Print


Accounting for Stock Options
Update on the Continuing Conflict

By Nicholas G. Apostolou and D. Larry Crumbley

AUGUST 2005 - In December 2004, a decade after bending to Congressional pressure and backing away from requiring the expensing of options on financial statements, FASB issued a revised standard to recognize stock-option compensation as an expense on income statements. Many in Congress may try to thwart the proposal before it becomes effective. A bill by Representative Richard Baker of Louisiana that would require expensing the cost of stock options for only the top five executives of a company has drawn the support of those groups still resolutely opposed to expensing.

This time, however, FASB is likely to prevail. Investors are demanding tougher accounting standards, and the International Accounting Standards Board (IASB) has already passed rules requiring the expensing of options. Many large U.S. corporations have already voluntarily agreed to expense options. Finally, there is more concern about, and less support for, Congressional interference in FASB’s standards-setting process.

History of the Debate

Accounting for stock options has been one of the most controversial topics in accounting during the last decade. The principal debate is whether compensation expense should be recognized for stock options and, if so, the periods over which it should be allocated. Before 1995, the provisions of Accounting Principles Board (APB) Opinion 25, issued in 1972, determined accounting for stock options.

APB Opinion 25 measured stock options using the intrinsic value method, whereby compensation expense was determined as the excess of the stock price at the measurement date (generally, the grant date) over the option exercise price. Because most stock options had exercise prices at least equal to current market prices, no compensation expense was recognized. This approach ignored any likelihood that the stock price would exceed the exercise price in the future.

In June 1993, FASB attempted to recognize the reality of stock-option value by issuing proposed SFAS 123, which required measuring the option value based upon the many factors that reflect its underlying value. Therefore, total compensation expense was to be based upon the fair value of the options expected to vest on the grant date. No adjustments would be made after the grant date in response to subsequent changes in the stock price. Fair value was to be estimated using Black-Scholes or binomial option-pricing models.

A groundswell of massive opposition to this fair value method resulted, led primarily by industries making significant use of stock options, particularly in the high-technology sector. Smaller high-tech companies were very vocal, arguing that offering stock options was the only way they could hire top professional management. Furthermore, they claimed that the losses that would result from forcing them to recognize stock options as compensation expense would impair their stock price and put them at a disadvantage compared to larger corporations better able to absorb the expense of stock options.

Opponents to the expensing of stock options included many members of Congress. In 1993, Senator Joseph Lieberman introduced a bill that would have mandated the SEC to require that no compensation expense be reported on the income statement for stock-option plans.

This bill would have set a dangerous precedent for interfering in the operations of FASB. The powerful interests aligned against it forced FASB to compromise. In 1995, FASB decided to encourage, rather than require, recognition of compensation cost based upon the fair value method and to require expanded disclosures. In other words, SFAS 123 requires companies that continued to follow APB 25 and did not include stock-option expenses in the income statement to disclose in the notes to financial statements what such expenses would have been. This compromise troubled many observers because the politicized rule-making process was less concerned with proper accounting and more influenced by the economic consequences of a new standard. Berkshire-Hathaway Chair and CEO Warren Buffett addressed this in the company’s 1998 annual report: “A distressing number of both CEOs and auditors have in recent years bitterly fought FASB’s attempts to replace fiction with truth and virtually none have spoken out in support of FASB.”

Critics of the failure to expense options (fair value method) on the income statement became particularly vocal in recent years because of the widespread concern over deceptive accounting practices at companies accused of fraud (e.g., Enron, Tyco, WorldCom). Few companies before 2002 chose to adopt the fair value method. Buffett was the most prominent critic of the failure to recognize stock options on the financial statements. Elsewhere in Berkshire’s 1998 annual report he stated: “Existing accounting principles ignore the cost of stock options when earnings are being calculated, even though options are huge and an increasing expense at many corporations.” He characterized this accounting policy as “outrageous” and an “egregious flaw in accounting procedure.” He also stated that he often had to adjust reported earnings per share (EPS) figures of other companies by 5%, “with 10% not at all uncommon.” Once he had made that adjustment, it affected his portfolio decisions, “causing him to pass on a stock purchase he might otherwise have made.”

To test Buffett’s estimates of the extent of dilution from the issuance of stock options, the authors surveyed 20 companies by examining their 2003 annual reports (Exhibit 1). Prominent high-tech companies were selected, along with several others used for comparison.

Buffett’s adjustment to reported earnings of 5% to 10% for stock option compensation expense is, in some cases, conservative. The difference between reported earnings and earnings under the fair value method (expensing stock options) substantially exceeds 10% for most of the companies in our survey. For Yahoo and Adobe, the percentages were 86% and 70%, respectively. For six of the companies, the expensing of stock options would have changed a net profit to a net loss.

Revised Statement of Financial Accounting Standards

The issue of expensing stock options returned to the front burner in October 2001 when Enron, then the nation’s seventh-largest company, disclosed more than $1 billion of accounting errors. The wave of financial fraud disclosures that followed stunned investors and increased the demand for transparency in corporate reporting. FASB responded to heightened interest in improved financial reporting with the release in December 2004 of SFAS 123 (Revised), Share-Based Payment. FASB would require public and nonpublic companies with calendar fiscal year-ends to recognize stock-based compensation in their income statements starting in 2006.

