1 Issues in Recent Banking Crises
  • 1 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund


This chapter examines recent experiences with banking crises in seven countries—Argentina, Chile, Malaysia, Philippines, Spain, Thailand, and Uruguay—focusing on the linkages between macroeconomic conditions, financial sector reforms and financial crisis, and the range and effectiveness of measures to deal with financial crises.1

This chapter examines recent experiences with banking crises in seven countries—Argentina, Chile, Malaysia, Philippines, Spain, Thailand, and Uruguay—focusing on the linkages between macroeconomic conditions, financial sector reforms and financial crisis, and the range and effectiveness of measures to deal with financial crises.1

The enormous diversity of experiences with crises, owing to crosscountry differences in macroeconomic conditions, the regulatory framework, the intensity of the crisis, and the approaches used to deal with it, makes it difficult to generalize and develop stylized descriptions or standardized prescriptions. Nevertheless, certain common features stand out from these experiences; in particular, three key lessons emerge:

  • Although macroeconomic instability can weaken the portfolio of financial institutions, weaknesses in the financial sector can have feedback effects on the economy and seriously complicate adjustment and growth policies. Therefore, correcting of the regulatory framework, central bank operating procedures, and portfolio quality in the financial system are important to ensure the effectiveness of adjustment policies.

  • Sound prudential policies and their proper enforcement are critical for minimizing major disruptions to growth and stability.

  • Measures to recapitalize banks and deal with problem loans and enterprises should be designed to preserve monetary policy independence, promote effective loan recovery and industrial restructuring, and minimize moral hazard. The detailed institutional arrangements that meet these criteria—such as whether to set up a separate agency for asset recovery—will vary according to factors specific to the situation. These factors include the legal environment, magnitude of loan losses, and public policy considerations that determine the distribution of losses among various constituencies. However, the chosen arrangement should preserve transparency and avoid shifting the losses to the central bank.

Section I discusses various definitions and models of banking crises, highlighting some key hypotheses with which to interpret the crisis episodes. Section II provides an overview of the case studies by considering the relationship between banking crises and macroeconomic conditions, the propagation of those crises, and the linkages between deregulation and the crisis. The effects of the crises on money demand and credit developments are then examined, and measures used to deal with the crises are analyzed. Section III contains concluding remarks, focusing on lessons for structural reforms and stabilization policies.

I. Definitions and Models of Crisis Situations

This section defines the terms “banking crisis” and “banking distress” and highlights the macroeconomic consequences of such situations. Some models of crises are surveyed to serve as background for the empirical analysis that follows.

Definitions of Financial Crisis

The terms financial crisis and banking crisis are used here interchangeably, an acceptable usage for countries where the banking system dominates financial intermediation.

The definitions of financial crisis in the literature vary with the specific manifestations of the crisis being studied. Samples of definitions are:

  • A demand for reserve money so intense that the demand could not be satisfied for all parties simultaneously in the short run (Schwartz (1985); Miron (1986); Wolfson (1986)).

  • A liquidation of credits that have been built up in a boom (Veblen (1904); (Mitchell (1941)).

  • A condition in which borrowers who in other situations were able to borrow without difficulty become unable to borrow on any terms—a credit crunch or a credit market collapse (Guttentag and Herring (1984); Manikow (1986)).

  • A forced sale of assets because liability structures are out of line with market-determined asset values, causing further decline in asset values—the bursting of a price “bubble,” for example (Fisher (1933); Flood and Garber(1981); Minsky (1982)).

  • A sharp reduction in the value of banks’ assets, resulting in the apparent or real insolvency of many banks and accompanied by some bank collapses and possibly some runs (Federal Reserve Bank of San Francisco (1985)).

All of the elements emphasized in these five definitions could be present in a financial crisis and some may be more important than others in a given episode.

For the purposes of this chapter, financial crisis is defined as a situation in which a significant group of financial institutions have liabilities exceeding the market value of their assets, leading to runs and other portfolio shifts, collapse of some financial firms, and government intervention. Thus the term crisis refers to a situation in which an increase in the share of nonperforming loans, an increase in losses (because of foreign exchange exposure, interest rate mismatch, contingent liabilities, etc.), and a decrease in the value of investments cause generalized solvency problems in a financial system and lead to liquidation, mergers, or restructuring. These events usually follow a shock to the economy, and reinforce the subsequent declines in output (or slowing of economic growth) and balance of payments problems.

Banking Distress Versus Banking Crisis: Some Real Consequences

Runs on individual banks—which depend critically on the confidence of their creditors and typically have large gearing ratios—can destabilize banking systems. Destabilization may result either because major macroeconomic or sectoral shocks affect depositors’ confidence in a wide range of banks or because the payment difficulties in one bank spread through the system, reflecting the financial interdependence among banks. A bank’s failure may jeopardize borrowers’ ability to service their loans with other banks, and the lack of adequate information on the relative soundness of various banks may lead bank creditors to lose confidence in the banking system as a whole when an individual bank fails.

The latter contagion effect is a peculiar feature that makes the banking business very different from other businesses. Such contagion is due to (1) the difficulty in knowing the market value of bank loans; (2) the fact that reductions in the value of bank assets do not simultaneously reduce bank liabilities; (3) the fact that depositors are paid back on a first-come-first-served basis; and (4) the relatively low cost (forgone interest) of withdrawing bank deposits compared with the probable cost of losing the capital value of the deposit. Measures to deal with bank runs have addressed one or more of these characteristics (see Section II).

A banking crisis has immediate economic effects. It disturbs normal credit relationships and raises the cost of credit intermediation.2 It induces a flight to quality by both banks and their creditors; it weakens monetary and budgetary control. For example, the need to support or recapitalize weak banks, and the possible instability in the demand for money—and hence in the price level and economic activity—complicate the task of regulating monetary growth and stabilizing the economy. Propagation of the crisis can be avoided by appropriate lender-of-last-resort intervention and, if the crisis involves only a few banks, by providing information about the true conditions of banks (in order to isolate the unsound from the sound banks).3 However, the crisis itself generates large and pervasive uncertainty (subjective as it may be), which lowers the perceived return on real assets, and depresses real investment and growth. Some economists have argued that a “crisis” has real effects only insofar as it affects the growth of money stock—either reducing it as in the 1930s, or raising it as in later experiences. However, the weight of evidence seems to suggest that the direct adverse effects of a financial crisis on credit markets, balance of payments, and real economic activity can be substantial, and these are in addition to the impact through changes in interest rates and monetary expansion.4

The advent of deposit insurance or implicit guarantees has allowed insolvent financial institutions to stay in business so long as their liquidity position remains manageable (sometimes thanks to central bank assistance). This situation, in which banks are insolvent but not illiquid, is properly called banking distress rather than banking crisis.

A distress situation has effects that resemble those of a crisis, but in a less acute form. Continuing distress perpetuates the resource misallocations that contributed to the distress; prolongs resource allocations that have become inappropriate following a change in the macro-environment; provides incentives for further risk taking, which could aggravate the ongoing losses of financial institutions; and raises the probability of a widespread banking crisis in the event of a major shock—such as a sharp change in relative prices, or a switch in policy regimes as with deregulation. As a result, stabilization and liberalization policies can be effective only if supported by structural reforms that reduce the distress and restore and preserve the soundness and stability of financial intermediaries. In the presence of widespread financial distress, however, the design of stabilization-cum-structural reform policies raises complex issues for the speed and sequencing of various structural reforms. These complexities and the cost of adjustment can increase if an actual financial crisis ensues.

Some Models of Financial Crises

In the last few years there has been a great resurgence of academic and popular interest in financial crises. Although a comprehensive review of the literature is beyond the scope of this chapter, it is useful to outline the major analytical approaches that can help in studying the crisis episodes discussed in this chapter. The theories differ regarding (1) the source of financial distress in the nonfinancial sector; (2) the factors affecting the demand and supply of credit; and (3) the specific manifestations of the crisis itself.

The business cycle approach,5 as modified and extended by present-day writers, holds that the financial environment responds endogenously to the state of the business cycle or to some “displacement” that opens up opportunities for profit. For example, such a displacement could be liberalization of the financial sector. The key hypothesis is that financial fragility—defined as vulnerability to economic shocks—increases over the course of the business cycle expansion, and in response to some displacement.

An economy’s financial fragility stems from factors such as the liquidity of the economy, the proportion of firms that need to borrow in order to honor outstanding debt obligations, debt/equity ratios, and the share of short-term debt in total debt. Financial fragility might also increase because of rising interest rates and overly optimistic expectations that prevail during an investment boom. This does not mean that financial crises occur at the peak of the business cycle as some writers have argued, but rather that crises result from systemic forces that develop near the peak. In particular, interest rates rise partly because of the increase in the interest-inelastic component of the demand for credit. Although velocity of credit rises initially, lenders start to restrict lending as cash flow problems of firms accumulate and nonperforming loans build up. Banks tighten their lending policies, but also attempt to meet loan demands from their prime customers by decreasing the growth of nonloan investments in relation to loans, and, if necessary, reducing excess reserves, and increasing the recourse to money markets. These developments make the economy more vulnerable to a crisis by reducing the capacity of the financial system to withstand a shock. A surprise event—either a new macro or institutional development or an unexpected bankruptcy (or the threat of one) disturbs these financing patterns, initiating a crisis (defined as a sudden, intense demand for reserve money) which is reflected in investor anxiety, bank runs, and portfolio shifts. Usually the central bank resolves the immediate crisis by acting as a lender-of-last-resort.

