Accounting for Stock Options: Should Bondholders Care?

Summary

The issue of how stock options should be accounted for has become particularly contentious in recent years as more companies have introduced option schemes as part of their compensation programs. The global proliferation of such plans in addition to their potentially large cost has led various accounting standards setting bodies to reexamine their standards for stock options. A recent example of this is in the U.S. where, on March 31, 2004, the Financial Accounting Standards Board (FASB) issued a long-awaited exposure draft (Proposed Standard) on accounting for stock options. Unlike previous standards, the Proposed Standard would require all public companies to recognize an expense for employee stock options using fair value estimates. Currently, companies are not required to expense employee options, although they must provide pro forma disclosure of the fair value of option grants.

Debate over mandatory expensing of stock options has largely focused on the earnings per share issues affecting equity investors. Fitch Ratings believes that bondholders also have an interest in understanding the proposal, including cash flow and balance sheet implications. Analysts at Fitch have traditionally considered the impact of employee option exercises but have viewed it primarily as a leverage issue, particularly when share repurchases are funded through debt issuance. With mandatory expensing of options imminent, Fitch’s analytical approach will evolve to capture the true economic cost and the cash requirements of option programs, provided the final standard contains adequate disclosure provisions.

In reviewing the Proposed Standard, Fitch has made several observations concerning issues that could be relevant to bondholders, including:

• Because of their dilutive effect, many companies have a high propensity to repurchase shares issued upon exercise of employee stock options. In this context, from a bondholder perspective, employee options often have a true cash cost and can be thought of as a form of deferred compensation, which has the effect of reducing available cash to service debt and increasing leverage.
• Expensing of options under the proposal results in a point-in-time estimate of compensation expense that may have little or no relationship with the actual future cost. Furthermore, there is no requirement to reconcile, either after the fact or period to period, the compensation expense to actual cost.
• Adjustments may be necessary to reconcile compensation expense reported in the income statement with free cash flow.
• There may be unintended balance sheet effects, including the creation of “phantom” or intangible equity via the creation of deferred tax assets.
• The Proposed Standard advances the FASB’s agenda with respect to convergence with International Financial Reporting Standards (IFRS), mirroring the provisions of IFRS No. 2 (IFRS 2).

Overview of FASB Proposal

Expense recognition would be required under the Proposed Standard, rather than voluntary, as is currently the case. The Proposed Standard takes a multistep approach in which an estimate of the compensation cost is made at the grant date and then recognized in earnings proportionally as the options vest. Expense related to options that do not vest is therefore not recognized, but no adjustment is made for changes in the price of the underlying shares or other variables subsequent to the grant date or upon exercise of options. The FASB recommends the use of a lattice model, such as the binomial model, which incorporates expected exercise and expected postvesting employment termination behavior instead of a single weighted average term, which is inherent in the closed-end version of Black-Scholes that is currently favored.

The concept of comparability, one of the primary tenets of U.S. Generally Accepted Accounting Principles (GAAP) as expressed in the FASB’s recently re-emphasized Conceptual Framework, is a key consideration behind this proposal. During 2002, beginning with companies like Coca-Cola, large corporations began adopting the income recognition provisions of Statement of Financial Accounting Standards No. 123, Accounting for Stock-Based Compensation (SFAS 123), which allows, but does not require, fair value recognition of options at the date of grant. Since then, over 500 publicly traded companies have chosen to voluntarily expense employee options under SFAS 123. However, there are still thousands of publicly traded companies that account for stock options under the Accounting Principles Board Opinion No. 25, Accounting for Stock Issued to Employees (APB 25), intrinsic value method, providing only pro forma disclosure of fair value option expense. This makes it necessary for investors to subtract the pro forma options expense from earnings, or add-back recognized options expense, as the case may be, in order to compare companies that are expensing to those that are not.

Furthermore, the newly promulgated IFRS 2 requires expense recognition for employee option grants, making the proposed standard consistent with the FASB’s objective of convergence. IFRS 2 also requires a mathematically derived estimate of option expense to be recognized in the income statement but differs in that it does not encourage the use of one model over the other. In addition, IFRS 2 requires that deferred taxes associated with option vesting be adjusted each period for changes in share price.

Should Bondholders Care?

From the perspective of equity investors and bondholders alike, persuasive arguments have been made in favor of expensing employee stock options as a clearly definable, if not easily measurable, operating item. For companies in highly competitive, technology-based industries, employee options have proven essential for attracting and retaining a scarce pool of talent, which can be critical to start-up enterprises. In many such cases, the use of options has made a material contribution to the expansion of shareholder value; Microsoft’s liberal use of options during its period of massive growth during the late 1980s and early 1990s comes to mind. That said, employee options do not involve an immediate cash expenditure and, as such, have been viewed by some as outside the realm of traditional fixed-income analysis.

