Commentary: Stock Options: The Fuzzy New Math

Shareholders rarely get a glimpse of the future. Yet in the past few months, they've been given a peek at what they may be in for if companies are required to include the cost of employee stock options on their financial statements, which now seems all but certain. It's not pretty. If 2002 is any indication, options will likely have a huge impact on profits, slicing 20% off reported earnings. Worse, in solving one problem by forcing companies to recognize that options have a cost, we've created something equally complex: Shareholders will have no way of knowing whether their companies are accurately estimating expenses or engaging in wishful thinking to burnish the bottom line.

The source of the problem, as any expensing opponent will tell you, is the Black-Scholes option pricing formula, the most widely used method for estimating option expenses disclosed in the footnotes of annual reports. Four of the six variables it uses to calculate option values require companies to peer as much as 10 years into the future. Trouble is, there's no accurate way to determine when employees will exercise their options, how wildly the stock price will fluctuate, or what will be the stock's dividend yield and risk-free rate of return on U.S. Treasury notes. What's more, there's a clear incentive to goose these numbers, especially among tech companies that are heavy issuers of employee stock options.

Even modest reductions in the assumed stock volatility and the life of an option -- or small increases in dividend yield and the risk-free interest rate -- can cut option expenses significantly. If the company is expensing, that boosts net income over several years as the full cost of option grants is recognized. Adjustments occur every year, as they should, to account for changing historical trends and option plans. The trick for investors is determining which are warranted.

As expensing becomes a reality and options become a cost to be managed like overhead or payroll, finance chiefs will likely find legitimate ways to reduce that expense. Faster vesting periods, for example, would justify shorter option lives, while boosting dividends would legitimately allow for a higher dividend yield assumption -- either of which would reduce option costs.

Shareholders should expect to spend more time drilling down into option pricing assumptions, as they do now with cash flow statements and the like. But there are things that regulators and auditors can do to make it easier -- and restore the trust that's critical to the success of option expensing. For starters, the Financial Accounting Standards Board needs to increase the transparency surrounding option values, requiring companies that depart significantly from actual historic trends to explain why. The Securities & Exchange Commission needs to crack down on finance executives who game the system. And auditors who value their reputations -- and their businesses -- shouldn't sign off on assumptions that would tarnish them, no matter how great the pressure from clients.

When deciding on Black-Scholes assumptions, companies are supposed to rely on historical experience, modified with a little common sense about what the future is likely to hold. For example, when estimating option life, companies are supposed to consider not only historical data on option exercises but also factors that may affect future exercises. Capital One Financial (COF ) Corp., for example, reduced its option life expectancy in 2002 to 5 years, from 8.5 years, after granting options with shorter vesting periods. That change cut option costs by $29.3 million.

Current rules give finance chiefs considerable leeway in determining which assumptions to use, and a growing body of evidence suggests they're using it. Jack T. Ciesielski, publisher of the The Analyst's Accounting Observer, found that one out of five companies in the Standard & Poor's 500-stock index reduced option life, stock volatility, or both, in 2002, increasing actual or pro forma earnings in the process. Derek Johnston-Wilson, an assistant accounting professor at Colorado State University, compared accounting assumptions used to value options in 2002 with actual historical trends and found that many companies underestimated both volatility and the risk-free interest rate.

To get a sense of how these little- noticed changes can affect profits, consider how Siebel Systems (SEBL ) Inc. dealt with volatility when calculating the value of 5.7 million options awarded in 2002. Siebel assumed volatility -- the degree to which stock price changes exceed their historical average -- of 65% for the life of the options, which it estimated at 3.4 years. At the time, volatility for the previous 3.4 years was inching past 100%. By reducing its volatility assumption to 65%, from 89% in 2001, the company trimmed $15.5 million from its option cost, a savings that is spread over several years. Siebel says its decision was based on lengthy analysis, including a survey of rival technology outfits. "The past five years was the biggest anomaly in the history of the equity markets," says Paul Gifford, Siebel's vice-president for finance. "To say that that history predicts the future without applying some sanity to it is absurd."

Other examples abound. Broadcom (BRCM ) Corp. reduced its volatility assumption to 70% last year, from 90% in 2001. Actual volatility at the time was nearing 112%. That trimmed $74 million from its option expense. The company concluded that volatility in the post-bubble years was likely to decline as the company became larger, more diversified, and more profitable. At Hilton Hotels (HLT ) Corp., management predicted that its options would be exercised faster -- a notion that compensation experts say seems unlikely with the stock down 23% from its post-September 11, 2001, high. And with five-year volatility at 43% and rising, Hilton predicted that volatility would decrease to 34%. Hilton maintains that the assumptions are based on empirical fact, and declined to elaborate. But Patrick S. McGurn, senior vice-president at proxy adviser Institutional Shareholder Services, says Hilton shows that "rosy scenarios are alive and living in Corporate America."

It's unclear why so many companies shifted their assumptions sharply downward last year. One possibility is that they're taking advantage of the lull before expensing takes hold. In the post-expensing world, such changes will be scrutinized closely for signs of tinkering. By making changes now, finance chiefs can lock in the more favorable assumptions for 2002 grants and lay the groundwork to justify low assumptions for future grants.

It should come as no surprise that companies are aggressively managing option values as the prospect of expensing becomes real. But that's no reason to abandon expensing. Regulators need to devise a way to value options that doesn't give companies an opportunity to engage in self-serving guesswork. Without that, the calculation will distort the real cost of stock options -- exactly the problem expensing was meant to solve.

Corrections and Clarifications "Stock options: The fuzzy new math" (Management, July 14) misstated the impact of increasing the risk-free interest rate assumption when calculating the Black-Scholes value of stock options. Increasing the assumption will increase option values.

By Louis Lavelle

    Before it's here, it's on the Bloomberg Terminal.