Credit market conditions—demand and supply—have received considerable attention among crisis theorists. Some argue that the demand for credit is interest inelastic or that the supply of credit is perversely elastic in certain situations of strong excess demand for credit, such as the peak of the business cycle. Wojinlower (1980) argues that at this peak credit is determined by the availability of funds and credit rationing, since demand is essentially insatiable at any conceivable rate of interest. In such an environment an interruption of the supply of credit triggers a business cycle downturn.

Some economists view credit rationing and credit market collapse as equilibrium phenomena that reflect market failures of various sorts.6 At certain levels of interest rates, or at certain levels of default risk, a rise in interest rates may not equilibrate the supply and demand for credit. Increases in interest rates will simply reduce the expected profits from lending—assuming additional deposits could be mobilized with higher interest rates—insofar as only higher risk borrowers are willing to borrow (“adverse selection” because of a worsening of the mix of applicants), and borrowers are induced to take on higher risks (“moral hazard”).7 In this framework the default risk is viewed as a positive function of interest rates, debt/equity ratio of borrowers, and the degree of uncertainty in the system. Because raising interest rates beyond a certain point is futile, there will be credit rationing, which could trigger the collapse of some nonfinancial firms. Thus, in the presence of major shocks that raise uncertainty and default risk, a bank has to decide what to do when it feels overexposed to a borrower whose solvency has become questionable. A solution would be to refuse all forms of credit to such a client, including the rollover of existing loans. However, the bank may decide to continue to provide finance—increasing its exposure—if it feels that the borrower’s situation is likely to improve in the future because of, for instance, anticipated government intervention (e.g., subsidies or a debt bailout).8 Market failure can also arise if, for many economically sound projects, the demand for and supply of funds do not meet at any interest rate-risk combination because of an exaggerated evaluation of risk on the part of the lenders.

To summarize, the foregoing examples indicate market failures may arise from moral hazard that is costly to monitor privately, adverse effects on the mix of borrowers when interest rates go up, or the inadequacy of lenders’ perception of default risk. These market failure theories provide a justification for government intervention in the credit market in the form of tighter banking supervision, loan guarantees, lender-of-last-resort facilities, and direct credit to certain segments of borrowers. 9

Whereas the above theories focus on the behavior of credit markets, the monetary approach emphasizes the central role of the growth of money stock and its variability in causing crises (Friedman and Schwartz (1963); Brunner and Meltzer (1988)). In this framework, a financial crisis need not occur at any particular stage of the business cycle but could develop whenever the central bank’s control of the money supply or reserve money is erratic and results in excessive monetary tightening. Banks are suddenly forced to sell assets in order to obtain needed reserves. This forced sale of assets reduces their price, raises interest rates, threatens bank solvency, and reduces confidence. Banking and debt crises are thus regarded as endogenous events conditioned by economic policy and the banking structure, and not as separate and independent exogenous shocks. In the absence of offsetting action by the central bank, debt and banking crises lead to an excess demand for money, which is an integral part of the monetary policy transmission mechanism.10 Although bank failures may arise because of exogenous factors such as poor credit decisions, the fall in the money stock propagates and deepens the crisis. Moreover, banking crises are regarded as important mainly because of their effects on money growth.

II. Crisis and Adjustment—An Overview of Case Studies

The analysis of sample countries considers the following questions that are suggested by the theoretical discussion above. (Appendix I summarizes in tabular form financial reforms and the financial crises in the seven countries.)

  1. Did the crises mainly reflect major macroeconomic shocks and macroeconomic instability? Were the banking problems exogenous events that aggravated the effects of macroeconomic shocks?

  2. What was the contribution of factors specific to the financial sector (such as financial reform and changes in prudential regulation) in mitigating, aggravating, or causing the financial crises? Did financial reform increase financial fragility?

  3. How did the crises alter the behavior of monetary and credit aggregates? What was the contribution of monetary policy in alleviating or aggravating the crises?

  4. How did the authorities respond to the crises? What were the key support operations and regulatory adaptations? What principles should govern the design of such support operations?

In most of the sample countries, the banking crisis occurred after a period of economic expansion and was associated with balance of payments problems and substantial changes in relative prices and, in some cases, major political uncertainties. In none of these cases could a monetary contraction have caused the financial crisis, although credit market conditions did play a role in propagating it. Interest rate deregulation and other regulatory reforms took place long before the crisis in some cases and contemporaneously in others. Weakness in bank supervision was a common factor in most cases, even in countries where prudential regulations appeared to be comprehensive. Measures to deal with the crises varied greatly, reflecting a complex set of objectives and institutional constraints.

Banking Crises and Macroeconomic Conditions

Tables 1 and 2 present selected economic indicators for the countries in the sample. In most countries, the banking crisis occurred after a period of rapid economic growth characterized by substantial variations in the relative performance of economic sectors. These variations, in turn, reflected major fluctuations in relative price and general business conditions.11 The period of the crisis itself was associated with strong reductions in real output in many countries and a sharp deceleration of output growth in others (Table 1). In all cases real investment ratios fell, although to varying degrees. The crisis periods were also marked by major external shocks, balance of payments difficulties, and sharp adjustments in exchange rates and interest rates, although the balance of payments crisis occurred before overt manifestations of the banking crisis in some cases and afterward in others. Owing to the consequences of import compression and other adjustments that typically accompanied the balance of payments problems, it is hard to separate the contribution of the banking sector problems to the severity of the recessions. Nevertheless, evidence from some countries points to the possibility that the credit market disturbances unleashed by the banking crises served to depress output to levels below what might have been expected from the relative price changes and real investment performance.12

Table 1.

Macroeconomic Conditions Before and During Banking Crises in Sample Countries

article image
Source: International Monetary Fund, International Financial Statistics, various issues, and IMF staff calculations.

Years in parentheses refer to periods of major bank liquidations, interventions, and restructuring.

Average of the three-year period before the year when the crisis started.

A reduction in the index means a depreciation of the exchange rate of the domestic currency.

There was a bill market crisis in 1981, with repercussions for the banking system.

Although the banking crisis began in 1978, the peak years of the crisis were 1982 and 1983. The figures in parentheses are the values when 1982–83 is considered the crisis period.

In most countries external imbalances were fairly severe just before the crisis, which in some cases aggravated them. Misallocation of foreign exchange resources underlay the external debt-servicing difficulties of many countries and was often reflected in large foreign exchange exposures and low quality of assets of the banking system, thereby creating financial fragility and vulnerability to crisis. When the crisis occurred, the growth in central bank credit needed to contain the propagation of the crisis worsened the balance of payments problems. The large devaluations associated with these problems deepened the banking crisis by impairing the debt-service capabilities of debtors with dollar-denominated loans, and by magnifying the losses of banks with large foreign exchange exposures.

Price behavior varied in most crisis episodes. Inflation decelerated—sometimes fairly sharply—particularly at the onset of the crisis (Table 2). In many cases this change was soon reversed, in part because of the expansionary effect of measures to deal with the crisis. Movements in key asset prices were important elements in some crises. A fall in property and share values played a role in the Malaysian crisis, owing to the concentration of the portfolios of many institutions in real estate and shares. A boom and a subsequent fall in land prices were partly responsible for the Uruguayan crisis. In other cases the banking crisis itself and the associated uncertainties seemed to have contributed to a decline in the value of enterprises, making it more difficult to gain access to equity markets just when lenders became more cautious.

Table 2.

Macroeconomic Indicators in Sample Countries, 1974—86

(in percent)

article image
Source: International Monetary Fund, International Financial Statistics, various issues.

Propagation of Financial Crisis: Bank Runs and Contagion

The buildup of depositors’ anxiety following the failure of a prominent firm or a financial institution, and the subsequent runs that concentrated in institutions regarded as weak or somehow linked to the initial failure, contributed to the propagation of financial crises, as did the difficulties of weak banks in accessing interbank markets. In some countries the lack of deposit insurance (e.g., the abandonment of full deposit insurance in November 1979 in Argentina) made deposits riskier and helped to propagate the crisis. In the Philippines insufficient resources of the deposit insurance agency and the resulting delays in settling depositors’ claims may have contributed to depositors’ anxiety and runs.13 In all cases confidence was restored by the emergency measures taken by the central bank. These measures encompassed lender-of-last-resort facilities, intervention in troubled financial institutions (including some that were initially outside the central bank’s jurisdiction), and the re-establishment of deposit insurance (sometimes retroactively).

In some cases the financial crisis spread across country borders. The international propagation of financial crisis could take place either through capital inflows or outflows affecting one country as a result of developments in others or through direct or indirect connections of distressed financial institutions or their customers. A case in point is the demise in March 1981 of the preannounced devaluation scheme for the Argentine peso. This event led to a loss of confidence in Uruguay’s similar scheme, causing a massive outflow of capital in that country, and contributing to the Uruguayan crisis. Also, the failure of two banks in Uruguay, Banco Pan de Azúcar and Banco de Italia y Rio de la Plata, was largely the result of the failure of their parent institutions abroad (a Chilean and an Argentine bank, respectively).

