Report condemns IMF ability to anticipate and prevent crises

21 July 2003

A US Government Accounting Office (GAO) report finds serious shortcomings in recent IMF efforts to improve its capacity to predict and prevent financial crises, which is at the core of its mandate. After the Asian crisis a new ‘vulnerability framework’ was designed which includes instruments to anticipate crises. The GAO, an investigative and audit arm of the US Congress, says it is too early to tell if this will work, but is concerned the framework relies heavily on elements such as the World Economic Outlook (WEO), and Early Warning Systems (EWS). The report says these “two major forecasting tools… have not performed well in anticipating prior crises”.

The World Economic Outlook is a six-monthly Fund publication describing trends in the world economy, including growth forecasts for member countries. The GAO emphasises the WEO’s “poor track record in forecasting recessions, including those directly associated with a financial crisis”. From 1991-2001, out of 134 recessions that occurred in developing countries, the WEO correctly forecast only 15, while actually predicting an increase in GDP in the other 119 recessions, including major ones leading to the Mexican and Asian crises.

The Fund’s Early Warning Systems models (EWS) examine a series of vulnerability indicators to compute the probability of a country having a crisis over the following 12 to 14 months. The GAO criticizes EWS because they have had a “high false alarm rate -a problem also pointed out by an IMF internal review- while failing to predict severe crises. The IMF argues that EWS are not designed to make final predictions of crises but are designed to identify vulnerable countries, and that users can set their own threshold at which to ‘call’ a crisis.

Whereas the IMF claims the US watchdog attaches too much importance to the WEO and EWS, the GAO argues they are “the only mature and quantifiable elements of the [vulnerability] framework”. The IMF argued in its formal response to the report that its responsibility to maintain financial stability could make its predictions less accurate so as not to contribute to a crisis. This led the GAO to retort that it “not only validates our finding on the WEO‘s weakness but also raises questions regarding the purpose and the credibility of the WEO forecasts”. Romilly Greenhill from Jubilee Research in the UK commented: “IMF economics remains rooted in neo-liberal economic analysis, analysis which fails to take note of the dangerously destabilising consequences of financial globalisation. It is time that the IMF moved to a new form of economic analysis – one which views reality as it really is.” Jubilee Research will release shortly a Real World Economic Outlook with contributions from prominent economists such as Joseph Stiglitz and Dani Rodrik.

Code warnings

New initiatives for crisis prevention introduced by the Fund since the late 1990s include joint assessments with the Bank of key elements of member countries’ financial sectors. These are known as Financial Sector Assessment Programmes (FSAP). Another component is the promotion of adherence to voluntary standards to ensure financial stability and reassure investors that a country’s policies and practices conform to standards and codes of good practices. Adherence to these standards related to transparency, financial sector regulation and corporate sector practices (see box) is measured in Reports on the Observance of Standards and Codes (ROSC). Participation of member countries in these two initiatives is voluntary. Contrary to IMF claims (based on surveys), the GAO says that private investors make little use of the information in these reports, because they often consider it outdated, untimely and too ‘dense’ (a euphemism for jargon). The ROSC for Argentina have not been updated since 1999 for example. The fact that participation is voluntary means that there is a lack of information for some important countries, such as China, Thailand and the US.

The GAO recommended that the Fund “improve the readability, timeliness, coverage, and frequency of updates of assessment reports, including the possibility of making participation mandatory for all members of the IMF”. The Fund and the Bank broadly agreed with these recommendations but while this would be legally possible they are officially reluctant to make assessments mandatory. However participants in a conference held in 2002 by the Overseas Development Institute in London noted that “while the Bretton Woods Institutions have tried to avoid a coercive element in their work, yet their need to achieve full coverage of countries that have a potential for causing instability has led them to exert a certain amount of pressure, even when the countries in question are not convinced of the benefits relative to the costs incurred”.

Indeed there is no decisive economic proof of these benefits. Many countries are concerned that standards and codes might become a new form of conditionality. They also fear that with the ‘outreach’ by the IMF to promote the use of standards and codes, conditionality will in effect be supplemented by additional market-based sanctions. At the heart of the problem is the fact that the 12 core standards have been set mostly by rich countries gathered in a ‘financial stability forum’. This means they are appropriate to these countries and not necessarily to others, and renders building ‘ownership’ of other countries around these standards difficult. While standards have been set by rich countries the Fund and the Bank have no leverage to ensure these countries actually comply with them.

Adding to these limitations are the capacity constraints for the Bank and the IMF to serve as global monitors. A viable alternative would be greater use of self-assessment with a peer review process, with the role of the Fund and the Bank limited to providing technical assistance and coordination.

The political significance of these debates, particularly of the Fund’s capacity to anticipate and prevent crises, is important because this aspect of the IMF’s work is considered its core area of competence. Weaknesses and failures give devastating ammunition to critics of the Fund, who argue that the IMF is not able to perform its core mission, and that it is destabilising countries more than ensuring financial stability.

IMF initiatives to anticipate, prevent and resolve financial crises


  • Vulnerability assessment framework
  • World Economic Outlook
  • Early Warning System models
  • Country external financing requirements
  • Market information
  • Financial sector vulnerability
  • Country expert perspectives


  • Long-term reforms
  • Financial sector assessment program
  • Reports on the Observance of Standards and Codes


  • Debt restructuring proposals
  • Sovereign Debt Restructuring Mechanism [on hold]
  • Collective Action Clauses
  • Strengthening of lending policies

source: GAO analysis of IMF data

Whose standards?

The Financial Stability Forum brings together senior representatives of national financial authorities (e.g. central banks, supervisory authorities and treasury departments), international financial institutions, international regulatory and supervisory groupings, committees of central bank experts and the European Central Bank. They have prioritised twelve key standards perceived as minimum requirements for good practice in three main areas:

  • macroeconomic policy and data transparency (transparency of monetary and financial policy, of fiscal policy ; data dissemination).
  • Institutional and market infrastructure (with standards of corporate governance, accounting, auditing, money laundering, insolvency systems, payment and settlement)
  • Financial regulation and supervision (banking, securities, insurance supervision)

The IMF designed standards on macroeconomic policy and data transparency, the Bank on insolvency systems. Other standard-setters include the OECD, the International Federation of Accountants, the Basle Committee on Banking Supervision etc. Jointly or separately the Fund and the Bank assess compliance of volunteering member countries with nearly all of these standards.

As of December 31, 2002, 343 ROSCs had been produced for 89 economies or 48 percent of Fund membership. About 71 percent of ROSCs were published.

More details on Key Standards and the FSF