Complementary association between real activities and accruals-based manipulation in earnings reporting

Abstract

Extant studies focus on examining the developed capital markets and reveal mixed results for the association between real activities versus accruals-based earnings management. This study establishes a set of simultaneous equations that captures managers’ behaviors to boost (or suppress) earnings performance to examine whether real activities and discretionary accruals play mutually complement roles in earnings reporting in Taiwan. The two-stage least squares regressions results reveal that the realactivities manipulation comprehensive measure is positively associated with discretionary accruals and supports the complement hypothesis. It suggests that managers jointly and simultaneously use these two tools in strategic earnings reporting decisions.

Keywords:

• complement hypothesis
• real activities manipulation
• discretionary accruals

JEL Classification:

• M40
• M41

Acknowledgments

The authors wish to express their heartfelt thanks to Professor Michael Connolly for his helpful comments and suggestions. The first author acknowledges the financial support generously provided by National Science Council in Taiwan (Grant number: NSC100-2410-H-224-007-MY2).

Notes

¹Duh et al. (2009) documented that the auditor litigation costs in Taiwan are lower than those in other developed economies. Thus, both the lower disclosure requirements and auditor litigation costs form a somewhat relative weaker financial reporting costs environment.

²The maximum VIF value for all equations is 1.560 for the NI variable in the AM equation.

³This Hausman test result is inconsistent with the finding of Zang (2007) who find that managers determine real manipulation before accrual manipulation. The result, to some extent, supports the conjectures that the reporting environment and/or litigation costs bring about distinct usage of RM and AM in managers’ strategic earnings reporting decisions.

⁴We winsorize the top and bottom 0.5% of outlier samples based on the RM and AM variables and re-examine the equations. The sample reduces to 9663 observations. The complement role between RM and AM in managers reporting decision is evidenced by the refined samples.

⁵The PROCOMP scandal in June 2004 aroused public awareness of auditor behavior in providing audit service. The severe and fast sanctions by FSC and the litigations against PROCOMP’s auditors were unprecedented in Taiwan. These may have led auditors to place more consideration on litigation risk and reputation loss inducing conservative auditing. Thus, the PROCOMP event may change the managerial earnings reporting behaviors.