II Systemic Crises: Causes, Cost, and Resolution
- Marc Quintyn, and David Hoelscher
- Published Date:
- August 2003
The Emergence of a Systemic Crisis
A systemic crisis emerges when problems in one or more banks are serious enough to have a significant adverse impact on the real economy. This impact is most often felt through the payment system, reductions in credit flows, or the destruction of asset values. A systemic crisis often is characterized by runs of creditors, including depositors, from both solvent and insolvent banks, thus threatening the stability of the entire banking system. The run is fueled by fears that the means of payment will be unobtainable at any price, and in a fractional reserve banking system this leads to a scramble for high-powered money and a withdrawal of external credit lines.
Loss of creditor confidence can result from the recognition of significant banking system weaknesses. Such recognition may be triggered by economic or political events. The events in the Asian countries in 1997 are among the more recent examples of such a crisis.5 Creditors can also lose confidence in sound banking systems due to poor macroeconomic policies, external shocks, or even improper sequencing of financial sector reforms. The resulting liquidity crisis, if mismanaged, can result in financial panic and insolvency. International capital market integration in the 1990s has added a new dimension to banking crises. Perceptions by foreign investors of macroeconomic weaknesses, concerns about the future path of economic policies, political turmoil, or fear of contagion from other crises can lead to a quick deterioration in international creditor sentiments (see Box 2 for a more in-depth discussion of the causes of banking crises).
Treatment of a systemic banking crisis contrasts in important ways with the treatment of individual bank failures in stable periods. Policies considered appropriate in stable periods may aggravate uncertainties in a systemic crisis, worsening private sector confidence and slowing recovery. In stable periods, for example, deposits have only limited protection, emergency liquidity assistance is given under very restricted conditions, and undercapitalized or insolvent banks are immediately intervened and resolved. In a systemic crisis, however, events can change rapidly, banking conditions may deteriorate quickly, and information on the true condition of banks tends to be limited and often outdated. In such an environment, policies should be aimed at limiting the loss of depositor confidence, protecting the payment system, restoring solvency to the banking system, and preventing further macroeconomic deterioration.
A successful restructuring requires decisive government action from the onset of the crisis. Determined actions to strengthen macroeconomic policies and implement structural reforms can reduce the magnitude of the crisis. Often, such early determination is lacking because the authorities (and market participants) need time to recognize the severity of the situation, or worse, are in denial. Delays in implementation may prevent or slow the return of both depositor confidence and access to international capital markets, deepening the crisis. Comparisons of recent crises in Argentina (2002), Brazil (1999), Russia (1998), and Turkey (2000) point to the importance of rapid and determined policy adjustments.
The Cost of Banking Crises
Measuring the cost of banking crises is a difficult task.6 One needs to distinguish between the total cost to the society, which is hard to measure, and the fiscal cost. The former concept is discussed in Section VI. Here we deal with the fiscal cost.
The costs of banking crises have varied sharply. Costs to the public sector have ranged from small amounts (close to zero) in Russia and the United States to over 50 percent of GDP in Indonesia (Figure 1 and Appendix I). Three main factors affect the gross cost to the public sector: the initial macroeconomic conditions and the financial sector, the authorities’ policy response (i.e., the specific measures taken), and the degree of success in recovering the value of assets acquired during the crisis. Fiscal costs may not represent a complete loss to the economy, as some part of the expenditure represents a transfer to the domestic private sector. A banking crisis will also result in costs to the economy in terms of macroeconomic instability and foregone growth, which have proven complex to estimate.
Figure 1.Fiscal Costs of Selected Crisis Countries
Sources: IMF, International Financial Statistics, World Economic Outlook, and national authorities.
1 Data on net cost unavailable.
Box 2.Typology of Banking Crises
A banking system crisis usually arises from multiple sources, each interacting with one another to aggravate and accelerate the crisis. For that reason, typologies are artificial and of only limited analytical use. At best, only broad categories can be identified that capture the most fundamental elements of a crisis.
