Employee stock options and efficient markets

The FASB (Federal Accounting Standard Board) has decided that companies must count employee stock options (ESOs) as an expense on their balance sheets. This policy is of course eminently sensible (so agrees Warren Buffet), since ESOs dilute the value of pre-existing shares in the company. However, thanks to lobbying on the part of Silicon Valley and tech companies in general, ESOs have been off balance sheet until now.

Of course, a believer in (anything but the weakest version of) the efficient markets hypothesis would claim that whether or not ESOs are included in reported earnings is of no consequence, since they do already appear in the text of every public company's quarterly report. (Surely any rational, intelligent investor reads the text of quarterly statements? ;-) But pretty much every tech CEO and VC has been predicting imminent destruction of our marvelous innovation engine due to expensing of ESOs, so I guess they don't believe in efficient markets.

There are some technical issues to be addressed here. How should accountants value ESOs? One could plug the historical vol of the issuing company into the Black-Scholes model, but the timescale of the ESOs (usually one to several years) is much longer than the period over which one usually trusts historical vol. Also, many employees leave the company before their options vest, letting companies recover part of the value, which means employee attrition rates have to be part of any model. I see a consulting opportunity here for quants who want to help CFOs and accountants with this problem :-) I also see a further reduction in the quality (reliability) of the earnings numbers that we investors have to rely on.