ASC 718-10-55-11 permits companies to select the option-pricing or equity valuation model that best fits their unique circumstances if the valuation technique:

As a result, for most employee stock options and other employee equity instruments, companies will have flexibility in selecting the option-pricing or equity valuation model used to estimate the fair value of their stock-based compensation awards.

The Black-Scholes model is relatively simple to use and well understood in the financial community. Its relative simplicity stems, in part, from the fact that when estimating the fair value of an employee option, all expected employee exercise behavior and post-vesting cancellation activity is reduced to a single average expected term assumption.

The principal advantage of lattice models, on the other hand, is that they can accommodate a wider range of assumptions about employees’ future exercise patterns, as well as accommodate other assumptions that may change over time. This approach may yield a more refined estimate of fair value.

A Monte Carlo model simulates a very large number (e.g., 1,000,000) of potential stock price scenarios over time and incorporates varied assumptions about volatility and exercise behavior for those various scenarios. A Monte Carlo model also has the ability to incorporate more complex market conditions. A fair value is determined for each potential outcome. The grant date fair value of the award is the average of the fair values calculated for each potential outcome.

For awards with typical service conditions and performance conditions, the Black-Scholes model will generally produce a reasonable estimate of fair value. Monte Carlo simulation and lattice models result in a more refined estimate of fair value. Additionally, companies that issue awards with market conditions or payoff conditions that limit exercisability should use either a Monte Carlo simulation model or a lattice model because those models can better incorporate assumptions about exercisability in relation to the price movements of the underlying stock and/or potential payoff outcomes related to achievement of market conditions.

Companies will need to weigh the advantages and disadvantages of each model in choosing a model that fits their particular circumstances. In deciding which model is most appropriate, some factors to consider are:

Companies may decide to change from one option-pricing or equity valuation model to a different one (e.g., from Black-Scholes to a lattice model). A change in option-pricing model is not a change in accounting principle—the underlying objective of estimating the fair value of the award is the same—and therefore does not require justification of preferability (or a preferability letter in the case of an SEC registrant). However, changes in valuation models should generally only be made when the new model will result in an improved estimate of fair value. Additionally, companies may use one model for certain awards and another model for different types of awards. For example, the fair value of a "plain vanilla" option could be estimated using the Black-Scholes model while a Monte Carlo simulation is used for an option with a market condition.

SAB Topic 14 requires companies to disclose any changes to the option-pricing model they use and the reasons for the change. Because Monte Carlo simulations and lattice models are generally considered to provide more refined estimates of fair value than the Black-Scholes model, we believe that once a company adopts a Monte Carlo simulation or lattice model to value a particular type of award, although this is not a change in policy that would require preferability, it would likely be difficult to support switching back to the Black-Scholes model to value that same type of award.