Based on their causes, banking crises can be classified broadly into one of two categories: predominantly microeconomic, or predominantly macroeconomic. Crises arising from microeconomic causes are largely related to poor banking practices. Lax lending practices may fuel asset price bubbles or lead to excessive concentration of banks’ portfolios. Weak risk-control systems have been a major factor in the emergence of a number of crises, leading to a variety of balance sheet deficiencies—large and undetected mismatches (either currency or maturity) on the balance sheets or poor asset quality, leading to large unrealized losses. Banks with balance sheet deficiencies are vulnerable to any shock, and even comparatively minor external events may be sufficient to provoke a loss of confidence and a generalized run.
Crises arising from macroeconomic causes are initiated by developments external to the banking system. Deterioration in the macroeconomic policy environment may cause banking crises in even well-managed banking systems. Well-run banking systems operating in a strong legal and regulatory framework can be overwhelmed by the effects of poor macroeconomic policies. While well-run banks may be able to absorb some macroeconomic shocks, continued unsound monetary, exchange rate, or fiscal policies can produce financial strains that overwhelm banks’ defenses, affecting their solvency or forcing them into risky operations. The way in which these policies might affect banks depends on balance sheet structures. For example, large exposures to the government—often the result of high yields on public bonds—increase bank vulnerability to unsustainable fiscal policies. Dollarization of financial contracts, while contributing to higher intermediation, makes banks vulnerable to unsound exchange rate polices, as even in the absence of currency mismatches banks are exposed to credit risk from unhedged borrowers. Even with overall sound balance sheet structure, banks are vulnerable to unsustainable monetary policies, leading to high and volatile interest rates. It goes without saying that the worst crises are those where macroeconomic shocks affect a weak banking system.
As long as the banks remain liquid, banking distress can persist for a prolonged period until some trigger leads creditors to lose confidence in the banking system, leading to a generalized run. Many kinds of events can trigger a loss of confidence, including emergence of illiquidity in an individual bank (local contagion), a loss of confidence in the government and its ability to implement its macroeconomic framework, or the emergence of a systemic crisis in a country related through financial and trade channels (international contagion). Lack of appropriate action or delays in addressing the emerging problems further fuels the bank run.
Gross costs can be measured as outlays of the government and central bank in terms of bond issuance and disbursements from the treasury for liquidity support, payout of guarantees on deposits, costs of recapitalization, and purchase of nonperforming loans. Net costs to the public sector deduct from gross costs recoveries from the sale of assets and equity stakes and repayment of debt by recapitalized entities.7
Gross costs have varied widely in recent crises. Initial macroeconomic and financial sector conditions are important determinants of the differences. For example, financial and corporate sector weaknesses in the Asian crisis countries made financial institutions more vulnerable to declines in asset values, devaluation, and reversal of capital flows. This vulnerability in turn helped foster speculative attacks and worsen capital flight, increasing losses and raising the ultimate resolution costs.
The policy response of authorities has also been key. Russia, for example, did not offer a blanket guarantee and did not recapitalize banks with public funds. While the costs to the economy may have been lower had there been more active public intervention, the direct costs to the state were low. The quality of the policy response, in terms of either macroeconomic adjustment or the handling of bank failure, affects the severity of the crisis. Lack of policy coordination or a clear, well-communicated strategy can raise costs to the economy by prolonging the crisis, undermining efforts to stem deposit runs and increasing uncertainty for borrowers, shareholders, and depositors. Late recognition of the systemic nature of the problems can also increase the final costs. For example, both these factors were important in Venezuela in the mid-1990s.
A final difference affecting net costs is how successfully the state recovers the value of assets acquired during the crisis. Two components are particularly important: sale of nonperforming loans and reprivatization (or sale of equity) of banks acquired during recapitalization. Recovery performance can vary considerably. While Sweden and Norway were able to recover all the costs associated with the crisis and still retain stakes in banks taken over, in general recovery rates have been low. The market value of nonperforming loans typically declines quite rapidly, while equity stakes in banks and enterprises might appreciate as they recover, or fall to zero. Also, in many cases, such as the United States, funds generated from asset disposals were recycled and used to recapitalize further institutions, thus lowering gross outlays.
