We investigate whether footnote disclosures under Statement of Financial Accounting Standards (SFAS) No. 123 are managed in 1996, the first year that the disclosure was required. The 1996 phase‐in of SFAS No. 123 provided firms with a unique opportunity to manipulate the pro forma disclosure in the initial years. SFAS No. 123 allows firms discretion in estimating the value of their stock option grants and in allocating that value across accounting periods. Although we find little evidence that firms manage the estimated value of their option grants, we find that firm‐specific incentives affect how that value is allocated. Specifically, firms that provide high levels of either CEO compensation or stock option compensation relative to performance allocate a smaller proportion of the options' value to the 1996 pro forma expense, apparently to reduce criticism of that compensation. Small firms and firms that recently went public also allocate a smaller proportion of option value to the 1996 pro forma expense, apparently to increase perceptions of their profitability.

We conjecture that firms were less likely to manage the value of the options granted than the allocation of that value in 1996 because the parameter estimates underlying the reported option value must be disclosed in the footnote, whereas the inputs to the allocation computation are not disclosed.

These results, which suggest that firms manipulated pro forma stock option expense when their estimate choices cannot be observed, have implications for both standard setters and financial statement users. In particular, the FASB's current deliberations on the transition from footnote disclosure to income statement recognition for stock options should consider additional disclosures to minimize unobservable choices. More generally, the FASB may reduce potential manipulation by requiring expanded disclosures about the choices used in computing both pro forma and reported numbers.

This content is only available via PDF.
You do not currently have access to this content.