This paper measures the performance of different metrics in assessing banking system vulnerabilities. It finds that metrics based on equity market valuations of bank capital are better than regulatory capital ratios, and other metrics, in spotting banks that failed (bad apples). This paper proposes that these market-based ratios could be used as a surveillance tool to assess vulnerabilities in the banking sector. While the measures may provide a somewhat fuzzy signal, it is better to have a strategy for identifying bad apples, even if sometimes the apples turn out to be fine, than not being able to spot any bad apples before the barrel has been spoiled.

• Cover
• Contents
• I. What Is the Best Way to Find Bad Apples?
  • A. Regulatory and Market-Based Capital Ratios
  • B. Using Equity Prices to Signal Distress
  • C. A Proposal
• II. Searching for Bad Apples: The Past
  • A. Suggested Metrics
  • B. Would These Metrics Have Helped Identify Problem Banks in the Pre-Crisis Period?
• III. Searching for Bad Apples, Reprise
• IV. Searching for Bad Apples: The Present
• V. What Have We Learnt from the Search?
• Tables
  • 1. Sample of Banks
  • 2. Proposed Metrics
  • 3. Median Value for the Proposed Metrics: Surviving and Failing Banks
  • 4. Area under the Receiver Operator Curve for Different Combinations of Metrics
  • 5. Thresholds and Associated Performance Statistics for Different Metrics
  • 6. Out of Sample Metric Statistics
• Figures
  • 1. Proposed Metrics: Sample Medians and Distributions in the Pre-Crisis Period
  • 2. Proposed Metrics: Time Series of Sample Distributions in the Pre-Crisis Period
  • 3. Receiver Operator Curves for Different Metrics
  • 4. Area Under the Receiver Operator Curve for Different Metrics
  • 5. Bank Metrics in the Pre- and Post-Crisis Period
  • 6. Banking Sector Distress Signals, Over Time
  • 7. Advance Warning Given by the Metrics
  • 8. Banking Sector Distress Signals, by Region
  • 9. Market-Adjusted Capital Ratios, May 2019
• References