Banking Crises and Financial Sector Reform

An often debated issue is whether financial sector reforms helped to trigger or aggravate financial crises. The timing and intensity of the crises and the timing and scope of financial sector reforms—interest rate deregulation and the liberalization of entry and portfolio regulations—varied considerably among the sample countries (Appendix I). In Argentina, Uruguay, and Chile the deregulation of interest rates had been completed and entry and branching restrictions relaxed by the mid-to-late 1970s, but the financial crises were concentrated in the early 1980s. In Spain a gradual deregulation of interest rates was begun in 1974, starting with the freeing of interest rates on long-term loans and deposits and ending in 1987, when rates on demand deposits and savings deposits of less than six months were freed.14 By the mid-1970s branching regulations and activity restrictions on various classes of banks had been liberalized. Various portfolio regulations—in the form of mandatory investment coefficients—were reduced beginning in 1974. The banking crisis in Spain was a protracted affair that began in 1977 and intensified during 1982–83 before subsiding. In the Philippines the deregulation of interest rates in 1981 just preceded the first episodes of the financial crisis. In Thailand interest rates remained subject to administrative ceilings, which were adjusted on several occasions. A higher ceiling applied to finance companies, which were the institutions initially affected by the 1983–84 crisis. In Malaysia interest rates were deregulated in 1978, although at various times strong moral suasion was exercised. From October 1985 to February 1987, interest rates on deposits of less than 12 months’ maturity were subject to restrictions.15 This was also a period of recession; authorities faced crises among deposit-taking cooperatives and strengthened a wide range of prudential regulations to contain and manage the risks in the banking system.

These chronologies of financial reforms and crisis episodes should be considered with caution, in part because in many cases the weaknesses of problem banks that surfaced during the crisis had originated well before these weaknesses became obvious. It took some time for the problems to be discovered by supervisory authorities because of the normal tendency of banks in distress to reduce the transparency of their accounts.16

To analyze the linkages between financial sector reform and financial crisis in the sample countries, it is useful to identify some ways in which financial sector reforms might increase the fragility of both financial and nonfinancial firms, thereby setting the stage for a crisis.

First, increased freedom of entry into the financial sector and freedom to bid for funds through interest rates and new instruments could lead to excessive risk taking, if such freedom were not tempered with adequate prudential supervision and regulation. For example, implicit guarantees of a government bailout of depositors (and, to some extent, bankers), together with weak prudential legislation and supervision permitting unsound lending patterns, could trigger excessive risk taking following deregulation. Deregulation could also facilitate a too rapid growth of some financial institutions and allow unqualified persons to enter into financial business. Examples of such effects were found in all the sample countries.

Second, the institutional structure of the banking system that emerged from regulatory changes could lead to concentration of power in banking, and interlocking ownership and lending patterns. Such an environment is particularly vulnerable to market failures—because of moral hazard, adverse selection, and oligopolistic pricing, all reinforced by the regulatory environment. This situation could favor excessive risk taking on the part of banks (particularly following deregulation of interest rates), and on the part of nonfinancial firms. This point is illustrated by the large share of credits to related firms in the private banking system of Chile and in the banking groups that failed in Spain. However, loans to directors and insiders, fraud and mismanagement, and loans to political interests have been a perennial source of risk concentration, banking difficulties, and bank failures, and there is no reason why deregulation per se should increase the incidence of such practices.

Third, deregulation could lead to excessive increases in interest rates if euphoric expectations coupled with unsound liability structures of firms cause a sharp increase in credit demand. With high debt/equity ratios, an initial increase in real interest rates, among other things, could lead to distress borrowing and hence fairly inelastic demand for credit (artificial demand for credit), which would perpetuate the high rates. This chain of events seems to have occurred in the Southern Cone countries.17 High debt/equity ratios of nonfinancial firms developed in the pre-reform period of negative real interest rates and directed credits, leading to distress borrowing when real rates turned strongly positive.18 Distress borrowing aggravated the financial crises in many of the sample countries (e.g., Chile and the Philippines), but its influence in starting the crises is unclear.

Fourth, following the deregulation of interest rates, the authorities might lack an adequate set of instruments of monetary control to influence interest rates or might follow a hands-off policy in the erroneous belief that domestic interest rates would automatically converge to international rates over time.19 Or, the authorities might base targets for monetary and credit aggregates on past behavior, despite massive shifts taking place in income velocity and the money multiplier in response to deregulation and other measures. These factors would cause excessive increases in real interest rates and precipitate a crisis, particularly if the policy error persists.20 Also, tighter credit policies could widen the interest margin and contribute to excessively high real lending rates, insofar as they are implemented by raising unremunerated reserve requirements, or by forcing banks to hold low-yield government securities. Such outcomes are readily corrected through appropriate refinements of instruments and timely reversals of policy stance (if macroeconomic conditions permit such reversals). There is little evidence that inappropriate monetary policies contributed to the crisis in the sample countries, except insofar as the authorities let interest rates become too high in some cases (e.g., Chile and Uruguay). In all cases monetary policy was complicated by the large portfolio shifts in the banking system (see the following subsection for details).

Fifth, following deregulation, instability in the credit markets could arise not only from an inelastic demand for credit, but also from credit rationing. For example, in times of perceived high risk and uncertainty or merely of tight financial policies, real interest rates could rise sharply. However, lenders will see a rise in interest rates beyond a certain point as counterproductive because of the higher risks associated with that increase, and the markets could resort to credit rationing. This inelastic or even perversely elastic supply of credit could result in instability, characterized by bankruptcies of firms and banks. In some cases this credit rationing behavior would show up in persistently high real interest rates, possibly rising interest margins, and a simultaneous fall in loan/deposit ratios of the banking system. There is evidence of these phenomena in some countries, notably in the Philippines.21

Sixth, the supervisory authority might be ill prepared to deal with a financial system that operates with much more freedom than in the past. Both the regulations and the administrative infrastructure might need to be overhauled to focus them on analyzing bank solvency and credit risk rather than on monitoring compliance with control regulations such as interest rate ceilings or selective credit regulations.22 Although supervision was inadequate in some of the sample countries, in others, despite an adequate supervisory apparatus, the enforcement was weak and indecisive because of political interference. In all the countries in the sample poor risk diversification, inadequate loan evaluation, and plain fraud were the main factors leading to financial institutions’ liquidation or intervention. Sometimes these factors were related to the fact that more liberal entry into financial intermediation allowed people with little or no experience in the field to set up or take over a financial institution.

Finally, the deregulation of interest rates could adversely affect financial institutions that have a large exposure to long-term assets funded by short-term liabilities, which carry fixed interest rates. This situation could precipitate a crisis for some segments of the industry, and for the whole system unless appropriate action was taken. However, this factor did not play a significant role in the sample countries.

In sum, the connection, if any, between financial reform and financial crisis derives from an unstable macroeconomic environment, the development of unsound liability structures of nonfinancial firms (before or after the reform), and weaknesses in the institutional structure for banking. Therefore, sound financial policies, vigilant supervision of banks, and well-designed prudential regulations would limit financial crises and help reduce the vulnerability of a financial system to the vagaries of the macro environment.

Banking Crises and Monetary Conditions

Banking crises are often associated with substantial portfolio shifts. The demand for money can rise because crises cause uncertainties and asset liquidations, or fall because savers shift to safer assets such as foreign currency assets, treasury bills, and nonmonetary instruments: the net effect of these two opposing forces is an empirical question. Moreover, different aggregates are likely to be affected differently. Demand will shift in favor of financial liabilities of institutions that are perceived as having no or negligible default risks—such as the central bank and state and foreign banks—and against liabilities of institutions perceived as risky. For instance, sharp—albeit temporary—increases in currency demand occurred in most of the sample countries at the time of bank liquidations. Also, the interest elasticity of the demand for various monetary aggregates might change, often permanently, because of the greater awareness of risks and returns built up in times of crisis. Such portfolio shifts complicate the conduct of monetary policy. The effect of these shifts on monetary aggregates will depend on the policy response. For example, implementation of interest rate controls will diminish the attractiveness of time and savings deposits.

Table 3.

Tests of Shifts in Demand and Interest Elasticities in Five of the Sample Countries

article image
Source: International Monetary Fund, International Financial Statistics.Note: Based on regressions, quarterly changes (1975–87) in the log of currency or M2 on prices, output, and interest rates, using general distributed lags in these variables and lagged values of the dependent variable. Figures in parentheses are at-values.

Direction of change in the absolute value of interest elasticity (sum of all lag coefficients).

Becomes insignificant when shifts in interest elasticities are allowed.

Some portfolio and parameter shifts took place following banking crises in the sample countries. There is evidence of a significant shift into currency or a decline in the interest elasticity of currency demand following the crises in Argentina, Chile, the Philippines, Thailand, and Uruguay (Table 3).23 Also, in several sample countries the demand for M2 showed an upward shift, and interest elasticities increased following the banking crisis. The shift in money demand in some cases reflected a portfolio switch from nonbank institutions and deposit substitutes into bank deposits. For example, in Thailand deposits switched from finance companies to banks following the crisis among the finance companies. A similar phenomenon was observed in the Philippines following the bill market collapse in 1981.

