Yes, employee stock options can be valued: recent history can provide clues for valuing employee stock options (ESOs). In fact, some experts say they can be modeled much like mortgage securities. (Stock Options).

Raise the subject of employee stock options (ESOs) to a group of financial executives, and you're likely to stir strong opinions: "They can't be valued." "How can you value something that is not traded in an open market?" "Why would you want to value them?"

Answering the last question is easy: The current move to provide more transparency in financial reports appears likely to result in companies being required to account for ESOs as an expense on their financial statements.

The answer to "how" may seem problematic, but based on "Valuing Employee Stock Options: A Comparison of Alternative Models," a recent study conducted by Financial Executives Research Foundation Inc. (FERF) and Analysis Group/ Economics (AG/E), it's not as difficult as some make it out to be -- and recent history provides clues to a solution.

"The financial community has already solved much more complicated problems than the valuation of employee stock options," says the study's author and lead researcher, John Finnerty, principal at AG/E and professor of finance at Fordham University.

Indeed, this situation mirrors the concern many had about mortgages back in 1979. Then, during the early 1980s, investment bankers began packaging them into different classes of mortgage pass-throughs that could be bought and sold like bonds. These asset-backed securities became known as "collateralized mortgage obligations" -- CMOs.

The value of these securities was based on the right to cash flows from mortgage payments, which include both interest and return of principal. To get around the problems of defaults, prepayments and variable long-term interest rates, investment bankers collected publicly available data on mortgage prepayments and defaults. From this data, they could then make certain assumptions about the relative probabilities of future cash flows.

Fast-forward to 2003: the mortgage-backed market has grown to more than $4 trillion outstanding, with as much as 70 percent of all mortgages now "securitized" into over 30 classes of asset-backed securities, including fixed- and adjustable-rate residential mortgages, commercial mortgages, automobile receivables and motor home loans.

Is It a Case of "Apples and Oranges?" Why compare mortgages with ESOs? To start, why not compare ESOs with publicly traded options, or exchange traded call options?

An exchange-traded call option gives the buyer the right -- but not the obligation -- to purchase an underlying security at a given strike price until the expiration date. Most valuation techniques consider six standard inputs: current stock price, exercise price, option duration, risk-free interest rate, dividend yield and volatility These options are completely transferable, do not have vesting requirements and are not forfeitable.

ESOs are long-dated call options granted to employees by their company to purchase shares of the company's common stock. They typically have contractual lives of up to 10 years, and usually require vesting periods of between one and five years. They almost never can be transferred or sold. Therefore, the only way for a holder to receive cash is to exercise them.

Traditional Valuation Models

Exchange-traded call options are usually valued using the Black-Scholes model, first published in 1973. Modified by Robert Merton to incorporate the payment of cash dividends, the resulting model (Black-Scholes-Merton, or BSM) has been widely used to value exchange-traded put and call options.

"The BSM model values call options on publicly-traded common stocks, assuming that the options, like the stocks, trade in a liquid market," explains Finnerty. "Additionally, it assumes that the risk-free interest rate and the stock's price volatility are constant during the life of the option. Neither of these assumptions is true for options with long lives, such as ESOs."

Statement of Financial Accounting Standards No. 123 (FAS 123) permits an adjustment to the BSM model to account for non-transferability after the vesting period. This adjustment is the substitution of the option's expected life for its contractual life.

The adjusted BSM model is currently the most common model used for valuing ESOs. However, FAS 123 does not allow companies to apply a discount to the BSM value to reflect the ESO's lack of transferability during the vesting period.

Prior studies by AG/E have estimated that even for low-volatility stocks, BSM overstates the value of ESOs by as much as 50 percent. For high-volatility stocks, the figure can be as high as 75 percent. As a result, CFOs are understandably reluctant to expense these inflated values on their income statements.

Two New Valuation Models

To address the unique features of ESOs that differentiate them from exchange-traded options (see chart on page 55), AG/E has developed two new valuation models. One is an enhanced BSM model; the second, a binomial tree model, is especially useful in valuing ESOs with more complex features such as indexed and performance-based options. Both models are calibrated to AG/E's proprietary database which contains extensive ESO data for more than 95 separate grants totalling over 30 million ESOs, and the client company's actual experience.

"We approach the ESO valuation problem in the same manner that Salomon Brothers addressed the mortgage issue in the early 1980s," says Rod Parsley of AG/E. "By modeling the behavior of individuals within the firm, we effectively derive the amount that an outside investor would pay at grant date for the right to obtain the actual cash flows realized by employees from a particular pool of ESOs."

