盈余管理
大裂谷
社会资本
高管薪酬
业务
收益
货币经济学
经济
会计
财务
公司治理
天文
社会科学
物理
社会学
作者
Paul A. Griffin,Hyun A. Hong,Liu Yun,Ji Woo Ryou
标识
DOI:10.1016/j.jcorpfin.2021.101920
摘要
We examine the role of CEO social capital as an important driver of the widespread practice of real earnings management ( REM ). Using the number of social connections to outside executives and directors to measure CEO social capital, we first find that well-connected CEOs associate with higher levels and volatilities of REM. The positive relation between REM and CEO network size is stronger when the CEO connects with more informed and influential persons, and when a more severe misalignment of interests can occur. Second, we find a contagion of REM among well-connected CEOs in an industry. Third, the level of REM induced by a large CEO social network associates negatively with future operating performance. This result is consistent with social capital circulating REM -related information ex-ante and increasing the power and influence for the CEO to deviate from optimal operating policies ex-post. Social capital shields the well-connected executive in the takeover and labor markets despite possible suboptimal future operating performance . While the prior literature finds that CEO social capital reduces accrual earnings management, our findings suggest a dark side of CEO social capital: it induces excessive levels and volatilities of REM costly to the firm in the long run while imposing relatively low personal risk on the top executive. • Well-connected CEOs associate with higher levels and volatilities of real earnings management. • Real earnings management is contagious among well-connected CEOs in a given industry. • Higher levels and volatilities of real earnings management induced by a large CEO network associate negatively with firm future operating performance. • CEO social capital has a dark side for corporate financial policy. It induces an excessive amount of real earnings management costly to the firm in the long run.
科研通智能强力驱动
Strongly Powered by AbleSci AI