Opportunistic earnings management or performance-related effects? Evidence from dividend-paying firms

Abstract

We propose that much of the variation in standard accruals and real-activities earnings management metrics can be explained by firms' performance trajectories. We test our proposition using dividend change as a natural setting that distinguishes high from low performance trajectory firms. Consistent with our proposition, we find that standard earnings management metrics have a stronger relation with performance trajectories than unexpected earnings at which firms' earnings management activities are supposedly targeted. Additionally, we find that firms with higher values of earnings management metrics, i.e., firms that appear to manage earnings to a greater extent, are more likely to increase their dividends; and standard earnings management metrics are unimportant in explaining changes in firm value around dividend change announcements. These results are consistent with the notion that applying standard earnings management metrics without taking performance trajectories into account can result in researchers mistaking managers' efforts to increase firm value with evidence of earnings management.