Auditor Sensitivity to Real Earnings Management: The Importance of Ambiguity and Earnings Context

Although real earnings management (REM) is increasingly common, limited research examines how auditors react when they know or suspect that their clients are using REM. In addition, prior research on whether auditors react to REM is not conclusive, nor does it provide strong insights into why auditors would react to REM, given that REM does not violate GAAP. We conduct a 2x3 between-subjects experiment (client Beat or Missed the analysts’ consensus forecast, with three levels of REM: Explicit REM, Potential REM, and No REM) with highly experienced auditors as participants to examine the conditions in which auditors are sensitive to REM. Informed by Correspondent Inference Theory, we predict and find evidence that, when a client beats an EPS target, auditors observing Explicit REM perceive the weakest management tone and assess the highest levels of risk, followed by those observing Potential REM, and then those observing no REM. However, we predict that, when a client misses an EPS target, auditors are less likely to view Potential REM as actual REM. Thus, we find that, when a client misses an EPS target, auditors observing Explicit REM perceive weaker management tone and assess higher levels of risk relative to when they observe Potential REM or when REM is absent. Further, when clients miss the EPS target, we find that auditor responses to Potential REM do not differ from when REM is absent. Finally, we find that auditors’ perceptions of management tone mediate the relation between REM and risk assessments. Overall, the results are consistent with Correspondent Inference Theory and demonstrate how and why auditors respond to REM.