Banking Crisis from an International Perspective, By Stefan Ingves, Director, Monetary and Exchange Affairs Department, IMF

April 8, 2003

Speech By Stefan Ingves
Director, Monetary and Exchange Affairs Department
International Monetary Fund
Given at the Seminar on Financial Safety Nets
SEDESA (Seguro de Depósitos Sociedad Anónima)
Buenos Aires, April 8, 2003

Ladies and gentlemen:

I am especially honored and pleased to be participating in SEDESA's seminar on safety nets. The role of safety nets in the management of banking crises is of critical importance. The design and function of safety nets affects the timing and the resolution of banking crises.

I hope I can bring to the discussions of this seminar a unique perspective, because I have been an "insider" on both sides of financial crises. First, during the Nordic banking crisis, I was a member of the Swedish government that struggled with controlling the evolution of the crisis and bringing the financial system back to soundness. More recently, I have been involved in the international fight against banking crises from the International Monetary Fund.

The organizers of the seminar have asked me to talk about the causes of financial crises. I will briefly review this subject generally, and then more specifically with reference to the recent crises in Latin America. Finally I will touch upon the unique crisis that has been facing Argentina, and what efforts are underway internationally to reduce the likelihood of future banking crisis.

I. The Causes of Banking Crises

Painting with a very broad brush, let me start by saying that, while each banking crisis has its own dynamics, most of the main ingredients are always present. Based on their most common causes, banking crises can be classified into one of two categories: microeconomic (or bad banking), and macroeconomic (or bad operating environment). Let me mention each of them briefly:

Bad banking

Banking crises often have their roots in poor bank operations: poor lending practices, excessive risk taking, poor governance, lack of internal controls, focus on market share rather than profitability, and currency and maturity mismatches in the banks themselves or among their borrowers. These conditions are aggravated if bank owners have little at stake in the banks—that is, do not have enough capital invested in the banks—and if bank managers carry little personal responsibility for the risks they take.

In some emerging countries these conditions may be worsened when bank ownership is very narrow and when banks are run as personal "piggy banks" or pyramid schemes of industrial groups or families. In these conditions, connected lending, insider operations, and outright fraud may go on with impunity. Similarly, state banks may be run as quasi-fiscal agencies based on political criteria with disregard for commercial principles, which undermines their solvency and the soundness of other better-run banks.

Bad banking can only persist in the absence of proper regulation and supervision, and of adequate market discipline.

  • Weak supervisory frameworks may include allowing for concentrated lending, portfolio mismatches, inadequate loan valuation that overstate bank profits and capital, incompetent management, etc. Supervision may also lack authority, and have an insufficient number of skilled staff who may be poorly motivated and compensated.

  • Poor transparency, limited financial disclosure, and poor accounting and auditing practices mean that the market—that is, bank creditors—will not have sufficient information to exert discipline on bank owners.

  • Market forces are further impeded by weak frameworks for dealing with problem banks, including weak legal, judicial and institutional frameworks for dealing with failing banks and companies. Expectations of depositor and creditor bailouts may overpower any policy to the contrary.

Bad operating environment

Crises can arise from macroeconomic causes that are external to the banking system. Even well-run banking systems operating in a strong legal and regulatory framework can be overwhelmed by the effects of a poor macroeconomic environment or inadequate policies. Macroeconomic difficulties may arise from lending booms, possibly stoked by excessive capital inflows or changes in tax rules; real estate and/or equity price bubbles that inflate and burst; slowdown in growth and/or exports, or the loss of export markets; growing excess capacity/falling profitability in real sector; lower overall investment; rising fiscal and/or current account deficits; weakened public debt sustainability; sharp changes in exchange rates and real interest rates; and so on. Not all these developments are under authorities' control—but governments must be ready to adapt macro policies that take the conditions of a systemically distressed banking system into account.

Crisis triggers

I should add, importantly, that as long as the banks remain liquid, banking distress can persist for a prolonged period—until some trigger leads depositors and creditors to lose confidence in the banking system. These can be market, policy, or political shocks that become the "wakeup call" for dealing with problems so far ignored, causing dramatic shifts in expectations and systemic bank runs.

