In all the countries discussed in this paper, urgent measures had to be taken to contain the crisis (Indonesia, Korea, and Thailand) or prevent growing pressures from developing into a full-blown crisis (Malaysia and the Philippines).

In all the countries discussed in this paper, urgent measures had to be taken to contain the crisis (Indonesia, Korea, and Thailand) or prevent growing pressures from developing into a full-blown crisis (Malaysia and the Philippines).

In systemic banking crises, major government intervention is required even in countries strongly committed to market-oriented policies. Such intervention is justified by negative externalities associated with widespread bank failures, such as a breakdown in the payment system, disruptions to credit flows, and depositor losses. Moreover, financial sector soundness facilitates macroeconomic stabilization and creates the conditions for the resumption of growth. In all the crisis countries and Malaysia, the authorities weighed the effects of these externalities against the potential fiscal costs of intervention and the moral hazard problems and decided to implement proactive restructuring strategies.

Strategies for dealing with the financial sector crisis have sought to incorporate good practices from international experience and have the following components: stabilization of the financial system; changes in the institutional framework to deal effectively with the crisis; resolution of nonviable financial institutions; strengthening of viable financial institutions; management of nonperforming assets; and restructuring of the corporate sector (see Box 5 for an overview of bank resolution procedures). While the broad strategies were similar across the crisis countries, each country adapted them to take into account national circumstances and preferences (see Appendices I–V for a detailed description of country–specific circumstances). These components are discussed in detail in Section V.

Bank Resolution Procedures: Terminology and Definitions

A variety of resolution procedures have been employed in the Asian crisis countries. This box defines the terminology used in this paper.

A bank closure is the act whereby a bank is physically closed to the public and, thus, prevented from doing business. A closure can be final or temporary (it may also be partial, involving continued management of existing assets and liabilities). In a legal final closure of an institution, there are several resolution options: the institution can exit the system either through liquidation or through a complete or partial transfer of its assets and liabilities to other institutions, as discussed below. In a temporary closure, the terms suspension and freeze may also be used. The purpose of a temporary closure is to allow time for a more careful evaluation of the institution’s situation, or to allow owners time to present recapitalization and restructuring plans.

Intervention by the authorities in insolvent or nonviable institutions refers to the authorities’ assuming control of a bank, taking over the powers of management and shareholders. An intervened bank usually stays open under the control of the authorities, while its financial condition is better defined and decisions are made on an appropriate resolution strategy. Resolution strategies include liquidation, nationalization, mergers/sales, purchase and assumption operations, and the use of bridge banks.

Liquidation is the legal process whereby the assets of an institution are sold, its liabilities are settled to the extent possible, and its license is withdrawn. A bank liquidation can be voluntary or forced, within or outside general bankruptcy procedures, and with or without court involvement. In a liquidation assets are sold to pay off the creditors in the order prescribed by the law. In a systemic crisis with several institutions to be liquidated simultaneously and quickly, special procedures or special institutions may be needed for the liquidation, as existing structures (e.g., the regular court system) cannot carry out the job in a timely manner.

Nationalization means that the authorities take over an insolvent bank and recapitalize it. It differs from the traditional use of the term “nationalization,” which describes a government takeover of a solvent private bank. Governments in the crisis countries distinguish such temporarily nationalized banks from other state-owned banks and often seek to divest/privatize the nationalized institutions at an early date.

In a merger (or sale) of an institution, all the assets and liabilities of the firm are transferred to another institution. Mergers can be voluntary or government assisted. A key issue is to avoid situations in which a merger of weak banks results in a much larger weak bank, or in which an initially strong bank is substantially weakened.

In a purchase and assumption (P&A) operation, a solvent bank purchases a portion of the assets of a failing bank, including its customer base and goodwill, and assumes all or part of its liabilities. In a publicly supported P&A operation, the government typically will pay the purchasing bank the difference between the value of the assets and liabilities. Variations of P&A operations could be a purchase of assets, entitling the acquiring bank to return certain assets within a specified time period, or a contractual profit/loss-sharing agreement related to some or all the assets. P&A operations in the context of bank resolution can involve the liquidation or transfer of bad assets to an asset management company.

