage limit to per capita GDP351.1421.73007.180
Were losses imposed on depositors? (Y/N)420.3100.46801
Monetary policy index40−0.0500.815−11
Change in reserve money (rate)351.6814.562−0.07020.47
Fiscal index400.6000.709−11
Fiscal balance (share of GDP)40−0.0360.030−0.1270.008
IMF program put in place (Y/N)420.5240.50501
Fiscal cost net (share of GDP)400.1300.13300.551
Gross fiscal cost (share of GDP)400.1570.15000.568
Recovery of fiscal expense400.0270.04800.261
Output loss (share of GDP)400.2010.26000.977
Source: Authors’ calculations.
Source: Authors’ calculations.

The data show that emergency liquidity support and blanket guarantees are two commonly used containment measures. Extensive liquidity support was used in 71 percent of crises considered and blanket guarantees were used in 29 percent of crisis episodes. Deposit freezes and bank holidays to deal with bank runs were less frequently used. In this sample, only five cases (or 12 percent of episodes) used deposit freezes: Argentina in 1989 and 2001, Brazil in 1990, Ecuador in 1999, and Uruguay in 2002. In all but one case—Brazil in 1990—the deposit freeze was preceded by a bank holiday. Bank holidays were used in only 10 percent of crises and only in the cases mentioned above. In all episodes in which holidays and deposit freezes were used, bank runs occurred. Bank holidays typically do not last long, about five days on average. However, deposit freezes last much longer, up to 10 years in one case, and about 41 months on average. The longest freeze corresponded to the Bonex Plan implemented in Argentina in 1989.13 After the conversion, the bonds traded with a discount of almost two-thirds and recovered to about 50 percent within a few months. Similarly, in Ecuador, depositors received certificates of reprogrammed deposits, which traded at significant discounts depending on the perceived solvency of the issuing bank. Moreover, bank runs resumed as soon as the unfreezing began (Jacome, 2004). It seems that at least in these cases, deposit freezes were highly disruptive, imposing severe losses on depositors, and therefore should be considered only in extreme circumstances. Bank holidays, however, may be used to buy time until a clear strategy is laid out; they were also used in the United States during the Great Depression in the 1930s.

Unlike the Bonex Plan in Argentina in 1989 and the deposit freeze in Uruguay in 2002—both of which covered dollar-denominated time deposits at public banks—the other episodes in which this instrument was used also covered deposits other than time deposits. The 2001 freeze in Argentina, for example, began with the corralito, which limited withdrawals to up to US$250 a week, prohibited transfers abroad unless trade related, introduced marginal reserve requirements, and limited transactions that could reduce deposits. However, soon after the corralito, the corralon was implemented, which reprogrammed time deposits over a five-year horizon. Similarly, in Brazil in 1990, the freeze included M2 plus federal securities in the hands of the public, except balances below 50,000 Brazilian cruzados novos (NCz$) for checking accounts and NCz$25,000 for savings accounts or 20 percent of the balance (whichever was larger) for deposits in the overnight domestic debt market, and 20 percent of the balance for mutual funds. The broadest freeze recorded in the sample was implemented by Ecuador, and included savings deposits up to US$500, half of checking account balances, repurchase agreements, and all time deposits.

Blanket guarantees also tend to be in place for a long period, about 53 months on average. A blanket guarantee is another policy tool that—if successful—may buy some time for policymakers to implement a credible policy package. Using the data set presented in this chapter, Laeven and Valencia (2012) examine the effectiveness of blanket guarantees in restoring depositor confidence and find that they are often successful. However, they also find that outflows by foreign creditors are virtually unresponsive to the announcement of such guarantees, despite being covered in most cases. They find that such guarantees tend to have high fiscal costs, confirming earlier results by Honohan and Klingebiel (2000), but argue that this correlation is driven mainly by the fact that guarantees are usually adopted in conjunction with extensive liquidity support and when crises are severe.

Peak liquidity support tends to be sizable and averaged about 28 percent of total deposits across the 42 crisis episodes considered. Liquidity support is clearly the most common first line of response in systemic crisis episodes, even in Argentina in 1995 when a currency board was in place. This was made possible by an amendment of the charter of the Central Bank of Argentina in February 1995, allowing it to lengthen the maturities of its swap and rediscount facilities, with the possibility of monthly renewal, and in amounts exceeding the net worth of the borrowing bank.

In severe crises, a positive correlation of about 30 percent between the provision of extensive liquidity support and the use of blanket guarantees is observed. Blanket guarantees are often introduced to restore confidence, even when previous explicit deposit insurance arrangements are already in place (this is so in about 52 percent of crises in which blanket guarantees are introduced). In some cases, guarantees were introduced to cover only a segment of the market, not all banks. Examples of such partial guarantees are provided in Table 2.4.

