United States: New Stock Option Accounting Must Be Applied Properly

The new stock option accounting is probably the most controversial accounting change in years. This is because (i) some companies view stock options as an integral part of their compensation strategy, and are resisting the substantial income change, and (ii) the methods of valuing employee stock options are not well established or accepted. To avoid overstating the new stock option expense, companies will need to carefully consider how this new pronouncement is implemented.

Since 1972, companies have generally treated options as an expense only if the option’s strike price was below the company’s share price when the options were granted. Because of this and similar tax provisions, virtually all employee stock options are issued with strike prices equal to the underlying stock’s then-current market price.

In 1995, the Financial Accounting Standards Board (FASB) attempted to change this accounting to expense the value of the options in the income statement. This proposed change reflected what everyone knows; namely, that stock options (i) compensate employees and (ii) dilute other shareholder interests. However, the FASB bowed to Congressional threats and made the proper accounting treatment optional, with only footnote disclosure of the options (See This Accounting Fraud is Sanctioned). Under the original Statement of Financial Accounting Standard (SFAS) 123, companies were required to expense only the intrinsic value of the option. This was practically always zero.

In December 2004, after heated debate, the FASB approved a revision to the flawed 1995 pronouncement, and issued SFAS 123(R). The new accounting considers future possibilities, so even if the stock’s market price is equal to the option’s strike price, the option still has a positive fair value. This fair value is recognized as a compensation cost in the income statement over the vesting period. Companies are also required to recognize a compensation cost for unvested stock options outstanding on July 1, 2005. This transition cost is equal to fair value of those options at the original grant date, amortized over the vesting period. The income difference between the old and new accounting is often substantial.

The new rule treats existing vested options as having no ongoing expense impact. This allows future costs to be eliminated by accelerating the vesting of "underwater" options. Underwater or out-of-the-money options are those whose exercise price is less than the option exercise or strike price. Under this approach, the company can get away with only reporting the vesting of these options in the footnotes before the footnote-only possibility disappears.

Some of the companies accelerating these options claim that the options are "worthless". If the options were really worthless (of course, they are not!), the same financial statement result could be achieved by canceling the options. While eliminating the direct income statement impact, vesting acceleration has an undesired effect on employee retention. This occurs because the employee would have otherwise needed to remain employed at the company to become vested.

According to Glass Lewis & Co., about 30 companies have so far taken advantage of this questionable technique. We suspect as the July 2005 implementation date arrives, more companies will similarly accelerate option vesting in order to avoid the pronouncement.

The accepted stock option valuation models were created to price exchange-traded options, rather than employee stock options. The following significant differences cause overstated employee stock option values:

1. Employee stock options are not transferable. In contrast, exchange-traded options can be bought and sold freely. Employee stock options can only be exercised by the original employee holder, and are worthless to anyone else.

2. Employee stock options often will have a life of up to ten years, while exchange-traded options generally have a life of six months or less. The model most widely-used to value options was never intended to address long-term options. Its mathematics mechanically overstate values when applied to longer periods.

3. Employee stock options vest over time and are forfeited if employment is terminated. Exchange-traded options have no comparable provisions.

Because of these differences, the value of employee stock options should be significantly lower than exchange-traded options. To avoid overstating options’ expense, modifications must be made to the traditional option pricing models, or custom models must be employed.

While SFAS 123(R) doesn’t mandate a specific valuation method, it does outline minimum criteria. As long as these minimum requirements are addressed, the company can identify the valuation approach and valuation assumptions. Since all accounting must be consistently applied from period to period, it is important that the initial application of SFAS 123(R) be done carefully.

The following two valuation methods are most common. Both meet SFAS 123(R) requirements:

1. Black-Scholes

Named after its Nobel Peace Prize winning authors, the Black-Scholes option-pricing model is the most widely taught and best-known option pricing model. The Black-Scholes model consists of a complex mathematical formula that takes into account the following inputs:

1. Strike price
2. Time to expiration
3. Volatility
4. Risk-free rate

For all its complexity, Black-Scholes assumes a number of items that are subject to obvious criticism:

1. Option exercise occurs only at the end of an option’s contractual term
2. Volatility, dividends AND risk-free interest rates are ALL constant over the option’s term
3. No dividends are paid during the life of the option

Over the years, several Black-Scholes variations emerged to adjust for some of its static assumptions, but the significant limitations remain.

2. Lattice / Binomial

Instead of the Black-Scholes models static assumptions, lattice models are able to address variable assumptions during the option’s life. The effect can be significant. For example, some employees will exercise their options upon vesting, while market conditions will cause more employees to wait in other cases. Lattice models attempt to capture these varying patterns.

Lattice models get their name because they can be shown as a decision tree that has a ladder or lattice appearance. Each branch of the decision tree represents a different assumption and related probability. For example, the following chart shows a two-year lattice model that charts the expected price changes of the option. Each number reflects the option’s potential year end price.

The lattice model uses data collected about employee exercise behavior and stock-price volatility to project an array of future possibilities. Compared to the Black-Scholes model, the lattice structure allows a range of assumptions. Data analysis must support the assumptions. The result is a more accurate (and often lower) option value and related lower expense.

The lattice approach also provides more flexibility in future measurements, since this method is not as mechanical as the Black-Scholes model. This flexibility can occur without violating the consistency requirement that pertains to all accounting applications.

The lattice structure models allow for multiple terms of uncertainty (i.e. - binomial, trinomial, multinomial). With more terms present, a Monte Carlo simulation may be preferable. Monte Carlo simulation entails randomly generating possible values using a specified probability function.

An appraisal professional can help address the following:

1. Regardless of the valuation method selected, stock volatility has an important impact on the calculated option value. Higher volatility generally translates into higher option value. Because it is difficult to predict the future, volatility estimates are usually based on historical results, at least as a starting point. However:

1. There are several options when measuring volatility, even when the calculation is based on historical results, and
2. Past volatility may not be the best estimate of future results. This is particularly true for young companies that, while currently recording highly volatile returns, expect decreased future volatility.

2. The use of a lattice method provides more control over assumptions. However, all assumptions must be based on documented analysis that will withstand outside auditor and regulator scrutiny.

3. If not explicitly addressed in a lattice model, discounts for marketability, vesting, and future terminations are needed to avoid overstating option values. However, there is little empirical data upon which to base these discounts. Consequently, use of an accredited valuation expert will be critical in having these discounts accepted.

Litigators face these same issues (and answers) in employment disputes involving stock options.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.