To tackle these questions, the IMF organized a consultative meeting on September 10-11 at IMF headquarters. The meeting, hosted by the IMF’s Monetary and Exchange Affairs and Statistics Departments, elicited a wide range of private and official opinions from outside the IMF. Participants included high-level experts from central banks and bank supervisory agencies, banking and investment officials from the private sector, academicians, and representatives from other international organizations. IMF senior staff, management, and Executive Board members also participated.
How to strengthen surveillance
In his opening statement, IMF Managing Director Mchel Camdessus emphasized the role of IMF surveillance over countries’ financial systems and the role of improved monitoring of financial systems by the countries themselves in strengthening the international financial architecture. The IMF is reinforcing its financial sector surveillance through the Financial Sector Assessment Program, which it is jointly developing with the World Bank. The objective of the program is to recognize potential problems at an early stage and develop responses promptly to avoid costly systemic crises. Identification of a core set of macroprudential indicators will support these efforts.
The main issues were whether it would be possible to define a core set of macroprudential indicators that the IMF could use in its surveillance work and whether macroprudential indicators could be included in the IMF’s Special Data Dissemination Standard or alternative vehicles. In the individual presentations, participants reported their experiences in identifying, collecting, using, and disseminating macroprudential indicators. The private sector participants contributed valuable insights drawn from their work in monitoring systemic financial sector risks as a basis for investment decisions.
Assessing financial sector soundness
There was broad agreement among participants that knowledge in the area of macroprudential indicators is still limited. In particular, there is no consensus yet on a model for determining the vulnerability of a financial system or on a set of universally accepted macropru-dential indicators. Using a single composite indicator was considered simplistic and even potentially misleading. Furthermore, analyses of financial sector vulnerability cannot rely on quantitative indicators alone. Qualitative information on institutional circumstances and informed judgments are also essential.
The meeting highlighted several important considerations that national authorities should bear in mind when monitoring and assessing financial sector vulnerabilities. These include the relative importance of macroeconomic variables and aggregated macropru-dential data as well as the role of cyclical factors in interpreting changes in macroprudential conditions. Participants also emphasized that financial institutions tend to exhibit herding behavior based on bail-out expectations and tend to underestimate the likelihood of shocks.
Selecting and measuring indicators
On specific issues related to the selection and measurement of macroprudential indicators, participants broadly agreed on the need to improve the quality of accounting practices in many countries, assess non-bank financial institutions and the health of the corporate sector, address the limitations of aggregating microprudential information to obtain macropruden-tial indicators, and develop benchmarks and norms for the indicators. In light of the complex questions raised about defining the scope of the work on macropruden-tial indicators and ascertaining the technical feasibility of compiling them, the group supported undertaking a survey of national supervisors, statistical authorities, and data users.
In his summary, Stefan Ingves, Director of the Monetary and Exchange Affairs Department, noted that macroprudential analysis requires well-developed linkages between theory and practice, between supervisors responsible for individual institutions and central banks responsible for systemic stability, and between a wide variety of backward- and forward-looking indicators for financial sector health. He emphasized the importance of stress testing in forward-looking approaches to macroprudential analysis. He also underscored the need for better indicators of asset price booms and cycles in order to understand their role in causing financial distress, the need for improved availability of data on corporate and household indebtedness and balance sheets, and the need for a better understanding of contagion issues.
Carol Carson, Director of the Statistics Department, indicated in her summing up that, given the substantial work that would be required to craft a core set of macroprudential indicators, it was premature to include the indicators within the Special Data Dissemination Standard. In light of a consensus that the public needs better access to information on financial sector soundness, Carson noted that consideration should be given to creating incentives for national authorities to compile and disseminate macropruden-tial indicators. Private sector participants favored the compilation of internationally comparable sets of data to facilitate cross-country comparisons. In this connection, they agreed on the need to establish a process leading to greater harmonization of data over time.
The comments received during this consultative meeting will be incorporated in a paper on macropru-dential indicators and data dissemination issues that will be presented to the Executive Board for discussion in October 1999.