Macroeconomic Context for Bank Restructuring
Coordination with the overall macroeconomic framework is a key factor in successful banking crisis management. Systemic financial crises affect most sectors of the economy and require adjustments in most aspects of economic policy implementation. While this paper concentrates on banking system restructuring, policymakers need to be aware of the broader policy context within which such restructuring must take place. Bank restructuring needs to be implemented in conjunction with supporting macroeconomic policies, taking into account existing macroeconomic constraints. Moreover, measures to contain the crisis and restructure banks may have macroeconomic consequences that need to be taken into account in the design of a bank resolution strategy.
At the outset of a crisis, macroeconomic policies may need to be adjusted to restore confidence in the banking system and the currency. The policy mix will depend on the nature of the macroeconomic imbalances and the state of the banking system. A temporary tightening of policies may be inevitable, however, when a creditor run from the banking system occurs with accompanying pressures on the price level, net international reserves, and the exchange rate. Comprehensive macroeconomic policy adjustments will be needed to restore stability, lower interest rates after an initial hike, and allow for a return of economic stability.
Authorities may consider capital controls when faced with runs not only on deposits but also on the currency. Capital controls must be tight enough to be effective, and their design should address the likelihood that means for evasion will quickly emerge. Such careful design is critical to ensure an effective system that minimizes the economic costs of the controls.8
Operations of the nonfinancial public sector often lie at the heart of a financial crisis. Persistent fiscal deficits, financed through central bank credit expansion, often fuel inflation and asset price bubbles. Moreover, high levels of government debt can undermine creditor and depositor confidence, making the banking system more prone to runs. Tax reform, combined with expenditure prioritization, is therefore a key aspect of the broad policy response to systemic financial crises in many cases. Design of the fiscal adjustment path must be clearly embedded in the broad strategic response to systemic financial crises.9 When sovereign debt dynamics are unsustainable, however, measures to achieve sovereign debt sustainability may sometimes cause banking system insolvency, thus increasing costs to the government and the economy that could reduce or eliminate the gains from debt reduction. Therefore, strategies for both bank and debt restructuring must be closely coordinated and consistent with each other.
As the macroeconomic policy stance and monetary policy tools are adjusted, the authorities may also have to strengthen their sovereign debt management. Banking crisis resolution has fiscal implications, and the techniques chosen will need to be consistent with the medium-term sustainability of public debt. The authorities should seek to ensure that the debt growth is not excessive and is closely coordinated with monetary and fiscal policy objectives. The changes in maturity structure, exchange exposure, and contingent claims must be carefully considered as part of the overall financial resolution strategy, as they can affect the costs of financing the resolution of the crisis.10
A Strategy for Managing a Systemic Banking Crisis
The strategy for managing a banking crisis must be tailored to country-specific conditions. Country-specific factors include the cause of the crisis, the macroeconomic conditions and outlook of the country, the financial position of the banking system, the risks of internal and external contagion, and the availability of resolution tools.11 In recent years, two polar examples have emerged. Banking crises have occurred in countries with weak banking systems dominated by local currency–denominated assets and liabilities, and where the governments had a relatively wide range of resolution tools. The Asian crisis largely reflected these conditions. In contrast, crises have also emerged where the banking systems have been relatively strong, and highly dollarized, but where the governments had limited resolution tools. The recent crises in Latin American countries reflect these conditions. While future crises may fall between these two extremes, the differences in approach illustrated by these polar cases provide a guide to adapting banking strategies to local conditions.
Any strategy typically includes three interconnected components. The first and most urgent component deals with an acute liquidity crisis. The liabilities of the banking system must be stabilized and deposit runs stopped. The second component seeks removal of insolvent or nonviable banks from the system and restoration of financial soundness and profitability. The third component of the strategy, which has a medium-term time horizon, focuses on the financial restructuring of nonperforming loans and the operational restructuring of bank borrowers.
The most immediate component of managing a banking crisis is the stabilization of bank liabilities—stopping depositor and creditor runs. The central bank, as the lender of last resort, should provide sufficient liquidity to the banking system to protect the payment system and give the authorities time to identify the causes of the crisis and design an appropriate response. In the face of sharp increases of liquidity, the central bank should use its monetary instruments to sterilize any resulting increase in the money supply. Moreover, such support should be limited to solvent banks. In the early stages of the crisis, however, it is difficult to distinguish between insolvency and illiquidity, and the government may have to recapitalize the central bank once the crisis has been resolved. In highly dollarized economies, the central bank’s ability to provide emergency liquidity is constrained by the level of international reserves, access to international capital markets, and support from international financial institutions. In such cases, the liquidity may not be available, and more aggressive policies on bank resolution may be needed.
