FASB's proposal, which will require companies to expense stock option compensation, contains more than just a hint of preference for a certain type of model to estimate the fair value of options.

According to the issuance by FASB, there are several valuation techniques that can be used to accurately determine the fair value of employee stock options. The proposal specifically mentions "lattice" models and closed-form models as acceptable.

These models, which are based on established financial economic theory, are also used to establish the fair value of other derivative securities including commodities and currencies.

Black-Scholes & Perceived Limitations

Black-Scholes, the most widely recognized closed-form model, was developed in 1973 as a pricing-model for stock warrants, and later extended to pricing other derivatives.

Black-Scholes is considered a closed-form model because it assumes that the option exercise occurs only at the end of an option's contractual term. The model further defines that volatility, dividends and risk-free interest rates are constant over the option's term.

Considerable scrutiny from critics regarding the entire issue of expensing options has focused on these underlying assumptions, declaring the model too static and theoretical for the dynamic fluctuations that occur in the real economic and business climate.

Chris McGee, a partner in the McLean, Va. office of Ernst & Young, believes the Black-Sholes option exercise assumption to be one of its biggest handicaps. "Black-Scholes assumes optimal option exercise, and the one thing we absolutely know for sure, especially about broad-based plans, is that exercise patterns are anything but optimal."

Also, keep in mind that many of the companies reluctant to adopt options expensing are technology companies, whose business conditions and stock values have been the polar opposite of the constant variables Black-Scholes assumes.

Lattice Preferred, But May Not Be Superior

According to FASB spokesperson Sheryl Thompson, FASB is favorably inclined to a lattice model because it can be designed to accommodate variable factors such as dividends and volatility over the option's contractual term. "The lattice model provides a greater number of data inputs and allows for a more precise, or higher-quality number."

The model derives its name from the patchwork pattern created by lines that define the model. In a lattice model, two possible price movements are assumed: up or down. Therefore, when estimating the value of an option over time, the price can go in either two directions: it can increase or decrease, similar to flipping a coin.

If one were to draw lines connecting each price to a possible future value—up or down— and then draw lines connecting the value to future price options, a patchwork pattern resembling a lattice would result.

The model also employes numerous relevant inputs to fine-tune the value of the options, including—for example—the existance of blackout periods, and the expected patterns employees actually utilize when exercising options.

Ernst & Young's Chris McGee believes FASB's focus on additional inputs recognizes important issues in determining fair value. "What FASB allows for is a combination of market data, such as interest rates, as well as peer analysis and stock movement," says McGee. "And individual company statistics—turnover, demographics—all these things that are good predictors of volatility, and of option exercise patterns," he adds. "All these kinds of issues that will ultimately affect value."

But many, including MIT Sloan School of Management professor S.P. Kothari, don't necessarily agree that the lattice model will derive a significantly better estimate. "Which model used will have some impact on the value of options," notes Kothari, "but in my assessment, the differences obtained in most cases is going to be fairly second order."

Black-Scholes Merits, Rigidity

According to Kothari, who heads MIT's accounting and economics department, there is one data input that is the most crucial to deriving an accurate fair value options estimate. "The key thing in all of this, which is subjective, is how soon employees will exercise their options."

A normal Black-Scholes model assumes that if an option is a 10-year option, the employee will not exercise before 10 years. This of course is contrary to a typical employee's behavior, who might execute the option well before its expiration date.

However, Kothari notes that Black-Scholes is not necessarily as "closed" as the name closed-model suggests. In fact, the formula can be adjusted to take into account the actual selling behavior of employees. "The usual way Black-Scholes valuation of options is done is by reducing the expiration from ten years, to something like four years," says Kothari. "Historically employees have been exercising in three or four or five years. Once you make that adjustment to Black-Scholes, I think it is fairly straightforward to implement."

In fact, FASB does recognize cases when a closed-form model is more practical. A newly public company, or a non-public company, would lack the historical data on employee exercise price patterns to be included in a lattice model.

Furthermore, companies that do not provide significant compensation in the form of stock options might find that a closed-model provides pricing estimates not materially different from a lattice model.

Nevertheless, barring these factors, FASB's announcement remains biased toward the lattice model in providing the most accurate estimate. "Public entities for which compensation cost from share option arrangements is a significant element of the financial statements may conclude, when inputs are available, that a lattice model would provide a better estimate of fair value," noted the Board's statement, "because of its ability to more fully capture and better reflect the characteristics of a particular employee share option or similar instrument in the estimate of fair value."

Choices, Changes And Costs

While companies have latitude in choosing which model they wish to employ, FASB cautions that the valuation technique selected should remain constant, and can only be changed to what the Board considers a more accurate one, meaning from Black-Scholes to a lattice model. Not vice versa.

Therefore, a company initially implementing the lattice model cannot later switch to a closed-form method.

According to FASB's issuance, "Once an entity changes its valuation technique to a lattice model, it may not change to a less preferable valuation technique."

Although FASB is fairly resolute in its preference for lattice models, when asked which model corporations will likely adopt, spokesperson Thompson is not certain. "I don't know. I think we'll have to wait and see. Companies have used lattice models in the past, but it's a lot less than the companies that use Black-Scholes."

Ernst & Young's McGee doesn't believe the current popularity of Black-Scholes will be an obstacle in getting firms to switch models. Now that FASB has stated its preference for lattice models, it leaves little room for auditing firms like Ernst & Young to encourage alternatives. "It would be difficult for an auditor who knows that FASB has basically said, 'We know that Black-Scholes does not do a good job predicting the value,'" McGee predicts. "How do I sign off on [a Black-Scholes determined valuation] unless it's not a material number?"

One might wonder if the current popularity of Black-Scholes is based on a capacity for the model to have less of an impact on the income statement, and thus be more advantageous to the corporate bottom line.

MIT's Kothari contends that the opposite is true. "[Lattice-based models] might be more accurate, but I can assure you, you put in more of these parameters, it gives more latitude. Corporations can try different assumptions, and the number you get will be a little more influenced by the choices of the management."

The current popularity of Black-Scholes likely exists because of two of the most fundamental driving factors in American business: It's cheap and easy. "Black-Scholes has virtually zero cost for companies internally, because they can plug it into an Excel spreadsheet and generate the value," McGee explains. "The lattice models are not easy. Most companies do not have that skill set and will be forced to outsource."

Since Ernst & Young is currently in deliberations regarding exact pricing for providing lattice numbers crunching, McGee refrained from providing a ballpark number. However, he hinted it would likely be more than a nickel and dime incremental cost to clients. "It is a cost in an area where companies did not have one before, and it can often be significant."

Governance Backfire?

What might be most remarkable in FASB's preference for lattice models vs. Black-Scholes is that it may lead to more massaging of numbers by corporate management. This potentiality is extraordinarily ironic considering the entire options expensing issue was created by the fallout of management-led accounting scandals.

When asked if utilizing a lattice model might tempt management to massage numbers, Kothari forecasts a strong likelihood: "In a way, yes," he affirms. "The reason is that the more parameters you give for management to choose, they are going to choose, in a fashion, what is more advantageous."

With FASB's preference plainly stated, and the auditing industry's likely bias toward congruence with FASB standards, it appears the lattice model will be picking up converts.

Still, McGee sees room for Black-Scholes on a more limited basis. "It gets down to what level of comfort the audit committee wants and the complexity of the company," McGee advises. "I think we'll see for companies with plain vanilla plans, and they're not that big, and the EPS hit isn't that big—they may actually stay with Black-Scholes."