For companies that elect to estimate forfeitures, a forfeiture assumption (considering forfeiture experience to date and estimating future forfeitures) should be applied to awards that are classified as liabilities as well. Liability awards are remeasured at fair value each reporting period, and any impact of forfeitures or updates to the forfeiture estimate, although not affecting the fair value measurement of the awards, should be reflected at that time as well.

For companies that elect to estimate forfeitures, the forfeiture-rate assumption is typically expressed as the estimated annual rate at which unvested awards will be forfeited during the next year, which may or may not differ significantly by employee group. Some companies estimate the total forfeitures for the entire grant or for each vesting tranche. The forfeiture-rate assumption can be based on a company’s historical forfeiture rate if known. However, management should assess whether it is necessary to adjust the historical rate to reflect its expectations. For example, adjustments may be needed if, historically, forfeitures were affected mainly by turnover that resulted from business restructurings that are not expected to recur.

Companies could use separate pre-vesting forfeiture assumptions for different employee groups when they believe those groups will exhibit different behaviors. For example, based on its history and expectations, a company may develop a 5% annual forfeiture estimate for senior executives and a 10% annual forfeiture estimate for all other employees.

Example SC 2-16 illustrates how a company could apply its estimated annual forfeiture rate to an option grant.

EXAMPLE SC 2-16
Estimated annual forfeiture rate applied to an option grant

SC Corporation grants to its employees a total of 400 stock options that (1) vest upon the employees’ completion of a service condition and (2) have a four-year graded vesting schedule (25% or 100 awards per year). SC Corporation estimates a 5% annual forfeiture rate, based on its historical forfeitures. SC Corporation uses the following calculations to determine the number of options that are expected to vest:

How much compensation expense should SC Corporation recognize in year 1?

In this example, 88% of the options are expected to vest (352 options expected to vest/400 options granted). As discussed in SC 2.8, for awards with graded vesting features, companies will use either a graded vesting (accelerated) or straight-line attribution approach to recognize compensation cost over the vesting period. If a company uses an annual forfeiture rate for awards with graded vesting, as illustrated above, and the straight-line attribution approach to recognize compensation cost, there could still be some compensation cost that needs to be front-loaded to the earlier portions of the requisite service period. In this case, SC Corporation would begin expensing 95 options in year 1 under the straight-line attribution approach, rather than 88 options, because of the requirement to expense at a minimum the number of awards actually vested at each vesting date.

As each tranche vests, a company should assess the actual number of awards vested in order to comply with the requirement that the amount of compensation cost that is recognized on any date should at least equal the grant-date fair value of the vested portion of the award. For example, if all 100 options vest in the first year in the above scenario (i.e., no awards are forfeited in the first year), the company should recognize compensation cost for those 100 awards. Additionally, the company will need to re-evaluate the number of unvested options remaining and the reasonableness of the forfeiture-rate assumption used for the remaining requisite service period.

For companies that elect to estimate forfeitures, the forfeiture estimate should generally start with an analysis of the company's historical data covering several years. The group of the employee and terms of an award could affect the likelihood of the award being forfeited; therefore, companies should evaluate the pre-vesting forfeiture rate of awards by employee group and grouping awards with similar terms and using a specific forfeiture rate for each group of similar awards. For each grant, actual forfeitures should be compiled by period (e.g., one year from the grant date, two years from the grant date, etc.), and the percentage of the remaining outstanding unvested award forfeited each year should be computed. The company should then average those forfeiture rates to compute an average historical annual forfeiture rate.

When analyzing forfeitures, companies should segregate forfeitures into two categories: (1) pre-vesting forfeitures and (2) post-vesting cancellations, as defined earlier. Assume, for example, that a company grants 500 options and that 100 of the options vest each year, over a five-year requisite service period. The employee terminates employment after two years. His vested options are underwater, and thus, are not exercised. Accordingly, the 200 vested options are not pre-vesting forfeitures but, instead, post-vesting cancellations; the 300 unvested options are pre-vesting forfeitures.

