Claremont Colleges Working Papers in Economics 2001-09
Employee stock options represent a significant potential source of dilution for many shareholders. It is well known that reported earnings tend to understate the associated costs, but an efficient stock market will show no such bias. If by contrast stock prices underestimate the future costs implied by stock option grants, option exercises will produce negative abnormal returns. We design and implement a stock-picking rule based on predictions of stock-option exercise using widely available data. The rule identifies stocks that subsequently suffer significnt negative abnormal returns using either a CAPM or the three factor Fama-French benchmarks. According to our point estimates, if the cost of employee stock options as a fraction of market capitalization is 10%, the stock will subsequently exhibit a negative abnormal return of between 3% and 5%. There is some evidence of market learning in that the abnormal returns tend to fall over time. We use a restricted sample of actual stock exercises and find that the reduced power of our trading rule does not reflect a reduced ability to predict stock option exercise. It also does not seem to reflect improved accounting disclosure since the portion of option costs recognized in diluted earnings per share appears to be priced by the market in all our sample years.