If credible, a blanket guarantee can restore investor confidence and stabilize banks’ liabilities. A blanket guarantee gives the government some breathing space to develop a comprehensive restructuring strategy. It also makes resolution of weak banks easier, as bank interventions and closures are less likely to prompt depositor panic. A blanket guarantee alone, however, cannot contain a liquidity crisis; to be successful, it must be part of a credible stabilization package. Moreover, the moral hazard risks in a blanket guarantee have to be balanced against the potential benefits from such a guarantee. A blanket guarantee may not be credible in the face of unsustainable public debt dynamics or in a highly dollarized banking system. If dollarization is moderate, a guarantee could cover foreign currency liabilities in their domestic currency equivalent at market exchange rates.
If market-oriented stabilization measures do not contain the crisis, the authorities may have to resort to administrative measures to avoid losing monetary control. Such measures include securitization of deposits, forced extension of maturities, or a deposit freeze. To varying degrees, these measures impose restrictions on the depositors’ ability to withdraw their funds. Administrative measures can cause major economic disruption and, therefore, must be viewed as a last resort to stop a run on banks if all other measures fail.
The second component of the restructuring strategy is to restore the profitability and solvency of the banking system. The first task is to identify the size and distribution of bank losses. Supervisory data may be outdated, so a process for collecting data based on uniform valuation criteria should be initiated. The next task is to classify banks into one of three categories: viable and meeting regulatory requirements, nonviable and insolvent, or viable but undercapitalized. In the latter classification, an additional assessment will be needed of the ability of the existing shareholders to recapitalize their banks within an acceptable period.
The determination of a bank’s viability is a critical aspect of managing the restructuring process. Financial statements and asset values of a bank are often distorted during a crisis, making it difficult to determine a bank’s financial position. Under such circumstances, viability may be determined by examining two factors. First, the bank must develop a medium-term business plan and cash flow projections, based on realistic macroeconomic assumptions that show future profitability and medium-term strength. Second, shareholders must be committed and financially sound. Any business plan can go wrong; the shareholders must stand ready to adopt corrective measures. In addition, the authorities must develop a view as to the future volume of activity that the economy can absorb and establish criteria for bank evaluation that aims at an appropriately dimensioned banking system.12
Resolution techniques will depend on the banks’ financial conditions and medium-term prospects. All solvent but undercapitalized banks should be required to present acceptable restructuring plans. Bank recapitalization may be phased in if accompanied by an acceptable restructuring plan. An insolvent bank should be intervened and transferred to the institution responsible for bank resolution, which will decide whether to close it or keep it open. If the bank is closed, decisions need to be made on how to manage assets and liabilities, including nonperforming assets, performing assets, and deposits. If the bank is kept open, the restructuring institution must decide on a range of options, including whether to recapitalize the bank with public funds; to offer it for immediate sale or as a merger partner to a private institution; or to merge it with another solvent, government-owned bank.
Explicit decisions about burden sharing must be made in the design of the restructuring strategy. A banking crisis reflects losses of banks and their borrowers. The costs must be shared by some combination of bank shareholders, depositors, other creditors, and taxpayers. How the costs are paid and the distribution of the costs among different agents is a political as well as a technical decision, which requires explicit consideration in the design of the strategy. This process is difficult because the determination of losses is extremely difficult, nonperforming loans have no clear market value, and the size of losses is constantly changing in response to changes in the business environment. Strong and cohesive political leadership is required to address these issues.
Public capital support of private banks may be justified in some situations. The authorities may help private owners achieve a least-cost resolution. In this case, injection of new funds by the shareholders could be supplemented with public funds. Public participation in the recapitalization may be justified when the economy does not have sufficient capital and foreign interest is limited. Public funds may also be justified when the banking problems are the direct result of public policy.