FASB cited four principal reasons for issuing SFAS 123(R):

  • Addressing concerns of users and others. FASB has received many complaints from users that using APB Opinion 25’s intrinsic value method resulted in financial statements that do not faithfully represent the underlying cost associated with the issuance of stock options. When employees exercise their options (i.e., buy the stock at the preset price), the company has to issue new shares, which reduces the earnings available for each share.
  • Improving the comparability of reported financial information through the elimination of alternative accounting methods. Beginning in summer 2002, many companies announced their intention to voluntarily adopt the fair value method and to expense stock options. Currently, more than 500 U.S.-listed companies have announced their decision to expense stock options, including Exxon Mobil, General Motors, and Coca-Cola. Most companies, however, continue to use APB Opinion 25’s intrinsic value method and do not expense options. FASB believes that similar economic transactions should be accounted for similarly, and favors the fair value method for all publicly traded companies.
  • Simplifying U.S. GAAP. SFAS 123(R) would simplify the accounting for stock options. FASB believes that U.S. GAAP should be simplified whenever possible. Requiring the use of the fair value method would eliminate the intrinsic value method and its many related rules.
  • International convergence. SFAS 123(R) would better harmonize U.S. accounting standards with international accounting standards. In February 2004, the IASB issued a rule that required the expensing of stock options as of January 1, 2005. Conformity with IASB standards is required of all publicly listed companies in the European Union and of all Australian entities.

Pricing Stock Options

All of the surveyed companies use the Black-Scholes option-pricing model, developed by Fischer Black and Myron Scholes in the early 1970s. This model calculates the present value of a stock option at the grant date, based upon specific information about the terms of the option and assumptions about future stock price performance. The value of an option reflects the estimate of the price that someone would pay in the market today for the option. It is the point at which an investor would be indifferent between receiving the option or the amount of cash equal to its value. The method is considered a probability model because it assumes that the underlying stock behaves in such a way that possible future prices can be modeled by a probability distribution.

Of the six variables in the Black-Scholes model (Exhibit 2), the estimated future volatility of the stock price is the most difficult to compute. This measure can be defined as the estimated future variance of the stock price based on historical stock price movement or expectations for future stock movement. Volatility measures the stock price fluctuation relative to itself and should not be confused with a stock’s beta, which measures the stock price fluctuation relative to a market average. Volatility is expressed as a percentage, and this variable is a relative measure of the expected difference between the stock price at the end of the stock option’s expected life and the stock price at the grant date. Volatility is, in effect, the standard deviation of the expected price of the stock. Because future volatility cannot be known, SFAS 123 suggested using stock prices over a historical period of time equal to the expected life of the options being granted.

The estimate of volatility can dramatically affect the value assigned to the options. Consider an option with the following characteristics: exercise price: $10; fair market value: $10; expected life: six years; and dividend yield: none. With an estimated stock price volatility of 30%, the option would have an estimated value of $4.18. With an estimated stock price volatility of 70%, the estimated value would be $6.75.
Concern about a company’s estimates of option value is sometimes expressed in annual reports. A typical example is PeopleSoft’s disclosure in its 2003 Form 10-K:

Limitations of the effectiveness of the Black-Scholes option-pricing model are that it was developed for use in estimating the fair value of traded options which have no vesting restrictions and are fully transferable and that the model requires the use of highly subjective assumptions including expected stock price volatility. Because the Company’s [stock-based] awards to employees have characteristics significantly different from those of traded options and because changes in the subjective input assumptions can materially affect the fair value estimate, in management’s opinion, the existing models do not necessarily provide a reliable single measure of the fair value of its stock-based awards to employees.

SFAS 123(R) does not specify which option-pricing model companies should use; however, it does suggest using either Black-Scholes or lattice models.

Valuing Stock Options with a Lattice Model

SFAS 123(R) provides new guidance on option valuation, including new emphasis on complex techniques that most companies have not used in implementing SFAS 123. The revised standard asserts that both the model used to value option expense and the related inputs into the model should be consistent with the value placed on them by willing parties. In the exposure draft issued in March 2004, FASB asserted that lattice-based models satisfied this criterion better than Black-Scholes. Lattice-based models use the same basic categories of inputs as Black-Scholes, but they can reflect post-vesting employment termination behavior and other adjustments designed to incorporate certain characteristics of employee share options and similar instruments. Furthermore, lattice models can accommodate changes in dividends and volatility over the option’s contractual term, estimates of expected option-exercise patterns during the option’s contractual term, and black-out periods (when options cannot be exercised).

As previously mentioned, SFAS 123(R) requires the same six inputs as used in Black-Scholes; however, the following changes are required in the measurement of these inputs:

  • Companies must take into consideration assumptions that may vary over the contractual term of the option.
  • The standard specifies the six inputs as the minimum number of factors to be included in the model. No guidance is offered on any other assumptions that could be incorporated into the model.
  • A range of reasonable estimates is anticipated for expected volatility, dividends, and option terms. If the likelihood within the range is similar, an average (expected value) of the range should be used. SFAS 123 permits selection of the low end of a reasonable range of assumptions.

Unlike the formula used to calculate option value under Black-Scholes, the lattice model does not use a straightforward mathematical equation. Instead, the lattice model uses an iterative approach that involves generating a large number of possible outcomes and assigning a probability to each outcome. The complexity of the calculations typically requires computer-based models to determine values.


Nicholas G. Apostolou, DBA, CPA, CrFA, is the LeGrange Professor, and D. Larry Crumbley, PhD, CPA, CFD, CrFA, is the KPMG Endowed Professor, both in the department of accounting at Louisiana State University, Baton Rouge, La.

Close