If CEO stock holdings were replaced with the same ex ante value of stock options, the pay-to-performance sensitivity for the typical CEO would approximately double.
CEOs of the largest U.S. companies now receive annual stock option awards that are larger on average than their salaries and bonuses combined. In contrast, in 1980 the average stock option grant represented less than 20 percent of direct pay and the median stock option grant was zero. The increase in these options holdings over time has solidified the link between executive pay -- broadly defined to include all direct pay plus stock and stock options revaluations -- and performance. However, the incentives created by stock options are complex. To the extent that even executives are confused by stock options, their usefulness as an incentive device is undermined.
In The Pay to Performance Incentives of Executive Stock Options (NBER Working Paper No. 6674), author Brian Hall takes what he calls a "slightly unusual" approach to studying stock options. He uses data from stock options contracts to investigate the pay-to-performance incentives that would be created by executive stock options if they were well understood. However, interviews with company directors, CEO pay consultants, and CEOs, summarized in the paper, suggest that the incentives are often not well understood - either by the boards that grant them or by the executives who are supposed to be motivated by them.
Hall addresses two main issues: first, the pay-to-performance incentives created by the revaluation of stock option holdings; and second, the pay-to-performance incentives created by various stock option grant policies. He initially characterizes the incentives facing the "typical" CEO (with typical holding of stock options) of the "typical" company (in terms of dividend policy and volatility, both of which affect an option's value). He uses data on the compensation of CEOs of 478 of the largest publicly traded US companies over 15 years, the most important detail being the characteristics of their stock options and stock option holdings.
His first question concerns the pay-to-performance incentives created by existing stock option holdings. Yearly stock option grants build up over time, in many cases giving CEOs large stock-option holdings. Changes in firm market values lead to revaluations - both positive and negative - of these stock options, which can create powerful, if sometimes confusing, incentives for CEOs to raise the market values of their companies.
Hall's results suggest that stock option holdings provide about twice the pay-to-performance sensitivity of stock. This means that if CEO stock holdings were replaced with the same ex ante value of stock options, the pay-to-performance sensitivity for the typical CEO would approximately double.
Moreover, if the current policy of granting at-the-money options were replaced by an ex ante value-neutral policy of granting out-of-the-money options (where the exercise price is set equal to 1.5 times the current stock price), then performance sensitivity would increase by a moderate amount - approximately 27 percent. However, the sensitivity of stock options is greater on the upside than on the downside.
Hall's second question is how the pay-to-performance sensitivity of yearly option grants is affected by the specific option granting policy. Just as stock price performance affects current and future salary and bonus, it also affects the value of current and future stock option grants. Independent of how stock prices affect the revaluation of old, existing options, changes in the stock price can affect the value of future option grants, creating a pay-to-performance link from option grants that is analogous to the pay-to-performance link from salary and bonus.
Stock option plans are multi-year plans. Thus different option-granting policies have significantly different pay-to-performance incentives built in, since changes in current stock prices affect the value of future option grants in different ways. Hall compares four options-granting policies. These create dramatically different pay-to-performance incentives at grant date. Ranked from most to least high-powered, they are: up-front option grants (instead of annual grants); fixed number policies (the number of options is fixed through time); fixed value policies (the Black-Scholes value of options is fixed); and (unofficial) "back door re-pricing," where bad performance this year can be made up for by a larger grant next year, and vice-versa.
Hall notes that because of the possibility of back-door repricing, the relationship between yearly option awards and past performance can be positive, negative or zero. His evidence, however, suggests a very strong, positive relationship in the aggregate. In fact, Hall finds that (even ignoring the revaluation of past options grants) the pay-to-performance relationship in practice is much stronger for stock option grants than for salary and bonus. Moreover, consistent with expectations, he finds that fixed number plans create a stronger pay-to-performance link than fixed value policies. In sum, multi-year grant policies appear to magnify, rather than reduce, the usual pay-to-performance incentives that result from CEO holdings of past options.
-- Andrew Balls