This paper highlights that for the IMF, July 2004 marked the 60th anniversary of the conference in Bretton Woods, New Hampshire, when delegations from 44 allied countries drafted and agreed upon the IMF’s charter. The IMF’s role and work have evolved in response, but like any large organization, its ability to change has been limited by its own rules and mandate and has been held back by inertia. This year’s anniversary offers an opportunity to reflect on how gaps between the reality and the ideal might be closed in the coming years.


This paper highlights that for the IMF, July 2004 marked the 60th anniversary of the conference in Bretton Woods, New Hampshire, when delegations from 44 allied countries drafted and agreed upon the IMF’s charter. The IMF’s role and work have evolved in response, but like any large organization, its ability to change has been limited by its own rules and mandate and has been held back by inertia. This year’s anniversary offers an opportunity to reflect on how gaps between the reality and the ideal might be closed in the coming years.

It is time to give supervisory authorities the independence to do their job

JULIO Maria Sanguinetti, former Uruguayan president (1985–95), once commented that “the banking system will never take you to paradise, but it can bury you in hell in an afternoon.” His words continue to echo in Latin America, which has been struggling to find ways to make financial systems more resilient and to avoid or minimize the costs of crises that do occur. After a series of banking crises hit the continent in the 1980s, many countries undertook significant reforms through bank consolidation, privatization, and allowing increased entry of foreign institutions. This transformation went hand in hand with the strengthening of the prudential regulatory framework and monetary stabilization. But banking crises have not disappeared, and the citizens of these countries have continued to bear the associated costs, fueling skepticism about market-oriented reforms.

That said, some countries have reduced their vulnerability to banking crises. The emergence of a consensus on the direction of financial sector policies and the need to reinforce vital institutions seems to have underpinned this transition. An analogy can be drawn with the Latin American experience of reducing inflation and establishing independent central banks: much of Latin America was plagued by very high and volatile inflation during the 1970s and 1980s. In response, a consensus emerged on the importance of establishing price stability and strengthening the institutional arrangements to carry out this task, such as granting independence to the central bank. The result, over time, has been lower and more stable inflation. A similar commitment to preserve financial sector stability is now needed. This article argues that granting the supervisory and regulatory authorities adequate independence can be a crucial element in making that commitment real.

Banking crises are costly

Latin American countries have been through many financial crises (see Table 1). In fact, they seem to have suffered a disproportionate number compared with countries at a comparable level of development in other regions. Furthermore, the frequency, severity, and tendency of these crises to recur are a concern. Nonetheless, much of what is argued here applies to other regions.

Table 1

Episodes of crisis and distress

Banking crises have been all too common in Latin America and the Caribbean.

article image
Sources: IMF publications; and World Bank database (Caprio and Klingebiel, 2003).

Defined as the exhaustion of much or all of bank capital and usually involves a run on deposits.

Defined as low or negative net worth without a systemic run on deposits.

In almost all cases, a banking crisis has implied a large redistribution of wealth and has led to an overall reduction in both income and wealth. The immediate aggregate cost can be measured in terms of loss of output, additional fiscal outlays, and public debt incurred (see Table 2). Additional costs may also be borne by bank shareholders, depositors, and future depositors and borrowers, who suffer wider interest rate spreads. A banking crisis in one country can also generate costs—including banking system distress—in other countries linked through trade or financial relationships.

Table 2

Costly crisis resolution

Output and fiscal losses have been large, and recovery can take years.

article image
Sources: Hoggarth and Saporta (2003); Hoelscher and Quintyn (2003); and IMF staff estimates.

Deviation from trend growth three years after the crisis.

Providing direct support to a failed banking system can hurt debt dynamics, making it even tougher to achieve debt sustainability. Indeed, the experiences of a sample of six Latin American countries (Argentina, Colombia, Dominican Republic, Ecuador, Mexico, and Uruguay) that suffered banking crises during the past 10 years illustrate how large and sudden the impact can be. Within one year of their respective crises, the ratio of public debt to GDP increased, on average, by about 40 percentage points. The initial fiscal costs of resolving banking crises accounted for about 20 percentage points of this increase. Thus, in a very short period, a banking crisis can undo much of the gains achieved through prudent macroeconomic (particularly fiscal) management over many years, as happened, for example, in the Dominican Republic (see box).

There are longer-term costs, too. The institutional features that interfere with good banking practices also interact with each other and with the shocks a country suffers—and may be one reason why certain countries tend to have repeated crises. Manifestations of this dynamic include, for example, low saving rates, very limited long-term financial relationships, reliance on external financing, high interest rate spreads, dollarization, and a heavy public sector debt burden. At root is an underlying lack of trust in the financial system, which can largely be explained by a history of substantially negative real interest rates and real depreciations, weak accounting and creditor rights, fiscal disincentives to save (such as financial transaction taxes), and banking crises followed by unsuccessful bank resolutions (for example, involving freezing and unfreezing deposits, as in Ecuador in the late 1990s).

