How to Forget the Footnotes
fortune.com
What are the best alternatives for measuring the cost of employee stock options? A FORTUNE cheat sheet.
FORTUNE.COM
Tuesday, July 23, 2002
By Justin Fox
The Stock Options Solution
CEO Perks That'll Drive You Berserk
The Great CEO Pay Heist
Special Report: Crisis of Confidence
Outside of the Silicon Valley true believers, and their friends in Washington, it's getting ever harder to find people who are willing to argue that employee stock options are in fact free, as current accounting practice says they are. But there are still lots of arguments over how best to measure the cost of options. Here's a cheat sheet to the most prominent alternatives.
Expense at "intrinsic value." This is how accounting for options works now, and has worked since the accounting profession first took a stand on the subject back in 1948 (the actual accounting standard now followed by most corporations dates to 1972). Options are treated as a compensation expense, and that expense is determined by subtracting the exercise price of the options from the market price of the stock on "the first date on which are known both (1) the number of shares that an individual employee is entitled to receive, and (2) the option or purchase price." So if a company gives an employee options to buy a set number of shares at a set price that happens to be at or below the stock price that day, the recorded expense is zero. This helps explain why virtually all employee stock options have an exercise price that's the same as the stock price on the day of the grant. Such plain vanilla options are "free," while any attempt to get clever with options grants--by making the options contingent on meeting performance targets, or by linking the exercise price to an index like the S&P 500--results in a charge against earnings.
Expense at exercise. This is what the IRS does, and has done since the Supreme Court ruled in 1945 that the profits people make by exercising their employee stock options should be taxed as income, not capital gains. Congress has created an exception to this known as the "incentive stock option," on which the profits are treated as capital gains as long as you hold on to the stock for a year after you exercise the option. But there are strict limits on incentive options, so the vast majority of options that companies give their employees are the kind on which gains are taxed as income. And because employees' options profits are taxed as income, corporations get to treat those profits as a tax-deductible compensation expense. Companies cannot, however, count this deduction toward reported earnings, unless for some reason the options costs showed up in reported earnings too. (Ain't accounting great?)
Treating options the same way in reported earnings as they're treated on tax forms would be simple, understandable, and consistent. But the people who make accounting standards have always rejected it. Why? Here's a pretty good explanation from a 1953 accounting standard: "Beginning with the time at which the grantee may first exercise the option, he is in effect speculating for his own account. His delay has no discernible relation to his status as an employee but reflects only his judgment as an investor."
Expense at "fair value." For decades, the accountants agreed that options are a compensation expense, but couldn't come up with a good way to measure that expense. Then, not long after the 1972 options accounting standard came out, Fischer Black and Myron Scholes published their world changing article on how to value an option using a formula that takes into account the duration of the option, the volatility of the underlying stock, and a few other inputs. In the decades that followed, options exchanges and zillion-dollar derivatives deals became an accepted part of the financial landscape. So in 1984, the Financial Accounting Standards Board (FASB) decided to revisit options accounting in light of the revolution Black and Scholes had wrought.
By 1993 (FASB isn't known for its speed) it had come up with a proposed standard in which companies were to use Black-Scholes or a similar options pricing formula, with some adjustments to reflect the fact that employee options can't be traded and that most people exercise them long before the exercise date, to estimate the "fair value" of the options on the day they were granted. Then that expense would be run through the income statement over the vesting period of the options (usually four or five years). This is when corporate America went ballistic and lobbied Congress into threatening FASB with annihilation. In the interest of self-preservation, FASB finally came out in 1995 with an accounting standard that says expensing options at fair value is "preferable" but not mandatory-although it does require that companies disclose their estimated options costs in an annual report footnote.
Until recently, Boeing and Winn-Dixie were the only S&P 500 companies to follow FASB's preferred accounting treatment. This month, however, Coca-Cola, Bank One, and Wachovia have announced that they will do the same. And the International Accounting Standards Board (IASB), whose pronouncements are followed by many big companies outside the U.S., will probably vote next year to require fair-value options expensing. At which point FASB will take the matter up again and, barring interference from Congress, do what it wanted to do back in the early 1990s.
Treat options as a liability. The accounting treatment preferred by the FASB and IASB has the advantage of being elegant, and consistent with accounting for other non-cash compensation--shares of stock, cars, houses, trips to Bora Bora, etc. It has the disadvantage, though, of generating options-expense numbers that strike a lot of people as downright loopy. This is especially true right now, as the options-cost footnote of a company like Cisco is still reflecting huge estimated costs from options granted a couple years ago at $60 a share--even though the stock is now at around $13 and those $60 options, while not entirely worthless, are certainly worth a lot less than they were when the stock was at $60. There are also concerns that companies might deliberately lowball the estimates of stock volatility and expected option duration that they feed into their options pricing models in order to report a lower options expense.
So why not treat options as a liability, re-estimating their value every quarter, running those changes in value through earnings, and eventually squaring up the estimated cost with the actual value realized by the employee upon exercise? This would eventually correct any errors in valuation, deliberate or otherwise, and would better reflect the volatility a company exposes itself to by issuing options. It has the added advantage of making reported options costs equal to tax return options costs over time.
The problem with this idea, say the accounting standard-setters, is that options aren't like debt, or pensions, or retiree health benefits. With options, "there is absolutely no potential cash liability down the road," says James Leisenring, who was a member of FASB during its options odyssey and is now on the IASB. Instead, he says, an option is "a ticket to the lottery that is the stock market. We're trying to value the ticket." Leisenring's view will almost certainly win out as far as accounting standards go--especially since, however much they might hate the FASB's and IASB's preferred way of expensing options, most CEOs and CFOs will prefer it to a liability that could unpredictably bounce earnings up and down for years to come. But the argument isn't going away.
Focus on dilution. There are those, among them President Bush, who contend that options costs are already fully reflected in earnings per share because options are counted as part of diluted shares outstanding. This is a bunch of hooey--outright grants of stock increase shares outstanding, too, but they're counted as an expense and nobody complains about that. But there is something to the argument from the perspective of investors trying to figure out how much of a company's future cash flows they're entitled to. In this case, measuring dilution is more important than estimating options costs. Existing measures of diluted earnings per share don't do the trick--they count options only as those options come into the money. An option to buy a share of a company's stock at $5 a share doesn't count toward diluted shares outstanding if the stock is at $5. It only begins counting as the stock price rises above five, and then only fractionally because it's presumed that the company will use the $5 it gets from the employee to buy back shares.
What this means is that diluted EPS today doesn't tell you anything about what kind of future dilution will occur if the stock price goes up. To get a better handle on dilution, you have to look at the options footnote in a company's annual report. There you can find out how many employee options are out there, and you can get at least an approximate idea of the exercise prices of those outstanding options. Some value investors simply add in all the options, regardless of exercise price, to shares outstanding to get something they call "fully diluted" shares outstanding. Others figure out just how many more shares there will be if the company's stock price reaches a certain level. This is a lot of work. But then, investing is supposed to be hard work, right? |
