New option expensing rulings spawn new options: to expense or not expense stock options is no longer the question, but new rulings have raised significant compliance questions, while simultaneously adding a new source of financial expense--two often-conflicting objectives.

An 11-year debate finally ended on Dec. 16, 2004. On that date, the Financial Accounting Standards Board (FASB) issued Statement of Financial Accounting Standards 123 Revised, FAS 123(R), mandating that companies expense the fair value of their employee stock options on the face of their income statements.

This new mandate, along with the requirements of the Sarbanes-Oxley Act of 2002, further increases the need for veracity and transparency in financial statements. For most companies and auditors, the conservative approach will be the preferred approach. But even with this course of action, there are still questions concerning what exactly this approach is and how to behave conservatively without drastically impairing bottom-line results.

Companies are now faced with a myriad of complex decisions that under the former standard, FAS 123, were given little if any scrutiny by companies or their auditors. Some of these key reporting issues include computing key assumptions (primarily volatility and expected term), choosing an option pricing model and considering overall tax impact. The following gives some insight into these key challenges, providing a high-level overview of some of the most pressing issues around FAS 123(R).

Implicit Preference for Lattice Models

In the initial Expose Draft (ED) related to share-based compensation, FASB stated an explicit preference for lattice models. Subsequently, the board apparently recognized the over-whelming evidence demonstrating why the Black-Scholes-Merton (BSM) formula (the previously accepted model) is a fully inadequate technique for deriving a reasonable estimate of fair value. This view is especially true for companies that have ample historical exercise data that facilitates a more rigorous analysis of employee exercise behavior and its changes due to stock price volatility.

However, two critical and flawed assumptions made by the BSM formula (vis-a-vis lattice) are important to understand:

* BSM values employee options as if everyone in the company exercised on the same exact day. Of course, this hardly corresponds to reality, since employees base their exercise decisions on where the stock price is at a given point in time.

* BSM presupposes that anyone with out-of-the money options would throw them away at the expected term because they have no belief that the options will ever have value.

Clearly these assumptions do not reflect how employee option-holders base their decisions, and as a result, preclude the BSM formula from reliably capturing the underlying economics of the option. Given how FAS 123(R) calls on companies to identify the best estimate of fair value (as stated in FAS 123(R), paragraph A17), the BSM formula may not be the typical answer. While there are no steadfast truths as to the perfect valuation technique, given the issues surrounding the BSM model, companies would be prudent to investigate the second category, which is the lattice model.

Lattice models--which include binomial and trinomial models--construct a tree of stock price paths that reflect the potential upward or downward movements the stock can take at each point ("node") in the lattice. At each node, employees must weigh the value of holding their options to the next period (the next node), versus exercising at the present one. The critical feature of a lattice model is that it contains an internal decision system to forecast when exercise will occur, drawing upon variables known to affect that decision.

Although FASB removed its explicit preference that companies use a lattice model, an implicit preference remains. The tone of FAS 123(R) reflects a belief and expectation by FASB that lattice models will typically produce a better estimate of fair value through their more comprehensive valuation framework and inherent flexibility.

Model Inputs--Volatility

Although the valuation technique is a critical decision a company will make as it considers compliance with FAS 123(R), an equally important decision concerns the methods of calculation chosen for the key inputs to any valuation model--primarily volatility and expected term.

Volatility is a required input to any valuation technique and is the single largest driver of the final value. Traditionally, most companies have calculated historical volatility values and used these (without much additional analysis) in their estimation of option fair value. In the new FAS 123(R) world, however, this technique is likely to be insufficient in deriving an acceptable volatility estimate.