Some software packages used to track stock option activity do not differentiate between pre-vesting forfeitures and post-vesting cancellations and, therefore, this data in some cases may be difficult to obtain. Additionally, startups and other companies that do not have a sufficient history to estimate the expected pre-vesting forfeiture rate might have to rely on surveys of, or disclosures by, other similar companies. However, ASC 718 does not require disclosure of the forfeiture-rate assumption; therefore, the ability to obtain public information on forfeiture rates may be limited.

Another factor that may be considered in developing a forfeiture assumption, or in adjusting historical forfeiture rates, is current human resources or industry near-term forecasts of anticipated employee turnover by employee group. An annual employee turnover rate and an annual forfeiture rate assumption may be comparable for this purpose.

Without proper recordkeeping, it will be difficult to accurately compute a historical pre-vesting forfeiture rate. Making accurate true-up adjustments to recognize actual forfeitures may also be difficult. Companies should review their recordkeeping systems to assess whether pre-vesting forfeitures can be separated from post-vesting cancellations; separating the two will ensure that companies sort the appropriate data to develop an accurate estimate regarding the pre-vesting forfeitures.

Example SC 2-17 illustrates how estimated forfeitures and actual forfeitures interrelate with different vesting conditions.

EXAMPLE SC 2-17
Accounting for actual and estimated forfeitures for each type of vesting condition

SC Corporation grants its employees 5,000 stock options on January 1, 20X1. The grant-date fair value is $8 per option.

All of the options cliff vest after three years of service. The company has elected a policy to estimate forfeitures. In 20X1 and 20X2, SC Corporation estimates that 95% of the options will vest. In 20X3, SC Corporation completes a significant restructuring, which results in only 45% of the options vesting because 55% of the options are forfeited prior to vesting. Because the actual pre-vesting forfeiture rate differs dramatically from management’s prior expectations, the company will recognize a credit to compensation cost in 20X3 as shown below.

The options are subject to a three-year service condition and a performance condition based on each employee achieving a specific cumulative sales target over the period from 20X1 through 20X3. In 20X1, SC Corporation estimates that 90% of its employees will achieve their targets and remain employed through 20X3 (i.e., 90% of the options will vest). At the end of year 2, however, SC Corporation reassesses the likelihood that the targets will be achieved and determines that 95% of the employees will achieve their targets by the end of 20X2 and remain employed through 20X3. Due to a new competitor’s product that is launched in 20X3, only 75% of employees actually achieve the cumulative sales targets.

The options become exercisable only if the employee remains employed by SC Corporation for three years and SC Corporation’s stock price outperforms the S&P 500 Index by 10% during that three-year vesting period. The requisite service period is three years because that is the explicit period for the market condition and the date that the employee must be employed in order to vest in the award. As a result of the market condition, the fair value of these options is $4.50. Ninety-five percent of the employees are expected to complete the requisite service period at the end of both 20X1 and 20X2.

At the end of the three-year period, SC Corporation's stock price has outperformed the S&P 500 Index by only 3%. Therefore, no awards are exercisable. Additionally, 10% of employees did not complete the three-year requisite service period as compared to the estimated forfeiture rate of 5%. In this scenario, the compensation cost should be adjusted to reflect actual forfeitures; however, compensation cost should not be reversed for the 90% of the employees who fulfilled the requisite service period of three years, even though the market condition was not met.

Assume the same facts as in Scenario 1, except that the company has elected to account for forfeitures as they occur. In 20X1, 20X2, and 20X3, actual forfeitures are 0, 750, and 500, respectively.

A “last man standing” arrangement is an agreement with more than one employee whereby if the employment of one of the employees is terminated prior to the end of a defined vesting period, the stock-based compensation awards granted to that employee will be reallocated among the remaining employees who continue employment. Because each employee has a service requirement, each individual grant of stock-based compensation awards should be accounted for separately. Generally, the accounting for a reallocation under a “last man standing” arrangement is effectively treated as a forfeiture of an award by one employee and regrant of options to the other employees. Therefore, if and when an employee terminates his or her employment and options are reallocated to the other employees, the reallocated options should be treated as a forfeiture of the terminated employee’s options and a new option grant to the other employees.