A. Potential Benefit from Changing the Features (e.g., Contractual Term, Strike Price, Vesting Schedule and Attribution Methods) of Traditional Employee Stock Options
Given that traditional “at-the-money” ESOs will be expensed under FAS 123R, we recommend that firms take a fresh look at the design of these instruments. Ideallly, this would include an evaluation of the impact that changing the features of traditional ESOs will have on the firms HR goals and compensation expense. Table 5 summarizes the expected effect on fair value of making various changes to the features of ESOs.
Table 5.
Effect of Changing Option Features
Option Features Proposed Change Impact on Fair Value
Contractual Term Shorten Reduce
Strike Price Increase Reduce
Length of Vesting Period Decrease Reduce Increase Vesting Frequency Adopt Reduce
Graded Vested Schedule Adopt Reduce
Reducing the option’s contractual term will reduce fair value by reducing the length of time that the stock price is allowed to increase, but it also reduces the value of the instrument to employees. Increasing the exercise price increases the level of the stock price at which the option is “in-the-money,” which reduces both fair value and the value of the ESO to employees. However, it also provides a strong incentive for employees to increase the stock price. Decreasing the length of the vesting period will reduce fair value and the option’s retention value, but it increases the value of the option to employees.
Table 6 shows the effect of modifying the base case shown in Table 1 by:
•
•
•
•
Shortening the contractual term from ten years to seven years;
Shortening the maximum length of the vesting period from three years to two years;
Increasing the strike price from $30 to $33.50; and Increasing the vesting frequency from annual to monthly.
Table 6.
Reduce Contractual Term $10.64 -2.5% $12.17 -7.7%
Increase Strike Price $9.95 -8.8% $12.61 -4.4%
Reduce Vesting Period $10.14 -7.1% $12.63 -4.2%
Increase Vesting Frequency $9.53 -12.8% $12.28 -6.9%
These changes reduce the fair value produced by the lattice model by roughly 3%, 9%, 7%
and 13%, respectively. For the MBS model, making the same changes to the features of the ESO reduces fair value by roughly 8%, 4%, 4% and 7% respectively.
In most instances, the reductions are greater for the lattice model than for the MBS model.
The sole exception occurs when the option’s contractual term is reduced by three years.
This reduces fair value by roughly 8% for the MBS model and only 2.5% for the lattice model.
The MBS model shows a greater reduction in fair value because the 30% reduction in the option’s contractual term results in a predicted reduction of 16% in the ET. The reduction in ET, in turn, leads to a greater reduction in fair value predicted by the MBS-based model than that predicted by the lattice model. This analysis suggests that the MBS should not be relied upon to accurately assess the effect on fair value of incremental changes to the features of an ESO.
The reduction in fair value associated with the increase in the vesting frequency may be surprising. The reason for the reduction is that instead of having one third of the options vesting at the end of each of three years (which occurs with an annual vesting frequency), 1/36 of the options vest at the end of each month for 36 months (with a monthly frequency).
As a consequence, each year, the majority (11/12) of the grant vest earlier under a monthly frequency than under an annual frequency.
The percentages shown above are the separate influences of each option feature. The results would have been greater if more dramatic changes were made or the changes were made in combination. For example, if all of the changes were made in combination the reduction in fair value would have been roughly 27% for the lattice model and 21% for the MBS model.
B. Impact of Changing Vesting Schedules and Attribution Methods
Table 7 shows the total compensation expense associated with the two basic types of vesting schedules (cliff and graded) and attribution methods (FIN 28 and straight line), assuming 1000 options are granted.
Table 7.