Table 2.4Selected Bank-Specific Guarantee Announcements
CountryDateCoverage
ChileJan. 1983Explicit guarantee announced to depositors of intervened-in banks.
Czech RepublicJun. 1996Deposit insurance coverage was raised substantially (from 100,000 Czech koruny to 4,000,000 Czech koruny) for 18 banks that had entered restructuring programs.
Dominican RepublicApr. 2003In their intervention, the authorities announced that all legitimate deposits of Baninter would be honored with central bank certificates. Later, the same treatment was applied in the resolution of two other banks.
LithuaniaDec. 1995The government passed a law extending full coverage to two closed banks.
ParaguayJul. 1995All recorded deposits in intervened in banks (unrecorded deposits were initially excluded, though in May 1996 a law was passed to compensate these, too).
United KingdomSep. 2007All liabilities of Northern Rock outstanding as of September16, 2007.
Sources: IMF Staff reports; and country authorities’ reports.
Sources: IMF Staff reports; and country authorities’ reports.

Crisis Resolution

Table 2.3 reports summary statistics for the crisis resolution policies of the 42 selected banking crisis episodes.

Regulatory forbearance is a common feature of crisis management. The policy objective aims at a gradual recovery of the banking system, or a gradual transitioning toward stricter prudential requirements. The latter is a common outcome whenever modifications to the regulatory framework are introduced. In Ecuador, for instance, banks were given two years to comply fully with new loan classification rules, among other requirements. In the 2001 crisis in Argentina, the authorities granted regulatory forbearance, which included a new valuation mechanism for government bonds and loans, allowing for gradual convergence to market value. Banks were also allowed to temporarily decrease their capital charge on interest rate risk, and losses stemming from court injunctions14 could be booked as assets to be amortized over 60 months. Prolonged forbearance occurred in about 67 percent of crisis episodes. In 35 percent of the episodes, forbearance took the form of banks not being intervened despite being technically insolvent, and in 73 percent of cases prudential regulations were suspended or not fully applied.

Forbearance, however, does not really solve the problems; therefore, a key component of almost every systemic banking crisis is a bank-restructuring plan. In 86 percent of cases, large-scale government intervention in banks took place as bank closures, nationalizations, or assisted mergers. The system survived a crisis without having at least significant bank closures in only a handful of episodes. For instance, in Latvia, banks holding 40 percent of assets were closed, but no further intervention by the government was implemented. In Argentina, in the 1995 episode, 15 institutions ran into problems: 5 of them were liquidated (with 0.6 percent of the system’s assets), 6 were resolved under a purchase and assumption scheme (with 1.9 percent of the system’s assets), and 4 were absorbed by healthier institutions. However, a significant consolidation process also took place through 14 mergers involving 47 financial institutions. Shareholders often lost money when banks were closed and were often forced to inject new capital in the banks they owned.

Closures were not limited to banks but included nonbank financial institutions. In Thailand, for instance, the problem began with liquidity problems at finance companies as early as March 1997, and 56 finance companies (accounting for 11 percent of the financial system’s assets) were closed. In Jamaica, a large component of the financial problem was in the insurance sector, whose restructuring cost reached 11 percent of GDP.

Sales to foreigners is often seen as a last resort before bank restructuring, though it has become common in recent crises. On average, sales of banks to foreigners occurred in 51 percent of the crisis episodes.

Bank closures seem to be associated with larger fiscal costs, with a positive correlation of 22 percent. However, bank closures are negatively associated with the issuance of a blanket guarantee, with a correlation of −22 percent. Because guarantees entail a sizable fiscal contingency, once the guarantee is in place, governments may try to refrain from closing banks to avoid activating the guarantee. Bank closures also seem to be positively associated with peak nonperforming loans, with a correlation of about 25 percent. One potential contributing factor to this correlation is that once a bank is closed, its asset quality may deteriorate because in the process of closing, any value attached to bank relationships with customers may be destroyed. Borrowers may delay payments, or the collection of loans may become less effective than before, which may also contribute to higher fiscal costs.

Special bank-restructuring agencies are often set up to restructure distressed banks (in 48 percent of crises), and AMCs were set up in 60 percent of crises to manage distressed assets. AMCs tend to be centralized rather than decentralized. Examining the cases in which AMCs were used shows that the use of AMCs is positively correlated with peak nonperforming loans and fiscal costs, with correlation coefficients of about 15 percent in both cases. These correlations may suggest some degree of ineffectiveness in AMCs, at least in those episodes in which AMCs were established. In line with these simple correlations, Klingebiel (2000), who studied seven crises in which AMCs were used, concludes that they were largely ineffective.

Another important policy used in the resolution phase of banking crises is recapitalization of banks. In 33 out of the 42 selected crisis episodes, banks were recapitalized by the government. Recapitalization costs constitute the largest fraction of the fiscal costs of banking crises and takes many forms. In 12 crises, recapitalization took place in cash; in 14 crises, in government bonds; in 11 episodes subordinated debt was used; in 6 crises, preferred shares were used; in 7 crises, it took place through the purchase of bad loans; in 4 crises, the government purchased ordinary shares of banks; in 3 crises, the government assumed bank liabilities; and in 2 crises, a government credit line was extended to banks. In some cases, a combination of these methods was used. Recapitalization usually entails writing off losses against shareholders’ equity and injecting either Tier 1 or Tier 2 capital or both. Recapitalization programs are usually accompanied by some conditionality. For instance, in Chile, a nonperforming loans purchase program was implemented, and during this period banks could not distribute dividends and all profits and recoveries had to be used to repurchase the loans. In Mexico, PROCAPTE (a temporary recapitalization program) had FOBAPROA (a deposit insurance fund) purchase subordinated debt from qualifying banks, but the resources had to be deposited at the central bank, bearing the same interest rate as the subordinated bonds. Banks could redeem the bonds if their capital adequacy ratio rose above 9 percent, but FOBAPROA had the option to convert the bonds into stocks after five years or if a bank’s Tier 1 capital ratio fell below 2 percent.