Once the banking system has stabilized and both corporate restructuring and asset resolution are under way, the authorities will need to turn to strengthening the financial system and fostering reintermediation. Tasks at this stage include determining the role of both private and public financial institutions, reinforcing prudential and regulatory oversight, and strengthening transparency. Market discipline has to be strengthened, as the safeguards put in place at the height of the crisis must be phased out at a safe but meaningful pace. Exit rules for failing banks will have to be enforced and legal, judicial, and institutional structures strengthened to promote an effective and competitive banking system.
Difficult decisions must be made concerning the treatment of any bank that was nationalized as a result of the crisis. Reprivatization of banks and bank assets should take place according to a carefully developed strategy. The government will have to determine how to maximize the bank’s net value in light of expected future economic growth, either through rapid divestment of nationalized assets or a slower approach.13 Similarly, the government must determine how best to increase market competition. Privatization of small banks may increase competition in the market, but it may be difficult to attract adequate buyers. A single large institution may be easier to sell at a cost of reduced competition in the market. The role of foreign investors must be defined. There is little experience to date on this stage of crisis resolution, and it remains a critical policy issue for the future.
The third component of crisis management is asset management and corporate debt restructuring. Banks have three options in dealing with nonperforming assets: they can restructure the loans, liquidate the loans, or sell the assets to a specialized institution for resolution. This process of asset management is inextricably intertwined with corporate restructuring. Poorly designed corporate debt restructuring can impede or even reverse progress in financial sector restructuring. The objective of this phase of crisis management is to seek arrangements that allow banks to maintain positive cash flows, deepen business relations with solvent borrowers, and encourage corporate debt restructuring.
Banks and corporations must seek the timely and orderly restructuring of corporate debt in a way that shares the burden equitably. Government action is often required to ensure that banks are not at a disadvantage in negotiating with borrowers. Formal insolvency rules often must be strengthened while, at the same time, institutions and mechanisms are established to encourage out-of-court settlement. The financial restructuring of corporate debt should proceed in the context of a broader operational restructuring of the companies, which is a prerequisite for renewed corporate profitability, new investment, access to bank credit, and economic growth. If loans cannot be restructured, they should be liquidated and any collateral foreclosed.
There are a number of institutional options for managing impaired assets. Nonperforming loans can be managed by the banks or sold to specialized asset management companies, either privately or publicly owned. While each setup has both advantages and disadvantages, experience suggests that, in general, private financial institutions can respond more quickly and efficiently. Government-owned centralized asset management companies may lack incentives for maximizing recovery values, and may be subject to political interference. On the other hand, such companies may be relatively more efficient when the size of the problem is large—hence, asset management may involve economies of scale—or the required skills are unavailable.
Implementation of Restructuring Strategies
Implementation of a wide-ranging restructuring agenda is difficult to coordinate. Crisis episodes have evolved at varying paces and with different results (see Appendix II). Such differences are caused by a variety of factors, including initial conditions, policy mix, international environment, and even unexpected exogenous events. Notwithstanding important differences among crisis cases, some general lessons have emerged. While the following practices are not always required for the successful implementation of a strategy, they do appear to make implementation relatively smoother and improve the probability of success (see Box 3 for an overview of guiding principles).
Successful restructuring efforts have used the sense of urgency created by the outbreak of the crisis to initiate quickly the reform process. Experience suggests that political resistance to reform measures is weakest early in the process when a broad consensus to address the causes of the crisis is highest. As the crisis continues, vested political or economic interests emerge or reemerge, and the reform process becomes slower and more difficult.
A second practice is the establishment of a coordinating unit or committee early in the crisis to design and oversee crisis management. To be effective, this committee should be composed of senior government officials that have both the responsibility and the authority to develop economic policy. Often such a group is composed of cabinet-level officials and senior staff of central banks and financial sector oversight agencies. While this committee should have responsibility for strategy design, it would normally oversee, but not implement, the strategy.
Box 3.Guiding Principles from Recent Systemic Banking Crises
- Political support is important for successful crisis management. Public disagreements or expressions of doubt among prominent government participants can undermine confidence in the containment and restructuring process.
- A single, accountable authority can facilitate crisis management. Strong leadership is needed to deal with vested interests, determine issues affecting wealth and income distribution (burden sharing), and shepherd the restructuring program through the legislative process. This authority should clearly communicate the strategy and decisions to the public.
