That's because the way companies value options is an inexact science. Current accounting rules let them do it in a number of different ways. And, it turns out, the method a business chooses could have a tremendous impact on how the compensation affects its reported earnings. Indeed, some studies suggest the impact could vary by as much as 100%.
The folks who write accounting rules could try to fix this by requiring a single method. But which one? And would a one-size-fits-all requirement really work for every public company in the country? Probably not.
MARKET VALUE. The problem has a solution: Let the Wall Street rocket scientists do with options what they have done with mortgages and credit-card receivables -- securitize them. Once the investment bankers create a derivative security based on the value of stock options, there'll be no more arguments about what the compensation costs the company. Instead, companies will be able to report an actual market value to shareholders.
Until then, as options expensing becomes more common, investors are looking at years of serious heartburn. It will be hard to compare options costs among companies, because each may be using a different valuation system. And it will be tough to figure if a company is low-balling its earnings hit. To understand why, let's walk though a few basics.
When a company reports options, it does so at the time the compensation is granted. But options usually can't be exercised for at least five years. During that time, all sorts of things may happen. The underlying stock price may rise or fall. Managers may quit before they can take their options. Or they may leave soon after five years and be forced to exercise the options (to buy company stock) right away.
EDUCATED GUESSES. All these uncertainties reduce the value of the options, but nobody can agree by how much. "We know it is worth less than a share of stock, and more than zero," says Tim Lucas, former research director at the Financial Accounting Standards Board, which writes the rules.
As a result, companies must make some educated guesses. Today, they use one of three ways to do that. The first, called the Black-Scholes method, was really designed to price exchange-traded options. It isn't very good at calculating the value of employee stock options, for which there is no market. Some companies try to improve the method by plugging in more accurate assumptions of, for example, how long execs hang on to options in the real world.
The second system is called the binomial tree model. This uses a different mathematical formula than Black-Scholes and attempts to better take into account how all the restrictions on the options affect price. Huge difference in value may result if a company uses this method rather than Black-Scholes.
DEPENDS ON ASSUMPTIONS. Analysis Group/Economics, a New York-based consulting firm, has developed one binomial system. Under its system, a company granting options today on a $100 stock would expense that compensation at as little as $34.70. Using Black-Scholes, by contrast, the company might have to take a noncash earnings hit of as much as $59.35 for each option it granted.
But a company can get wildly different results if it uses the same mathematical formula but alters the assumptions. Mark Rubenstein, a finance professor at the University of California at Berkeley, found that merely by adjusting some of the key assumptions, a company could produce huge differences in costs and, therefore, in the amount it would have to trim earnings. In one test, Rubenstein found the value of an option for a $100 stock ranged from $18.68 to $36.32.
Other approaches exist as well. Options could be expensed when they're exercised, rather than at the time they're granted. That would solve the valuation problems since the cost to the company would be there for all to see: the difference between the exercise price of the option and the market price of the stock.
WARREN'S WAY. Another method, favored by Rubenstein and others, would require companies to expense options at the time they're granted, but make companies adjust the value of the paper over time. That way, a company could gradually recalculate the cost of the options to reflect real-world changes in their value.
All these approaches merely simulate the market value of options. Coca-Cola, pressed by über investor Warren Buffett, has devised a third way. Under this approach, four investment bankers come up with binding quotes on options to buy 10,000 shares of Coke and options to sell the same number of shares. The idea: The average price would be a pretty good proxy for the market value of those options.
Buffett and Coke are on to something. But getting quotes on 10,000 options is not the same as making a real market. The better idea is for Wall Street to create an options-based security that would tell shareholders exactly what all this executive comp costs their companies. That won't be easy, but they tackle projects like this for clients. And it may be the only way that investors will ever really know what all those options are worth -- and what they mean for the bottom line. Gleckman is a senior correspondent in BusinessWeek's Washington bureau. Follow his views every Tuesday in Washington
Watch, only on BusinessWeek Online