New Studies Look at Option Exercise Patterns

Two new studies by Watson Wyatt, one using public filings and the other using data provided by Charles Schwab, look at the patterns of employee option exercises. In the first, Watson Wyatt evaluated footnote disclosures on options in 823 S&P 1500 companies, looking at the number of options exercised in 2003 as a percentage of total options outstanding. The results indicated 12.2% of all options were exercised. If, overall, the average price at which outstanding options could be exercised was 50% of the actual share price or better, the expected exercise rate would be 15.6% of all outstanding options per year. If the average exercise price is out of the money, the rate drops to 10.5%. Companies that grant more options per year than average see the exercise rate go up as well, by about 2 percentage points over what it would otherwise be. In contrast, higher volatility lowers the exercise rate, as economic theory suggests it should (because options on more volatile stick are worth more).

In a separate study, Watson Wyatt used exercise data provided by Schwab to analyze individual exercise patterns. (All data identifying companies or individuals was removed before Schwab gave the data to Watson Wyatt.) The data consisted of more than 65,000 actual option exercises over five years for 36 companies in technology and financial services, by far the largest analysis to date. The companies were large, with an average employment of 21,500 in financial sectors, 3,900 in technology, and 8,700 overall. The study found that employees with larger grants wait longer to exercise. While the mean waiting period before exercise was three years, the mean for those optionees with 10,000 or more shares was four years, compared with 33 months for those with fewer than 1,000 shares. As expected, people who waited longer made more. Those who made a 200% or better return had waited a mean of 41 months to exercise, while those who made less than a 50% return had waited a mean of 32 months.

Like the study that used public data, the study using Schwab data found a strong inverse correlation between volatility and the probability of exercise. Similarly, where the recent stock price performance has been good, the likelihood of exercise declines. Longer vesting periods increase the chance of exercise soon after vesting. Longer terms and larger grants lower the exercise rates. Based on these exercise patterns, the study then looked at how the options were valued according to Black-Scholes and how employees actually behaved as an indicator of how much they thought the options were worth. The results indicate employees discount options 30% to 50% below what the Black-Scholes value provides.

A paper on the findings with detailed data will be released early next year.