FASB appreciates the relevance of historical volatility in estimating future volatility. However, it states that this measure should constitute a point of departure for a rigorous analysis of volatility that considers a myriad of other crucial data points. For example:

"Historical experience is generally the starting point for developing expectations about the future." (FAS 123(R), Paragraph A21)

"Expectations based on historical experience should be modified to reflect ways in which currently available information indicates that the future is reasonably expected to differ from the past." (FAS 123(R), Paragraph A21)

Some of the more obvious factors warranting a conclusion that historical data alone will fail to reliably forecast future volatility would be merger and acquisition activity or the disposition of a key line of business (FAS 123(R), Paragraph A21). It may help to segment the types of historical events that would likely bias a historical volatility estimate (being applied forward) into four categories:

1. Company-specific price shock events: corporate scandals or significant litigation, unexpected regulatory changes or similar exogenous events causing a price-shock;

2. Structural company changes: steady increases (or decreases) in market capitalization relative to peer companies, a dramatic change in the firm's employee base (resulting from M & A activity) and other similar structural changes;

3. Capital structure and dividend policy changes and major strategy revisions: changes in financial leverage, dividend policy or long-term growth prospects through significant strategy additions or reversals;

4. Accounting policy changes: changes in the company's revenue recognition policies, inventory valuation approach, etc.

Model Inputs--Expected Term

One of the most popular lattice models, and the one referenced in FAS 123(R), is a simple barrier model hypothesizing that exercises occur and are triggered by the option attaining some in-the-money level (also called the "suboptimal exercise factor"). The reasoning is simple: employees are after some percentage gain on their options, and once achieved, they are induced to exercise.

There are many ways to derive this input, the simplest being a historical average in-the-money level at which exercises have occurred in the past. However, companies must use caution to scrutinize the values produced through this simpler method. Any company that has had significant stock price run-ups is likely to calculate a value that is not representative of the future. This could easily cause the company to over-expense its options, warranting the investigation of other, more robust methods.

Model Inputs--Forfeiture Estimate

Companies are called to estimate at the time of grant how many options are likely to be canceled prior to vesting, and only recognize expense on those expected to vest. This estimated forfeiture level is checked against actual forfeiture levels at appropriate intervals, resulting in a "truing-up" of any differential.

It pays to do a good job on this estimate up front to avoid discontinuous peaks or troughs in the expense recognized during any one period. This starts with an analysis of historical forfeiture rates and the rate of growth or decline over the years. Consideration of new markets being entered, labor conditions in the broader industry and the company's own growth strategy are useful in moderating the historical value up or down.

Income Tax Effects

The dynamics of income taxes related to option accounting is perceived to be one of the most complicated areas of FAS 123(R). However, with the rise of software products that can offload much of this complexity, the key is simply to understand what the tools are doing.

Internal Revenue Service (IRS) tax regulations value an option differently from FAS 123(R). Tax law assumes the option's value is its intrinsic value, whereas generally accepted accounting principles (GAAP) now call for a fair value measurement (equal to the intrinsic value plus the "time value" of the option). Because the intrinsic value of the option fluctuates over time and the GAAP value is fixed as of the grant date, the differential between these two measurements is dynamic.

Since an option's intrinsic value is typically zero at the time of the grant, the total GAAP-calculated compensation cost will exceed the IRS deduction. This creates a "deductible temporary difference" in the form of a deferred tax benefit that hits the income statement (with exceptions for instruments not resulting in tax deductions).

Should the opposite occur, where the IRS deduction exceeds the cumulative GAAP cost recognized, then any tax benefit net of the previously recognized excess tax benefit from those instruments is recorded as additional paid-in capital. This is to guard against changes in the fair value subsequent to the grant date impacting income statement deductions. So if the actual (realized) tax benefit exceeds the deferred tax asset for other reasons, this differential would appear on the income statement.

Clearly, the complexity of income tax effects arises from the difference between IRS and GAAP valuation frameworks. The challenge is primarily one of truing up this dynamic discrepancy, which is why a robust software tool, combined with an intuitive understanding of the fundamentals, offers the best strategy to ensure compliance.

FAS 123(R) raises significant compliance questions, while simultaneously adding a new source of financial expense. Although these two objectives can and will conflict at times, a strong grasp of the central issues underlying FAS 123(R) is the financial executive's best weapon.

David Roberts is President and CEO of Equity Methods LLC, a firm that creates expert products and related services in the equity compensation field. He can be reached at [e-mail].

Thomas Roberts is Vice President of E*Trade Financial Corporate Services, a firm that provides corporations with employee stock-plan solutions. He can be reached at [e-mail].