Using Annual Report Sentiment as a Proxy for Financial Distress in U.S. Banks

abstract

Current measures of bank distress find marginal value in predictive variables beyond a capital adequacy ratio and tend to miss extreme events impacting the entire sector. The authors advocate a new proxy for bank distress: sentiment measures from banks’ annual reports. After controlling for popular forecasting variables used in the literature, they find that more negative sentiment in the annual report is associated with larger delisting probabilities, lower odds of paying subsequent dividends, higher subsequent loan loss provisions, and lower future return on assets. The findings suggest that regulators could augment current early warning systems for banks and the banking sector—where the measures are based exclusively on financial statement data—by using the frequency of negative words in banks’ annual reports.

Keywords:

• Financial distress
• Textual analysis
• Negative sentiment
• Distressed delisting
• Financial institutions

Acknowledgments

The authors thank Thomas Cosimano, Paul Gao, John Holland, Peter Kelly, Jerry Langley, an anonymous referee, and seminar participants at the University of Miami, HKUST, University of Minnesota, and York University for helpful comments.