- Speed of intervention is essential. Decisive steps need to be taken in the initial stages to stop creditor and depositor withdrawals. The best opportunity for significant progress is in the early stages when political support to resolve the crisis is normally at its highest.
- A coherent and comprehensive package of measures should be implemented. Such a package may include credible macroeconomic adjustments, emergency liquidity support, a blanket guarantee where credible, and early closure of clearly insolvent banks. Should such measures not be effective, the authorities may need to resort to administrative measures as a last alternative—securitization of deposits, lengthening of deposit maturities, or a deposit freeze.
- Protection of depositors and other creditors will ease the restructuring process. Where credible, a blanket guarantee can ease creditor fears and facilitate the closure of weak banks. When a blanket guarantee is not credible, the authorities may have to rely on administrative measures.
- Burden-sharing decisions should be explicit. Existing shareholders should be the first to either inject additional capital or lose their investment. If capital continues to be insufficient, other stakeholders need to take losses. With a blanket guarantee, the government—and thereby taxpayers—assume the losses of depositors and other creditors. Under administrative measures, depositors and other creditors typically may have to take a share of the losses.
- The bank resolution strategy should include a thorough diagnosis and bank resolution plan. A process of bank triage should take place. Nonviable institutions should be liquidated or merged with viable banks, and adjustment periods could be provided for viable but undercapitalized banks.
- Bank resolution should follow a principle of equity and fair treatment. Restructuring policies should be applied to all banks on a uniform basis, that is independent of their ownership (public, domestic private, or foreign private) or type (wholesale, retail, or niche).
- The banking strategy should be designed to minimize the present value costs of the crisis. The likely costs of different options should be identified, and the strategy should be based on the lowest present value of costs to the economy. Costs estimates should include an estimate of the impact of each option on the banking system, economic growth, and the sustainability of government debt.
- Recapitalization with government funds may be justifiable in some circumstances. Private funds should be used first, and the terms for public assistance should be uniform and transparent.
- Asset resolution is an essential complement to bank restructuring. An early and active involvement in impaired asset management would prevent credit discipline from eroding. A variety of institutional arrangements and techniques are available. They should be chosen in order to achieve the desired trade-off between rapid resolution and recovering the value of the impaired assets.
- Corporate restructuring should go hand in hand with bank restructuring. Without corporate restructuring—both financial and operational—economic activity will slow down further, and the banking system may not get out of, or fall back into, a state of distress. Corporate debt restructuring should be based on market principles and reinforce payment discipline.
A related practice is to avoid excessive centralization of policy implementation and give government agencies the authority and responsibility to implement the restructuring strategy. Often, existing agencies with established credibility may be given such a responsibility. Existing agencies—including the central bank, the supervisory agency, the ministry of finance, and the deposit insurance agency—often have adequate staff, organization, and infrastructure. Given the need to act quickly, it may be cost effective to rely on these existing institutions. On the other hand, establishing new bodies with specialized staff may be necessary. Such new agencies may be staffed with specialists not normally available or may be given special responsibilities in the conduct of the restructuring strategy. The decision to rely on the existing institutional framework or establish new institutions will depend on the credibility and experience of existing institutions, the need for specialized institutions, and the speed with which new institutions can be created and staffed.
The fourth practice is communicating with the broader public. An often-neglected aspect of crisis management is the importance of strengthening market confidence by informing the public about the direction of public policy. The authorities should use all forms of communication to explain their views on the causes of the crisis, their understanding of how the crisis will be resolved, and where the reform strategy will lead. The announcements should be consistent, regular, and ideally be given by a single official spokesperson throughout the containment and restructuring efforts. This communication strategy can be effective in forging and then maintaining support for the reform efforts and can help reduce the influence of entrenched special interest groups.
Finally, the authorities can benefit from the demonstration effect of quick and successful wins. Restructuring can be a prolonged process, with difficult adjustments and costs. Reform fatigue can be reduced by taking advantage of opportunities for short-term successes. The authorities should seek to identify steps that produce some immediate results as an indication that the benefits from the reform process are not only to be achieved over the medium term, but also immediately. These steps, when placed in the context of the broader strategy, can help to mobilize support for the entire reform process.14