Similar conditions were applied to recapitalization programs in Turkey in 2000 and Thailand in 1997. In the former, SDIF (the Turkish deposit insurance fund) matched owners’ contributions to bring banks’ Tier 1 capital to 5 percent, but only for banks with a market share of at least 1 percent. SDIF could also contribute to Tier 2 capital through subordinated debt, to all banks with Tier 1 capital greater or equal to 5 percent. Similar to Mexico, if Tier 1 capital fell below 4 percent, the subordinated debt would convert to stocks. In Thailand, the recapitalization plan involved Tier 1 capital injections, with the government matching private contributions and the requirement that the financial institution make full provisions up front, in line with new regulations. Additionally, the government and the new investors had the right to change the board of directors and management of each participating financial institution. The government also had the right to appoint at least one board member to each financial institution. The program also included Tier 2 capital injections equal to a minimum of the total writedown exceeding previous provisioning or 20 percent of the net increase in lending to the private sector, among other criteria.

On average, the net recapitalization cost to the government (after deducting recovery proceeds from the sale of assets) amounted to 6.0 percent of GDP across crisis countries in the sample, though in the case of Indonesia it reached as high as 37.3 percent of GDP.

It is interesting that about half the countries that experienced a systemic banking crisis had an explicit deposit insurance scheme in place at the outbreak of the crisis (and several countries adopted deposit insurance during the crisis). Losses were imposed on depositors in a minority of cases. Table 2.5 provides brief descriptions of those circumstances in which depositors faced losses. Simple correlations show that episodes in which losses were imposed on depositors faced higher output losses, with a correlation of about 8 percent.

Table 2.5Episodes with Losses Imposed on Depositors
CountryCrisis yearLoss severityDescription
Argentina1989LargeBonex Plan converted time deposits into long-term bonds at an exchange rate below that prevailing on the market.
Argentina2001LargeDollar deposits were converted into domestic currency at Arg$1.4, which was below the prevailing market rate.
Bolivia1994Minor to moderateLarge depositors of the two closed banks received as compensation non-interest bearing bonds.
Chile1981Minor to moderateIn 1983, depositors at banks forced into liquidation were paid 70 percent of face value.
Côte d’Ivoire1988LargeIn the liquidation of BDN, 85 percent of depositors were compensated fully.
Ecuador1998LargeFrozen deposits were significantly eroded by accelerating inflation and depreciation of the currency, and some payments to depositors are still pending (despite the blanket guarantee)
Estonia1992LargeDepositors of Tartu Commercial Bank were partially paid.
Latvia1995LargeWith the collapse of Baltija Bank the government compensated depositors for LVL 500 ($1,000) per depositor (LVL 200 in 1995 and LVL 100 over next three years).
Lithuania1995Minor to moderateDepositors of Litimpex Bank had their deposits turned into equity. Furthermore, depositors of Innovation Bank received some cash (LTL 4,000 in 1997 and LTL 4,000 in 1998 per person) and the difference in five-year, non-interest-bearing government bonds; legal entities received ten-year, non-tradable, non-interest bearing notes for the entire claim; certain public organizations, embassies, charities, and the like received cash during 1998; other creditors received their pari passu share of residual funds left from collection of Innovation Bank’s assets. Public sector deposits were written off.
Russia1998Minor to moderateSome depositors (those whose savings were not transferred to Sberbank) sustained losses at insolvent banks. Even those who benefited from the transfer faced some losses because the exchange rate used in the transaction was less than half of the market exchange rate prevailing at the time.
Thailand1997Minor to moderateDepositors of the closed finance companies received certificates yielding below-market interest rates.
Ukraine1998LargeDepositors were not fully compensated.
Venezuela1994Minor to moderateDepositors at Banco Latino with more than 10 million bolivars received long-term nonnegotiable bonds with interest rate below market, for the amount exceeding the 10 million.
Sources: IMF staff reports; and country authorities’ reports.Note: Arg$ = Argentine peso; LTL = Lithuanian litas; LVL= Latvian lats.
Sources: IMF staff reports; and country authorities’ reports.Note: Arg$ = Argentine peso; LTL = Lithuanian litas; LVL= Latvian lats.

Monetary policy tended to be fairly neutral during crisis episodes, whereas the fiscal stance tended to be expansive, arguably to support the financial and real sectors, and to accommodate bank-restructuring and debt-restructuring programs. On average, the fiscal balance was about −3.6 percent of GDP during the initial years of a banking crisis.

The IMF participated in programs in about 52 percent of the episodes considered.

Fiscal Costs and Real Effects of Banking Crises

Fiscal costs, net of recoveries, associated with crisis management can be substantial, averaging about 13.3 percent of GDP, and can be as high as 55.1 percent of GDP. Recoveries of fiscal outlays vary widely as well, with the average recovery rate reaching 18.2 percent of gross fiscal costs. Although countries that used AMCs seemed to achieve slightly higher recovery rates, the correlation is very small, at about 10 percent.