  • The emergence of illiquidity in one bank can quickly spread to others through contagion, as bank or payment system weaknesses destroy credibility of all banks, and lead to creditor and depositor runs regardless of the soundness of individual banks. Contagion is in this case local, but can also be international, when it is caused by the emergence of a systemic crisis in a country related through financial and trade channels.

  • Premature financial liberalization, together with inadequate preparation among bankers and supervisors, has also in the past triggered banking crises. Bankers may not have the needed skills to manage and price risk, and supervisors may lack adequate resources and competencies to monitor the more complex new risks. This can easily create a situation where liberalization unleashes the effects of pure ignorance about the risks involved among relevant parties.

  • A loss of confidence in the government and its ability to implement its macroeconomic framework can trigger a systemic crisis. Such concerns erode confidence in the banking sector as well as, often, the currency.

  • In addition, if the medium-term sustainability of a country's macroeconomic policies, including its external debt, comes under question, are in doubt, private confidence in the economic outlook can erode, resulting in the emergence of banking pressures and, in unaddressed, banking distress.

But maybe the most important issue is not the specific causes of a banking crisis, but the ability of the political authorities to come together to develop a strategy and then implement it. This is a point that I would like to emphasize.

Political decision makers must recognize that there is a problem and make quick and resolute actions. Developing a political consensus on the path forward is a critical step. Most importantly, the process must be seen as fair and transparent. Bank owners and borrowers must be prevented from exerting inappropriate interference.

Framework for crisis resolution

By the late 1990s, the Fund had significant experience in addressing banking sector crises in a variety of economic environments. The experiences of the Nordic countries were among the first, intensive areas of involvement and, in the late 1990s, the Fund, and MAE in particular, was intensively involved in resolving the financial crises in Asia.

These crises emerged in an environment quite different from the crises in Latin America today. The roots of the crises lay, in large part, with the weak financial sector and were only triggered by macroeconomic difficulties. Our response was to address those root causes, implementing important legal and regulatory reforms and strengthening bank supervision. These efforts were supported by reforms to address the deep insolvencies in the banks and institutional steps to manage the system's nonperforming assets.

The results of the Nordic and Asian crises have been quite good. The countries involved have largely emerged from the crises with vibrant financial systems. However, the lessons learned so well in those countries are now being modified in light of the new challenges faced in the more recent financial crises in Latin America.

Let me now turn to those new challenges.

II. Recent Challenges

If these issues where not sufficiently difficult and challenging, recent banking crises have dealt with a combination of high dollarization and high levels of government debt that have imposed particularly hard constraints on crisis management.


Crises have emerged recently in highly dollarized economies such as Argentina and Uruguay. This aspect raises issues not previously encountered and special factors must be considered.

A high degree of dollarization particularly makes banking systems more prone to runs and makes runs more difficult to stop. Dollarized economies face liquidity constraints early in the crisis, as bank runs must be addressed using limited international reserves. Furthermore, as depositors know that the supply of dollars is limited, they may be more prone to leave the banking system.

These conditions—the constraints on liquidity support and on depositor protection—may make the use of administrative measures—such as restrictions on deposit withdrawals or securitization of bank liabilities—more likely. As we all know, both Argentina and Uruguay were forced to design administrative measures to contain a run on their banks that could not otherwise be contained.

Sovereign debt restructuring

Another factor that has recently become evident is the constraints on banking crisis management when this is combined with sovereign debt restructuring. Although we have limited experience, available evidence suggests that the impact of sovereign debt restructuring on bank soundness will depend on a large number of issues, including the size of banks' exposure to the government, the currency of denomination of the debt, and the terms and modalities of the debt restructuring.

Large-scale restructuring can result in immediate undercapitalization or even systemic insolvency when banks have sizeable holdings of government debt. At the same time, policymakers should be aware that the use of fiscal resources to deal with banking problems are inherently limited when the debt is already unsustainable and needs to be restructured—hence the government's role as creditor, guarantor, or owner of last resort is severely limited. For these reasons, any strategy for sovereign debt restructuring must be designed to explicitly consider the impact on the banking system.

