The Consequences of Audit-Related Earnings Revisions

2017 
In this paper, we investigate the consequences that auditors and their clients face when earnings announced in an unaudited earnings release are subsequently revised, presumably as a result of year-end audit procedures, so that earnings as reported in the 10-K differ from earnings as previously announced. Specifically, we examine whether the likelihood of an auditor “losing the client” is greater following such revisions, and whether the likelihood of dismissal is influenced by revisions that more negatively impact earnings, that cause the client to miss important earnings benchmarks, by greater local auditor competition, or by auditor characteristics. We also examine audit pricing subsequent to audit-related earnings revisions for evidence of pricing concessions to retain the client. Finally, we examine whether client executives experience a greater likelihood of turnover following an audit-related earnings revision. Consistent with expectations, we find that auditor dismissals are more likely following audit-related earnings revisions. We also find that dismissals are more likely when revisions cause clients to miss important benchmarks and when there is greater local auditor competition. Among non-dismissing clients, we find that future audit fees are lower when the effect of the revision on earnings is more negative, consistent with auditors offering price concessions to retain clients when revisions are more displeasing. We also find a greater likelihood of future chief financial officer (CFO) turnover as the effect of the revision worsens. Our findings offer important insights into the consequences that auditors face when balancing their responsibility for high audit quality and client satisfaction, as well as into the consequences that CFOs face when releasing inflated but not fully audited earnings. This article is protected by copyright. All rights reserved.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    48
    References
    16
    Citations
    NaN
    KQI
    []