Finally, output losses (measured as deviations from trend GDP) of systemic banking crises can be large, averaging about 20 percent of GDP during the first four years of the crisis, and ranging from a low of 0 percent to a high of 98 percent of GDP.

Global Liquidity Crisis of 2007–08

During the course of 2007, U.S. subprime mortgage markets melted down and global money markets were under pressure. The U.S. subprime mortgage crisis manifested itself first through liquidity issues in the banking system owing to a sharp decline in demand for asset-backed securities. Hard-to-value structured products and other instruments created during a boom of financial innovation had to be severely marked down because of the newly implemented fair value accounting and credit rating downgrades. Credit losses and asset writedowns worsened with declining house prices and accelerating mortgage foreclosures, which increased in late 2006 and degenerated further in 2007 and 2008. Profits at U.S. banks declined to $5.8 billion from $35.2 billion (83.5 percent) during the fourth quarter of 2007 compared with the previous year as a result of provisions for loan losses. As of August 2008, subprime-related and other credit losses or writedowns by global financial institutions stood at about US$500 billion.

This section briefly compares the ongoing global liquidity crisis and its policy responses to the other crises included in the database. Given that the global liquidity crisis is still very much unfolding at the time of this writing, this analysis is obviously preliminary and incomplete.

Initial Conditions

The underlying causes of the global 2007–08 financial crisis are still being debated, and most likely can be attributed to a combination of factors. However, from the perspective of describing its initial conditions, it is useful to classify the underlying factors into two groups: macroeconomic and microeconomic.

The macroeconomic context is characterized by a prolonged period of excess global liquidity induced, in part, by relatively low interest rates set by the U.S. Federal Reserve Bank and other central banks following the 2001 recession in the United States. The excess liquidity fueled domestic demand, particularly residential investment, triggering a significant increase in house prices, which more than doubled in nominal terms between 2000 and mid-2006.15 During this period, the U.S. economy faced high current account deficits, reaching 7 percent of GDP in the last quarter of 2005, induced primarily by household expenditure but also by sizable fiscal deficits.

However, microeconomic factors related to financial regulation (and lack thereof) and industry practices by financial institutions also appear to have played a crucial role in the buildup of the bubble. The “originate-and-distribute” lending model (see Bhatia 2007 for a description) adopted by many financial institutions during this period seems to have exacerbated the problem. Under this approach, banks made loans primarily to sell them on to other financial institutions, which, in turn, would pool them to issue asset-backed securities. The underlying rationale for these loan sales was to transfer risk to the ultimate buyer of the security backed by the underlying mortgage loans. These securities could then be pooled again and new instruments would be created and so forth.

A mispricing of risk of mortgage-backed securities linked to subprime loans led the market to believe that there was an arbitrage opportunity. This market perception fueled demand for these instruments and contributed to a deterioration in underwriting standards by banks in an attempt to increase the supply of loans to meet the demand for securitized instruments. Regulatory oversight missed the buildup of vulnerabilities induced by this process because the risks were being transferred to the unregulated segment of the market. The premise was that heavily regulated banks would only be originators, and the ultimate holders of securities were beyond the scope of regulation. In this process, however, spillover effects and systemic risks seem to have been neglected by regulators, and the regulated segment ended up being significantly affected. The crisis reached a global dimension as it became apparent that foreign banks, mainly European, had also played a significant role in the demand for mortgage-related (particular subprime mortgage–linked) securities. For U.K. banks, this shock coincided with a homegrown house price bubble.

In addition to a move toward the originate-and-distribute lending model, many banks, particularly in the United Kingdom, increasingly relied on wholesale funding. As the crisis unfolded, banks that relied heavily on wholesale markets for their funding, such as Northern Rock in the United Kingdom, were hit particularly badly, causing stress in global money markets. Given ongoing concerns with counterparty risk, notably regarding adequacy of banks’ capital, money market strains have continued.

At first glance, the buildup of this crisis episode in the United States and the United Kingdom does not seem to differ significantly from the traditional boom-bust cycles observed in the other crisis countries in this database. Many of these historical crisis episodes experienced buildups of asset price bubbles, especially real estate bubbles, often originating from financial liberalization. In many cases, deregulation of financial systems led to rapid expansion of credit, but with deficiencies in risk management and pricing as the financial system was evolving and prone to abuse. In the United States, it was not financial liberalization in the conventional sense, but innovation of financial instruments that the market and regulators did not fully understand. Supported by these new financial products and asset securitization, mortgage credit markets expanded rapidly but then virtually collapsed in some segments as the financial crisis unfolded. In 30 percent of the episodes included in the database in this chapter, the crisis was preceded by a credit boom. In the United States and the United Kingdom, however, although credit rose rapidly—mortgage lending, in particular—the pace of expansion did not satisfy the criteria to be labeled as a credit boom in this chapter.

What is different from many previous financial crises, especially in developing countries, is that the United States and the United Kingdom have thus far not suffered from a sudden stop of capital flows, which has caused major economic stress in other countries. The dollar did depreciate against the euro in the years preceding the 2007 turmoil, but demand for U.S. assets did not contract sharply, possibly because of the dollar’s use as a reserve currency. Also, the speed and breadth with which stress in U.S. mortgage markets spread to other continents, financial institutions (especially securities firms), and financial markets (notably money markets) seem to have been fueled by uncertainty about the unfolding of the subprime crisis, as it became more clear that risk had been mispriced and exposures had not been transparent.