A variation of a P&A operation involves the use of a temporary financial institution—a bridge bank—to receive and manage the good assets of one or several failed institutions. The bridge bank may be allowed to undertake some banking business, such as providing new credit and rolling over existing credits. Bad assets would be liquidated or transferred to an asset management company.

The initial priorities of such a strategy were to stabilize the financial sector and lay out a restructuring strategy. They were complemented by a macroeconomic stabilization plan. Stabilization of the financial sector was accomplished by providing central bank liquidity support and a blanket guarantee on depositors and most creditors.17 To stabilize foreign funding, countries used voluntary debt restructuring where feasible (notably Korea), and capital controls on outflows (Malaysia). To cut the flow of central bank liquidity support, prevent further losses, and demonstrate their commitment to implement necessary reforms, authorities closed institutions judged to be insolvent or nonviable. These measures were the first elements of broader restructuring plans.

Macroeconomic Policies

A credible macroeconomic stabilization program was essential to restore depositor and creditor confidence.18 After the initial shocks—that is, withdrawals of foreign funds and exchange rate depreciation leading to further withdrawals of capital—all countries sought to implement macroeconomic stabilization policies. Monetary policy was used to dampen overshooting of nominal exchange rates and avert depreciation-inflation spirals. Following the initial depreciations, however, uncertainty over the success of stabilization efforts and continued efforts of foreign creditors to cut their exposure in Asia led to continued capital flight, further exchange rate depreciation, and higher interest rates, all of which aggravated problems in the corporate and financial sectors. In Indonesia, confidence was further undermined by policy reversals.

A well-designed, comprehensive, and credible financial sector restructuring strategy was necessary for a sustainable macroeconomic stabilization and resumption of high growth. Progress in structural reforms was also critical for improving domestic and foreign confidence in these economies. Questions arose over which to put first: should the economic programs have focused exclusively on macroeconomic policies, leaving the structural reforms for a later time, or should structural reforms have been made at a slower pace? Several considerations regarding the financial sector argued against delay. First, a banking system saddled with large amounts of nonperforming loans would have maintained an excessively cautious lending policy, which would have caused an even greater credit slowdown and further delayed the restoration of normal credit flows. Second, where banks (and companies) were insolvent, allowing them to continue operating without restructuring would have allowed market distortions and moral hazards to build. Third, bank and corporate restructuring was necessary to facilitate the rollover of maturing foreign loans and new private investment that was crucial to ensuring the necessary financing of the economies; it would have been difficult for such flows to resume if domestic banks and corporations were perceived to remain financially shaky and inadequately supervised. Fourth, keeping insolvent banks (and companies) in operation could have entailed higher fiscal costs and further complicated monetary management.

The crises themselves created a demand for structural reforms. There was a widespread perception domestically and abroad that serious structural flaws in banking and corporate practices had been key determinants of the crisis. Thus, economic programs that failed to address those flaws and practices would likely have been viewed to be incomplete or have only a temporary success. Moreover, there was a momentum and a social pressure for reform. For instance in Korea, labor unions demanded that the chaebols—the large diversified industrial groupings—be reformed not only on economic grounds but also on equity considerations. In particular, the union view that the unavoidable social cost of the crisis be borne also by the owners of the chaebols helped to mobilize government support for the reforms. In Thailand, there was also widespread acceptance of a need for changes to the financial and corporate infrastructure to ensure that excessive risks and vulnerabilities of the kind that had led to the crisis would not be repeated. The sharp reduction in demand put pressure on corporations in all the countries to restructure their business. For some time, the authorities had identified and prepared many of the structural reforms that became part of IMF–supported programs; the crisis brought the pressure for many such reforms to be implemented. For instance, in Korea, a program of financial reform had been prepared but was only implemented when the crisis broke.

