The Effects of Audit Partners on Financial Reporting: Evidence from U.S. Bank Holding Companies

Abstract

This paper uses confidential data on audit engagement partner names from regulatory filings of bank holding companies (BHC) to investigate whether partners display individual style that affects the financial reporting of the BHCs. We focus on loan loss provisioning. We construct an audit partner-BHC matched panel data set that enables us to track different partners across different BHCs over time. We employ two empirical approaches to investigate partner style. The first approach tests whether partner fixed effects are statistically significant in loan loss provisioning models. The second approach tests whether a partner's history of loan loss provisioning predicts future practices for the same partner. Our empirical evidence does not support systematic differences in loan loss provisioning across audit engagement partners, suggesting that the audit firm's standards and quality control constrain personal partner style.