A Study on CEO Overconfidence and Corporate Social Performance


Student thesis: Doctoral Thesis

View graph of relations


  • Shen XU

Related Research Unit(s)


Awarding Institution
Award date4 Aug 2014


Despite the fact that CEO overconfidence is acquiring considerable importance, most research has focused on how it affects corporate financial or investment decisions. In this study, I analyze the implications that overconfidence has on corporate social responsibility (hereafter CSR), a non-financial decision. Using financial data and ratings of CSR performance from the KLD STATS database from 1996-2010, I study a large sample of firms in US, to see whether firms with an overconfident CEO (measured by late exercise of options) are more socially responsible.

The main regression shows that CEO overconfidence has a negative influence on CSR performance in general, after controlling for firm-specific factors and CEO-characteristics, which is supportive to the argument that overconfident CEOs are less likely to invest on CSRs to hedge potential future risks. In cross-sectional analysis, the increases of ex-ante litigation risk weaken the negative relation between CEO overconfidence and CSR performance of the firm. I also find that CSR can increase future firm value (as measured by ROA and Tobin's Q) when the firm is managed by overconfident CEO, which implies that firms can benefit from the joint effect of CEO overconfidence and CSR because CSR partly mitigates potential loss from future risk inherent in overconfident CEO's investment decisions. In addition, I show that firms try to improve their CSR performance after they experience significant bad events such as financial restatement announcement. Finally, a number of robust checks are conducted. I exclude first years of CEOs' tenures to partly address the self-selection concerns. The main results are not affected.