The authors study the relationship among CEO personal risk taking, corporate risk taking, and total firm risk. As a proxy for personal risk taking, they use CEO possession of a private pilot license, which gives the authors an observable measure of risk taking that is unrelated to firm activities. Their results demonstrate that CEO tolerance for risk in nonpecuniary contexts has explanatory power for corporate project selection and overall firm risk.

The authors’ results could be relevant to practitioners in several ways. First, the authors find that firms led by pilot CEOs have higher return volatility, which they say can be explained by these firms’ higher acquisition activity. This result is inconsistent with prior research that suggests that pursuing acquisitions is a means to reduce risk.

Second, the results suggest that the acquisition activity of pilot CEOs leads to significant positive value creation, which indicates that pilot CEOs mitigate agency costs. Prior research on agency costs has shown that shareholders bear the cost when risk-averse managers forgo risky projects with positive net present value.

Third, the authors find that pilot CEOs are associated with compensation structures that are more likely to have high performance-based pay. The option-like payoff of such types of compensation contracts benefits from a greater sensitivity to the level of equity volatility.

The authors find CEOs with pilot licenses by identifying an initial sample of 4,012 CEO–firm combinations from the Compustat ExecuComp database for CEOs who took office on or after 1 January 1991. They match those CEO names against those in the Federal Aviation Administration (FAA) Airmen Certification database. Names that match are further matched using personal information on the CEOs from public records to validate the FAA certificate information. The result is a final sample of 179 pilot CEOs (1,016 firm-years) and 2,931 nonpilot CEOs (14,611 firm-years).

Next, the authors perform a series of regressions and use an indicator variable that equals 1 if the CEO has had at least one certificate in FAA records and 0 otherwise. All the specifications contain independent variables that proxy for CEO characteristics and firm characteristics. Many of the specifications use firm, industry, geographical, and year fixed effects, and the results are robust to these specifications. In addition, the authors test the robustness after adding controls for such other managerial characteristics as overconfidence, upbringing during the Great Depression, military experience, tenure, and age. They find no significant differences when including these additional control variables.

The authors acknowledge that flying airplanes is relatively rare. As such, their results depend on how well pilot licenses serve as a proxy for personal risk taking. Moreover, measuring an individual’s appetite for risk is challenging, and it is an issue that arises in other areas of finance as well. Notably, this issue is central to investment management because the individual’s degree of risk aversion is a critical input in determining the appropriate investment policy to pursue.