Are Options Hurting Your Stocks?

After much debate and delay, stock options will soon be treated as an expense on companies' financial statements--as they always should have been. Companies with fiscal years beginning after June 15, 2005, will be required to expense options in the first quarter of the next fiscal year, which means that companies that just started a new fiscal year on July 1 will be expensing options on income statements in fall. We will then find out whether moving this cost from footnotes to income statements will cause the financial Armageddon long prophesied by Silicon Valley.

My prediction: A Nasdaq meltdown is not at hand (at least, not one caused by expensing options), but we are in for several months of confusion as option-happy companies try to whitewash results by emphasizing pro forma earnings results that do not include options expense and complicit Wall Street analysts that follow the companies' lead. For investors who rely on price/earnings ratios as their main valuation tool, this might result in some uncomfortable moments. (Do I use the pro forma number? Do I increase my earnings multiple? Get me an aspirin!)

We'll be watching the debate with some interest here at Morningstar, but it won't affect how we value companies, since we've been including the cost of options in our fair value estimates for the past couple of years. I'll walk you through how we do it--and show you which companies take the biggest valuation "hits" from option policies. (For some companies, options reduce our fair value estimate by 20%-30%.)

The Ghost of Options Past
When we estimate the economic value of options, we separate the cost into two parts: options that have already been issued, and options that have not yet been granted. Although it might seem odd to take a stab at the latter, we think it's vital--stock options are an ongoing cost of doing business for companies that rely on options for employee compensation; ignoring options' future cost would underestimate how much value is being transferred from shareholders to employees.

The biggest cash cost for companies that grant lots of options is managing share dilution. Most firms try to mask the impact of their stock-option policies by buying back shares on the open market, so their share count stays steady, or at least increases much more slowly than it would otherwise. When we estimate the cost of options that have already been issued, we assume that companies keep their share count constant and buy back all options that are exercised.

Further, we assume that options are exercised when they are in the money, which we define as any price above our fair value estimate. So, if our fair value estimate is much lower than the bulk of the option exercise prices, past options won't have much of an impact on what we think the company is worth. This makes a lot of sense when you think about it: Those options are unlikely to ever be in the money, so they're unlikely to be exercised, which means there won't be any excess shares for the company to buy back.

The Ghost of Future Options
Of course, we don't know the exercise prices of options that have not yet been granted, so we can't use the "assumed repurchase" method to estimate value. So, we model the economic value transfer to employees by taking the footnoted option expense in past years, looking at how it shakes out as a percentage of corporate overhead, and then projecting the expense into the future.

In making this projection, we use the past as a guide, but nothing more. For one thing, the footnoted expenses in some past years are inflated by the very high volatility level of the equity market in 2000-2002. (Volatility has a big impact on the valuation method used by most companies to report option expenses.) Moreover, many firms have recently gotten religion now that options will finally be expensed on corporate income statements, so they're drastically reducing the amount of options they plan to issue. We take factors like these into account when estimating the economic cost of future option grants.

Options' Effect on Fair Value Estimates
So how much do options affect our fair value estimates? More than you might think. On average, options reduce our fair value estimates for the nonfinancial companies that we cover by about 3%. About 30 companies with heavy options use see fair value estimates reduced by 15% or more, while another 70 take hits of between 10% and 15%. About 200 companies fall into the 5%-10% category, and 400 get hit by 1%-5%. Below, I've listed the 20 companies with the greatest fair value estimate deductions stemming from stock options.

Percentage of Fair Value Eaten by Options
CNET Networks	30%
Asyst Technologies	24%
Agere Systems	23%
BEA Systems	22%
Multimedia Games	21%
Neurocrine Biosciences (NBIX)	21%
Amer. Axle & Mfg. Holdings (AXL)	20%
Siebel Systems	20%
Corinthian Colleges	18%
Veeco Instruments (VECO)	18%
Credence Systems	18%
NVIDIA (NVDA)	17%
Citrix Systems	17%
Electronics for Imaging	17%
Sapient	16%
Maxim Integrated Products	16%
Triad Hospitals (TRI)	16%
Scientific-Atlanta	16%
Lifepoint Hospitals	16%
Macrovision	16%

Not surprisingly, we see a lot of smaller tech companies on the list, with six from the chip industry alone. (In case you're curious about some of the larger tech bellwethers, Cisco's stock options reduce its fair value estimate by 14%, while Intel's result in only a 6% reduction.) A few non-tech companies make our "biggest option ding" list, but it's not hard to see why Silicon Valley has been fighting hardest against option expensing--they're the ones who've benefited most from the previous policy by keeping a sizable portion of compensation expenses off the income statements.

Conclusion
Is our method for estimating the economic cost of options perfect? Hardly--few things in finance are precise to the degree that we'd often like. But we do think our method does a reasonable job of capturing the economic value of stock options, and we firmly believe that value needs to be taken into account when deciding how much to pay for a firm's shares. As hedge fund manager--and occasional academician--Cliff Asness wrote last year in a very pithy article about option expensing, "…there is only one non-negative option value that we know to be demonstrably false--that is, zero. Saying we cannot calculate option values accurately so let us choose the one and only value we know for a fact cannot be correct is simply nuts."

And that's really all there is to it. We can argue all day long about how to best measure the cost of stock options--and some interesting ideas have been run up the flag pole recently--but that's a far more logical debate than whether options should be expensed. They should, and soon they finally will be.