Although the Exposure Draft characterized lattice models as “preferable", neither FAS 123 R nor the Exposure Draft prescribes the use of a specific option-pricing model. However, both documents require that an option-pricing model take into account the first 6 inputs listed in the above table, at a minimum. The directional impact of changes in each of these assumptions is described in the following table:

FAS 123 R - Measurement Objectives of Fair Value

The chosen valuation model must meet all three of the requirements stated in Statement 123R, paragraph A8, in order to meet the fair value measurement objective:

1. the model is applied in a manner consistent with the fair value measurement objective and other requirements of Statement 123R,
2. the models is based on established principles of financial economic theory and generally applied in that field, and
3. the model reflects all substantive characteristics of the instrument.

General Practices

Service Based Options: For a share option in which the exercisability is simply based on passage of service time, either Black-Scholes or Binomial model meets the above three measurement objective.

Performance Based Options: For a share option in which the exercisability is conditional, e.g. on a specified increase in the price of the underlying shares, the Black-Scholes model would generally not be an appropriate valuation model because, while it meets both the first and second criteria, it is not designed to take into account that type of market condition.

SAB 107 - SEC's Position

SEC understands that estimates of fair value of employee share options, while derived from expected value calculations, cannot predict actual future events. If a valuation technique or model is used to estimate fair value, entities may concern about its estimates of fair value to be materially misleading because the estimates of fair value do not correspond to the value ultimately realized by the employees who received the share options. Per SAB 107, it clearly states taht "as long as the share options were originally so measured, changes in an employee share option’s value, no matter how significant, subsequent to its grant date do not call into question the reasonableness of the grant date fair value estimate."

SEC would not object to a company’s choice of a technique or model as long as the technique or model meets the fair value measurement objective. A change in the valuation technique or model used to meet the fair value measurement objective would not be considered a change in accounting principle. As such, a company would not be required to file a preferability letter from its independent accountants as described in Rule 10-01(b)(6) of Regulation S-X when it changes valuation techniques or models. However, the staff would not expect that a company would frequently switch between valuation techniques or models, particularly in circumstances where there was no significant variation in the form of share-based payments being valued. Disclosure in the footnotes of the basis for any change in technique or model would be appropriate.

Option123 provides three models for valuation. Entities may use one or all three models that meet the fair value measurement objective before making their selection as to the appropriate technique or model. 

FASB's Exposure Draft Does Indicate Lattice Models are Superior to Closed-Form Models.

The Exposure Draft defines a lattice model as a model that produces an estimated fair value based on the assumed changes in prices of a financial instrument over successive periods of time. The binomial model is an example of a lattice model. In each time period, the model assumes that at least two price movements are possible. The lattice represents the evolution of the value of either a financial instrument or a market variable for the purpose of valuing a financial instrument.

Historically, most companies have used the Black-Scholes model to value employee stock options. However, as a result of the FASB’s research and discussions with members of the FASB’s Option Valuation Group, it concluded that the use of the Black-Scholes model or other closed-form option-pricing models that require static assumptions usually are not the best methods to estimate the fair value of a typical employee stock option. While closed-form option-pricing models may be modified to address some of the characteristics of employee stock options (e.g., specifying an expected term for the option rather than the contractual term to address observed early exercise behavior of option holders), such modifications typically require simplifying assumptions that the FASB believes may result in measurement error. Because of the limitations of Black-Scholes model, the Exposure Draft indicates that the use of a more complex lattice model (e.g., a binomial model) that takes into account employee exercise patterns based on the dynamics of a company’s stock price and other salient variables and provides for other dynamic input assumptions will result in a better estimate of fair value.

The concepts that under lattice models and the Black-Scholes model are the same, but the key difference between a lattice model and a closed-form model is the flexibility of the former. For example, a lattice model can explicitly use dynamic assumptions regarding the term structure of volatility, dividend yields, and interest rates, as demonstrated in the above example. It can also specify a relationship between volatility and past returns whereby negative stock returns are associated with increases in volatility and positive stock returns are associated with decreases in volatility. Further, it can incorporate assumptions about how the likelihood of early exercise of an employee stock option may increase as the intrinsic value of that option increases or how employees may have a high propensity to exercise options with significant intrinsic value shortly after vesting. In addition, it can incorporate market conditions that may be part of an option’s design such as a requirement that an option is only exercisable if the underlying stock price achieves a certain level (“target market price” awards under FAS 123 R). Because of the versatility of lattice models, the FASB believes that they can provide a more accurate estimate of an employee stock option’s fair value than a value based on a closed-form, Black-Scholes formula.

It is anticipated that more entities will switch to use Binomial model as more options will be granted based on performance/market conditions when they are required to expense options.

The Black-Scholes Option Pricing Model

As indicated above, while the Black-Scholes model is complex, the application of the formula in practice is relatively easy and thus it has been widely used by financial professionals in the world. The formula has been programmed into Option123 which make the calculation as simple as clicking a mouse. However, a number of assumptions underlying Black-Scholes model are such that the formula may be better suited to valuing short-term, exchange-traded stock options than employee stock options. Black-Scholes model is less effective as a valuation technique for employee stock options that have long term to expiration and subject to vesting provisions, term truncation, or blackout periods (a specified period around the release of earnings information during which no options may be exercised and sold). These requirements would significantly reduce the likelihood of option exercise during a blackout period. Blackout periods are not readily incorporated into a valuation using the Black-Scholes model but, as discussed below, can be incorporated into a lattice valuation.