Nevertheless, our study has several limitations, such as the difficulty in measuring managerial overconfidence, the use of an index-based measure of audit committee strength, and selection bias due to correlated omitted variables. However, our findings are robust across multiple measures of overconfidence and model specifications.