Lattice Model-Based Compensation Expense
Under Alternative Vesting Schedules and Attribution Methods Timing of Compensation Graded Vesting - FIN 28 $6,299.49 $3,228.77 $1,384.53 $10,912.79 0.00%
Graded Vesting -
Straight Line $3,637.60 $3,637.60 $3,637.60 $10.912.79 0.00%
Cliff Vesting $4,153.59 $4,153.59 $4,153.59 $12,460.78 14.19%
The rows labeled “Graded Vesting - FIN 28” and shows total compensation expense based on the tranche-by-tranche calculation of compensation expense. That is, each tranche is treated as if it were a separate award. Tranche-by-tranche compensation expenses are computed by multiplying the values in Table 3 by the number of options allocated to each tranche (333.33 [1000/3]). In addition to the tranche-by-tranche method, FAS 123R (Footnote 86) also allows firms to compute total compensation expense by using a single weighted-average expected life to value the entire award.29
Under the FIN 28 method, total compensation expense is allocated on a straight line basis over each tranche’s vesting period (e.g., total compensation expense for tranche one is allocated over one year, total compensation expense for tranche two is allocated over a two year period etc.). With the “Straight Line” method, total compensation for all three tranches is allocated on a pro-rata basis over the maximum vesting period. .
Total compensation expense is roughly 14% greater for a three-year cliff vested option than for a three-year graded vested option. This occurs because as shown in Table 3, fair value increases with increases in the length of the vesting period. Hence, total compensation expense for a three-year cliff vested option, which allocates 100% of the grant to the third tranche, will generally be greater than the total compensation expense for a three-year graded vested schedule, which allocates options to all three tranches.30 For the MBS model, total compensation is 7.58% greater for the year cliff-vested option than for the three-year graded vested option.
29 As shown in Paragraphs 303 and 304 or FAS 123, this method tends to result in a greater compensation expense than the tranche-by-tranche method.
30 The fair value for a three-year cliff vested option is the same as the third tranche of a three-year ratable graded vested option. The only difference between the third tranche of a three-year ratable graded vested schedule and a three-year cliff vested schedule is that a fraction of the grant is allocated to the third tranche with a three-year graded vested schedule, whereas 100% of the award is, in essence, allocated to the third tranche for a three-year cliff vested schedule.
It should be noted that graded vesting does not always produce the lowest compensation expense. The exception can occur in situations where the departure rate is very high or the length of the vesting period is very long. In these situations, the number of options expected to vest (product of the number of options granted and the probability that the options vest) can become sufficiently small so that total compensation expense (the product of the number of options expected to vest and fair value) is lower for cliff vesting than for graded vesting.
With respect to the allocation of total compensation expense, firms are allowed to make a one-time election as to the attribution method they will use for options that are subject to graded vesting. As stated in Footnote 85, “The choice of attribution method for awards with graded vesting schedules is a policy decision that is not dependent on an enterprise’s choice of valuation technique.” However, there are important nuances that must be considered when making this election. For example, the amount of compensation expense recognized as of a particular point in time must be at least as great as the vested portion of the award up to that point. Also, the straight line attribution method is only applicable to options with service conditions. As a consequence, the straight line method can not be used for options that are subject to either performance or market conditions.
As shown in Table 7, compensation expense is the same for both the FIN 28 or “tranche-by-tranche” method and the straight line attribution methods; however, the timing of compensation expense is front loaded for the FIN 28 method. Hence, total compensation expense will be greater, on a present value basis, for the FIN 28 attribution method.
C. Impact of Using Nontraditional Instruments to Accomplish Attraction, Retention and Shareholder Alignment Goals
1. The new standard is expected to create a more level playing field for non-option equity-based instruments
Under FAS 123, nontraditional instruments were expensed and were generally subject to variable accounting. With the new standard, all equity-based instruments will be expensed and only liability instruments (e.g., cash settled stock appreciation rights) will be subject to variable accounting. As a result, it is expected that under the new standard, firms will make greater use of nontraditional instruments because all instruments will be treated the same and nontraditional instruments have the potential to allow firms to better accomplish their EBCP goals and reduce compensation expense. For example, instruments with market conditions (measures derived from the firm’s stock price) have the potential to better align employee and shareholder interests than traditional ESOs. Also, stock appreciation rights settled in stock appear to dominate traditional ESOs because they have the same fair value as traditional ESOs, produce less dilution and have a greater perceived value to employees.
