CEO Personality and Firm Policies

Steve Kaplan, University of Chicago Booth

Anastasia Zakolyukina, University of Chicago Booth 

Recent research in finance and economics has explored the relations between executive traits and firm policies. One stream of research, led by Bertrand and Schoar (2003), has used manager fixed effects to discern the effect that managerial “style” has on firm outcomes without a limited focus on specific executive characteristics. Another stream of research has taken a more direct approach to executive characteristics, with a focus on risk aversion, time preference, ability, and optimism or overconfidence (e.g., Malmendier et al., 2005, Kaplan et al., 2012, Graham et al., 2013). At the same time, psychology research has somewhat converged on a model that posits the existence of five personality factors (the Big Five), which provide a basis for more thoroughgoing analysis of the relations between executive traits and firm outcomes. Personality psychology views traits as the “relatively enduring patterns of thoughts, feelings and behaviors that reflect the tendency to respond in certain ways in certain circumstances.” (Roberts, 2009, p.140). One challenge to examining the relation between executives’ personality traits and firm outcomes is that measurement of such traits typically involves administration of costly instruments or detailed interviews, which is unlikely to be feasible even if executives were in principle willing to share such information with researchers. In this project, we use linguistic features from conference calls, including those identified in prior research on language and personality, along with the personality data from O’Reilly et al. (2014) and Kaplan, Klebanov, and Sorensen (2012), as training data for a statistical learning algorithm that seeks to classify CEOs in terms of personality traits. Having developed a model for each Big Five trait, we then use data on linguistic features of CEOs’ utterances on conference calls to extend our measure of personality to 65,759 conference-call-level observations on 4,372 CEOs for whom we do not have personality data. We then examine relations between this measure of personality and a variety of firm policies and outcomes. The policies and outcomes we consider follow Bertrand and Schoar (2003), who examine the effect of managerial style of firm policies.