The impact of executive stock options on managerial risk taking

I examine the hypothesis that executives undertake more risky investment projects after they are awarded company stock options by testing the ratio of post-award volatility of return on investment to pre-award volatility of return on investment (referred to as the risk relative), and volatility increase from pre-award period to post-award period (referred to as the risk difference). Compared to the literatures, my test design has two advantages. First, I examine the impact of stock options on the risk profile of the firm's capital investment projects directly. Second, I allow for the “time-to-build” characteristic of most investment projects by examining the risk relative based on a variety of lags between option awards and the change in volatility of ROI. The results exhibit a significant increase in investment risk for firms granting stock options to executives during the period from 1992 to 1995. To address the argument that risk increases are due to other economic factors, three control methodologies are employed. First, the nonparametric two-sample tests are done to test if the risk increases for the option awarding firms are higher than the risk increases for the zero option awarding firms. Second, volatility of return on investment is adjusted with the industry median. Third, the volatility change from period to period is analyzed. The results all appear to support the hypothesis that executives undertake more risky investment after they are awarded stock options. In addition, I investigate the relation between the risk relative (risk difference) and a measure of incentive effects of stock options (referred to as the incentive ratio). The evidence shows a positive relation between the risk relative (risk difference) and the incentive ratio, implying that executives undertake more risky projects when their compensation contain a higher proportion of stock options. The causal link between higher option compensation leading to higher risk increase is further examined in terms of firm-specific factors, including firm size, investment opportunity, and managerial discretion. For example, without managerial discretion, option awards may not really impact managerial risk taking. The evidence from statistical tests and three different regression models (logit, OLS, and separate-slopes) supports the hypothesis that the relation between option compensation and capital investment risk is affected by firm-specific factors. The separate-slopes model also indicates a significant industry effect in the relation of option compensation to capital investment risk choice. In addition, the evidence of two more logit regression models validate the lag consideration in test designs. I validate the importance of the time-to-build characteristic, since in the absence of lags there is no persistent relation between option compensation and risk taking. Finally, as a robustness check, I adjust the each firm's volatility with autocovariance, and re-examine the relation between option compensation and risk-taking with the estimates of autocovariance-adjusted standard deviation. The results support the hypotheses that the relation of higher option compensation and lagged impact on capital investment risk is constrained by firm-specific factors.