Effect of managerial ability on the investment policy of overconfident chief executive officers

Previous research documents that overconfidence is a double-edged sword to firms. In this study, we investigate the relationship between managerial overconfidence and managerial ability. This study examines whether the successes of overconfident managers are attributed to their superior ability or pure luck because both managerial overconfidence and ability would be related to superior past performance. We adopt the widely used measure, the delay exercise behavior of chief executive officer (CEO) options, to proxy for managerial overconfidence (Hirshleifer, Tech, and Low (2012)). Managerial ability, according to Demerjian, Lev, and McVay (2012), is based on the managers’ efficiency in transforming corporate resources into revenue. They use data envelopment methods to compute the measure. First, it calculates total firms' efficiency to generate total revenue using a well-known set of inputs and divides them into groups based on the characteristics of the industry. However, total firm efficiency might associate with the firms' resources and manager's ability. Hence, they separate the total efficiency between firm and manager by regressing total efficiency on six firm-related variables that definitely influence firm-specific efficiency and included the year fixed effect as well. Then, they classify that the unexplained part, the residual, of the regression belongs to the managerial ability score. The data on managerial ability is publicly available and can be downloaded from Demerjian’s website. Both of these measures are robust and could not be explained by alternative explanations. The data format in this paper is panel data and the methodology used in this paper is the fixed-effects model. The empirical results suggest that overconfident CEOs tend to invest more. Superior managerial ability can restrain their overinvestment behavior but the correction effect is not strong enough. From the perspective of the risk-taking behavior of overconfident CEOs, the influence of managerial ability mitigates the increased risk weakly. Moreover, we find that overconfident CEOs with higher ability increase the value of capital expenditures and therefore enhance firm value. Banerjee, Humphery-Jenner, and Nanda (2015) find that passage of the Sarbanes-Oxley Act (SOX) and Kolasinski and Li (2013) also suggest that better governance structure, defined as a moderate board size (board size between 4 and 12 people) and independent director-dominated board can restrain overconfident CEOs. Therefore, we investigate whether boards of directors would select more capable managers after passage of the SOX or better corporate governance structure. However, we find neither selected overconfident CEOs with higher ability.