Under these conditions of “short termism” and poor financial intermediation, deposit withdrawals translate quickly into credit contractions, which may further worsen the situation as firms are starved of working capital and investment is constrained. Countries in the region have often experienced credit crunches, when net credit to the private sector has declined sharply, increasing the probability and expected severity of a banking crisis and creating a vicious circle.

Dominican Republic’s rapid unraveling

Before a major banking crisis unfolded in the Dominican Republic in 2003, the economy had been considered one of the top performers in the region, with real annual GDP growth averaging 6 percent during the 1990s, inflation under control, and the debt-to-GDP ratio stable at around 25 percent. The banking crisis sparked a major depreciation of the local currency, the peso, which fell by nearly 29 percent in real terms in less than a year. During the year following the crisis, the country suffered a recession and the debt-to-GDP ratio skyrocketed to 57 percent—more than double the level at the time the crisis hit (see chart). Foreign as well as domestic debt increased, although largely because of the large currency depreciation in the former case.


Soaring debt

Strong public finances built up over years can be undone in a few months by a banking crisis.

(Dominican Republic’s public sector debt stock as percent of GDP)

Citation: Finance & Development 0041, 003; 10.5089/9781451922530.022.A011

Source: IMF staff estimates.

Origins of crises

The experience of banking crises in Latin America—and elsewhere—suggests that these episodes typically display similar pathologies, even while their causes and consequences differ. Crises arise from bad banking practices or from the effects of poor macroeconomic policies; most of the time they arise from a combination of both. The main disturbances facing Latin American countries have shifted over time. Nonetheless, the most common triggers of banking crises in Latin American countries can be identified as

  • a boom in credit to the private sector, for both investment and consumption (Mexico, 1994; and Colombia, 1999). A particular form of boom and bust cycle is generated by the end of hyperinflationary episodes (Bolivia, 1986);

  • wholesale liberalization in the absence of an appropriate and effective prudential regulatory framework (Mexico, 1994; and Chile, 1984). It is worth stressing, however, that highly regulated systems have also suffered crises (Peru, 1987);

  • direct effects of fiscal difficulties on the domestic banking system, a factor that seems to have become an increasingly important source of strain on Latin American banks (Argentina, 2001);

  • contagion and spillovers, where a crisis in one country induces economic agents to reassess their expectations and thus reduce investment in other countries (Argentina, 1995), or where a crisis in one country has a direct effect on economic conditions in another country (Uruguay, 2001);

  • terms of trade shocks and movements in real exchange rates (Venezuela, 1994; and Ecuador, 1998); and

  • political instability, unrest, and, in some cases, civil conflict.

These sources of disturbance to which Latin American economies were subject interacted with an institutional setting that was often unable to cope or that added to the instability and costs. The institutional setting is taken here to comprise a range of more structural and legal features of an economy, such as the organization of the banking system, the state of prudential regulation and supervision, and the wider framework of accounting practices, corporate governance, the administration of justice, and the enforcement of property rights.

The institutional setting has not been constant. Many Latin American economies have made substantial progress in these areas over the past 30 years. Nonetheless, poor banking practices and a weak institutional framework have played a major role in making Latin American countries vulnerable to financial crises following shocks—and in compounding those crises.

Deficiencies in the following areas have detracted from good banking practices and increased vulnerability to crisis in some Latin American countries:

  • inappropriate and ineffective prudential regulation and supervision;

  • inefficacy of bank intervention and resolution;

  • policy-induced distortions, and, in particular, government influence over public sector banks;

  • poor structure and composition of government finances;

  • inadequate accounting practices, property rights, and corporate governance; and

  • inefficiency of the judicial system and poor observance and enforcement of laws.

Assessments carried out for a range of Latin American countries since 1999 as part of the Financial Sector Assessment Program (FSAP)—an IMF–World Bank diagnostic tool for evaluating the soundness of countries’ financial systems—suggest that weaknesses are common in such areas as consolidated supervision (whereby all the risks run by a banking group as a whole are taken into account, wherever they are booked) and timely measurement of true capitalization (especially the early recognition of when a borrower gets into trouble and a full repayment of a loan becomes unlikely).

One reason why regulations and supervisory practice may have been weak is that supervisors have often lacked independence and supporting institutional arrangements. Even if the supervisors were fully aware of what they should have been doing, they may have met opposition from interest groups or lacked either the legal powers or resources to carry out their intentions. Indeed, it is notable that Latin American countries rate especially poorly on compliance with the Basel Core Principle relating to the independence of the supervisory authority. The region’s poor performance with respect to this principle might help explain its rather poor compliance with the remaining Basel Core Principles—even though the regulatory framework in many countries was strengthened during the period between the onset of crisis and the relevant FSAP.

