2013年-ECB欧洲央行_Predicting_Financial_Vulnerabilities_to_Guide_the_Set-Up_of_Counter-Cyclical_Capital_Buffers_11页_1mb

Summary of "Predicting Financial Vulnerabilities to Guide the Set-Up of Counter-Cyclical Capital Buffers"

Core Content

This document explores the use of macro-financial and banking sector indicators to predict financial vulnerabilities and guide the setting of counter-cyclical capital buffers (CCBs). It emphasizes the need for a multivariate early warning model framework to better assess systemic risk in the financial cycle, especially in the context of the EU's Capital Requirements Directive (CRD) IV and the Basel Committee on Banking Supervision (BCBS) guidelines.

Main Viewpoints

• Systemic Risk and CCBs: The financial crisis highlighted the importance of counter-cyclical capital buffers to increase the resilience of the banking system against systemic risk. These buffers are designed to be built up during periods of credit expansion and drawn down during financial stress.

• Credit-to-GDP Gap as a Key Indicator: The credit-to-GDP gap, defined as the deviation of the credit-to-GDP ratio from its long-term trend, is considered a strong early warning signal. However, it may not be early enough for policy implementation due to its limited ability to capture turning points.

• Multivariate Analysis: The document advocates for a multivariate approach to early warning modeling, which includes not only credit variables but also other macro-financial and banking sector indicators such as equity prices, house prices, nominal GDP growth, and inflation.

• Global vs Domestic Variables: Global variables, such as global credit growth and global credit-to-GDP gap, tend to outperform domestic variables in terms of predictive power. This is attributed to the increasing integration of the global financial system, where vulnerabilities in one region can rapidly spread to others.

• Evaluation Metrics: The effectiveness of indicators is evaluated using several metrics, including relative usefulness, adjusted noise-to-signal ratio, percentage of vulnerable states predicted, and conditional probability of a crisis.

Key Information

Vulnerable State Definition

• A vulnerable state is defined as the period between twelve and seven quarters before the onset of a systemic banking crisis.
• This period allows for a sufficient time lag for policy implementation.

Data and Methodology

• The analysis uses a panel dataset of macro-financial and banking sector data from 23 EU Member States, covering the period from the second quarter of 1982 to the fourth quarter of 2010.
• Data sources include the Bank for International Settlements (BIS), Eurostat, the International Monetary Fund (IMF), the Organisation for Economic Cooperation and Development (OECD), and the European Central Bank (ECB).
• The models are estimated in a quasi-real-time framework, using backward-looking trends to ensure the use of only available information up to a given point in time.

Key Indicators and Their Performance

• Domestic Credit-to-GDP Gap: Performs best among domestic indicators, with a relative usefulness of 0.256, correctly predicting 81.3% of vulnerable states.
• Domestic Credit Growth: Also shows strong performance, though slightly less than the credit-to-GDP gap.
• Global Credit-to-GDP Gap: Outperforms domestic variables, with a relative usefulness of 0.443 and a higher percentage of vulnerable states predicted.
• Interaction Terms: Include combinations of domestic and global credit variables, as well as other macro-financial indicators, to capture complex relationships between variables.

Policy Implications

• Policy-makers should consider a wide range of indicators when setting CCB rates, not just the credit-to-GDP gap.
• The inclusion of global variables may improve the accuracy of early warning signals, especially in the context of global financial crises.
• The evaluation framework helps balance the risk of false alarms and missed signals, ensuring that policy decisions are informed and credible.

Limitations and Considerations

• The BCBS study did not account for the 12-month implementation period, which may affect the timeliness of the credit gap as an indicator.
• Global variables may be influenced by the clustering of crises in the same year, particularly the 2008 crisis, which could skew their predictive performance.
• The models include country fixed effects and lagged variables to account for unobserved heterogeneity and data availability lags.

Conclusion

This special feature demonstrates that while the credit-to-GDP gap is a valuable indicator, a multivariate model incorporating a range of macro-financial and banking sector variables provides a more comprehensive and accurate framework for predicting financial vulnerabilities. It underscores the importance of considering both domestic and global factors in setting CCBs, given the interconnected nature of the global financial system.