Containment

Average house prices in the United States reached a peak in mid-2006 and began to decline after the initial signs that a financial crisis could be around the corner. Losses at financial institutions began to appear as early as February 2007 with HSBC Finance, the U.S. mortgage unit of HSBC, reporting more than US$10 billion in losses from its U.S. mortgage lending business. Bad news continued in April 2007 with the bankruptcy filing of New Century Financial, one of the biggest subprime lenders in the United States, followed by the rescue of two Bear Stearns hedge funds in June 2007. Problems intensified when, on August 16, 2007, Countrywide Financial, the largest mortgage lender in the United States, ran into liquidity problems because of the decline in value of securitized mortgage obligations, triggering a deposit run on the bank. The Federal Reserve Bank “intervened” by lowering the discount rate by 0.5 percentage point and by accepting $17.2 billion in repurchase agreements for mortgage-backed securities to aid in liquidity. On January 11, 2008, Bank of America bought Countrywide for US$4 billion. Up to this point, containment policy in the United States was limited to alleviating liquidity pressures through the use of existing tools.

During this time, the United Kingdom experienced its own banking sector problems, in light of tight conditions in money markets. In September 2007, Northern Rock, a mid-sized U.K. mortgage lender, received a liquidity support facility from the Bank of England, following funding problems related to turmoil in the credit markets caused by the U.S. subprime mortgage financial crisis. Starting on September 14, 2007, Northern Rock experienced a bank run, until a government blanket guarantee—covering only Northern Rock—was issued on September 17, 2007. The run on Northern Rock highlighted weaknesses in the U.K. financial sector regulatory framework, including the maintenance of adequate capital by financial institutions, bank resolution procedures, and deposit insurance (IMF, 2008). Commercial banks in the United States did not seem to have experienced runs among retail customers, but as mentioned earlier, many institutions faced significant stress in wholesale markets. The blanket guarantee issued on Northern Rock was perhaps the first significant step away from the usual tools used to resolve liquidity problems. However, unlike in other episodes in which blanket guarantees were used, it was introduced at an early stage. In the sample in this chapter, 29 percent of episodes used a blanket guarantee. However, in the majority of them, they were put in place in the midst of a financial meltdown.16 In the Asian countries, for instance, blanket guarantees were announced when markets were under significant stress and the crisis was already of systemic proportions with widespread runs throughout the financial system.

The next significant policy measure adopted by authorities in both countries was an increase in the range of tools available to provide liquidity. The Federal Reserve introduced the Term Securities Lending facility in March 2008 by which it could lend up to US$200 billion of treasury securities to primary dealers secured for a term of 28 days (rather than overnight, as in the program in place at the time) by a pledge of other securities, including federal agency debt, federal agency residential mortgage-backed securities (MBS), and nonagency AAA/Aaa-rated private-label residential MBS. Similarly, it increased its currency swap lines with other central banks in an attempt to reestablish calm in money markets. The Bank of England took similar steps on April 21, 2008, when it announced it would accept a broad range of MBS under the new Special Liquidity Scheme and swap those for government paper for a period of one year to aid banks with liquidity problems. The new scheme enabled banks to temporarily swap high-quality but illiquid mortgage-backed assets and other securities. These steps are common measures in other documented episodes. Central banks usually increase the tools for providing the system with additional liquidity at both longer and more flexible terms.

Following the Fed’s announcement of the expansion of liquidity facilities, a major event took place: the collapse of Bear Stearns, the fifth largest investment bank at the time. Mounting losses caused by its mortgage exposure triggered a run on the bank requiring emergency financial assistance from the government. Bear Stearns was to be purchased by JP Morgan Chase with federal guarantees on its liabilities in March 2008. It was a controversial measure because Bear Stearns was not subject to regulation by the Fed, yet the Fed’s guarantee on its liabilities was crucial to prevent Bear Stearns’ bankruptcy. To some extent the case is similar to the failures of Sanyo Securities and Yamaichi Securities in the Japanese crises (Nakaso, 2001). Neither fell under the scope of the deposit insurance system but were supervised by the Ministry of Finance. However, the collapse of Sanyo caused the first default ever in the Japanese interbank market, resulting in a sharp deterioration in market sentiment. However, Yamaichi was unwound gradually. Because of large counterparty risks, it was believed that an intervention was justified in the case of Bear Stearns, perhaps to avoid a disruption similar to the one that followed the collapse of Sanyo. Although there was no explicit blanket guarantee announced for Bear Stearns, there was a de facto protection of all its creditors. Shareholders of Bear Stearns, however, suffered significant losses.

The containment measures used thus far by U.S. and U.K. authorities to deal with the ongoing financial turmoil were not that different from those used in previous crisis episodes. Almost all crises have used generous liquidity support to deal with illiquid banks. What is different in this episode is that liquidity support has been extended not only to commercial banks but also to investment banks. Blanket guarantees are also not uncommon, although thus far they have mainly been used in developing countries to deal with systemic financial crises in which depositors have lost confidence in the ability of banks to repay depositors.

