Real earnings management and future profitability: Measurement of real earnings management's current net income impact, cross-sectional variation in real earnings management's effect on future profitability, and market participant response

Prior research demonstrates real earnings management (REM) is associated with decreased future profitability (Cohen and Zarowin 2010; Gunny 2010) and that there is cross-sectional variation in this relation (Gunny 2010; Zang 2012). This study makes three contributions to the REM literature. First, the cross-sectional variation of REM's effect on future profitability is examined for different REM methods and different firm characteristics. While a firm can increase current net income by one dollar through either overproducing inventory or decreasing discretionary spending, the effect on future profitability is not necessarily the same for the two REM methods (Cohen and Zarowin 2010). To compare the impact on future profitability in a general setting, traditional REM measures are transformed to their current net income effect and included in the same model. This study's evidence is mixed regarding REM methods; compared to discretionary spending reduction REM, overproduction REM leads to more detrimental net income in some but not all settings. This study provides compelling evidence that the cross sectional variation of REM's effect on future profitability can be explained by the following firm characteristics: inventory turnover, risk of inventory obsolescence, financial health, competitiveness within an industry, equity overvaluation, corporate governance, and external monitoring. Second, investor and analyst responses to REM are examined. To the extent that observable firm characteristics are associated with REM that is more detrimental to future profitability, market participant responses to REM should depend upon these characteristics. However, REM can be difficult for market participants to unravel (de Jong et al. 2012) and therefore reaction may be delayed. This study provides some evidence that stock returns and analyst earnings forecast revisions for REM depend upon firm characteristics, but the evidence is primarily for stock returns in the subsequent year. Finally, this study provides methodological adaptations for measuring REM on a net income basis that allows REM to be examined in broader contexts.