Thus, just as CMOs pool mortgage pass-throughs into specific tranches, AG/E models the expected cash flows to be realized by specific groups of employees from specific grants. "The market price of this hypothetical ESO-backed obligation represents the opportunity cost to the firm, and, therefore, the fair market value of the ESO grant," explains Parsley.

Because both models account for the unique characteristics of ESOs, both typically produce values substantially below those derived from the BSM model or the BSM model with FAS 123 modifications.

* The AG/E Enhanced BSM Model. Extending the BSM Model, this approach reflects ESOs' vesting requirements, transfer restrictions, early exercise and forfeiture features. Variables representing ESO exercise and forfeiture rates are inserted as separate parameters to enhance the model.

With these modifications, the model measures the fair value of ESOs based on the hypothetical price a fully-diversified outside investor would pay in an arm's-length transaction to purchase the right to receive the ESO's cash flow (as distinct from the ESO itself, which is not transferable).

Forfeiture before vesting is handled by calculating the fraction of ESOs that can be expected to never vest (and as a result, to expire worthless) based on actual ESO data. Actual ESO data are also used to calculate the fraction of ESOs that will be forfeited without exercise after the vesting date (but prior to the expiration date). In the event that no ESO history exists for a particular company, data from a sector basket of like companies are used to estimate the parameter values.

Transfer restrictions are taken into account in two ways: 1) Actual ESO exercise data are used to calculate the fraction of ESOs that holders would be expected to exercise each year prior to the vesting date were it not for the vesting restrictions. 2) The holder's inability to exercise or transfer the ESOs during the vesting period is modelled as the loss of the value of a put option, and the value of this put option is calculated and subtracted from the value of the (unrestricted) ESO.

AG/E also uses this analytical approach to calculate the discount for lack of free transferability when valuing unregistered privately placed shares of common stock.

* The AG/E Binomial Model. Designed to both meet the requirements of FAS 123 and incorporate the unique features that differentiate ESOs from exchange-traded options, this model explicitly accounts for such standard ESO features as early exercise, forfeiture and vesting requirements.

The model assumes that the price of the underlying stock evolves according to a binomial process and that exercise decisions are made so as to maximize the expected utility of the option holder's terminal wealth.

This model applies a two-step process: In the calibration step, it determines the expected life of the stock option. This value is then compared with the expected life calculated from the observed data on forfeiture and early exercise, and the model parameters are adjusted to equate the two, hi the valuation step, the calibrated model is used to value the ESO.

The model can address such nonstandard features as indexed strike prices (indexed ESOs), performance-vested ESOs (vesting can occur only if the stock price exceeds a preset level), repriceable ESOs (where the strike price may be reset if the ESO is too far "underwater"), purchased ESOs (where the employees must pay a certain percentage of the strike price at the grant date and the remainder when the option is exercised) and blackout periods.

"When ESOs are valued at grant date, it cannot be known for sure how many will be exercised, nor how many will be forfeited and therefore never exercised," says Finnerty.

"However, you can value ESOs if you make some reasonable assumptions -- based on probabilities -- about how many ESOs will be exercised, when and at what price. ESOs are much less complex than many of the asset-backed products used regularly in the financial world. As with mortgages, if you can estimate future cash flows, you can model the timing of those cash flows to calculate their present value," adds Finnerty.

RELATED ARTICLE: ESO Features

ESOs have the following features that differentiate them from exchange-traded options.

Vesting requirements -- Can only be exercised at the completion of the vesting period. If the employee leaves the company before ESOs vest, they automatically expire and are worthless.

Forfeiture provisions -- Subject to forfeiture or forced early exercise after vesting if an employee terminates employment due to dismissal, retirement, death, disability or voluntary termination.

Non-transferability provisions -- Must be held by the employees, they cannot sell of gift their options to a third party.

Other stated restrictions -- Could include blackout periods for directors and officers (described in Section 16 of the Securities Act of 1934), similar to non-transferability provisions.

Early exercise behavior -- Employees tend to exercise prior to expiration for a variety of reasons, including immediate cash needs risk aversion and portfolio diversition.

William M. Sinnett ([e-mail]) is Manager of Research for Financial Executives Research Foundation (FERF). The study can be purchased at www.fei.org/rfbookstore/. See AE/G at www.analysisgroup.com/eso.htm.