The behavior of the money multiplier and the stability of the monetary base might also be significantly affected by a banking crisis. The money multiplier would fall because of increased demand for currency and a precautionary rise in banks’ excess reserves prompted by greater volatility in deposits and increased riskiness in lending.24 Although further work is needed on this topic, preliminary examination of data suggests that multipliers (for M2) showed significant changes during crisis periods (Chart 1). For example, in Spain, despite the progressive reductions in statutory cash reserve ratios, the multiplier declined fairly sharply between 1977 and 1979 before resuming its upward trend. Sharp reductions in multipliers occurred in Uruguay and the Philippines following their crises, and a mild decline and greater volatility were evident in Argentina (where a large drop occurred in 1982). These declines are partly attributable to changes in reserve requirements. However, even after allowing for such changes, a notable fall or a pause in the growth of the multiplier could be detected in many of the sample countries. In the Philippines the sharp increase in currency demand contributed to the decline in the multiplier.

Quarterly growth rates for reserve money showed no significant changes in their level, variability, or seasonality following the crises in most countries, although a brief acceleration in the growth of reserve money or a breakdown of normal seasonal patterns was evident during the crises in some countries (Chart 2).

An important monetary development in most countries, at least for brief periods, was the change in the sources of growth in reserve money. In several cases, central bank credit to banks and other financial institutions rose sharply as the crises unfolded, and its share in reserve money rose (Chart 3). This increase of central bank credit typically served a dual purpose: to assist ailing institutions or borrowers, and to offset the contractionary effect of reductions in the money multiplier. In some cases this growth in central bank credit jeopardized the attainment of monetary and balance of payments targets, forcing the authorities to use other instruments of monetary control to absorb excess reserve money. Although in some countries the initial spurt in central bank credit was reversed within a short period, in others the central bank credit’s share in reserve money increased permanently (Table 4).

Banking Crises and Credit Markets

This subsection discusses the relationship between the banking crises and credit markets by focusing on three key questions: (a) What was the role of credit market conditions in initiating or aggravating the crisis? (b) Did the causality run from credit to the real sector or in the opposite direction? (c) What were the effects of financial crises on credit demand?

Chart 1.
Chart 1.
Chart 1.

Banking Crises: Money Multiplier

Note: Shaded areas indicate crisis periods.
Chart 2.
Chart 2.
Chart 2.

Banking Crises: Rate of Growth of Reserve Money

(In percent)

Note: Shaded areas indicate crisis periods.
Chart 3.
Chart 3.
Chart 3.

Banking Crises: Central Bank Credit to Nongovernment Agents

(In real terms)

Note: Shaded areas indicate crisis periods.
Table 4.

Total Central Bank Credit as a Percentage of Reserve Money in Six of the Sample Countries, 1974–86

article image
Source: International Monetary Fund, International Financial Statistics, various issues.Note: Although Chart 3 shows central bank credit to nongovernment agents, this table focuses on total central bank credit. In many countries the central bank supported the problem institutions indirectly, with the budget providing the actual support but financing it through central bank credit to the government. This was a particularly significant factor in the Philippines.

The Role of Credit Market Conditions

The role of credit market conditions in crises differed across the sample countries. The extent to which caution on the part of bankers, and the resulting credit rationing, accelerated business bankruptcies in the sample countries is not fully clear, although there is some evidence to this effect. In all cases nonperforming loans rose sharply just prior to and during the crises. The effect of high real lending rates on the financial conditions of already highly leveraged borrowers, and the perpetuation of such high rates because of the resulting distress demand for credit and increased riskiness of lending, was an important factor in Argentina, Chile, the Philippines, and Uruguay. High nominal lending rates in relation to growth of credit—a result of tight monetary policy to attain adjustment—aggravated the crises in both the Philippines and Thailand.25 The behavior of lending rates had no significant effect on the crisis in Spain, in part reflecting the gradual pace of liberalization there. In the Philippines bank credit to the private sector declined precipitously in nominal terms between 1983 and 1986 (by more than 50 percent in real terms), while nominal bank lending rates averaged 23 percent. Real lending and deposit rates were significantly positive during most of the 1983–86 period, which was marked by asset liquidations and a fall in output. The decline in credit to the private sector in the Philippines could not be attributed to demand factors alone or to simple crowding out by the Government, but also reflected credit rationing by bankers, owing to the increased riskiness in lending.26

Comparisons of credit aggregates before and after the crises emerged show that, with the exception of the Philippines and (marginally) Spain, credit continued to grow strongly in real terms after the crises began, despite falls in real output or real growth during the crises (Table 5).27 In some cases the increase in real domestic credit was accompanied by a sharp fall in the country’s total foreign reserves, which suggests that in some countries the central bank ensured the provision of credit to the economy at the cost of losing substantial amounts of reserves.28 The strong growth in credit in some countries also reflected the rollover of nonperforming loans, interest accrual on them, and new work-out loans to try to rescue ailing borrowers. In addition, the increased demand for M2, noted earlier, might have facilitated a continued strong credit expansion following the crisis.

Table 5.

Change in Real Credit Balances of the Sample Countries

(Percentage change)

article image
Source: International Monetary Fund, International Financial Statistics, various issues.Note: For each country the change was calculated using the average levels for the three years immediately before and the three years immediately after the year the crisis began. See Table 1 for the dates of the crises.

Causality Between Credit and GDP in the Short Run

Short-run disruptions in credit flows—a common phenomenon during banking crises—can have a serious effect on short-run economic performance. When real credit and real growth fall, the question arises whether the fall in real credit was supply determined (and thus a possible cause of the fall in growth) or demand determined (and thus a possible result of the fall in growth).

To shed some light on this question, Table 6 shows Granger causality tests to analyze the relationship between real domestic credit and real GDP (quarterly data, seasonally adjusted) for Argentina, Chile, and the Philippines.29 The data correspond to total domestic credit and domestic credit to the private sector. Four lags for each variable were used in the tests.

The results in this table are generally inconclusive. Only in the case of the Philippines is there evidence of statistically significant causality, which is unidirectional for total credit (running from credit to GDP) and bidirectional for private sector credit. These results are somewhat surprising, given the importance of credit in financing economic growth and the substantial variability in both credit and output in the three countries under consideration. It may be concluded that shocks from other variables overshadowed the importance of credit as a determinant of output—and of output as a determinant of credit—in the countries and the periods considered here. This interpretation appears plausible, because the timing of some of these other shocks (e.g., the collapse of the exchange rate regimes in Argentina and Chile) did not coincide with those of shocks to the supply of credit.30

Table 6.

Granger Causality Tests of the Relationship Between Real Domestic Credit and GDP in Three of the Sample Countries

(In percent)

article image
Source: International Monetary Fund, International Financial Statistics, various issues.

Statistically significant at the 5 percent level.

Effects of Financial Crisis on Demand for Credit

Do financial crises affect the demand for credit? As mentioned earlier, some authors have suggested that borrowers become unwilling or unable to pay their loans and are happy to roll over these loans, even at high real rates, because the borrowers expect to be bailed out eventually. A way to test this hypothesis is to estimate a simple model of the demand for credit and see what effect, if any, the crisis had on interest elasticities. Although such analysis is not available for all countries, studies with Argentine data suggest that the demand for loans rises in real terms when the real cost of credit goes up, confirming the hypothesis that the demand for credit in Argentina included many loans that were insensitive to interest rates and on which interest was accrued and capitalized but not paid (i.e., effectively nonperforming loans).31

Dealing with Banking Crises

The measures to deal with failing and ailing institutions during a banking crisis should aim at arresting the propagation of the crisis, restoring depositors’ confidence, and protecting the payments system immediately; and at bringing about an orderly restructuring and recapitalization of problem banks later. Evaluation of any measures should take into account the following factors: monetary and budgetary effects; moral hazard (effects on future risk-taking behavior); distribution of losses among borrowers, depositors, the banking system, and the government; effectiveness of loan recovery; and side effects on solvent borrowers and banks. The measures undertaken in the sample countries varied a good deal, owing to differences in legislative framework and macroeconomic and political constraints. These measures fall into the following categories: (1) emergency measures; (2) measures to deal with failing banks; (3) measures to assist borrowers; and (4) reforms of banking regulations and legislation. Appendix III shows a stylized framework of these measures and the objectives and constraints governing them. Appendix IV outlines the approaches the sample countries used to deal with failing banks and borrowers.

Emergency Measures

In each country emergency measures were aimed at stabilizing the financial system as rapidly as possible. They always included making available lender-of-last-resort facilities, often to replace the loss of deposits of the affected institutions. In many instances normal lender-of-last-resort facilities were not adequate to handle a major crisis, and special credit facilities had to be set up. In some cases the government had to pass emergency legislation to broaden the institutional coverage of the lender of last resort and to put in place other arrangements to on-lend funds through stronger institutions (Philippines) or through special ad hoc funds (Thailand). Central bank intervention in the management of ailing institutions also helped to restore confidence (Argentina, Chile, Malaysia, and Spain). Explicit deposit guarantees, sometimes with retroactive effect, were found useful in Argentina and Chile.

In none of the countries studied could the disclosure of information about individual institutions have helped, because insolvency on a wide scale, rather than isolated instances of failure, was a problem.

The traditional prescription that a central bank should lend freely at a penalty rate in times of panic fully applies to situations of illiquidity arising from sudden surges in the demand for reserves. By itself, the willingness to lend freely is often enough to prevent the propagation of a financial crisis. The penalty rate serves to reduce moral hazard, and in open economies helps to induce capital inflows. However, in the crisis episodes considered here, insolvency, rather than illiquidity, was the main problem. Moreover, it was of such a scale that in most countries initial emergency lending at market or penalty rates soon had to be replaced by longer-term lending at concessional rates. Thus, the central bank or the government ended up actively subsidizing many financial institutions. Therefore, to minimize moral hazard, nonpecuniary penalties—such as replacing management, requiring the surrender of shares, or preventing dividend distribution—were often employed.