It is doubtful that the crisis economies could have been stabilized and confidence returned—even temporarily—without implementing major structural reforms. Malaysia, which did not have an open liquidity crisis, also found it essential to implement far-reaching reforms of its bank and corporate sectors. Without structural reforms, forbearance regarding loss recognition would have allowed inefficient and unsound enterprises and banks to continue operations, leading to growing distortions, discouraging new private investment, and constituting a major burden on economic growth. As a result, the fiscal costs of the restructuring—which already raise medium-term sustainability concerns—would have been even higher. Also, not addressing the key sources of the crisis would have cast a cloud over the success of any program. All these elements suggest that, had the structural reforms been delayed or very weak—for example, the continued financing of nonviable institutions—this would have cast strong doubts on the sustainability of any macroeconomic adjustment.

Liquidity Support

The central banks in all five countries provided liquidity to financial institutions to offset the withdrawal of deposits and credits at some institutions. Many banks were subject to withdrawals both from domestic depositors and creditors, as well as external creditors. Central banks provided liquidity under various emergency lending and lender-of-last-resort facilities. The amounts were especially large in Indonesia and Thailand (Table 3). Most liquidity support was in domestic currency except for Korea, where the Bank of Korea also provided support in foreign currency ($23.3 billion) to commercial banks. In Korea, Malaysia, and Thailand, support was also provided to nonbank financial institutions, such as merchant banks and finance companies.

Table 3.

Liquidity Support Provided to Financial Institutions, June 1997 to June 1999

article image
Source: IMF staff estimates.

Excluding commitments from former shareholders of banks that received liquidity support to make repayments over four years.

The total would be 561 billion baht, if debt-equity conversions were included.

To preserve monetary control these massive amounts of liquidity support had to be sterilized. Sterilization enabled central banks to recycle liquidity from banks gaining deposits to those losing deposits and credit lines. Sterilization had to take place amid underdeveloped money and interbank markets.19 Sterilization was largely effective in Korea and Thailand but not in Indonesia where, for several months, protracted political and macroeconomic uncertainties resulted in continued deposit withdrawals and capital outflows from the system as a whole, making it impossible for the central bank to recycle liquidity. The resultant highly expansionary monetary policy led to a continued flight from the currency and to the collapse of the rupiah. Since July 1998, when overall conditions stabilized, monetary policy exercised through market-based auctions became more effective. In Malaysia, sterilization was partial, because of concerns about the effect of high interest rates on economic activity.

Blanket Guarantees

To stabilize banks’ funding and prevent bank runs, Indonesia, Korea, Malaysia, and Thailand announced full protection for depositors and creditors. In all four countries this blanket guarantee was introduced as soon as the severity of the crisis became apparent. Such a guarantee entails a firm commitment by the government to depositors and most creditors of financial institutions that their claims will be honored.20 A blanket guarantee generally aims at providing confidence in the banking system; stabilizing the institutions’ liability side; buying time while the restructuring work is being organized and carried out; and preserving the integrity of the payment system. Thailand had announced the major elements of the guarantee in July 1997, which was reconfirmed under the IMF-supported program in August 1997. Korea established a full guarantee in November of that year, before negotiations with the IMF had started. In Indonesia, the blanket guarantee was established as part of an IMF-supported program (January 1998), after a limited guarantee had failed to stabilize the situation. Although never faced with a full-blown crisis, the government in Malaysia introduced a blanket guarantee in January 1998. The Philippine authorities, in contrast, have not seen a need for a blanket guarantee. The country had a well-established limited deposit insurance scheme that had been tested in the precrisis period. In none of the countries was any sort of government guarantee extended to entities or shareholders in the nonfinancial sector.

