Peer Effect in Bank Loan Accounting

We document new evidence that peer banks play a significant role in determining a bank’s accounting for loan loss provisions. For example, a 10% increase in peer banks’ provisions is associated with a 11.5% increase in a bank’s own provision. Our finding is robust to alternative classification of peer banks, various model specifications, and exogenous shock of regulatory scrutiny. With respect to heterogeneity of peer effect, we show that the peer influence is more pronounced for banks with less capable managers, more homogeneous loans, and when banks are under greater regulatory scrutiny. Channel tests indicate that peer effects work through observational learning. Finally, our analyses of economic consequences suggest that peer effects increase the timeliness of bank accounting and decrease bank risk.