Resolution

As of the time of this writing, it is too early to discuss how exactly the crisis will be resolved because it is still ongoing and its consequences have not fully materialized. However, some insights can be extracted from the events that have taken place so far.

The vast majority of bank failures observed in the United States since the start of the crisis have been handled through traditional purchase and assumption schemes. This, of course, is no different from what has been done in bank failures in the past. A large fraction of failures included in the database were handled this way, with only 31 percent of episodes imposing losses on depositors.

The most notable failures so far, however, have been those of three major U.S. investment banks: Bear Stearns, Lehman Brothers, and Merrill Lynch. Bear Stearns collapsed on March 16, 2007, after facing major liquidity problems, and was sold to JP Morgan Chase after the Federal Reserve Bank of New York agreed to take over Bear Stearns’ US$30 billion portfolio of MBS. Lehman Brothers filed for Chapter 11 bankruptcy protection on September 14, 2008, after failed attempts to sell the bank to private parties. Merrill Lynch was acquired by Bank of America on September 15, 2008.

Another significant event was the placement under conservatorship of Fannie Mae and Freddie Mac, the two largest U.S. housing government-sponsored entities (GSEs). As part of the plan announced on September 7, 2008, the Federal Housing Finance Authority (FHFA) was granted direct oversight of the GSEs, the U.S. Treasury was given authority to inject capital into the GSEs in the form of senior preferred shares and warrants (while dividends on existing common and preferred stock were suspended), and senior management and the boards of directors at both enterprises were dismissed. Effectively, this resulted in nationalization of the two entities. The treasury was also granted temporary authority to purchase agency-backed MBSs, and a short-term credit facility was established for the housing GSEs. The rescue of Fannie and Freddie came shortly after legislation was approved in late July 2008 that gave the U.S. Treasury the power to use public funds to recapitalize them. The bill also contained a tax break of as much as $7,500 for first-time homebuyers, created a new regulator to oversee Fannie Mae and Freddie Mac, and allowed the federal government to insure up to US$300 billion in refinanced mortgages. These measures came after severe declines in the stock prices of Fannie and Freddie following market perceptions of a significant capital shortfall.

Recapitalization measures have been widely used, with 76 percent of episodes in the data set implementing them, but in most cases such measures were implemented only after major insolvency problems at banks. In the United States, gross recapitalization costs reached, as of 2012, 4.5 percent of GDP while those for the United Kingdom, 8.8 percent of GDP.

The crisis at Northern Rock, which was triggered by illiquidity, but in which solvency concerns led to a loss of depositor confidence, was contained at first through a government guarantee on deposits, but when a private sector solution on acceptable terms was not identified by the government, the bank was nationalized on February 22, 2008. Nationalizations are last-resort measures commonly used in previous crises, with 57 percent of episodes in the sample using them. However, they have been more common in developing countries where it may be hard to find new owners for failed banks. In developed economies such as the United Kingdom, where capital is abundant, nationalizations are rare and generally considered something to be avoided. Other U.K. banks that have reported major losses have sought private sector solutions to restore bank capital, mostly by attracting new capital from existing shareholders through rights issues, but also through asset sales and reductions in dividends. Another mortgage lender that experienced stress, Alliance & Leicester, was bought in July 2008 by the Spanish bank Banco Santander.

A noteworthy difference from previous crisis episodes is the role that sovereign wealth funds have played in the 2007–09 crisis in providing new capital to restore the health of banks’ capital positions. Globalization in conjunction with asset securitization provided an international dimension to this crisis by allowing many investors around the world to take a piece of the U.S. mortgage pie. Sovereign wealth funds injected capital in major banks in both the United States and the United Kingdom as part of their recapitalization efforts.

Conclusion

This chapter presents a new database of information on the timing and resolution of banking crises. The data show that fiscal costs associated with banking crises can be substantial and that output losses are large. Although countries have adopted a variety of crisis management strategies, emergency liquidity support and blanket guarantees have frequently been used to contain crises and restore confidence, though not always with success.

Policy responses to financial crises normally depend on the nature of the crises, and some unsettled issues remain. First, fiscal tightening may be needed when unsustainable fiscal policies are the trigger of the crises, though crises are typically attacked with expansionary fiscal policies. Second, tight monetary policy could help contain financial market pressures. However, in crises characterized by liquidity and solvency problems, the central bank should stand ready to provide liquidity support to illiquid banks. In the event of systemic bank runs, liquidity support may need to be complemented by depositor protection (including through a blanket government guarantee) to restore depositor confidence, although such accommodative policies tend to be very costly and will not necessarily speed up economic recovery. All too often, intervention is delayed because regulatory forbearance and liquidity support are used for too long to deal with insolvent financial institutions in the hope that they will recover, ultimately increasing the stress on the financial system and the real economy.

The preliminary analysis based on partial correlations indicates that some resolution measures are more effective than others in restoring the banking system to health and containing the fallout on the real economy. Above all, speed appears of the essence. As soon as a large part of the financial system is deemed insolvent and has reached systemic crisis proportions, bank losses should be recognized, the scale of the problem should be established, and steps should be taken to ensure that financial institutions are adequately capitalized. A successful bank recapitalization program tends to be selective in its financial assistance to banks, specifies clear quantifiable rules that limit access to preferred stock assistance, and enacts capital regulation that establishes meaningful standards for risk-based capital. Government-owned AMCs appear largely ineffective at resolving distressed assets, largely because of political and legal constraints.

