Earnings Management. An overview of the relative literature

Abstract

This article aims to present a critical overview of the traditional studies in earnings management, focusing on the impact on the making decision process. This overview in the literature provides numerous aspects of this topic, in line with the firms’ motivations. Earnings management procedure includes smoothing and opportunistic practices and illustrating accounting rules with a significant effect on accounting information quality. Several researchers have shifted their attention to real activities as a primary method or supplementary to accrual-based methods to obtain a complete view of the earnings management levels. Earnings management is considered an opportunistic instrument, and it can be part of the aggregate long-term business strategy. This review study provides some guidelines, to academics and professionals, in line with the models and the motivations that lead managers to engage in this procedure. This overview creates new research avenues enhancing the existing knowledge. which undoubtedly is a purpose of earnings management activities. The first hypothesis examines the linear relation between earnings management proxies (accrual-based and real earnings) with the firm performance proxy, which is the change of dividends. This stage's results, presenting a significant positive relation with the change of dividends, support that the real earnings management measures provide more information for firm performance trajectories than for unexpected earnings. The second hypothesis is whether the higher levels of earnings management measures are associated with a higher probability of increases in dividends. The logistic regression model uses a dummy variable that controls for firms that present an increase in dividends regressed by earnings management metrics, the unexpected earnings variable. This stage's findings demonstrate a significant positive coefficient of all earnings management measures, which indicates that the firms that intensively engage in earnings management procedures are more likely to increase their dividends. Furthermore, the research states a third hypothesis that examines whether the announcement of dividends change conveys more information to investors than earnings management activities. This regression model's dependent variable is the three-day cumulative stock return around the dividend declaration for the first dividend change of the year. The key independent variables are the change of dividends and earnings management measures, while the set of control variables of the previous stages is also included. The findings are shown to be consistent with the hypothesis and, in particular, indicated a significant positive coefficient of dividend change and an insignificant coefficient of earnings management metrics.