Summary of Key Points Made at the Users Conference on the Financial Crisis and Information Gaps
In March 2009, the Group of 20 (G-20) Working Group 2 on "Reinforcing International Cooperation and Promoting Integrity in Financial Markets" recommended that the IMF and the Financial Stability Board (FSB) identify data gaps and provide appropriate proposals for strengthening data collection before the next meeting of G-20 Finance Ministers and Central Bank Governors. This recommendation was endorsed by the IMF's International Monetary and Financial Committee at its Spring Meetings in April 2009.
In order to ensure these efforts are informed by users, the IMF and the FSB hosted a users conference during July 8-9, 2009 in Washington, D.C. International experts on financial stability and statistics from national authorities (primarily the G-20) and international agencies attended. The purpose of the conference was to solicit views of users as to which data gaps were most relevant in the build up to the crisis, and to inform the staff of the IMF and the FSB Secretariat as they prepare their report for G-20 Finance Ministers and Central Bank Governors.
There was broad agreement that to meet the challenges ahead there needs to be a strengthening of available data to enable a better, forward-looking and targeted identification of risks for global financial stability. There was also an understanding that designing and implementing changes to data collection frameworks are costly and that priorities must therefore be established among the various potential improvements. The key points made at the User Conference are presented below.
- There is a need to strengthen the analytical/conceptual framework for financial stability analysis and the global monitoring of financial stability risks. A more robust conceptual framework will help clarify data priorities, while better data will contribute to testing and refining the framework.
- As part of strengthening the conceptual framework, the evidence of increasingly global financial transmission mechanisms and strong feedbacks between the financial system and the real economy are considered very important topics for further investigation. This in turn reinforces the importance of international and inter-agency co-operation and information-sharing in data collection, and in making progress in data harmonization.
- There was a call for greater coordination of data collection between supervisory authorities and economic statisticians. Encouraging the large number of agencies that collect data to work together can pay important dividends in terms of reducing the burden on business of providing data, and in terms of better harnessing analytical synergies. Some countries noted their positive experience in this regard.
- Data gaps should be prioritized based on cost benefit analysis, especially considering the need for capacity building in some countries.
- As the inter-connected financial sector was the main bearer and carrier of risks globally during the recent crisis, and given that this transmission of risks tended to be concentrated in a relatively small number of systemically important global financial institutions, there was seen to be a clear need to monitor their exposures and activities at the global level. One idea was for an international agency to compile data on the largest institutions (such as the top 50) for such a purpose. Acknowledging confidentiality issues and legal restrictions on the sharing of information faced by some national supervisors, one option would be to rely on publicly available information, which would put the emphasis on strengthening consistency in financial accounting standards and disclosures. The specific data series required would need to be investigated.
If common disclosure or reporting templates are used across countries this would facilitate aggregation, institutional comparison, measurement of concentration, and identification of network links. It was noted that measuring exposures implies many difficult issues, but would in return offer benefits in understanding patterns of activity across markets and institutions.
- On Financial Soundness Indicators (FSIs), there was support for the compilation of these data, their dissemination as part of the data needed for financial stability analysis, and maintaining their principal focus on banks. Different indicators performed differently during the crisis in terms of providing warning signals across countries. Also, there were some ideas for reprioritization. A cautious review of the list is advisable. Expectations for near-term progress on FSIs were discussed, including work currently underway to incorporate them into Special Data Dissemination Standard (SDDS) and to expand country coverage to encompass all G-20 members.
- The capturing of data on the shadow banking sector was stressed, broadly defined as financial intermediation by institutions, markets, and products outside of the banking sector and traditional securities markets. This definition covers nonbank financial institutions (NBFIs), such as hedge funds and SIVs, financial products such as asset backed securities, and markets such as repo markets.
More, and more timely, data are needed on the size and risk exposures of NBFIs, on their relationship and connections with banks, and on international capital flows and cross border investments for this heterogeneous group of institutions. As a start, systemically important NBFIs could be required to disclose assets, capital, maturity mismatches, and large exposures. Work is progressing in a number of jurisdictions to monitor hedge funds.
- One of the practical data issues that need to be addressed is the representation of financial institutions in any analytical framework. Specifically, while for some analytical purposes there is a need to treat the activities and obligations of large conglomerates on a consolidated basis, it is also important to acknowledge institutions as complexes of organizations (i.e., a holding company and multiple subsidiaries). This is particularly relevant when performing firm-level analysis related to derivatives (e.g., analyzing sectoral trends, cross-border credit-risk transfer, etc.).
- The need to monitor leverage across the system was raised. Balance sheet data are a good starting point but account also needs to be taken of leveraged products, such as derivatives and securitized products, and off balance sheet items and contingent obligations, such as contingent liquidity lines of credit.
- Data on cross-border exposures and foreign activities of local financial and nonfinancial entities were another area of concern. Despite improvements in home-host arrangements and supervisory colleges, some participants still felt not adequately informed on the activities of large institutions headquartered abroad.
Sectoral and Other Economic Data Sets
- The need to monitor the vulnerabilities of the non-financial corporate sector and households was stressed. The Balance Sheet Approach is considered a good starting framework, with its focus on maturity, leverage, and currency composition. Broader country coverage, including emerging market economies, appears to be necessary for many categories of non-financial sectoral data.
- The increasingly strong feedbacks between the financial system and the real economy are well-covered by integrated-financial and non-financial-sector accounts. But distributional and other indicators are also important to allow potentially informative tail risk analysis to be carried out (e.g., a breakdown of the household sector accounts by income and wealth classes).
- The need for data on external positions and for improved flow of funds data was also noted. The from-whom to-whom cross border data (such as the BIS International Banking Statistics (IBS) and the IMF's Coordinated Portfolio Investment data) are important both to monitor who is investing in the domestic economy and to help track activities of domestic entities in foreign markets. The "hidden" positions of nonfinancial corporates and households in foreign markets have been a source of vulnerability for some economies that has been difficult to monitor.
- Further improvement of BIS IBS data was suggested, including: a further sectoral breakdown for non-bank financials, e.g., hedge funds and insurance corporations, and an improvement in country coverage to include at least all G-20 economies, including the largest emerging market countries.
- The need for information on the ultimate risk of external debt was also raised.
Housing and Related Data
- There was a call for international efforts to improve data on residential and commercial real estate, and the analysis of such data, in view of the importance of understanding and monitoring patterns of the financing of these assets by banks.
Risk Transfers and Markets
- Participants noted the need for better information on the nature of the assets underlying securitized products, and on the risk exposures, both relating to the underlying item and the counterparty (by sector and geography), arising from credit risk transfer instruments, such as credit default swaps. Centralized clearing systems, central data warehouses, and standardized contracts would ease data collection constraints in these areas.
- Better dissemination of market liquidity data should be encouraged, although it was recognized that the past was not necessarily a good guide for the future. A quick win could be through the release of data on volumes in inter-bank markets as these data are readily available to many central banks.
- Although simple averages and ratios of asset prices are not always good signals of building imbalances, they should be monitored as one part of a broader financial stability analysis.
General Comments on Data
- The conference stressed the importance of analyzing tail risks; aggregate data could be misleading as to the vulnerabilities hidden in the financial system. There was a general call for greater granularity/disaggregation of data, while at the same time linking micro data to macro risks.
- Soft signals, especially the existence of practices that are country specific, and market intelligence, are all important for using/interpreting data appropriately.
- The need to protect confidentiality was stressed.
- Participants agreed on the need for data that are comparable across countries and timely, while accepting the limitations imposed by institutional differences.