Are Extensive Audits 'Good News'? Market Perceptions of Abnormal Audit Fees and Fair Value Disclosures

Abstract

Although the role of auditing is to increase the reliability of financial statements, surprisingly little is known about addressees' perceptions of the auditor-client relationship. Using a sample of more than 1,000 U.S. bank-years from 2008 to 2011, we analyze the economic consequences of the joint announcement of audit fees and the level breakdown of fair value assets. We confirm prior findings that audit fees are higher for banks with larger proportions of Level 3 fair values. Moreover, we find that the market perception regarding fair value assets depends on the extent of managerial discretion in the estimation process. Most importantly, we find evidence consistent with the notion that addressees' interpret abnormal audit fees as an indication of additional risk. Thus, unexpected audit fees are not interpreted as a signal for a more reliable audit, but further increase the market discount on Level 3 fair values. Bank managers have to consider these negative consequences in their reporting strategy.

Client's business risk, public‐interest entities, and audit fees: The case of German credit institutions

“…The first industry‐specific studies report mixed evidence on whether an association between audit fees and a client's business risk is empirically observable (Hill, Ramsay, & Simon, ; Stein, Simunic, & O'Keefe, ). However, recent research (Boo & Sharma, ; Chen, Lam, Smieliauskas, & Ye, ; Cullen et al, ; Doogar et al, ; Ettredge, Xu, & Yi, ; Kanagaretnam, Krishnan, & Lobo, ; Kanagaretnam, Krishnan, Lobo, & Mathieu, ; Mohrmann, Riepe, & Stefani, ) broadly supports the findings of Fields et al (), which demonstrate that a client's business risk—as reflected by the bank's credit risk, market risk, operating risk, liquidity risk, and capital risk—is significantly associated with audit fees (Fields et al, ) . Further analyses consider the audit pricing of European banks (e.g., Altmann, ; Cameran & Perotti, ; Sipple, ), and they also find empirical support for such an association.…”