Finally, nontraditional options generally have lower fair values than traditional ESOs.31 The nontraditional instruments that will be analyzed in this report include:
• Nontraditional options - Premium options32
31 As demonstrated later in this section, the instrument with the lowest fair value will not necessarily produce the lowest total compensation expense. The instrument that produces the lowest compensation expense will be the one for which the product of fair value and the number of instruments expected to vest is the lowest.
32 We could have also discussed the potential benefits or using discount options. A discount option is similar to a premium option, but the strike price is set below the grant date stock price. However, because of an expected change in the Internal Revenue Service Code (IRC), it is expected that this type of option will cease to be used by most firms.
Under the recently proposed IRS ruling, discount options will be both subject to, and in violation of, the provisions of
- Maximum value options - Purchased options - Indexed options
- Performance-based options - Market-based options
• Non-performance-based and performance- or market-based versions of non-option instruments
- Restricted stock - Restricted stock units
- Stock appreciation rights settled in stock or cash
Because of their complexity, more flexible models, such as lattice models, will generally required to value such nontraditional instruments as capped, indexed and performance-based options with market conditions as well as restricted stock, restricted stock units and stock appreciation rights where vesting or the number of instruments granted is contingent on market conditions. More flexible models, such as lattice models, are also required to determine the appropriate “exchange ratio” that will make employees indifferent between a particular nontraditional instrument and traditional ESOs.
2. Descriptions and evaluations of nontraditional instruments This section provides descriptions, objectives, pros and cons and fair value estimates for the nontraditional instruments discussed above. All of the valuations are based on inputs shown in Table 1 and use lattice-based and MBS-based models have been specifically designed to reflect the features of each of these nontraditional instruments. While it is possible to value all of the instruments with lattice-based models, because of its complexity, it is not generally possible to use a MBS-based model to value options with market conditions.
a. Premium options
With Premium Options (POs) the strike price is set above the grant date stock price. POs are designed to reduce fair value and to provide a stronger incentive than traditional ESOs for employees to increase the firm’s stock price (because these options begin under water).
As a consequence, more options will need to be awarded to make employees indifferent between POs and traditional ESOs. Assuming that the strike price is set $3.50 above the grant date stock price, the fair value of the PO will be about 9.0% lower than that of the traditional ESO analyzed in Table 3 (from $10.91 to $9.95) for the lattice model and about 4% lower (from $13.19 to $12.61) for the MBS model. Also, the value produced by the MBS model is 27.0% greater for the PO than that produced by the lattice model.
b. Maximum value options
Maximum Value Options (MVOs) are designed to reduce fair value by capping the “spread”
between the stock price and the strike price, where the maximum spread allowed is usually expressed as a multiple of the strike price. The objective of MVOs is to reduce the cost of the option, compared to a traditional ESO, without significantly reducing the value perceived by employees. A key advantage of MVOs, over other nontraditional options, such as an indexed option, is that they are easy to understand, have a lower fair value than a traditional
Section 409A of the Internal Revenue Code (IRC). Under the proposed regulations, both employee stock options and stock appreciation rights are generally exempt from Section 409A of the IRC, as long as the strike price is never less than the price of the underlying stock at the grant date.
ESO and can be designed so that the value to employees is similar to that of a traditional ESO. Table 8 shows fair values for MVOs, where the maximum intrinsic value is based on multiples of one, two or three times the strike price.
Table 8.
The fair value estimates above are based on both lattice-based and MBS-based models that have been specifically designed to reflect the features of an MVO. The fair value estimates produced by the lattice model are roughly 6% to 26% lower than that of a traditional ESO, while the MBS-based model shows reductions in fair value of 27% to 58%. As shown in the table, the MBS model tends to understate the fair value of MVOs. This occurs because the MBS model assumes that an MVO can be exercised only at its expiration date.
Consequently, the fair value of the MVO will be based on cash flows that range from zero to the maximum intrinsic value allowed and these cash flows will be heavily discounted (from the MVO’s expiration date to the grant date).
