FASB EXPOSURE DRAFT
Specific Issues Upon Which FASB Is Seeking Comment:
Recognition of Compensation Cost
FASB seeks comments on whether employee stock options are an expense.
This is a radical change from their complete dismissal of this issue in the Invitation to Comment and the IASB’s refusal to seek comment on this issue in ED 2. Companies should continue to express their strong views that employee stock options do not constitute an expense and explain the basis for their position.
FASB concluded that a non-expensing standard with pro forma (“as if”) disclosure of what the income statement would have looked like if expensing had occurred is not an acceptable substitute for mandatory expensing. They seek comment on whether constituents agree with this conclusion.
Given that there are so many issues surrounding valuation and considerable debate still exists as to whether employee stock options represent an expense in the first place, disclosure of both (with non-expensing being the rule and expensing being the pro forma) would make sense. Those who believe that the purported expense is a meaningful number would have it and see its impact on the financial statements. Those who believe the number is meaningless could simply ignore it. Although the SEC has generally disfavored pro forma financial statements, FASB sanctioned pro forma financial statements would recognize the considerable valuation problems that exist in this area.
Measurement
FASB also is seeking comment on whether the grant date is the appropriate measurement date.
Potential alternative measurement dates would be the vesting date or the exercise date – each of which has its own problems.
Fair Value
FASB reaffirmed that the determination of fair value requires a willing buyer-willing seller analysis. App. B, ¶4. FASB also believes that market prices are the most reliable measure of fair value. However, if market prices are unavailable, FASB would require that option pricing models be used. Their preference is for a lattice, or binomial, model. FASB seeks comment on whether they have provided sufficient guidance to “ensure that the fair value measurement objective is applied with reasonable consistency” and whether constituents agree with the conclusion that the fair value of options can be measured reliably with options pricing models.
NVCA believes that the answer to this is no – the models actually will result in less transparency and less consistency across financial statements.
FASB requires that the model used “reflect[ ] any and all substantive characteristics of the instrument.” App. B, ¶5. Neither binomial models nor Black Scholes can reflect “any and all substantive characteristics” of an employee option.
FASB has said that it wants flexibility in the standard so that new models can be used when, and if, developed. However, upon a close reading, their proposed standard would seem to preclude the use of any model that is not based on Black-Scholes or a lattice models. Not only must any model used contain the Black-Scholes inputs, but the model must be based on established principles of financial economic theory that are “generally accepted by experts in that field (paragraph B9).” The cross reference to paragraph B9 refers only to lattice and closed and models such as Black Scholes. Thus, as a practical matter, new models could not be used unless they were simply further derivations of Black-Scholes or binomial models.
FASB believes that no restrictions that exist during the vesting period should be taken into account in valuing the options. They believe that this treatment is appropriate because only instruments that vest are ultimately expensed (as a result of the accounting for options that are forfeited before they vest). Again, however, not accounting (discounting) for vesting restrictions overstates the value of an employee option and violates FASB’s fair value principle.
FASB is seeking comments on whether a specific method for estimating volatility should be used and if so, what method should be used and why. The Exposure Draft states that historical volatility should not be used in most instances. Instead, companies would be required to consider the extent to which “future experience is reasonably expected to differ from historical experience.” Essentially, companies will be required to predict the impact of future events on their future volatility and convince their auditors to sign off on their assumptions.
FASB requires that such predictions reflect the life of the option. Thus, companies could be required to make predictions up to 10 years in the future.
Aside from the clear unreliability of any such predictions and any valuation based thereon, unless stock options were used sparingly by the company (i.e., any expense number is immaterial), it is doubtful that any auditor would sign off on such predictions.
For newly public entities, FASB would allow volatility predictions to be based on the volatilities of “entities that are similar except for having publicly traded securities.” App. B, ¶25. This is problematic in that (1) many times it is impossible to find a “similar” public entity and (2) that a public entity’s experience may bear no relation to the experience of a non public entity.
FASB has concluded that the nontransferability attribute of employee stock options is accounted for in a valuation model by adjusting an option’s contractual terms for expected early exercise and post-vesting employment termination behavior. In essence, they believe that shortening the life of the option accounts for this attribute. They seek comment on whether this gives appropriate recognition to the unique characteristics of employee options.
Adjusting the “life” of the option from the contractual life may account, in part, for the fact that employee stock options vest. However, this adjustment does not address the other restrictions that exist such as the fact that an option cannot be transferred, hedged, pledged, or sold.
FASB also asks for comments on alternative methods for reflecting the impact of these restrictions.
FASB has stated that where it is impossible to estimate the fair value of employee stock options, companies be required to use an “intrinsic value” method where the value is adjusted each reporting period. Non-public entities may elect this method, but if they do so, they must apply this method for the life of the option. FASB seeks comment on whether this is an appropriate method and, if not, what alternatives might exist. This purportedly “new” intrinsic value method is nothing other than variable accounting under APB 25.
Even staunch expensing advocates like John Biggs, formerly with TIAA-CREF, have stated that this type of accounting makes no sense and is confusing to investors. In essence, it will bring the stock price right into the income statement. The expense will increase or decrease each quarter depending on the price of the underlying stock. In some instances, the expense might even be negative. Try explaining that to investors.
