Stock option grant vesting terms: economic and financial reporting determinants

In contrast to these models where extended vesting is seen as the critical contract feature that extends the horizon of managers and prevents myopia, Laux (2012) cautions that, if there is a risk that the CEO will get fired before the options vest, the cost of forfeiting these options may induce managers to myopically boost interim performance to prevent their being fired and losing all their options. His optimal contract separates the two roles of vesting: ownership transfer and the timing of exercisability. In the presence of long-term projects, it may be optimal to have a transfer of ownership of the options after only a short period, while still preventing option exercises until the uncertainty about the project has been resolved. Thus the prediction from this model for vesting terms is ambiguous.

Firms with considerable growth opportunities must make investment decisions about projects where the cash flows will not be fully revealed for longer periods and management has a considerable information advantage that cannot be credibly disclosed. Thus equity values may not fully reflect the expected value of the investment. The price adjusts to reflect the investment choices as information about investments is revealed over time.

For example, when Ford Motor Company hired Alan Mulally from Boeing, Ford paid him an $18.5 million dollar bonus, which was largely composed of $11 million to “make up for bonuses and stock options forfeited by leaving Boeing” (Ford 2006 proxy statement).

In contrast, direct monitoring and contractual provisions such as vesting periods could also act as substitutes. In that case, more closely monitored firms can allow executives to exercise their options earlier in the term, because these more closely monitored managers impose fewer agency costs.

We focus on stock option grants for several reasons. First, stock option grants are the largest component of (equity) pay during our sample period. Second, since our focus is on equity grants with service-based vesting, we would like to exclude any equity grants with performance vesting conditions. Gerakos et al. (2007) find that option grants with performance-based vesting restrictions are relatively rare, while Bettis et al. (2007) find that performance-based vesting awards are more common for restricted stock grants than they are for option grants.

Before August 2002, beneficial owners were not required to file their Form 4 until the 10th day of the month following the change in beneficial ownership. This leads to errors in the Form 4 files in the early part of the period. We begin our sample in 1997 because our ability to match ExecuComp grants with Form 4 filings significantly improves from this year.

Although SFAS 123(R) is not truly exogenous, it imposes option grant expensing on all firms. Thus it provides an opportunity to investigate how firms alter their contracts when required to recognize option grants as an expense. Nonetheless, we address this concern empirically when examining the change in the vesting periods in Sect. 4.5.

While we use this as our primary measure of growth opportunities, our results are consistent when we use the conventional market-to-book measure.

Abnormal cash compensation could alternatively be a proxy for CEO power as CEOs with higher bargaining power could negotiate higher current compensation. If this variable captures CEO power rather than retention incentives, we would expect a negative relation with vesting terms because powerful CEOs would not want to be constrained by longer vesting terms.

Industry homogeneity is calculated using a two-factor regression model of a firm’s stock return on the returns for the industry and the overall stock market. The measure is the average across all firms in the two-digit SIC industry of the partial correlation coefficient on the industry return index. Industries where returns are more correlated are considered more homogenous.

In cases where the CEO is not the highest paid executive, we measure the difference between CEO compensation and the highest paid executive, which yields a negative value. Recent work by Bebchuk et al. (2011) uses a slightly different but related metric, namely CEO total pay scaled by the total pay of the top five executives; they find that CEOs with a large pay slice tend to be associated with greater agency problems. They include equity pay, long-term incentive payouts, and other compensation in their metric. Because equity pay is linked with the vesting term, it is important to exclude this from our measure of CEO power, and we therefore use the measure in Hayward and Hambrick (1997) and Chatterjee and Hambrick (2007).

Our tests examining the influence of CEO power on vesting terms has potentially differential predictions for the CEO versus other top executives, while all the other determinants suggest variations across firms rather than across executives within a firm. A within-firm comparison has the advantage of using the firm as its own control. However, a significant drawback is that, in about 75 % of firm-years, the vesting schedules are identical for other executives. Nevertheless, we compare the average vesting of the CEO versus the CFO using a paired t test for all firm years in our sample where the CFO also received options and data could be matched between ExecuComp and Thomson Financial, as discussed in Sect. 3.1. We pick CFOs for the comparison to keep the executive role constant and because most companies have a CFO and the position can be easily identified in ExecuComp. While the difference between CEOs and CFOs is not statistically significant in the overall sample, CEOs have lower Vesting Duration relative to the CFO in the sub-sample of powerful CEOs; this difference is significant at the 5% level.

In untabulated tests, we repeat the analysis in Table 5 using Vesting Period as the dependent variable. The results are generally similar. In particular, the two reporting variables and the measure of growth opportunities are still positive and statistically significant. The effect of CEO Power is positive and significant. Similar to the result in Table 5, the main effect and the interaction effect of the monitoring index are positive, although now the main effect is significant and the interaction effect is not. In contrast to the results in Table 5 using Vesting Duration, the retirement indicator is negative and significant, while the coefficient on ROA is no longer significant.

For example, an executive may be indifferent between an option grant that vests ratably over 6 years and one that cliff vests at the end of 4 years. This is because the equal vesting grant has a lower probability of forfeiture of the full grant and affords the executive opportunities to exercise portions of the grant earlier.

Note that for cliff-vesting grants the vesting period and the Vesting Duration should be the same since the full grant vests at the end of the vesting period. In this table, the two are not always identical because we have firms that have multiple cliff-vesting grants in the same year. In that case, the Vesting Duration is the weighted average of the grants, but the vesting period is the maximum of the vesting periods. Results are similar if we exclude those observations.