Measures to Deal with Failing Banks

The fate of failing banks chiefly depended on each country’s legislative framework (particularly on the range of enforcement actions available in the legislation), the structure of the banking system, the presence or absence of a deposit insurance agency, and the magnitude of each bank’s losses. In Argentina most failed institutions were eventually liquidated, even after the legislation had been amended to allow the Central Bank to arrange for their merger or sale. In Thailand a variety of disposition decisions were employed for finance companies, including liquidations, mergers, restructuring of activities, recapitalization, and support through long-term, soft loans from the central bank. No bank was liquidated, but most troubled banks received subsidies in the form of soft central bank loans, and the government assumed the ownership of one bank.

In Chile, following some initial liquidations of small institutions, troubled banks were recapitalized and assisted through various subsidy schemes. In the Philippines troubled rural and thrift banks, which were numerous and small, were commonly liquidated. For the larger commercial banks, however, the preferred method was takeover by a government-owned financial institution, or acquisition by the government with central bank support and intervention. In Spain, the deposit insurance agency (or a similar agency) typically took control of the problem bank by buying shares at a nominal price, assumed the bad debts, and sold off the clean bank through competitive bidding.32 As an exception, a large banking group was temporarily nationalized, together with other enterprises belonging to the same conglomerate (RUMASA).

The disposition method and the legal environment affected the distribution of losses among bankers, depositors, borrowers, and the government. Owners experienced their greatest losses when the bank was liquidated or they had to surrender ownership for a nominal price, and their smallest losses when they were allowed to hold on to their shares and the bank received subsidies from the central bank or the government. The loss to depositors was minimal in most countries, except where negative real interest rates were reinstated to reduce the debt burden.33 Malaysia implemented a scheme to pay off depositors partly with equity in the institutions being recapitalized, and in one case Uruguay followed the same course.34 In most cases, however, the government or the central bank assumed the bulk of the losses. These losses are difficult to measure, depending inter alia on the chosen method of disposition.35 In countries with deposit insurance or other arrangements requiring contributions and loans from other banks (e.g., Thailand’s Rehabilitation Fund to which all banks contributed), other banks—and indirectly their clients—shared the losses to the extent that the deposit insurance fund or the equivalent agency had insufficient resources to cover the losses.

The methods for recapitalizing banks typically involved various forms of subsidization. Under one approach the central bank (or a separate government agency) purchased bad loans at par, paying with central bank securities (or government bonds) that carried market rates.36 Such purchases entail recovery risks for the central government or the central bank. To minimize these risks, in Chile the selling bank had to buy back the bad loans from the Central Bank according to a schedule stretching over ten years.37 Thus, the Central Bank did not assume the commercial risk or management of these loans, but the approach still carried the risk that the continuation of old linkages between the bank and the debtor could weaken loan recovery. In Uruguay the Central Bank assumed the bad loans, while collection responsibility kept switching between the Central Bank and a state commercial bank, thereby hampering loan recovery. In the Philippines, to facilitate loan recovery, some staff of the commercial banks were seconded to the fiscal agency that assumed the bad loans.

Central banks are ill equipped to assume the administration of bad loans of banks being recapitalized or restructured.38 Borrowers have little incentive for repaying outstanding loans to an institution that cannot make new loans. Moreover, a central bank usually lacks the expertise in loan administration of a commercial bank or a specialized agency. Also, central bank losses on account of bad loans compromise monetary control, in addition to making the losses less transparent.

The case studies clearly point toward certain significant advantages—and some disadvantages—in shifting the bad assets to a separate agency with explicit funding sources. Such a shift, which took place in the Philippines and Spain, had the advantage of not only making the losses transparent, but also breaking off the linkages between problem banks and problem borrowers, linkages that had contributed to the problems in the first place. The resulting restructuring of banks’ balance sheets also facilitated the implementation of monetary policy. Moreover, a separate entity—such as the so-called bad or collecting banks, those found in the United States—could offer greater flexibility and capability to mobilize staff with expertise in loan administration and recovery than a central bank or in some cases even commercial banks.

However, setting up a separate agency also has significant disadvantages. Generally, incentives for repayment by the debtor, and loan recovery efforts by the creditor, are likely to be much higher if the owner of the loan is a functioning commercial bank or a separate specialized agency with its own resources to lend, or if the commercial bank is given incentives to work with the debtor in asset restructuring and recovery. The debtor has greater incentives to restructure and resume normal credit relationships if a lending institution, which offers opportunities for new credits, is involved in loan recovery than if a pure asset recovery agency or a central bank is involved. Moreover, a separate loan recovery agency may lack the depth of knowledge of the debtor’s situation and the loan recovery experience that the originating commercial bank is likely to have. In addition, staffing such an agency may deplete the human resources available to commercial banks in countries where loan recovery skills are in short supply.39

In some cases, rather than purchasing bad loans, the central bank offered a soft loan with which the troubled bank could acquire a government bond or a central bank security at a market-related rate, and thereby start receiving a stream of subsidies. Chile offered cheap government loans to facilitate the purchase of bank stock and tax credits for subscriptions of new issues of bank stock. In Chile and the Philippines, a government agency was charged with underwriting issues of new stock for the intervened banks.

Measures to Assist Borrowers

Measures to assist borrowers included financial support, technical assistance, and debt-equity conversions. Under a typical arrangement, the central bank provided medium-term refinance credit to commercial banks at subsidized rates (or other forms of subsidies) to encourage the banks to consolidate and reschedule their clients’ loans.40 In most cases the rate charged to final borrowers was not concessional. Uruguay enacted special legislation requiring banks to refinance all “insolvent” borrowers, with the terms of refinance depending on the classification of each borrower. A special government commission was to classify each borrower using various solvency and sectoral criteria, and a debt moratorium applied while classification was pending. Loan recovery and administrative costs with this approach have not been satisfactory in Uruguay.

Measures assisting borrowers that had contracted loans denominated in foreign exchange included preferential exchange rates, interest subsidies on foreign exchange swap transactions, and exchange insurance schemes. These measures often resulted in sizable subsidies to borrowers—with corresponding losses to the central bank—because of large devaluations. Malaysia also provided technical assistance to borrowers for restructuring their operations. The Philippines relied extensively on the takeover of ailing nonfinancial companies by government financial institutions (which converted part of the outstanding debt into equity), an action that postponed hard solutions and even led to the insolvency and restructuring of some of the rescuing institutions.

A key issue in formulating an assistance program for borrowers is whether it should be case by case or a blanket program that covers all who request assistance in a given industry or sector. Although care and selectivity can maximize the cost effectiveness of assistance, on occasion practical considerations might require choosing a blanket program. Ideally, debt restructuring should be left to the individual banks and their clients. But in the uncertainty surrounding a major banking crisis, market decisions may be influenced by subjective fears and more than the socially optimal amount of caution; such a situation would justify some temporary government action.

The most powerful but probably also the most costly method to help insolvent borrowers and banks is to reimpose interest rate controls and produce negative real interest rates that transfer wealth from depositors to borrowers. This was the solution adopted in Argentina, and for a limited time in Uruguay. But this solution discourages depositors, and therefore over time results in a shrinking of the banking system in real terms.

Regulatory and Legislative Reforms

The emergence of crises highlighted the weaknesses in legislative and regulatory frameworks and triggered substantive regulatory reforms in most of the sample countries. For example, significant changes in regulations occurred in Thailand: central bank supervision was extended to finance companies in addition to banks, the central bank’s powers of intervention and enforcement options were broadened, and regulations to limit concentration of ownership and portfolio were strengthened. In Chile comprehensive measures to tighten bank supervision were adopted in late 1981 and 1982; major legislative reforms were accomplished beginning in 1986. These changes provided a more precise definition of the limit on loans to a single borrower, taking into account the interlocking ownership of firms; a formal rating system for financial institutions; public disclosure of information on the nature and quality of the assets of financial institutions; tighter capital requirements; a formal deposit insurance for small savers to replace ad hoc guarantees, and full insurance protection of sight deposits, with the restriction that sight deposits exceeding two and a half times a bank’s capital should be invested in central bank and government securities.

In Malaysia formal guidelines on suspension of interest on nonperforming loans and provisions for bad and doubtful debts were more forcefully implemented and industry practices standardized in order to promote consistent and prudent lending policies. Other measures included giving the central bank expanded powers to intervene in all deposit-taking institutions, establishing the Board Audit and Examination Committees to strengthen supervision of bank management,41 and tightening lending limits per borrower. In the Philippines regulations on bill market practices and dealer supervision were strengthened following the 1981 crisis, and prudential accounting standards and bank liquidation procedures began to be strengthened.

III. Concluding Remarks

A key question in the study of financial crisis is what can be done to prevent it, or at least to minimize its consequences. The answer must address three main problems posed by crisis: how to keep the system liquid, how to restore its solvency, and how to keep it solvent. The structural measures to deal with these problems cannot be fully effective in the presence of major macroeconomic instability and relative price distortions. At the same time, however, eliminating the major portfolio weaknesses in the financial sector can significantly reduce the cost of macro-economic adjustment measures and make them more effective.