For many more mature companies, however, it is perhaps myopic to view options programs as being an entirely noncash expense and, therefore, not relevant to bondholders. The reality is that many companies manage their capital structure to meet certain leverage and earnings per share targets. Issuance of shares on exercise of options, of course, expands the number of shares. This makes it critical that the issuance of these shares be offset with active, ongoing stock repurchase programs to manage the dilutive effect of employee option exercises. Thus, within many of these more mature companies – the type of companies that are likely to have highly rated debt – employee options are viewed as being akin to a deferred expense.

Fitch believes this analysis to be correct. Essentially, the economic cost of employee options can be thought of as the opportunity cost of issuing the associated shares at a below-market price, or simply the difference between the exercise price and the market price at the date of exercise. Recognizing that issuance and repurchase may not occur simultaneously, the price at which the company actually executes the buyback may, in many cases, act as a reasonable proxy for market price at date of exercise. Indeed, in such cases the difference between the proceeds from employee option exercises and the cost of reacquiring an equivalent number of shares in the open market is the real cash cost of the option compensation. The share repurchase can be viewed as the last step in the option compensation cycle or, alternatively, the actual payment of deferred compensation. Fitch notes that this cost is, in fact, recognized in the financial statements under the current intrinsic value accounting rules. The “expense,” however, makes its way straight to equity, effectively going around the income statement (see table at right).

Further, one could argue that these share buybacks represent in substance, if not in form, an operating outlay that should be deducted from free cash flow. If the options are compensation whose cash cost is measured as the incremental cost of management’s expressed or implied decision to maintain the level of outstanding shares, this incremental cost is most properly classified within cash flow from operations.

These effects can be seen very clearly in the financial statements of Intel for the year ended Dec. 27, 2003. Intel uses the “par value method” of accounting for shares reacquired. Under this method, par value and paid-in capital are adjusted only for the amounts at which the shares were originally issued, and the remaining cost to reacquire goes to reduce retained earnings. A large part of this reduction in retained earnings associated with the repurchase of shares – $1.3 billion in fiscal 2003 for Intel – can be viewed as the payment of deferred compensation to employees. Further, if this amount were reclassified to cash flows from operations, Intel’s free cash flow to total debt ratio would be reduced by approximately 10%.

It is interesting to note that while the Proposed Standard partially rectifies the above-mentioned income statement sleight of hand, it does not acknowledge the linkage between employee options and the actual cost associated with stock repurchase programs. Nor does it recognize that funds expended to repurchase shares issued in connection with employee option exercises are properly classified as cash flows from operations. The proposal would have companies recognize an expense at time of vesting based on a model-driven estimation, but never reconcile to the actual economic cost at the date of exercise or, alternatively, to the proxy of repurchase cost. Equity would continue to be adjusted to the proxy, however, as any cash outlays to repurchase shares over and above the estimate determined at the option grant date would reduce equity directly, bypassing the income statement. This combination of factors could create permanent and growing imbalances between cumulative net income and cash flows from operating activities for companies that continue to grant options.

Implications for Bondholder

Fitch believes that the Proposed Standard has several potentially important implications for bondholders. These include the effects of reduced option grants on corporate leverage, the use of subjective estimates in the determination of net income with no mechanism for adjusting to actual cost, the impact on cash flow, and the potential for long-term cumulative equity distortions.

Potential for Fewer Options May Affect Leverage

The requirement to record an incremental compensation expense under this proposal may reduce the number of options granted by many companies. This has the potential to decrease leverage, all other things being equal. Consider the typical option cycle for a large corporation, by way of a simplified example that ignores the effects of taxes (see table above):

XYZ Corporation grants 1.0 million options at market to its employees on Jan. 1, 2005. The price of a share of XYZ stock at the grant date is $100. The options vest on Jan. 1, 2006. On Jan. 2, 2006, the share price for XYZ is $200. All the options are exercised, and XYZ receives cash proceeds of $100 million.

Further, on Jan. 3, 2006, XYZ repurchases 1.0 million shares in the open market for $200 million, fearing dilution of its equity and the effects on earnings per share.

As a result of issuing options to employees and then repurchasing shares to offset the effect of the exercise of those options, XYZ’s leverage has increased. Why? Because the cost of reacquiring shares exceeds the proceeds from exercise (otherwise, the options would not typically be exercised). XYZ must either take this cash out of its operations, which presumably means that it must increase borrowings elsewhere, or borrow it outright. The $100 million premium that the company paid to repurchase the shares can be viewed in this context as deferred compensation, although under current accounting standards (APB 25), it flows directly to equity.

