발생을 통한 이익조정과 실제활동이익조정 간의 상충관계에 대한 재조명：재벌과 비재벌을 중심으로

[Purpose] The purpose of this study is to examine whether the research of Zang(2012) that costs associated with real activities manipulation(RM) affects the trade－off between accrual－ based earnings management(AM) and real activities manipulation(RM) is also applied to Korean Chaebol. [Methodology] The first hypothesis is for firms without market－leader status among the costs associated with RM. As a result of examining the effect of market share on AM, it was found that market share had a significant positive effect on AM in the non－Chaebol, but did not affect AM in the Chaebol. The second hypothesis is for firms with poor financial health among the costs associated with RM. Zang(2012) was reported that firms with poor financial health prefer AM rather than RM, which has a declining corporate value. [Findings] As a result, in the sample of non－Chaebol companies, it was found that the RM level was lower in firms with poor financial health, but it was not found to be related to AM. On the other hand, in the case of Chaebol, even if they are firms with poor financial health, it was expected that the costs associated with RM would be weakened or eliminated by utilizing internal capital market within affiliates. Consistent with this hypothesis, it was found that the degree of healthy financial condition had no effect on RM and AM. [Implications] This study is expected to help regulators accurately understand the earnings management of Chaebol and control earnings management for the pursuit of private interests by controlling shareholders. In addition, it was revealed that Zang(2012)’s research cannot be generalized in Korea due to the special governance structure of the Chaebol. Therefore, it provides a suggestion that Korea’s specificity should be reflected in citing previous studies. However, this study has a limitation in that it was partially modified and studied due to data collection problems and differences in economic environment, as it did not use the same variables and methodologies as those of Zang(2012).