The modalities of the guarantees differed slightly from country to country. In Thailand, the guarantee was preceded by the announcement that the operations of 58 finance companies would be suspended pending acceptable recapitalization proposals and that depositors and some creditors in those companies would be compensated in full or in part, in line with the government’s earlier announcement.21 All depositors and creditors of remaining finance companies and commercial banks were fully guaranteed. In Indonesia, delays in recognizing the systemic nature of the crisis slowed the introduction of the blanket guarantee. Thus, the Indonesian government initially attempted to control the crisis by extending liquidity support to problem banks and instituting a limited deposit insurance scheme.22 However, such limited depositor protection was ineffective, and when a large number of banks experienced runs, making apparent the systemic nature of the problem, the government announced a blanket guarantee for all depositors and creditors. The Korean government announced a full guarantee on all depositors and most creditors of financial institutions. Malaysia offered the blanket guarantee on deposits to all commercial banks, finance companies, and merchant banks, including the overseas branches of domestic banking institutions. In all countries, the guarantees were announced to be temporary and meant to maintain public confidence during the period of restructuring.23 All the countries’ central banks announced that they would provide the necessary liquidity to make it possible to honor the guarantee.

A blanket guarantee must be credible to stop the need for liquidity support and the run on banks. Credibility can be enhanced by stating the terms of the guarantee explicitly, and by confirming the government’s commitment by law and making the guarantee part of a comprehensive restructuring strategy and part of the macroeconomic program. Most countries faced credibility issues initially. In Thailand, where the full guarantee was announced as part of the IMF-supported program, markets did not trust the government’s commitment until the legal status of the guarantee had been strengthened, and the guarantee had been tested following the intervention of some banks at the beginning of 1998. In Korea, the comprehensive IMF-supported program bolstered the credibility of the authorities’ rehabilitation plan, including the blanket guarantee. In Indonesia, a blanket guarantee was introduced as part of a new bank restructuring and macroeconomic program in January 1998. As a result of these measures, the exchange rate stabilized and deposit runs subsided slowly and stopped after the guarantee was tested during the closure of seven banks in April 1998.24

While a blanket guarantee may be a necessary condition to stop bank runs by depositors, it is not a sufficient one. A blanket guarantee—backed by a willingness to provide the necessary liquidity—can restore market confidence in a bank’s ability to pay back deposits and other protected liabilities in local currency. However, if people are fleeing the currency (e.g., because of political uncertainties or economic turmoil), bank runs will continue because a blanket guarantee cannot restore confidence in the currency or prevent capital flight. Also, external credit lines may not be rolled over despite the guarantee, even at higher interest rates.

If the authorities wish to impose losses on depositors and creditors of failing financial institutions, they must do so before the blanket guarantee is extended. In systemic crises, however, drawing clear distinctions between categories of institutions in the initial stages may not be possible due to a lack of information or equity considerations. In Thailand, the authorities inflicted losses on some creditors of the suspended and subsequently closed finance companies.25 The other countries did not take such a measure and covered depositors and creditors of all financial institutions still operating at the time of the announcement. Indonesia even applied the blanket guarantee retroactively to the 16 banks closed in October 1997.

Countries have introduced a variety of measures to limit moral hazard problems. These measures include intensifying the supervision of banks; capping deposit rates at a maximum premium above the average levels being offered by the “best” banks to prevent weak banks from bidding too aggressively for deposits (in Indonesia and Thailand); covering the principal of the deposit only, above a specific threshold amount (and not the interest); explicitly announcing that the blanket guarantee was a temporary measure (Indonesia, Korea, and Thailand); requiring institutions to contribute a guarantee fee (Indonesia, Korea, and Thailand);26 and, in case of insolvent banks, completely writing down current owners’ shares and removing existing management.

Blanket guarantees entail a very large contingent liability for the government. Initially, the guarantee is mainly a confidence booster, but by giving a blanket guarantee, the government acquires a sizable contingent liability against assets of uncertain value—which most often will be insufficient to pay for the contingent liability that the government will be called to honor. Even though the blanket guarantees entail large costs, these may well be lower than the potential economic and social cost of a collapse of the banking system. But since the blanket guarantee protects not just small depositors, it may entail a regressive wealth distribution effect because taxpayers’ funds are also used to protect large depositors and creditors, including external creditors. All these factors suggest that in each situation the costs and benefits of the blanket guarantee have to be weighed carefully. In case of a systemic crisis, however, a blanket guarantee will be necessary, provided that the government can make such a system credible and that the financial sector is deemed to be sufficiently large and of major importance to the economy.