The implication of the above example for bond analysis is that if expensing of options causes companies to reduce the amount of option grants, there will be less use of leverage in the future to buy back shares. Naturally, this assumes that all other factors remain the same, and equity investors act rationally, assigning the same valuation to the company irrespective of whether options are directly expensed or just disclosed on a pro forma basis as is currently the case. To the extent investors assign a lower valuation to reflect the negative earnings of expensed options, this would raise the after-tax cost of equity capital and could shift the dynamic back toward a higher reliance on debt financing. Further, this is a simplified example that specifically assumes a rising stock price.

Subjective, Point-in-Time Earnings Estimates with No Mechanism for Adjustment to Actual Cost

Regardless of the method chosen by companies to estimate the earnings effect of option grants, the resulting expense will be just that, an estimate; and it will vary by company. Thus, on an earnings basis comparability will remain an issue. Furthermore, the value of the estimates in predicting future cash flows associated with option exercise will vary. For example, if XYZ Corporation estimated the value of the options granted at Jan. 2, 2005 at $50 million, this amount would be added to expense at the vesting date, rather than the “real” cost of $100 million (the difference between the exercise price and the cost to repurchase these shares). (See table above, right.)

Under the proposal, there is no adjustment of the model-driven estimated cost of the options and their actual economic value at the time of exercise. This results in a misstatement of income of $50 million in the example. Fitch believes that the Proposed Standard should contemplate an adjustment to actual expense upon option exercise or expiration. For example, at the exercise date of Jan. 2, the actual value of the XYZ employee options becomes known with absolute precision – $100 million in the example above (the difference in the fair market value and the employee exercise price). At that point, XYZ should recognize additional compensation expense of $50 million. Alternatively, the Proposed Standard could require periodic adjustments of compensation expense to reflect the most current market data. Conversely, if the options were to reach their expiration date worthless, then the prior period expense should be reversed.

Debt Coverage and Cash Flow Subject to Distortion

For fixed-income investors, the implication of the Proposed Standard is that the estimated expense should be stripped out of the current period earnings before interest, taxes, depreciation, and amortization calculation. Further, analysts may want to consider deducting an estimate of stock repurchase expenditures from free cash flow projections. Because of the nature of the expense – i.e. deferred compensation – it should be considered an element of cash flows from operating activities.

What amounts should the analyst deduct from cash flows from operations? Should it simply be the estimated expense over the vesting periods? Probably not, as these are point-in-time estimates and may have no relationship with the actual cost to repurchase shares in the future. In the example above, the company’s actual cost was $100 million, compared with an estimated cost of $50 million at the time of vesting. A more reliable estimate of the cash outlays associated with option expense will, in most cases, be discernable by looking at historical propensity and/or stated desire to match share repurchases with option exercises, and projected prices of the entities’ equity shares.

In the U.S., cash flows from operations as portrayed in the statement of cash flows will change under the new proposal. Under the Proposed Standard, tax benefits from option exercises will flow through the financing section of the statement of cash flows. Proceeds from option exercises and expenditures related to the repurchase of resulting shares will continue to flow through the financing section. However, the option expense itself will presumably be a reconciling item in cash flow from operating activities; i.e. an add-back. As such, cash flow from operations will continue to ignore the effect of deferred compensation resulting from share repurchases. The upshot of this is that credit analysts may want to consider reclassifying net repurchase cost to operating cash flows, presuming sufficient information exists to make these adjustments.

Potential Equity Distortions

From a bondholder perspective, it is notable that a company’s balance sheet and equity may be prematurely inflated under the FASB’s proposal. How so? Through the generation of a deferred tax asset that represents the timing difference between the book cost of the options and any future tax benefits that may accrue if the options are exercised. The direct offset is a commensurate uplift to paid-in capital, despite the fact no tangible equity has been raised at the time the options vest. Over time, this timing difference and the associated “phantom” equity it creates could become material, necessitating a writedown if the options expired worthless. Obviously, the degree to which this could be material will vary company to company.

Outlook

Most of the controversy around this topic has been focused on the direct impact on earnings per share and potential unintended consequences for cost of capital. While much of the attention has been on the earnings per share impact, which is largely a matter for equity investors, Fitch believes that bondholders have a stake as well.

Fitch believes that employee options have a real cost for many companies, particularly those in more mature stages of growth. The difficulty is that the true economic cost is often not known with certainty until future periods. Moreover, under the new proposal, the full economic cost may not be captured in income and operating cash flow. Analysts need to understand the implications of yet another complex, estimate-driven accounting standard. As is often the case, appropriate analytical adjustments may be needed to reconcile the bondholder’s perspective with what is being reported in the financial statements. This complexity combined with the potential need to make analytical adjustments underlines the key role disclosure will play in ensuring the Proposed Standard is ultimately effective in providing fixed-income investors and analysts with the information they need.

Note: Joseph St Denis, Eileen Fahey and Julie Burke were co-authors of this report. This report was originally published by Fitch on 20 April 2004.