Attribution of Compensation Cost
The exposure draft proposes that compensation expense generally be recognized over the vesting period. They seek comments on whether this is appropriate.
To the extent that an expense is recognized, the vesting period is an appropriate period to use.
FASB has provided additional criteria to follow if an award is subject to additional conditions. These new criteria relate to, for example, performance based options.
We have not yet studied this because few, if any, of our members actually issue these types of options.
In essence, FASB would require that where options vest on a “graded schedule,” that each set of options constitutes a separate award that must be valued separately and accounted for separately. For example, if a company issued 144 options on 1/1/x, and 36 of those options vested on 12/31/x and 3 options vested at the end of each of the next 36 months, FASB would view this as 37 different options grants that the company would be required to value and account for separately. Additional issuances to, for example, new employees, also would presumably have to be accounted for separately to the extent they were issued other than on a company-wide grant date.
Simply put, this would be a nightmare. It would be extremely costly to do, very time consuming, and unauditable.
Interestingly, all of FASB’s examples relate to options granted to one or only a limited number of executives. NVCA believes that companies with broad-based plans would encounter significant difficulty and cost.
In addition, the exposure draft would require companies to group their employees for purposes of predicting exercise behavior. Again, in practice this would be costly and complicated and ultimately useless. For example, would each compensation grade have to be analyzed separately? Age group? Sexes? Races? Employees with college-aged children? Employees with sick relatives? Newly married employees who may want to buy their first house? The possibilities are endless.
Modifications and Settlements
FASB has provided additional guidance on how to account for awards that are modified or settled, other than through exercise by the employee. This has not historically been something that our member companies do. We have not yet studied the impact of the proposals.
Income Taxes
FASB has proposed a new method for accounting for the income tax effects of options. This proposed method differs from that contained in FAS 123 and the IASB’s proposed standard. We have not yet studied the impact of the proposal and companies may have different views on the appropriateness of FASB’s proposals.
Transition
In essence, the proposal would require unvested options to be expensed based on the Black Scholes value that was contained in prior footnote disclosures. Newly issued options would have to be valued based on the new standard. Thus, during the transition period, outstanding options would be valued in radically different ways. FASB seeks comment on the transition rules and asks whether companies should be allowed to apply the rules retrospectively to prior financial statements (for period to period comparability purposes).
This clearly will result in apples and oranges being in the income statement and none of the values, whether computed under Black Scholes or a binomial model, will be reliable.
In addition, estimates made years before the rule change was even contemplated will now affect financial statements going forward even though FASB has acknowledged that those estimates are inaccurate.
Nonpublic Entities
FASB states that it is attempting to alleviate the concerns of nonpublic companies by, for example, allowing them to use the intrinsic value method (adjusted each period) and extending the proposed effective date by one year. FASB seeks comments on whether these decisions are appropriate and whether any additional modifications be made for nonpublic entities.
The intrinsic value method and problems therewith is discussed above.
A one-year extension of the effective date is simply not long enough. Bottom line, the compliance costs that would be imposed on nonpublic companies would far outweigh any possibly perceived benefits resulting from a mandatory expensing standard.
Differences Between the Exposure Draft and the IASB Proposed Standard
FASB notes that there are several differences between the exposure draft and the IASB’s proposed standard. They seek comments on whether constituents prefer the IASB approach over the approach in the exposure draft and why. They also seek comments on whether, if a constituent prefers the treatment in the exposure draft, the constituent believes that FASB should nevertheless consider the IASB treatment in the interest of achieving convergence.
We have not fully analyzed these differences, but is seems doubtful that we would prefer very much in the IASB document given the complete lack of due process afforded by the IASB and the general unfamiliarity in other countries with employee stock options.
Understandability of the Exposure Draft
FASB states that its goal is “to issue financial accounting standards that can be read and understood by those possessing a reasonable level of accounting knowledge, a reasonable understanding of the business and economic activities covered by the accounting standard, and a willingness to stud the standard with reasonable diligence.” They then seek comment on whether “you believe that this proposed Statement, taken as a whole, achieves that objective?”
FASB has generally failed to recognize the essential elements of employee stock options that differentiate them from freely tradable options. Where they attempt to recognize these differences, for example, by stating that black out periods should be taken into account in the valuation, they do not provide sufficient guidance to explain how they believe this should be done. FASB also has proposed a standard that is wholly unworkable in the real world (e.g., treating each separate vesting as a separate grant) and, because of all the predictions of future events that would be required, unauditable.
ESPPs
FASB has taken the position that any ESPP is compensatory, and an expense must be recognized, unless all holders of the same class of stock are entitled to purchase shares of stock on no less favorable terms than the employee. This is a significant departure from current rules, but is not that significant of a departure from what is contained in FAS 123 (5% safe harbor).
FASB’s Exposure Draft position is that any discount that is not offered to all holders of that class of stock is compensatory. This is a difficult issue to deal with, except on public policy grounds.
The comment period deadline is June 30, 2004.
