Principles for Accounting for Stock Options
So what is the best and fairest way to account for the cost of stock options? Let's start with two principles.
First, the cost of options should be treated as a compensation expense, and appear on the income statement. As
Warren Buffett asked in
his 1998 annual letter to Berkshire Hathaway shareholders: "If options aren't a form of compensation, what are they? If compensation isn't an expense, what is it? And, if expenses shouldn't go into the calculation of earnings, where in the world should they go?"
Second, the cost of options should be calculated based on their value at the time they are issued. Let's say a company issued options and then the stock did poorly, such that the options expired worthless. Does that mean the options had no value when they were issued? Of course not. Had the company initially sold the options to investors rather than giving them to employees, it would have raised cash that could have been reinvested in the company, paid out to shareholders as dividends, etc. Similarly, if the stock did exceptionally well, such that the options became extremely valuable, is it fair to go back and deduct this high "cost" from a company's earnings? Of course not.
So why doesn't the
Financial Accounting Standards Board, or FASB, require this? Primarily because options are difficult to value -- not to mention the enormous opposition from companies that use options (and benefit from the bogus accounting for them). So the FASB compromised and, beginning with options granted in 1995, required all companies once a year in their 10-Ks to estimate a value for the stock options issued (using the
Black-Scholes model) and show what earnings would be had this expense appeared on the income statement.
Here are the data Dell reported in its most recent 10-K.
| Dell's Calculation of Its Stock Option Costs |
 | FY 99 | FY 00 | FY 01 |
| Reported EPS | $0.53 | $0.61 | $0.81 |
| EPS deducting stock option expense | 0.48 | 0.53 | 0.65 |
| % decline | 9% | 13% | 20% |
| Source: Dell's fiscal 2001 10-K |
The table shows that Dell's earnings last year would have been 20% lower (equal to $620 million before tax and $434 million after tax) if the cost of stock options had been expensed, and that this percentage has been rising rapidly.
Adjusting earnings to reflect the cost of options would have a material impact on Dell's price-to-earnings ratio. Analysts are projecting Dell's earnings this year at 70 cents a share, giving the stock a P/E of 34.6 (based on Monday's closing price of $24.25). If we reduce earnings per share by 20%, however, the P/E jumps to 43.3.
This table shows how much earnings per share would have declined for some other representative, widely held companies last year, had the cost of stock options been expensed.
Options Haircut
How much earnings would decline if stock option costs were expensed |
| EPS Decline |
| General Electric | -2% |
| Intel | -7% |
| Oracle | -9% |
| IBM | -10% |
| Microsoft | -13% |
| Sun Microsystems | -16% |
| Starbucks | -29% |
| Cisco | -42% |
| Compaq | -67% |
| eBay* | -317% |
| Yahoo!* | -2017% |
* Had the cost of stock options been expensed, eBay's EPS in FY 00 would have been -36 cents vs. the reported 17 cents and Yahoo!'s would have been -$2.30 vs. the reported 12 cents.
Source: Companies' most recent 10-Ks |
All the figures shown above are based on estimates calculated using the Black-Scholes option valuation method, as required by the FASB. This approach, in my opinion, significantly understates the true cost of options. Consider that many companies reprice options or make special, extra option grants if their share price falls (Dell and
Microsoft, among others, did the latter last year), which has the effect of greatly reducing the risk that options granted to employees might expire worthless. Also, the cost of options is spread over the vesting period. In Dell's case, for example, "the Black-Scholes value at the date of grant is assumed to be amortized over the vesting period [five years], subject to revisions for later forfeitures," according to a spokesperson. This makes sense, but keep in mind that spreading out the cost has the effect of understating the true cost of Dell's massive option grants last year, which were more than triple those of the previous year.
Finally, the Black-Scholes model is dependent on the assumptions that companies make, which may not be 100% accurate. For example, Dell assumes that the "expected term" of the options it grants is five years, though they don't expire for 10 years, because, according to a spokesperson, "our experience indicates that options generally are outstanding only for five years when you factor in historical exercise and forfeiture actions." But will the future look like the past? Until recently, Dell's stock went pretty much straight up for a decade, which undoubtedly led employees to exercise their options long before expiration. But if Dell's stock were to be flat for the next five years, employees could still wait another five years for the stock to rise before they expired worthless. The difference in value between five-year vs. 10-year options is substantial. For example, according to
Bloomberg, a 10-year Dell call option at today's price, using volatility of 54.85 (Dell's assumption in its most recent 10-K) is worth 35% more than an identical five-year option ($16.88 vs. $12.46).
Can the cost of options be calculated in a better way?
Read the conclusion of this column.
Click here to return to the first part of this column.
Whitney Tilson
RSS
