Bank Insolvency Risk and Z-Score Measures: Caveats and Best Practice
35 Pages Posted: 5 Jan 2017 Last revised: 27 Jul 2018
Date Written: July 18, 2018
We highlight caveats arising in the application of traditional ROA-based Z-scores for the measurement of bank insolvency risk, develop alternative Z-score measures to resolve these issues, and make recommendations for best practice for the US/Europe based on the experience of the financial crisis of 2007-2008. Using a probabilistic approach (i) our novel regulatory capital Z-score dominates traditional Z-score measures for both US/Europe; (ii) Z-scores computed with exponentially weighted moments dominate those with moving moments for the US sample, but not for Europe. For both US/Europe, using a multivariate logit approach (i) allows computation of augmented Z-scores that provide probabilities of distress that better discriminate between distressed/surviving banks than the probabilistic approach; (ii) suggests that the ROA-based Z-score using current values of the capital-asset ratio is best, calculated either with moving or exponentially weighted moments.
Keywords: Z-score; bank; insolvency risk; risk measure
JEL Classification: G21, G28, G32, G33
Suggested Citation: Suggested Citation