Capital Controls and Debt Rescheduling

To stabilize foreign funding, specific measures had to be taken. The five countries followed different paths to stabilize and reverse the capital outflows. Korea continued to keep its capital account open and renegotiated the country’s short-term foreign debt. In response to declining rollover rates of short-term foreign debt, Korea reached an agreement with foreign banks in January 1998 to reschedule some $22 billion in interbank deposits and short-term loans due in 1998. This marked the beginning of the stabilization of capital flows and of the rapid reduction in central bank liquidity support. Indonesia, the Philippines, and Thailand imposed temporary capital controls measures to fight currency speculation (Box 6). These controls were lifted in Indonesia and Thailand once an IMF-supported program was in place. Malaysia introduced more comprehensive controls over capital flows. Indonesia, Korea, and Thailand all liberalized foreign ownership rules during the crisis to attract additional foreign capital to the financial and corporate sectors.

Emergency Capital Control Measures

Indonesia imposed limits on forward sales of foreign exchange contracts by domestic banks to nonresidents (excluding trade and investment related transactions) in August 1997.

Malaysia attempted to minimize the impact of short-term capital flows on the domestic economy by first restricting (August 1997) and later (September 1998) eliminating the offshore ringgit market. As such, the measures eliminated practically all legal channels for transfer of ringgit abroad; required the repatriation of ringgit held offshore to Malaysia; blocked the repatriation of portfolio capital held by nonresidents in Malaysia for a 12-month period; and imposed tight limits on transfers of capital abroad by residents. In February 1999, the 12-month holding period rule was replaced with a graduated system of exit levy on repatriation of portfolio investments, with the rate of the levy decreasing with the duration of investment.

In July 1997, the Philippines began to require prior approval for the sale of nondeliverable forwards to nonresidents and lowered the limit on residents’ foreign currency purchases from banks for nontrade purposes. The latter limit was further reduced in April 1998.

As soon as the pressure on the exchange rate started to build up (May–June 1997), Thailand took a series of measures to limit baht lending to nonresidents through transactions that could facilitate a buildup of baht positions in the offshore market. Genuine underlying business related to current international transactions, FDI flows, and various portfolio investments were exempt. These measures in reality led to the creation of a two-tier exchange market with separate exchange rates for investors who buy baht in domestic and overseas markets (the spreads between both rates were narrow).

Immediate Closing of Financial Institutions

Closing of insolvent or nearly insolvent financial institutions in the three crisis countries was needed to stem accumulating losses and rapidly growing liquidity support and to give markets a signal that there was a break from the past practice of extensive forbearance (Box 7, see page 21). In Thailand, 58 finance institutions had their operations suspended— 56 of which were later closed (for liquidation) after failing to present acceptable recapitalization plans. The Korean government initially suspended 14 merchant banks, 10 of which were later liquidated. Subsequently, seven more merchant banks would be closed. In Indonesia, the closure of 16 small, deeply insolvent private banks was soon followed by intensified bank runs. The partial nature of guarantees provided to depositors, the perception that other weak institutions remained in the system, a loss of confidence in the government’s overall economic management, and a flight from the currency fueled the runs. Indonesia’s experience, contrasted with that of Korea and Thailand, underscores the fact that bank closures are only successful if all clearly nonviable institutions are closed; the action is part of a comprehensive and credible restructuring strategy; appropriate macroeconomic policies are in place; and the process is clearly and credibly explained to the public (Box 8).

Considerations Regarding the Immediate Closure of Banks in a Systemic Crisis

During a systemic crisis, deciding when to close banks, and which banks, is not easy. The benefits and drawbacks of immediate closure are outlined below.


