This paper studies the information content of bank accounting fundamental data in the prediction of bank distress using an international sample of banks from 15 Western European countries and the U.S. during the financial crisis of 2007-12. We assemble an exhaustive and unique set of bank distress events, and model bank distress as a function of accounting-based fundamentals, while controlling for country-year fixed effects, and the type of resolution in the distressed entity. The analysis of our bank distress models reveals a substantial cross-country variation in the ability of accounting fundamentals to discriminate between distressed and non-distressed banks within countries. We examine the extent to which the variation in informativeness and accuracy of accounting fundamentals is explained by proxies of country-specific bank disclosure requirements and the enforcement thereof. We show that the association between accounting fundamentals and bank distress is attenuated in jurisdictions with relatively lax bank disclosure laws and their implementation. Accounting ratios, whose information value is the most sensitive to the quality of regulatory disclosure include regulatory capital ratios, loan loss provisions, and unreserved loan losses. The evidence in this paper supports the oft-voiced concern that excessive flexibility in financial reporting undermines the ability of accounting signals to accurately capture the underlying financial health of banks. Obliqueness of the distressed bank’s accounting signals makes such information less useful for investors and regulators, and thus has negative regulatory implications.

Discussant: Mark Dijkstra (University of Amsterdam)