5. Discussion and Conclusion
This study provides more insight into a recent finding by Gilliam et al. (2015) concerning the disappearance of zero-earnings discontinuity after the passage of the Sarbanes-Oxley Act. We suggest that benchmark-beating earnings management could have contributed to the disappearance of the discontinuity. Given the evidence that benchmark-driven manipulation has decreased due to the recent market scepticism about beating benchmarks by a slight margin (Koh et al., 2008; Bartov and Cohen, 2009) which implies less attention is attracted by firms that miss earnings benchmarks, firms with a pre-manged loss or earnings decrease could enjoy a quiet room to improve their earnings up to just behind earnings benchmarks. That is, restricting benchmark-beating earnings management could have two effects: (1) a decrease in the frequency of beating benchmarks by a small margin, and (2) an increase in the frequency of missing benchmarks by a small margin. This in turn could narrow the gap between the number of small profit and small loss firms resulting in the fade of the discontinuity. Therefore, the disappearance of the discontinuity could be due to upward earnings management by small loss firms. Our findings are consistent with this and suggest further that small loss firms and small profit firms show a very similar earnings management behaviour both in terms of accruals management and real activities manipulation. This finding also explains why prior studies could not find any significant difference between accruals management of small profit and small loss firms (Dechow et al., 2003). Despite the prevalent notion in the literature that small loss firms are less likely to manipulate earnings, they are actually not. In other words, failure to document a difference between firms to the right and left of zero is not due to small profit firms not managing their earnings, but it is instead due to small loss firms being similarly engaged in upward manipulation.