  • Reduces bank losses and minimizes cost to depositors, creditors, or provider of guarantee.

  • Ceases central bank liquidity support.

  • Allows distribution of losses to shareholders, holders of subordinated debt, and other creditors.

  • Eliminates moral hazard and adverse selection problems associated with insolvent institutions remaining in operation.

  • Removes excess capacity from the financial sector.

  • Helps to restore confidence in other banks and in government strategy, if done quickly and effectively.


  • Restricts access to deposits and disrupts payment system.

  • Severs relationship between lender and borrower, causing a credit contraction.

  • Fuels contagion and runs on other banks, if not executed properly (for example, without blanket guarantee).

  • Leads to further loss of value of bank assets.

  • May entail a base money expansion when deposits are paid out.

For closures to be successful

  • Must be accompanied by a credible blanket guarantee.

  • Owners and subordinated debt-holders must absorb losses.

  • The right set of institutions must be included; markets should be reassured that all nonviable institutions are dealt with.

  • Must include measures to protect payment system and minimize disruptions to the credit market.

  • Must be accompanied by clear and consistent public announcements.

  • Must be part of a broader strategy to make future banking system more efficient.

Indonesia: Closure of 16 Banks

At the outset of the crisis (October 1997), 16 banks (3 percent of assets of banking system) were closed. These banks had been deeply insolvent for several months and had been subject to fraud. Depositor protection was limited to the equivalent of $2,000 (this covered over 90 percent of depositors, but only 20 percent of the deposit base).


The closure of these 16 banks, unlike in Korea and Thailand, did not assist in stemming the general loss of confidence in the government and the banking sector. The main reasons for this failure appear to be:

  • Some politically well-connected banks known to be insolvent were kept open (despite the recommendations of IMF staff).

  • The announcement of the bank closures suggested that more banks would be closed later. This news, in conjunction with the lack of a full guarantee on depositors, led to a flight to safety because depositors expected to incur further losses.

  • One politically well-connected closed bank was allowed to reopen under a new name within weeks, showing ineffectiveness in pursuing the resolution process.

  • Failure on part of the government to implement key reform measures in its IMF-supported program.

  • After an initial fall-off of bank runs, depositors started to withdraw funds in sizable amounts from many banks, with a tendency to redeposit in state banks.

The selection of nonviable institutions to be closed relied largely on liquidity indicators, such as borrowing from the central bank. This was necessary given the typical lack of meaningful information on bank solvency and the realization that insolvent banks can operate as long as they remain liquid. The liquidity triggers typically included the size of central bank credit as a multiple of bank capital.27 Only later, as more information became available either through special audits or the supervisory process, could solvency indicators be used as criteria for choosing nonviable institutions (see Section V). In Korea, solvency data were available from the beginning and insolvency was the criterion for the suspension of merchant banks and for corrective action vis-á-vis the two most distressed commercial banks.

The combined emergency measures reduced the need for central bank liquidity support (see Figure 5). In Korea, these measures, in conjunction with the macrostabilization plan, and the announcement of a restructuring strategy, had an immediate effect on the demand for central bank liquidity support. A similar effect can be observed from the announcement of the blanket guarantee in Malaysia as part of a restructuring plan. In Thailand, the demand for central bank credit from finance companies abated in the aftermath of the suspensions, although several small- and medium-sized banks required support until they were intervened in early 1998. As explained earlier, the impact in Indonesia was different; even after the introduction of the blanket guarantee, banks’ demand for central bank liquidity did not subside until after the closing of another seven banks in April 1998.

Figure 5.
Figure 5.

Selected Asian Countries: Central Bank Liquidity Support

(Monthly change, as a percentage of reserve money in the preceding month)

Source: IMF, World Economic Outlook.

Cited By

Lessons from Asia: Lessons from Asia
  • View in gallery

    Selected Asian Countries: Central Bank Liquidity Support

    (Monthly change, as a percentage of reserve money in the preceding month)