There has been a surge of support recently for expensing stock options, led by sages like Warren Buffett, John Bogle and Arthur Levitt. A small number of well-known companies, including Coca- Cola, BankOne, Procter & Gamble and the Washington Post, have said they will adopt this practice. Undoubtedly, there will be more.
Advocates of expensing argue as follows: Cash salaries and bonuses have value to the employee, and are treated and accounted for as current period expenses by the corporation. Stock options also have value to executives — often a value that far exceeds their cash compensation. Therefore, the thinking goes, stock options also should be thought of and accounted for as current period expenses. This is very seductive logic. Unfortunately, it happens to be wrong on both a conceptual and practical level.
In truth, a number of others who have engaged in this debate have a different motive for wanting to expense stock options. They believe that the use of stock options has been abused and, in too many cases, very large option grants have led to unjustified wealth for undeserving CEOs. They want those abuses to end. They further argue that very large stock awards have lead to behaviors by CEOs that are in the CEOs’ personal interests, but not in the interests of shareholders in general.
In some cases this judgment is correct. Indeed, there have been abuses: Some CEOs have been granted too large stock awards by overly-compliant boards, and the use of options may have led to unethical behavior in some cases (although there have been no studies that support that contention). Critics believe that by expensing stock options, their use will be reduced. And, in their minds, if that end is achieved by the simple means of forcing companies to expense options at the time of grant, then they should be expensed.
Clearly, any abusive practices should end; no sensible person would argue otherwise. However, expensing stock options is the wrong way to achieve this worthwhile goal. If stock options are a current-period cost, they should be accounted for as such, even if they are perfectly aligned with shareholder interests, provide a marvelous incentive and are never abused. Similarly, if they are not a current period cost, they should not be expensed, no matter how badly abused they are.
It is obviously true that stock options have value. If they didn’t, no executive would want them. Over the years, I received a lot of stock options, for which I was grateful, and which I certainly viewed as valuable. However, while stock options have value to the executive — here is the crucial point — they cost the corporation nothing. They are never a cost to the company and, therefore, should never be recorded as a cost on the income statement.
The value inherent in a stock option, when exercised, is value taken from other shareholders – – without that value flowing through the company’s books. In other words, their cost is entirely borne by the company’s shareholders. This is an important distinction. That corporations would have to use cash, were it not for options, is beside the point.
When people try to understand the performance of a company, assess its prospects and estimate its value, they understand no single measure will suffice. Thus, they use a number of different measures. Obviously, one of these, and indeed a very important one, is “net income” — a number for which stock options granted and shares outstanding have no relevance or impact. Net income is the same whether a company has one million or one billion shares. A different number, earnings per share (EPS), is an even more significant performance measure, and it is this number that does reflect the value of options and does so accurately and completely.
Suppose a company has 10 million shares outstanding with a current market price of $40 per share, and it earned $50 million in the past year. Its EPS is therefore $5.00. Assume further that the company makes an option grant to executives at the current market price, with immediate vesting. Under this plan, the executive can buy shares from the company at $40 per share at any time for the next 10 years. No cost yet to the company.
Next year, the company’s stock rises to $50 per share. As a result, each of the shares granted as options now have a discernible value to the executive — $10 per share or $10 million in total. Still, no cost to the company. But there is now an accounting cost to the shareholder — a cost created by a reduction in earnings per share. In this hypothetical example, utilizing what is known as the “treasury method” of accounting, earnings per share would fall from $5.00 to $4.90 — not because the earnings are lower, but because the number of shares is higher.
There is value transferred to the executive through stock options. That value is captured as a transfer of wealth from the old shareholders to the new. The transfer is larger as the stock price goes up. However, there is never a cost to the company of the grant that should require an expense on the income statement. The effect of options is accurately reflected in the EPS number — where it belongs.
The current system works fine. There are, however, a few rules to follow when devising a fair and effective options program:
? Require executives to hold company stock at a fixed ratio to their current cash compensation. ? Require executives to hold stock bought as a result of exercised options for a period of time. ? Require executives to limit sales to a fixed percentage of their ownership in any quarter. ? Require options to be at market prices or higher at time of grant. ? Do not reprice options for the CEO if the stock price decreases.
There should be ways to stem stock-option abuses, but requiring that companies expense all option grants is clearly not one of them.
Mr. Golub, retired chairman and CEO of American Express, is currently chairman of TH Lee Putnam Ventures, Client Logic and AirClic. He is also a senior adviser to Lazard and serves on the board of directors of Dow Jones, publisher of this newspaper.