Next, the adverse impact of the stress on the real economy needs to be contained. To relieve indebted corporations and households from financial stress and restore their balance sheets to health, intervention in the form of targeted debt-relief programs for distressed borrowers and corporate restructuring programs appear most successful. Such programs typically require public funds, and tend to be most effective when they are well targeted with adequate safeguards attached.

Future research based on this data set needs to discuss in more detail how policymakers should respond to financial system stress in a way that ensures that the financial system is restored to health while containing the fallout on the economy. Such research should establish to what extent fiscal costs incurred by accommodative policy measures (such as substantial liquidity support, explicit government guarantees, and forbearance from prudential regulations) help to reduce output losses and to accelerate the speed of economic recovery, and identify crisis resolution policies that mitigate moral hazard problems.

References

    BatunanggarSukarela2002Indonesia’s Banking Crisis Resolution: Lessons and the Way Forward” (unpublished; London: Center for Central Banking Studies [CCBS], Bank of England).

    BhatiaAshok2007New Landscape, New Challenges: Structural Change and Regulation in the U.S. Financial SectorIMF Working Paper 07/195 (Washington: International Monetary Fund).

    BeimDavid and CharlesCalomiris2001Emerging Financial Markets (New York: McGraw-Hill/Irwin Publishers).

    CalomirisCharlesDanielaKlingebiel and LucLaeven2005Financial Crisis Policies and Resolution Mechanisms: A Taxonomy from Cross-Country Experience” in Systemic Financial Crises: Containment and Resolution ed. by PatrickHonohan and LucLaeven (CambridgeUnited Kingdom: Cambridge University Press) pp. 2575.

    CaprioGerard and DanielaKlingebiel1996Bank Insolvencies: Cross-Country ExperiencePolicy Research Working Paper No. 1620 (Washington: World Bank).

    CaprioGerardDanielaKlingebielLucLaeven and GuillermoNoguera2005Appendix: Banking Crisis Database” in Systemic Financial Crises: Containment and Resolution ed. by PatrickHonohan and LucLaeven (CambridgeUnited Kingdom: Cambridge University Press) pp. 30740.

    ClaessensStijnDanielaKlingebiel and LucLaeven2003Financial Restructuring in Banking and Corporate Sector Crises: What Policies to Pursue?” in Managing Currency Crises in Emerging Markets ed. by MichaelDooley and JeffreyFrankel (Chicago: University of Chicago Press) pp. 14780.

    ClaessensStijn2005Crisis Resolution, Policies, and Institutions: Empirical Evidence” in Systemic Financial Crises: Containment and Resolution ed. by PatrickHonohan and LucLaeven (CambridgeUnited Kingdom: Cambridge University Press) pp. 16996.

    CollynsCharles and G. RussellKincaid2003Managing Financial Crises: Recent Experience and Lessons for Latin AmericaIMF Occasional Paper No. 217 (Washington: International Monetary Fund).

    Demirgüç-KuntAsli and EnricaDetragiache2002Does Deposit Insurance Increase Banking System Stability? An Empirical InvestigationJournal of Monetary Economics Vol. 49 pp. 1373406.

    Demirgüç-KuntAsliEdwardKane and LucLaeven eds.2008Deposit Insurance around the World: Issues of Design and Implementation (CambridgeMassachusetts: MIT Press).

    DjankovSimeonCaraleeMcLiesh and AndreiShleifer2007Private Credit in 129 CountriesJournal of Financial Economics Vol. 84 No. 2 pp. 299329.

    DooleyMichael and JeffreyFrankel eds.2003Managing Currency Crises in Emerging MarketsProceedings of a National Bureau for Economic Research conference. (Chicago: University of Chicago Press).

    DreesBurkhard and CeylaPazarbasioglu1998The Nordic Banking Crisis: Pitfalls in Financial LiberalizationIMF Occasional Paper No. 161 (Washington: International Monetary Fund).

    EnochCharlesBarbaraBaldwinOlivierFrécaut and ArtoKovanen2001Indonesia: Anatomy of a Banking Crisis: Two Years of Living Dangerously 1997–1999IMF Working Paper 01/52 (Washington: International Monetary Fund).

    FrankelJeffrey and AndrewRose1996Currency Crashes in Emerging Markets: An Empirical TreatmentJournal of International Economics Vol. 41 pp. 35166.

    HoelscherDavid and MarcQuintyn2003Managing Systemic Banking CrisesIMF Occasional Paper No. 224 (Washington: International Monetary Fund).

    HonohanPatrick and DanielaKlingebiel2000Controlling the Fiscal Costs of Banking CrisesPolicy Research Working Paper No. 2441 (Washington: World Bank).

    HonohanPatrick and LucLaeven eds.2005Systemic Financial Crises: Containment and Resolution (CambridgeUnited Kingdom: Cambridge University Press).

    HonohanPatrick2003Uruguay—Third Review Under the Stand-By Arrangement and Request for Modification and Waiver of Applicability of Performance Criteria—Report on the 2002 Banking CrisisCountry Report 03/247 (Washington).

    HonohanPatrick2008United Kingdom: 2008 Article IV ConsultationStaff Report 08/271 (Washington).