Minimizing costs of crises

Although each banking crisis is unique, there are certain measures that policymakers can take to foster the development of a financial system that contributes to growth rather than becoming a fiscal burden. The usual policy prescription includes pursuing stable macroeconomic policies, enhancing the regulatory and supervisory framework, and improving creditor rights and related institutional arrangements, such as insolvency procedures and bankruptcy legislation. But for Latin America, certain measures deserve special attention. Their effective implementation requires a shift in approach to financial sector policies and, in particular, enhanced independence of the supervisory and regulatory authority.

What should be the priority measures for Latin America? First, the risks of exposure to government need to be fully recognized. Recent experiences of Argentina, Ecuador, and Uruguay have demonstrated that potential sovereign debt distress can undercut banking system soundness. Therefore, bank regulators may need to adopt prudential measures to mitigate the transmission of risks or losses from sovereign debt distress to the banking system. Some countries have imposed risk-based capital requirements that are above the international norms on banks’ foreign or even domestic currency government debt holdings, such as Brazil in 1997–99. Some countries, such as Argentina, have imposed limits on bank exposure to government—similar to sectoral limits or limits on large exposures.

Second, governance of state banks must be stronger. The regulatory agency needs to be able to impose prudential requirements on public sector banks and sanction them as necessary. Such requirements may make it more difficult for a public sector bank to achieve its noncommercial objectives or may increase explicit costs. For example, loans to priority sectors often have poor repayment rates. Imposing prudential rules would require that provisions for loan losses be made promptly and in full, which would reveal any implicit subsidy at an early stage.

Third, supervision and regulation must adapt to new challenges. Supervisors are currently facing certain challenges, such as strong growth in off-balance-sheet exposures by banks—in several Latin America countries, off-balance-sheet liabilities were larger than balance-sheet liabilities—and expansion of offshore banking operations, conglomerates, and cross-border financial institutions. To meet these challenges, supervisors must be able to adapt regulations as market practices evolve, upgrade the skills of their staff, and have sufficient staff for effective oversight.

Fourth, crises must be managed effectively. Even countries that have effective prudential supervision may—for understandable reasons—suffer policy paralysis when faced with a banking crisis. In this environment, it may become apparent that existing institutions and legal provisions are not up to the task. Contingency planning and the establishment of mechanisms to resolve problem institutions can be highly valuable, if only to avoid the gross policy errors that are often evident in the immediate response to banking sector difficulties.

Build on consensus

For Latin America, many of the policy decisions that now appear to have been unfortunate were not made out of a lack of technical skills or ignorance of the true state of affairs. Rather, they were the outcome of a political process that too often placed the near-term interests of specific groups over financial sector stability and efficiency and over longer-term development needs. Therefore, prudential regulation and supervision, including the technical aspects of dealing with bank crises, need to be protected from immediate political influences and have the resources and legal authority to act for the general good. This is not to deny the need for accountability and ultimate democratic control, especially where large sums of the public’s money are involved. Rather, what is required is to establish institutional arrangements that enshrine what should by now be a broad consensus on the need to pursue policies that foster financial sector soundness.

Agustín G. Carstens is IMF Deputy Managing Director and former Governor of the Central Bank of Mexico. Daniel C. Hardy is Deputy Division Chief, and Ceyla Pazarbasioglu is Technical Assistance Advisor (and former Vice President of the Banking Regulation and Supervision Agency of Turkey) in the IMF’s Monetary and Financial Systems Department.


  • Barth, James R., Gerard Caprio, and Ross Levine, forthcoming, “Bank Regulation and Supervision: What Works Best?” Journal of Financial Intermediation.

    • Search Google Scholar
    • Export Citation
  • Caprio, Gerard, and Daniela Klingebiel, 2003, “Episodes of Systemic and Borderline Financial Crises,” World Bank data set available via the Internet:

    • Search Google Scholar
    • Export Citation
  • Das, Udaibir S., Marc Quintyn, and Kina Chenard, 2004, “Does Regulatory Governance Matter for Financial System Stability? An Empirical Analysis,” IMF Working Paper 04/89 (Washington: International Monetary Fund).

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hardy, Daniel C., and Ceyla Pazarbaşioğlu, 1999, “Determinants and Leading Indicators of Banking Crises: Further Evidence,” IMF Staff Papers, Vol. 46 (September), pp. 24758.

    • Search Google Scholar
    • Export Citation
  • Hoelscher, David, and Marc Quintyn, 2003, Managing Systemic Banking Crises, IMF Occasional Paper 224 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Hoggarth, Glenn, Ricardo Reis, and Victoria Saporta, 2002, “Costs of Banking System Instability: Some Empirical Evidence,” Journal of Banking and Finance, Vol. 26 (May), pp. 82560.

    • Crossref
    • Search Google Scholar
    • Export Citation
Finance & Development, September 2004
Author: International Monetary Fund. External Relations Dept.