The problem of maintaining liquidity has been resolved through the lender-of-last-resort function of the central bank. However, given the unstable macroeconomic conditions before and during the banking crises in many sample countries, this solution posed difficulties for the design of financial policies—particularly monetary policy. Specifically, central banks had to balance the objective of preserving monetary stability with the fulfillment of their lender-of-last-resort obligations and, at the same time, had to contend with significant shifts in money demand and the money multiplier in times of crisis.

The restoration of solvency of the system in crisis involved a variety of disposition decisions for problem banks, problem loans, and problem borrowers. Because the widespread losses had to be dealt with in some fashion—to protect the depositors and ensure the sound future functioning of the financial intermediaries that would remain—the key issue was how to apportion the losses among depositors, borrowers, the banking system, and the government. In most cases, the government or the central bank assumed the bulk of the losses, because banks could not be expected to outgrow such large losses with only limited initial support. This approach complicated the design of monetary and fiscal policies. However, to ensure monetary policy independence, promote effective loan recovery and industrial restructuring, and minimize moral hazard, there are advantages in solutions that encompass the following elements: make the losses transparent rather than hide them in the books of the central bank or commercial banks; work out institutional arrangements that maximize professionalism in loan recovery, asset liquidations, and asset restructuring; and ensure that original shareholders absorb losses to the maximum extent possible under the law and that the old management is replaced in the problem financial institutions by new management.

After the initial recapitalization and restructuring of weak institutions, complex issues arose in the development of the regulatory and supervisory framework to preserve solvency of the financial system. First, the effectiveness of banking supervision was influenced by the uncertainties in the quality of the loan portfolio of financial institutions. Sharp changes in general business conditions (relative prices, policy regimes, etc.) caused cash-flow problems to businesses and raised loan defaults in some of the crises discussed here (Argentina, Chile, the Philippines, and Uruguay). Even loans that were sound when they were granted became bad loans; thus, major macroeconomic instability and sharp changes in relative prices constrained the effectiveness of vigilant bank supervision and made the identification and treatment of problem loans more complex. Nevertheless, in all the sample countries, factors specific to some institutions (e.g., weak management, fraud, risk concentration) also contributed to the poor quality of the assets of the financial system42 and highlighted the importance of tighter prudential regulations and enhanced bank supervision.

Second, reforms of bank supervision and regulation raised difficult issues in defining the appropriate roles for the government and the market in maintaining the stability and soundness of the financial system. One area of debate has been the need for and form of deposit insurance. Some authors have argued that stricter supervision entails excessive government intervention in private business; others consider stricter supervision ineffective. These concerns have been reinforced by the moral hazard problem that most deposit insurance schemes pose. Reflecting these concerns, the Chicago plan for monetary reform—which originated in the 1930s and was espoused subsequently by Friedman (1959)—suggested splitting banks into two types of institutions, one allowed to receive demand deposits but subject to a 100 percent reserve requirement, and the other totally free from government regulation that would be able to receive any type of deposit from the public except demand deposits and invest in loans or any other assets. The authors of this plan argued that it would ensure that demand deposits (and hence the narrow money definition they were interested in) would not be subject to the contractions that take place during financial crises. Depositors in the second type of institution would bear the risk of the investments because the government would not bail out these institutions or their depositors.

A similar idea lies behind proposals for financial reform that have been triggered by financial crises in the eighties. For instance, Fernández (1983) advocated changing the nature of banks so that they would no longer undertake to guarantee a certain yield to their depositors. Also, the recent reform of the banking legislation in Chile distinguishes demand deposits from other deposits. The former have a state guarantee, and banks must invest demand deposits in securities issued by the government or the Central Bank. Other deposits have no guarantee and no restriction over their investment. In the event of insolvency, a bank may enter into agreements with the holders of nonguaranteed deposits to settle claims on the bank; these depositors are treated like any other creditors of the bank. Reforms along these lines aim to preserve the integrity of the payments system by severely restricting the assets in which banks can invest demand deposits, but other types of deposits bear the full market risk.

Some writers have suggested ways to reduce moral hazard problems and minimize losses to deposit insurance schemes without reforming the financial system. For example, banks could be required to value their assets at market prices; large depositors could share in the losses of a liquidation or reorganization mandated by the regulator; banks could be required to partly fund themselves by issuing subordinated debt;43 deposit insurance premiums could be differentiated on the basis of banks’ riskiness; and more information could be provided to the market on the financial condition of individual banks.44 These proposals reflect the view that with enough information the market will be able to assess, largely on its own, the relative soundness of the financial system.45 Moreover, market reactions would help the regulator to detect problems in financial institutions and act before their net worth becomes zero, thus avoiding losses to the deposit insurer. Unfortunately, there is little evidence so far regarding the market’s ability to judge a bank’s soundness.46

To summarize, prompt support from the monetary authorities and the government during a financial crisis can prevent problem banks from causing major and lengthy disruptions to the payments system and the economy at large. However, effective restructuring and recapitalization of the problem banks and the banking system may require significant fiscal adjustment, prompt structural reforms of the nonfinancial sector, and progress toward macroeconomic stability. Such stabilization and structural policies should be supported by adequate supervision and prudential regulations, which are also necessary to make financial crises less likely and less costly. But regulatory authorities must balance their concerns for the financial system’s safety (which requires appropriate regulation and supervision) against the need to maximize its efficiency (which requires that market forces play the main role in shaping the structure and operations of the system). Discussion on ways of attaining this balance is still wide open.


Financial Reforms and Financial Crises in the Seven Case Studies: An Overview

article image
article image
article image
article image
article image
article image
article image
article image
article image
article image
article image
article image

APPENDIX II Demand for Credit in Argentina

A simple partial adjustment model was postulated for the private sector’s demand for credit:


where L is the outstanding stock of loans; P is the wholesale price index; y is real CNP; r is the bank nominal lending rate; inf is the inflation rate (measured by the wholesale price index); and v is the speed of adjustment. The basic equation estimated was


In addition, dummy variables were introduced for the crisis period and for the period in which interest rates were subject to controls. The dummy for the crisis period was statistically insignificant, but the dummy for interest rates controls (D) was significant. The data used corresponded to the period beginning in the third quarter of 1975 and ending in the fourth quarter of 1987. The reduced-form equation was estimated using instrumental variables and yielded the following results (t-values shown in parentheses):


DW=1.89 R2=0.976

The signs for the interest rate and inflation coefficients are the opposite of what would normally be expected; the estimated coefficients indicate that demand for loans rises in real terms when the real cost of credit goes up. This finding is consistent with the hypothesis that a large fraction of the demand for credit in Argentina consisted of loans that were insensitive to interest rates and on which interest was accrued and capitalized but not paid—that is, effectively nonperforming loans. Thus, when nominal interest rates went up, the real stock of credit increased because interest accrued at a faster pace on these loans; when inflation increased, the value of these loans was eroded—which would account for the negative sign for the inflation rate. Estimates also reveal that interest rate controls had a negative effect on the real value of this debt, insofar as real interest rates were negative over the period. Finally, the low speed of adjustment (0.084) suggests that the outstanding stock of debt was determined largely by the previous period’s stock—as could be expected when banks make few new loans and simply accrue interest on a large fraction of their portfolios.


A Stylized Framework for Dealing with Banking Crises

article image

A Stylized Framework for Dealing with Banking Crises

article image

APPENDIX IV Measures to Deal with Banks and Borrowers in Distress in the Sample Countries

The authorities took various measures to deal with financial crises in the sample countries. As noted in the text, these measures differed widely from country to country. The following is a list of the measures taken in one or more of the sample countries, classified according to whether they dealt with banks or with bank borrowers.

I. Banks

  • Liquidation

    • Bad loans managed by the central bank or supervision agency (acting as receiver and liquidator)

    • Bad loans assumed by the deposit insurance agency

    • Bad loans taken over by a separate government agency

  • Merger

    • Provision of tax and regulatory incentives

    • Provision of special lines of subsidized credit from the central bank

    • Information service to facilitate mergers

  • Sale

    • Takeover by government

    • Takeover by another bank or government financial institution, sometimes with medium-term credit from the central bank (for acquisition and investment)

    • Sale after bad loans are assumed by the central bank, or deposit insurance agency (purchase and assumption)

  • Recapitalization

    • Cash infusion from the government or deposit insurance agency, the central bank, or a special fund created with subscriptions from financial institutions

    • Capital contribution paid with government securities or central bank securities yielding market rates

    • New stock issues underwritten by a government agency (sometimes the central bank) and supported by fiscal and financial incentives

      • —cheap loans to buy banks’ shares

      • —tax credit based on value of shares purchased

    • Purchase of bad loans at par by a government agency with government funds or government securities

    • Unconditional purchase of bad loans by the central bank which pays with central bank securities at market rate (or with central bank obligations in foreign currency)

    • Purchase of bad loans by the central bank, subject to various conditions (on dividend distribution, repurchases of bad loans, provision of foreign currency loans to the central bank)

    • Soft loans from the central bank to be invested in government or central bank securities

  • Restructuring

    • Change in management, in internal controls, and in operating procedures

    • Authorization of new product or service lines

II. Borrowers

  • Central bank credit to banks at low rates to encourage consolidation and stretching out of business firm debt

  • Government subsidies to banks, to facilitate rescheduling of business debt

  • Credit guarantee schemes

  • Preferential exchange rate for foreign currency debtors, interest subsidies on foreign exchange swap operations, exchange insurance programs

  • Controls on interest rates, to bring about negative real rates for borrowers and depositors (a reversal of previous deregulation)

  • Technical assistance to borrowers

  • Conversion of business debt into equity held by the commercial bank

  • Legislation on debt refinancing and temporary debt moratorium

  • Partial write-off of accrued interest or principal by the central bank (or the government agency) after assuming the problem loan


  • Altman, Edward I., and Arnold W. Sametz, eds., Financial Crises: Institutions and Markets in a Fragile Environment (New York: John Wiley and Sons, 1977).

