The impact of Chief Financial Officer overconfidence on corporate decisions

This thesis investigates the effects of Chief Financial Officer (CFO) overconfidence on corporate decisions. The thesis specifically explores: if overconfident CFOs use earnings management, what influence CFO overconfidence has on conditional accounting conservatism, and how overconfident CFOs impact stock price crash risk. Through the lens of upper echelons and overconfidence theories and based on a large sample size of U.S.-listed firms, this thesis finds that overconfident CFOs use accrual-based earnings management (AEM) due to their high risk-taking incentives and low career concern. Overconfident CFOs engage in real earnings management (REM) as they underestimate the cost of deviating from optimal business strategies by using REM and overestimate their managerial ability to mitigate this cost of REM (Paper 1). Moreover, overconfident CFOs, particularly those with significant power, tend to reduce conditional accounting conservatism (Paper 2). Furthermore, firms led by overconfident CFOs increase future stock price crash risk (Paper 3). To increase the robustness of these findings, this thesis carries out a series of robustness tests, including using the entropy balance method, conducting difference-in-difference tests based on the propensity score matching sample, and changing the measure of main variables. After that, each paper, relying on power circulation and false consensus effect theories, further discusses the effect of overconfident CEOs on the relationship between CFO overconfidence and corporate decisions. Therefore, this thesis makes incremental contributions to the literature, theory, and practice. Specifically, these findings add a new determinant, CFO overconfidence, to earnings management, conditional accounting conservatism, and stock price crash risk literature. Besides, this study responds to calls from prior studies for more research to provide a deeper insight into CFO overconfidence and the joint effect of CEO overconfidence and CFO overconfidence, extending the overconfidence literature. The thesis also demonstrates the applicability of upper echelons, overconfidence, power circulation, and false consensus effect theories in the CFO overconfidence study. Lastly, this thesis has important implications for academia and practice in recognizing the pros and cons of CFO overconfidence in corporate decision