    JácomeLuis2004The Late 1990s Financial Crisis in Ecuador: Institutional Weaknesses, Fiscal Rigidities, and Financial Dollarization at WorkIMF Working Paper 04/12 (Washington: International Monetary Fund).

    KlingebielDaniela2000The Use of Asset Management Companies in the Resolution of Banking CrisesWorking Paper No. 2294 (Washington: World Bank).

    LaevenLuc and FabianValencia2012The Use of Blanket Guarantees in Banking CrisesJournal of International Money and Finance Vol. 31 No. 5 pp. 122048.

    La PortaRafaelFlorencioLopez-De-Silanes and AndreiShleifer2002Government Ownership of BanksJournal of Finance Vol. 57 pp. 265301.

    LindgrenCarl-Johan2005Pitfalls in Managing Closures of Financial Institutions” in Systemic Financial Crises: Containment and Resolution ed. by PatrickHonohan and LucLaeven (CambridgeUnited Kingdom: Cambridge University Press) pp. 76108.

    LindgrenCarl-JohanGillianGarcia and Matthew I. Saal1996Bank Soundness and Macroeconomic Policy (Washington: International Monetary Fund).

    NakasoHiroshi2001The Financial Crisis in Japan During the 1990s: How the Bank of Japan Responded and the Lessons LearntBIS Paper No. 6 (Basel: Bank for International Settlements).

    SanhuezaGonzalo2001Chilean Banking Crisis of the 1980s: Solutions and Estimation of the CostsWorking Paper 104 (Santiago: Central Bank of Chile).

    SturzeneggerFederico and JerominZettelmeyer2006Debt Defaults and Lessons from a Decade of Crises (CambridgeMassachusetts: MIT Press).

    ValenciaFabianforthcomingBanks’ Precautionary Capital and Credit CrunchesMacroeconomic Dynamics.

    World Bank2002Global Development Finance (Washington: World Bank).

For a review of the literature on the macro origins of banking crisis, see Lindgren, Garcia, and Saal (1996); Dooley and Frankel (2003); and Collyns and Kincaid (2003).

For an overview of existing literature on how crisis resolution policies have been used and the trade-offs involved, see Claessens, Klingebiel, and Laeven (2003); Hoelscher and Quintyn (2003); and Honohan and Laeven (2005).

Bank runs are defined as a monthly percentage decline in deposits in excess of 5 percent. Demand deposits (International Financial Statistics [IFS] line 24) and time, savings, and foreign currency deposits (IFS line 25) are combined for total deposits in national currencies (except for the United Kingdom, Sweden, and Vietnam, where IFS 25L is used for total deposits). Extensive liquidity support is defined as claims from monetary authorities on commercial banks (IFS line 12E) as a ratio of total deposits of at least 5 percent and at least double the ratio compared with the previous year.

In some cases, nonperforming loans are built up slowly and financial sector problems arise gradually rather than suddenly. Japan in the 1990s is a case in point. Although nonperforming loans had been increasing since the early 1990s, they reached crisis proportions only in 1997. Also, initial shocks to the financial sector are often followed by additional shocks, further aggravating the crisis. In such cases, these additional shocks can sometimes be considered as part of the same crisis. For example, Latvia experienced a systemic banking crisis in 1995, which was followed by another stress episode in 1998 related to the Russian Federation’s financial crisis.

Note that estimates of output losses are highly dependent on the method chosen and the time period considered. In particular, this measure tends to overstate output losses when there has been a growth boom before the banking crisis. Also, if the banking crisis reflects unsustainable economic developments, output losses need not be attributed to the banking crisis itself.

Throughout this chapter, t is the starting year of a crisis, and time is measured in years, for example, t – 1 is the year before the start of a crisis.

Whenever general government data were not available, central government data were used.

Regulatory forbearance often continues into the resolution phase, though it is generally viewed as a crisis containment policy.

As a result, output loss estimates are not available for many transition economies that experienced crises in the early 1990s.

The electronic version of the banking crisis database is available at http://www.imf.org/external/pubs/cat/longres.aspx?sk=22345.0. The electronic version also contains a slightly larger set of variables than described here, including a brief description of each crisis; the name of the administering agency of the blanket guarantee (if introduced) and the coverage of the guarantee; and the name of the entity in charge of the AMC (if set up), its funding, and the type of assets transferred to the AMC.

The freeze converted time deposits—except for the first US$500, special accounts such as charitable foundations, and funds meant to be used in tax or salary payments—into dollar-denominated bonds at the exchange rate prevailing on December 28, 1989. The measure was announced on January 1, 1990, after the exchange rate dropped from 1,800 to more than 3,000 australs per dollar between December 28 and 31, 1989.

In 2002, the Argentine government introduced an asymmetric pesification of assets and liabilities of banks. However, the exchange rate used for deposits—1.4 Argentine pesos per US$1—was substantially below market rates. Depositors initiated legal proceedings and some obtained additional compensation through court injunctions.

Measured as the percentage change in the Case-Shiller 20-city composite index between January 2000 and its peak in July 2006.

Mexico is one example in which an implicit blanket guarantee was already in place before the crisis, beginning end-1993. However, the guarantee was reaffirmed at end-1994, during the burst of the Tequila crisis.

    Other Resources Citing This Publication