This situation is to be contrasted with a lattice-based model, which assumes that an MVO can be exercised any time after the option vests. In fact, it can be shown that MVOs tend to be exercised earlier than traditional ESOs, typically when the maximum allowed stock price is reached. As a result, the cash flows predicted by a lattice model for a MVO will generally be greater than those predicted by a MBS-based model (typically equal to the maximum intrinsic value allowed) and, because they are exercised earlier, these cash flows will generally receive less discounting than occurs with the MBS-based model.
Because MBS-based models tend to understate the fair value of MVOs, it is generally not advisable to use them to value this type of instrument. The same conclusion has been reached by PricewaterhouseCoopers (PwC).33 PwC states: “However, only lattice models should be used for certain alternative awards, including certain performance awards (those with market conditions), as well as options with payoff functions limited in certain ways (such as maximum value options) …”34
c. Purchased options
With Purchased Options (PUROs), the employee pays a fraction of the strike price at the grant date and the remainder when the option is exercised. PUROs provide a strong incentive for employees to increase the firm’s stock price and to remain with the company,
33 PricewaterhouseCoopers, 2005, Page 3-2.
because PUROs require employees to invest their own money. It should be noted that because the employee puts money at risk, the perceived value of the PURO will be less than that of a traditional ESO. Consequently, more options will be required to make employees indifferent between PUROs and traditional ESOs.
By requiring employees to pay a fraction of the strike price at the grant date, PUROs reduce the fair value of the option. For example, if the employee must pay 5% of the strike price at the grant date and the remaining $28.50 at vesting, then the fair value of the PURO would be
$9.88, or 9.4% less than the fair value of the traditional ESO ($10.91). The estimate produced by a MBS model, which has been modified to reflect the features of a PURO, is
$12.98. This is 1.6% less than the fair value of a traditional ESO ($13.19). Lastly, the fair value produced by the MBS-based model is 31% greater than that produced by the lattice-based model.
d. Indexed options
With Indexed Options (IOs), the strike price is not fixed, but instead varies according to an index that reflects either market, industry or peer group performance (e.g., S&P 500 index).
IOs are designed to reward employees when the company performance exceeds that of the index. As such, employees can be rewarded even when company performance declines, as long as it declines less than that of the index. The benefits of IOs are:
•
•
•
They provide a strong incentive for employees to improve performance;
They can provide incentives even when the company’s current stock price is less than the grant date price; and
The fair value of an IO can be significantly lower than that of a traditional ESO.
To make employees indifferent between IOs and traditional ESOs, additional options will need to be awarded. One potential problem with traditional IOs is that they may not be exempt from Section 409A of the IRC, because the strike price can become less than the stock price at the grant date. The problem could be alleviated by preventing the strike price from dropping below the grant date stock price. However, this change would reduce the IO’s value to employees.
Using lattice and MBS models that have been modified to reflect the features of IOs, the fair value produced by the lattice model is $6.15, which is 44% lower than the fair value of a traditional ESO ($10.91).35 Similarly, the fair value produced by a MBS model is $9.01, which is 32% less than the fair value produced by the MBS model for the traditional ESO ($13.19). Lastly, the fair value produced by the MBS for an IO is 47% greater than that produced by a lattice model.
e. Options with performance or market conditions
The new standard is expected to lead to an increase in equity awards whose payoff depends on attaining performance targets. As previously noted, under the new standard we expect that firms will make greater use of instruments with performance targets because they have the potential to provide incentives that will better align employee and shareholder goals and to reduce compensation expense. In fact, according to a recent article in the Wall Street Journal (2/21/2006), the use of instruments with performance targets has greatly increased.
In 2003, 17% of the major U.S. companies granted instruments with either vesting or the number of instruments awarded tied to performance targets. This percentage increased to
35 In addition to the inputs shown in Table 1, it is assumed that the volatility and dividend yield of the index are 40% and zero percent respectively and the correlation between the index and the stock price is 70%.