مدیریت درآمد پس از ناپدید شدن گسستگی سود صفر Earnings management in the aftermath of the zero-earnings discontinuity disappearance

1. Introduction Detection of earnings management has long been the subject of heated debate in the accounting literature. In order to uncover earnings manipulation, a common methodology is to associate it with a motivation for earnings management. Earnings management induced by management compensation packages (e.g. Healy, 1985; Skinner, 1993; Baker et al., 2003; Cheng and Warfield, 2005), debt contracts (e.g. Defond and Jiambalvo, 1994; Sweeny, 1994; Dichev and Skinner, 2002; Gupta et al., 2008), targets set by industries (e.g. Beaty et al., 1995; Petroni, 1992), earnings benchmarks (e.g. Burgstahler and Dichev, 1997; Burgstahler and Eames, 2006; Daske et al., 2006; Gunny, 2010; Das et al., 2009), and cost of capital (e.g. Dechow et al., 1996; Strobl, 2013; Kim and Sohn, 2013) is established in the literature. For decades, the attempts to document earnings management had substantially focused on accruals and discretionary accruals, but a study by Burgstahler and Dichev (1997) embarked on a different route. Instead of calculating discretionary accruals to observe earnings management, they employed natural earnings benchmarks such as avoiding reporting of losses and earnings decreases in order to uncover earnings manipulation. Through this approach, manipulation of earnings is inferred from the pattern of earnings distribution. To this end, the frequency of earnings level or earnings change scaled by a size variable such as total assets or market value is plotted and observations are grouped into intervals according to their level of scaled net income. Any distributional irregularity is considered as evidence of earnings management. The only assumption here is that “under the null hypothesis of no earnings management, the cross-sectional distribution of earnings changes and earnings levels are relatively smooth” (Burgstahler & Dichev, 1997, p. 102). Simply stated, it is generally expected that in the absence of earnings management the distribution of earnings would be symmetric. Therefore, if earnings are manipulated to reach an earnings target, it is expected that there will be “too few” frequency of earnings sitting just below the target and “too many” at or above it (Degeorge et al., 1999). Comparing discretionary accruals of small profit and small loss firms, Dechow et al. (2003) report that the discontinuity in the earnings distribution cannot be associated with accruals management. Hansen (2010) points out that the reason behind earnings management not differing between small loss firms and small profit firms could be that loss avoidance is not the only motivation for earnings management and that incentives other than reporting positive earnings, such as to meet/beat analysts’ forecasts or to continue reporting earnings growth, could also encourage earnings management.