From a banking system perspective, we have come to believe that, if restructuring is necessary, and if the authorities have a choice, a preferable approach is to avoid nominal reductions (haircuts) in bank assets. Haircuts would result in reduction of bank capital and, likely, the imposition of remaining losses on depositors. Instead, the authorities should try to achieve a restructuring through reductions in interest rates or extension of deposit maturities that may not entail immediate losses.

III. The Case of Argentina

I will not review the history of the Argentine banking crisis. I know that all of you in this room are more familiar with recent developments than I am.

Let me just acknowledge that, in 2000, the Argentine banking system was in sound financial condition following the substantial consolidation, privatization, and increased foreign entry in the second half of the 1990s. This strengthening of the system was supported by a tightening of regulation and supervision.

The build up of the crisis began in 2001 and emerged in full in 2002.

During 2002, progress in stabilizing the financial conditions was delayed by the lack of basic information on banks' financial conditions, uncertainties concerning compensation for policy-induced losses, as well as delays in issuing necessary prudential and accounting rules.

Banking sector strategy

The policy agenda for 2003 is both clear and urgent.

As I mentioned above, the most important challenge now is for the Argentine authorities—both the Ministry of Economy and the Central Bank of Argentina—to agree on a coherent banking strategy and then implement that strategy as quickly as possible. Further delays will only increase the costs of restructuring and reduce the already limited options available.

Second, as we have seen in numerous other crises, the policies adapted must be transparent and uniformly implemented. Special interests will always try to gain in period of economic distress. The policy challenge is to build a broad domestic consensus for the difficult challenges ahead without giving in to limited, special interest groups.

Third, progress must be made in understanding the true condition of the banking system. While banks appear reasonably liquid, their solvency remains unclear. This reflects uncertainties about compensation for policy-related losses as well as concerns about the quality of banks' portfolios. These issues must be quickly addressed. Specifically:

  • The mechanism for compensating banks for policy-related losses should be finalized;

  • The regulatory framework should be updated;

  • Banks must provide detailed audits, business plans, and monthly cash flow projections.

Fourth, the central bank must determine the state of the banking system. Using financial statements and the business plans, viable and nonviable banks should be identified. Viability can be determined by evaluating banks' capital, the strength and commitment of shareholders to recapitalize the bank, the business plans, their client base, the quality of the loan portfolio, and future cash flow.

Fifth, nonviable banks must be resolved, either through recapitalization and restructuring by shareholders, or through least cost resolution techniques including merger, sale, or liquidation.

Finally, I believe that, regardless of the resolution option chosen, if feasible, additional losses to depositors should be minimized. If depositors lose confidence further, it will be difficult to prevent a resurgence of bank runs.

Special issues for public banks

A priority task is to accelerate the restructuring of the large public banks, which were weak even before the crisis and have accumulated a substantial debt to the central bank. Uncertainties remain about the true condition of these banks, which must be ascertained through a detailed due diligence process. After that, a difficult restructuring process must begin, leading to a reasonable financial solution to these banks' problems, and a clear definition of the future role for public banks in the Argentine system.

Remaining institutional agenda

I would like to highlight two more issues that should have priority in the emerging strategy for the financial sector.

We have seen in all countries that progress cannot be made in the restructuring of the banking system without clear protection for supervisors. Dealing with weak or insolvent banks touches at the heart of a country's financial system. Pressures can be put on supervisors to make inappropriate decisions or not to make decisions at all. Without protection, supervisors will have little ability to withstand such pressures. Legal protection is a key element of any successful strategy.

Second, the autonomy of the central bank has increasingly become an accepted international standard. In the region, as well as in Europe and Asia, more and more central banks are gaining their autonomy. This allows them to act in the best interest of the country, aiming monetary policy at controlling inflation and ensuring financial stability. I would urge the Argentine authorities to continue their movement towards the full autonomy of the central bank.

These are challenging times and I wish the best for Argentina and the Argentine authorities.

Thank you!


Public Affairs    Media Relations
E-mail: E-mail:
Fax: 202-623-6278 Phone: 202-623-7100