As you can see, the call value nearly triples if estimated volatility is assumed at 35% instead of 15%. If that sounds like an awfully large range for observed volatility, consider the relatively blue chip Johnson & Johnson. JNJ's twelve-month rolling volatility has ranged from 12% to 42% since 1995. For a more extreme case, Computer Associates has exhibited volatility in the range of 20% to 120%. The market's assessment has varied as much or even more. Looking at the prices of options on JNJ and CA common stock, we see that the "implied" volatility (i.e., solving for volatility using the current market price of the option) in traded options varies even more than the observed volatility in the stock price. CA's implied volatility ranged from 50% to 150% in the last quarter, while JNJ's has ranged from 20% to 65%. One of these companies could easily vote on an option plan in one month at an estimated cost of $10 million and find that the plan costs $25 million at the time of the grant.

This unpredictability in option valuation will be problematic for company boards, particularly those looking to change their senior management teams. Computer Associates, for instance, may well need to attract new management to improve its credibility within the financial markets. Yet the market behavior of CA's stock has made an attractive option package prohibitively expensive. Tyco International, having just hired a new CEO to stem sales of their stock, would find the associated 2 million option grant extremely difficult to value.  Ironically, many public companies seeking new management to arrest a falling stock price would find themselves in the same situation. 

The Black Scholes method uses measures of central tendency to define the expected value of an option, but that value is merely the average of many potential outcomes. For instance, the expected value of a lottery ticket is a tiny fraction of one cent. No lottery ticket holders actually receive this amount, as the distribution consists of one big winner and millions of worthless tickets. A distribution of option outcomes will not be as extreme, but some options will still expire worthless.  One study estimates that 33% are "underwater" even after recent market declines. Another, authored in 1995, suggests that the Black Scholes method would have overestimated compensation expense by 39% when compared with an ex post certain measure of options values.  The chance that any option will be exercised at  exactly  the Black Scholes expected value is extremely remote.

If option expense is booked as compensation expense at the time of granting, no corresponding cash flows will occur. Therefore, when the option expires (regardless of its value), the company will be carrying a deferred option expense that is now overstated. Should the company take the expense back as earnings? Will companies learn to create an earnings cushion of deferred options expense through complex grant programs?

The troubling implication of these accounting and valuation realities is that if option costs are estimated and expensed, the market price of the stock itself will have a significant effect on earnings. In a world where earnings are supposed to drive market performance, we find that outcome disturbing.

There is no doubt that too many members of the investment and corporate community fail to consider the impact options may have on shareholders. But even though options are not expensed through the earnings statement, options grant data and valuation parameters have been completely disclosed since the mid-1990s under SFAS 123. Potential dilution is easily calculated and presented in corporate earnings statements. Analysts or investors may have chosen to ignore them, but the facts are there for all to see unlike, for instance, the notorious Enron special purpose entities. It is unclear what proponents of options expensing would do with the dilution calculation, as putting the option cost in the P&L and earnings dilution (the numerator and the denominator of fully diluted earnings per share) strikes us as redundant. Furthermore, if the market has seized on the undiluted earnings as the primary valuation metric, should we pass legislation to alter the content of that metric?  In the absence of one "true" number, it is unclear we should legislate a preferred choice.

FASB describes an Effective Yield Test for determining when the embedded options in convertible bonds become Common Stock Equivalents for accounting purposes. Prior to 1997, this meant that the underlying shares were included as if converted in the primary earnings per share calculation. Unfortunately, SFAS 128, issued in February of 1997, replaced primary EPS with "basic" EPS, including no immediate dilution. Perhaps a reasonable solution to the stock option valuation dilemma will be to re-introduce options that are close to exercise in EPS, along the lines of the old "primary" EPS guidelines. Companies could include executive options within a given range of their strike price as CSEs. This would address the majority of the objections to current practice, including the "distorted" P/E multiples reported on the market. It is of more than passing interest that older accounting rules might well have handled this problem more effectively, as opposed to the options expensing controversy highlighting outdated conventions.

The moral hazard argument carries more weight, as it is clear that the payoff in executive stock options is misaligned with investors interests. An executive owning stock has a risk profile more similar to that of a shareholder. Unfortunately, a security that provides a perfect alignment of interests between shareholders and management does not exist. If an executive receives common shares as a grant, his cost of ownership differs from that of a shareholder. The existing restrictions on insider stock sales further exaggerate the difference between executives and passive shareholders. While we would encourage boards to attempt to align executive and shareholder interests, changing SEC accounting rules on options seems like an indirect way of tackling the problem.

Unfortunately for the "moral hazard" camp, there is little evidence to support the view that options are responsible for the most recent cases of creative accounting. Virtually every recent accounting scandal - Xerox, WorldCom, Adelphia, Enron, Global Crossing, Tyco, Qwest and Computer Associates, occurred at companies where senior executives owned primarily large chunks of company stock, complete with the accompanying downside risk. While it seems tidy to suggest that executives holding stock rather than options are less likely to employ "pump and dump" practices through misleading accounting, critics must confront recent market evidence to the contrary.

Today's options discussion is a distraction in a world where even seasoned financial analysts were denied the full measure of off-balance sheet transactions and sham revenue deals between companies. There is still room to improve disclosure on a number of fronts without further obfuscating options expense. Finally, we in the financial professions need to take some medicine as well. As Robert Bartley says in a recent Wall Street Journal editorial:

Would that all the energy now going into this accounting crusade were directed at another message. To wit, there is not and never will be a perfect number to sum up a company's results. Earnings per share is a good starting point, but you have to go deeper for real understanding. Getting this message home to Wall Street and beyond would do more than a peck of laws and bushel of accounting changes to prevent future WorldComs and Enrons.

We have said on many occasions that the ultimate reform is the market itself.