This paper develops a model for the valuation of executive stock options (ESOs) considering two sources of early exercise: forced exercise due to executive turnover and voluntary exercise due to personal considerations. Using data of about 4000 US executives, I estimate separate hazard rate factor models for both sources of early exercise. In a second step, I combine both conditional hazard exercise models for the valuation of a representative ESO in a Monte Carlo simulation. Analysis of the individual valuation impact of each source of early exercise shows that turnover induced exercises are responsible for most of the valuation discount of ESOs to market traded options. This result is important as most of the current literature on ESOs concentrates solely on voluntary exercises. I further find that the common practice valuation approach suggested by the Financial Accounting Standards Board (FASB) consistently underestimates ESO values.

We analyze how 14, 000 top U.S. executives exercise their stock options. Factors suggested by existing theories have low or moderate explanatory power. Variables that model executives’ motive to diversify fare particularly poorly, whereas variables that capture reference-dependent preference, such as past highs and lows of stock prices, perform better. By contrast, characteristics of option portfolios are of first-order importance and suggest that managers have target ownership levels. Institutional features like vesting restrictions or blackout periods also have a first-order impact. We conclude that executives’ main motivations for exercising stock options early seem to be outside the scope of extant models. (JEL G30, M52)Received August 20, 2018; editorial decision December 28, 2019 by Editor Andrew Ellul.