    • Search Google Scholar
    • Export Citation
  • Argentina,Banco Central de la República Argentina, Memoria Anual, 1980.

  • Baliño, Tomás J.T., “The Argentine Banking Crisis of 1980,” IMF Working Paper No. 87/77, International Monetary Fund, November 17, 1987.

    • Search Google Scholar
    • Export Citation
  • Batcheldor, Roy A., “The Avoidance of Catastrophe: Two Nineteenth-Century Banking Crises,” in Financial Crises in the World Banking System,ed. byForrest Capie and Geoffrey E. Wood (New York: St. Martin’s Press, 1985), pp. 4173.

    • Search Google Scholar
    • Export Citation
  • Benston, George J., and others, Perspectives on Safe and Sound Banking—Past, Present, and Future (Cambridge, Massachusetts: MIT Press, 1986).

    • Search Google Scholar
    • Export Citation
  • Bernanke, Ben, “Nonmonetary Effects of the Financial Crisis in the Propagation of the Great Depression,” NBER Working Paper No. 1054:1-54 (Cambridge, Massachusetts: National Bureau of Economic Research, January 1983).

    • Search Google Scholar
    • Export Citation
  • Blinder, Alan, and Joseph Stiglitz, “Money, Credit Constraints and Economic Activity,” American Economic Review, Papers and Proceedings, Vol.73 (May 1983), pp. 297302.

    • Search Google Scholar
    • Export Citation
  • Bordo, Michael D., “Financial Crises, Banking Crises, Stock Market Crashes and the Money Supply: Some International Evidence, 1870–1933,” in Financial Crises in the World Banking System,ed. byForrest Capie and Geoffrey E. Wood (New York: St. Martin’s Press, 1985), pp. 190239.

    • Search Google Scholar
    • Export Citation
  • Brunner, Karl, and Allen H. Meltzer, “Money and Credit in the Monetary Transmission Process,” American Economic Review, Papers and Proceedings, Vol.78 (May 1988), pp. 44651.

    • Search Google Scholar
    • Export Citation
  • Capie, Forrest and Geoffrey E. Wood, eds., Financial Crises in the World Banking System( New York: St. Martin’s Press, 1985).

  • Cikató, Manfredo, “La Crisis Bancaria en el Uruguay,” in Saneamiento de Bancos,ed. byJuan Carlos Casas (Buenos Aires: El Cronista Comercial, 1989).

    • Search Google Scholar
    • Export Citation
  • de Juan, A., “From Good Bankers to Bad Bankers: Ineffective Supervision and Management Deterioration as Major Elements in Banking Crisis” (unpublished; 1987).

    • Search Google Scholar
    • Export Citation
  • Diamond, Douglas W., and Phillip H. Dybvig, “Bank Runs, Deposit Insurance, and Liquidity,” Journal of Political Economy, Vol.91, No. 3 (1983), pp. 40119.

    • Search Google Scholar
    • Export Citation
  • Dwyer, Gerald P., Jr., and R. Alton Gilbert, “Bank Runs and Private Remedies,” Federal Reserve Bank of St. Louis Review (May/June 1989), pp. 4361).

    • Search Google Scholar
    • Export Citation
  • Federal Reserve Bank of San Francisco,“The Search for Financial Stability: The Past Fifty Years,” proceedings of a conference held June 2325, 1985.

    • Search Google Scholar
    • Export Citation
  • Fernández, Roque B., “La Crisis Financiera Argentina: 1980–1982,” Desarrollo Económico, Vol.23, No. 29 (April-June 1983), p. 55.

  • Fisher, Irving, “The Debt-Deflation Theory of Great Depressions,” Econometrica, Vol.1 (October 1933), pp. 33757.

  • Flood, Robert P., and Peter M. Garber, A Systematic Banking Collapse in a Perfect Foresight World, NBER Working Paper No.691: 141 (Cambridge, Massachusetts: National Bureau of Economic Research, June 1981).

    • Search Google Scholar
    • Export Citation
  • Friedman, B.M., and F. H. Hahn, eds., Handbook of Monetary Economics (Amsterdam: North-Holland, 1990).

  • Friedman, Milton, A Program for Monetary Stability(New York: Fordham University Press, 1959).

  • Friedman, Milton, and Anna J. Schwartz, A Monetary History of the United States,1867–1960 (Princeton, New Jersey: Princeton University Press, 1963).

    • Search Google Scholar
    • Export Citation
  • Guttentag, Jack, and Richard Herring, “Credit Rationing and Financial Disorder,” Journal of Finance, Vol.39 (December 1984), pp. 137.

    • Search Google Scholar
    • Export Citation
  • Guttentag, Jack, “Disclosure Policy and International Banking,” Journal of Banking and Finance, Vol.10 (1986), pp.7597.

  • Kindleberger, Charles P., Manias, Panics, and Crashes: A History of Financial Crises (New York: Basic Books, 1978).

  • Kindleberger, Charles P., “Bank Failures: The 1930s and 1980s” The Search for Financial Stability: The Past Fifty Years, proceedings of the conference sponsored by the Federal Reserve Bank of San Francisco, Asilomar, California, June 1215, 1985.

    • Search Google Scholar
    • Export Citation
  • Kuprianov, Anatoli, and David L. Mengle, “The Future of Deposit Insurance: An Analysis of the Alternatives,” Federal Reserve Bank of Richmond, Economic Review,Vol. 75 (May/June 1989), pp. 315.

    • Search Google Scholar
    • Export Citation
  • Long, Millard, Crisis in the Financial Sector, World Bank,EDI Working Papers, 1988.

  • Manikow, N. Gregory, “The Allocation of Credit and Financial Collapse,” NBER Working Paper No. 1786 (Cambridge, Massachusetts: National Bureau of Economic Research, January 1986), pp. 119

    • Search Google Scholar
    • Export Citation
  • Massad, Carlos, and Roberto Zahler, “Another View of the Latin American Crisis; Domestic Debt,” CEPAL Review, No. 32 (August 1987), pp. 1125.

    • Search Google Scholar
    • Export Citation
  • Minsky, Hyman P., “The Financial-Instability Hypothesis: Capitalist Processes and the Behavior of the Economy,” in Financial Crises,ed. byCharles P.

    • Search Google Scholar
    • Export Citation
  • Kindleberger and Jean-Pierre Laffarge (Cambridge, England: Cambridge University Press, 1982), pp. 1339.

  • Kindleberger and Jean-Pierre Laffarge, “A Theory of Systemic Fragility,” in Financial Crisis: Institutions and Markets in a Fragile Environment,ed. byEdward I. Altman and Arnold W. Sametz (New York: John Wiley and Sons, 1977), pp. 13852.

    • Search Google Scholar
    • Export Citation
  • Miron, J. A., “Financial Panics, the Seasonality of the Nominal Interest Rate, and the Founding of the Fed,” American Economic Review, Vol.76 (March 1986), pp. 12540.

    • Search Google Scholar
    • Export Citation
  • Mitchell, Wesley Clair, Business Cycles and Their Causes (Berkeley, California: University of California Press, 1941).

  • Nascimento, Jean-Claude, “The Crisis in the Financial Sector and the Authorities’ Reaction: The Case of the Philippines,” IMF Working Paper No. 90/26, International Monetary Fund, March 1990.

    • Search Google Scholar
    • Export Citation
  • Randall, Richard E., “Can the Market Evaluate Asset Quality Exposure in Banking?” New England Economic Review, Federal Reserve Bank of Boston (July/August 1989).

    • Search Google Scholar
    • Export Citation
  • Schwartz, Anna J., “Real and Pseudo Financial Crises,” in Financial Crises in the World Banking System,ed. by Forrest Capie and Geoffrey E. Wood (New York: St. Martin’s Press, 1985), pp. 1137.

    • Search Google Scholar
    • Export Citation
  • Sheng, Andrew, “Financial Adjustment in a Period of Disinflation: The Case of Malaysia,” in Visiting Specialists’ Papers, 17th SEANZA Central Banking Course, Sydney, Australia, October/November 1988.

    • Search Google Scholar
    • Export Citation
  • Sinai, Allen, “Credit Crunches—An Analysis of the Postwar Experience,” in Parameters, Policies in the U.S. Economy,ed. by Otto Eckstein (Amsterdam: North-Holland, 1976), pp. 24474.

    • Search Google Scholar
    • Export Citation
  • Stiglitz, J. E., and Andrew Weiss, “Credit Rationing in Markets with Imperfect Information,” American Economic Review, Vol.71 (June 1981), pp. 393410.

    • Search Google Scholar
    • Export Citation
  • Taylor, Lance, and Stephen O’Connell, “A Minsky Crisis,” Quarterly Journal of Economics (1985).

  • Tybout, James, “A Firm-Level Chronicle of Financial Crises in the Southern Cone,” Journal of Development Economics, Vol.24(1986), pp. 371400.

    • Search Google Scholar
    • Export Citation
  • Veblen, Thorstein, The Theory of Business Enterprise (New York: Charles Scribner & Sons, 1904).

  • Velasco, Andrés, “Financial Crises and Balance of Payments Crises—A Simple Model of Southern Cone Experience,” Journal of Development Economics, Vol.17 (1987), pp.26383.

    • Search Google Scholar
    • Export Citation
  • Velasco, Andrés, “Liberalization, Crisis, Intervention: The Chilean Financial System 1975–1985,” IMF Working Paper No. 88/66, International Monetary Fund, July 21, 1988.

    • Search Google Scholar
    • Export Citation
  • Wachtel, Paul, ed., Crisis in the Economic and Financial Structure (Lexington, Massachusetts: D.C. Heath and Co., 1982).

  • Wall, Larry D., “A Plan for Reducing Future Deposit Insurance Losses: Puttable Subordinated Debt,” Economic Review, Federal Reserve Bank of Atlanta (July/ August 1989).

    • Search Google Scholar
    • Export Citation
  • Wojinlower, Albert M., “The Central Role of Credit Crunches in Recent Financial History,” Brookings Papers on Economic Activity, Vol.2 (Brookings Institution: Washington,1980), pp. 277326.

    • Search Google Scholar
    • Export Citation
  • Wojinlower, Albert M., “Private Credit Demand, Supply and Crunches: How Different Are the 1980s?” American Economic Review, Vol.75 (May 1985), pp. 35156.

    • Search Google Scholar
    • Export Citation
  • Wolfson, Martin H., Financial Crises: Understanding the Postwar U.S. Experience (Armonk, New York: M.E. Shoorpe, Inc., 1986).


Except for Malaysia and Spain, the analysis in this chapter relies heavily on the country studies included in this book.


This effect includes, for example, rising interest spreads, collapse of individual credit markets, borrowers forced into nontraditional credit sources at high cost, and untimely recall of loans.


Batcheldor (1985) analyzes the value of providing information to contain a crisis.


For a discussion of the effect of a banking crisis on real economic activity, see Bernanke (1983). Velasco (1987, 1988) discusses the linkages between domestic banking crisis and external debt problems. Misallocation of foreign exchange resources is often reflected in the large foreign exchange exposures and low quality of assets in the banking system. The resulting insolvency and illiquidity of banks weakens monetary control, or induces excessive foreign borrowing, and can precipitate a balance of payments crisis.


This approach is developed in Minsky (1977), Kindleberger (1985), Wojinlower (1980, 1985), Taylor and O’Connell (1985), and Wolfson (1986), by extending and modifying the classical business cycle analysis contained in Veblen (1904) and Mitchell (1941).


In the context of creditor-borrower relationships, moral hazard usually refers to the risk that “the borrower after obtaining the loan may act in such a way as to increase the probability of default in an effort to raise the probability of very high returns.” (Guttentag and Herring (1984), p. 1369.) In the same context, other aspects of moral hazard include the risk that borrowers may not use the cash flow from their projects to service the debt, and the risk that borrowers may pledge their assets to another creditor. The importance of each type of moral hazard will depend on the nature of the loan contract.


Borrowers will tend to engage in riskier behavior as their solvency deteriorates because potential losses arising from this behavior are limited to the enterprise’s net worth but the gains are not.


The role of banking supervision is to develop procedures for both creditors and supervisors to monitor financial fragility and vulnerability to shocks, identify exposure to moral hazards, and formulate standards (e.g., capital adequacy) for evaluating the appropriateness of particular levels of risk exposure.


The supply of money falls, owing to the rise in the currency/deposit ratio as deposits are reduced in failed banks, and the public want to hold more currency relative to deposits. The demand for money also falls because currency is an imperfect substitute for deposits. The fall in the supply of money will exceed the fall in demand unless the central bank increases base money sufficiently.


For example, sharp increases in real lending rates as a result of financial liberalization and the introduction and subsequent abandonment of preannounccd schedules of devaluation all contributed to major variations in general business conditions in Argentina, Chile, and Uruguay.


This point is discussed further below. Bernanke (1983) examines the independent contribution of credit market conditions in causing output declines during the Great Depression.


In practice, the deposit insurance agency in the Philippines seems to have relied heavily on loans from the Central Bank.


The rates on deposits of one year or more had been free since 1977 and those on time deposits of six months to one year were set free in 1981.


Some restrictions still apply to tending rates. The central bank approves the base lending rate (BLR) of banks and finance companies after analyzing their cost of funds. Actual lending rates can deviate from the BLR only by a margin set by the central bank.


For an elaboration of this point, see de juan (1987) and Long (1988).


Sometimes this was coupled with borrowers’ stripping their enterprises of most valuable assets, leaving creditors little on which to foreclose.


Distress borrowing typically leads to strong liquidity problems of individual banks. These problems, in turn, are a factor weakening monetary control.


Following deregulation, domestic interest rates would depend upon both domestic monetary conditions and external factors (foreign interest rates plus exchange rate expectations), and the relative importance of the domestic and external factors would be a function of the extent of openness to capital flows. Even in fairly open economies, monetary policy can have a strong effect on interest rates in the short run, and an automatic and immediate convergence to foreign rates cannot be assumed.


Nominal interest rates could remain high even in times of price stability, owing to temporary uncertainties and stubborn exchange rate expectations. If this is the case even after major corrections to the real exchange rate, nonaccommodating credit policy would soon alter exchange rate expectations and reduce real rates. I lowever, if high real rates persist long enough, they could precipitate a crisis.


Evidence on this topic should be evaluated carefully to distinguish between shifts in the credit supply function and changes in its shape. For instance, a shift in the supply of credit could also lead to most of the phenomena attributed to credit rationing here.


Several authors have studied the decline in income velocity of money during the Great Depression in the United States.


In the case of Argentina, the behavior of excess reserves during the crisis did not seem to deviate significantly from historic patterns.


Also, a high or rising nominal loan rate (as a result of inflation) on adjust able-rate loans implies a faster loan repayment, which caused hardship to borrowers in some countries. In addition, uncertainty tends to reduce the maturity of loans, which means that most financing carries a de facto adjustable rate.


See Nascimento (1990) for further details.


This finding does not preclude the possibility of short-term credit disruptions in the early months of the crises. See the next section for further discussions.


As discussed in the next section, central banks provided large sums of credit to ailing institutions and, in some cases, to final borrowers.


Data availability allowed the analysis of only these three countries.


A recent study concludes that the data do not support the hypothesis of a simple or unidirectional relationship between bank runs and economic performance for the case of the United States (Dwyer and Gilbert (1989)).


See Appendix II for further details.


This process was similar to the purchase and assumption procedure used in the United States.


In Argentina holders of foreign currency deposits in failing banks lost everything. In Thailand, depositors of failed finance companies were to be paid over a 10-year period, without interest.


For Uruguay, see Cikató (1989).


Baliño (1987) gives partial figures for the Argentine Government’s losses.


For some operations the Central Bank of Uruguay issued U.S. dollar claims to purchase bad loans, some of which were denominated in domestic currency.


This condition essentially converted the operation into a rediscount credit against the collateral of bad loans, with the proceeds reinvested in central bank securities. Thus, the objective was to support the profits of the bank for an extended period of time, with a pan of the profits being used to repurchase bad loans. This method, based on a prespecified schedule of repurchases, could pose difficulties for bank operations if the profits were not adequate to cover the repurchases. This consideration was not relevant in the case of Chile, however, because the repurchases were stretched out over a long period.


In the cast of liquidations, however, it is common for the central bank or the deposit insurance agency to assume the bad loans.


In the Philippines, commercial bank staff with knowledge of debtors were used by the specialized agency, with administrative safeguards to ensure professional independence of such staff from the influence of debtors.


In many developing countries, banks are reluctant to provide long-term loans, given the short maturity of deposits and the lack of liquidity facilities.


Every financial institution was required to establish such a committee, chaired by a non-executive director, to evaluate internal and external audit reports and pursue follow-up action to remedy any inadequacies discovered.


The importance of these specific factors is clear from the fact that quite a few financial institutions in all the sample countries emerged from the crises virtually unscathed.


Proponents of this requirement believe that it can reduce moral hazard because holders of subordinated debt would suffer a loss in the event of liquidation (unlike insured depositors) and would not gain from extraordinary profits arising from risky behavior (unlike bank shareholders).


Kuprianov and Mengle (1989) discuss some of these ideas. For detailed proposals on “puttable” subordinated debt, see Wall (1989). Benston and others (1986) present a comprehensive discussion of options to make the hanking system more resilient.


Chile’s recent legislation mandates the supervisory authority to publish indicators of the financial condition of individual institutions as a way of facilitating the market’s assessment.


In a recent article, Randall (1989) casts some doubt on the market’s effectiveness in this regard. He suggests that the stock market and bond-rating agencies were too late in identifying the problems in large bank holding companies in the 1980s, and even then underestimated the seriousness of the problems. Moreover, he argues that it would be impractical to give the market the kind of information required for a proper evaluation of the quality of bank assets. In his view many of the market-oriented proposals discussed here would only make the financial system more vulnerable and would fail to protect the deposit insurer.

Cases and Issues