Blogroll Me!

Monday, December 19, 2005

Expensing Options is such a bad idea...

I knew the option-expensing rules would eventually disintegrate upon itself. This article in the Washington Post illustrates why expensing options is such a bad idea. In short, what people don't get is that options are NOT AN EXPENSE. Rather, options are a CONDITIONAL FUTURE DILUTION. I know big words have a hard time resonating with the general public and most people prefer easy-to-understand solutions, but option expensing illustrates why simple answers don't always work for complex situations.
I know that Warren Buffett has come out in favor of expensing options, but I think everyone is a little gaga over Warren anyways. Maybe the guy has just been lucky. Have you seen how much money he lost investing in The Gap? But I know that what he says is gospel so many people have parrotted him on this issue as well.
See, here's the deal. The GAAP income statement tries to present a balanced view of earnings, not an accurate one. Why not accurate? Because there is no accurate way to measure earnings. Some people think that cash is a good way to measure earnings, other feel that you need to adjust for "temporal shifts" (i.e. you buy an income producing asset (a building) and you spread the expense of that asset over its "useful life"). All of these temporal shifts require making assumptions (how long is the useful life of a building?). Assumptions are an inexact science at best (e.g. SWAG). Therefore the generation of financial statements as dictated by GAAP necessarily implies inaccuracy.
So, how does this apply to options? Well, options aren't a cash expense of the company. It doesn't cost a company anything to issue options, so therefore they don't really effect cash flow. And if you're like every newly minted MBA, you know that in the long run cash flow is what really matters. If you're an economics undergrad, you also know that in the long run, we're all dead. But I digress.
So, the common response by expensing advocates is that the company is giving away value, so therefore that value should be reflected as an expense. Well, the problem is that the value is highly speculative and extremely variable. The only thing I know about option valuations is that they're often wrong. The guys who won a nobel prize for their work on options valuation also oversaw the largest hedge fund catastrophy to date. That's right; Myron Scholes and Robert Merton, two of the creators of the Black-Scholes options valuation model got hammered when it came to applying their theory in the real world of investing. My point here is that no one really knows how to value options. Why? Because their valuation also rests on a ton of assumptions. (Don't even get me started on normally distributed populations! I love heteroskedasticity.)
So, what you have is a second-order of assumptions at work. And we all know the only true thing about financial assumptions is that they're wrong.
So why would we try and affect company earnings using such an exact science?
Well partly because the general public is stupid. Unfortunately, the general public also gets to vote, which leads to all sorts of disastrous consequences and option expensing is no exception.
I think the market has a hard time wrapping its mind around the concept of a conditional future dilution. Let me explain this furthur.
If options don't have an expense, where does their value come from? The value of an option is a result of the fact that shareholders' slices of the pie (as determined by the number of shares they hold over the total) shrink -- if the options are exercisable.
The best way to really illustrate the value lost by shareholders through options is not through an expense but rather through a factor that would illustrate the shrinkage in slice (ownership) size to the increase in stock price.
Assume a company with a stock price of $10 and 1,000,000 shares outstanding. Also assume you own 10% of the company, or 100,000 shares. Now consider the following two scenarios and consider the impact to your ownership:
1. 10,000 options outstanding with a strike of $12. If the stock price goes above $12, you would be diluted by 1%. (100K/1MM = 10% v. 100K/1.01MM = 9.9%).
2. 100,000 options outstanding with a strike of $20. If the stock price goes above $20, you would be diluted by 9%. (100K/1MM = 10% v. 100K 1.10MM = 9.1%)
Which would you rather have? Well, it kind of depends on a lot of things - where the price will end up, how variable the price is normally, how much do you think management deserves to be richly rewarded, etc...
So, you can see why thinking about how the exercise will effect your position is more useful that trying to sum all of it up and putting it as an expense on the income statement.
Further proof that it's a bad idea: professional investors (ex-Buffet) often exclude non-cash charges like this one to focus on core earnings power when trying to guage a company's value. So now you have a situation where the professionals (once again!) play by a different (and more advanced) set of metrics than your average investor.
Guess who often wins these contests?

2 Comments:

At 7:43 PM, Blogger Bill said...

Warren Buffett has never identified the Gap as an investment he made. I think if you're going to take a cheap shot that you might be better off finding something that has a higher likelihood of having actually happened.

I for one am happy that all compensastion expenses, including stock options, are now going to be recorded on the income statement. I think that BubbleVision and the retail investors are too wedded to the primacy of the income statement rather than cash flows, but nonetheless compensation is an expense.

And stock options used for any other purpose besides compensation are, and have been, expensed. It is a bizarre loophole and abortion of logic that the exact same instrument would be expensed for one use, and not expenses for another.

For example, here's what Intel had to say in its 2003 10-K:

From its 2003 10-K. P 57.

Fair values of cash equivalents approximate cost due to the short period of time to maturity. Fair values of short-term investments, trading assets, long-term investments, marketable strategic equity securities, certain non-marketable investments, short-term debt, long-term debt, swaps, currency forward contracts, equity options and warrants are based on quoted market prices or pricing models using current market rates. Debt securities are generally valued using discounted cash flows in a yield-curve model based on LIBOR. Equity options and warrants are priced using a Black-Scholes option pricing model. For the company's portfolio of non-marketable equity securities, management believes that the carrying value of the portfolio approximates the fair value at December 27, 2003 and December 28, 2002. This estimate takes into account the decline of the equity and venture capital markets over the last few years, the impairment analyses performed and the impairments recorded during 2003 and 2002.

The arguments against stock options as an expense are bankrupt. The fact that even the most ardent supporter of the status quo, Intel, expensed them for options transactions for every other use besides employee compensation makes that bankruptcy indisputable.

 
At 7:17 AM, Blogger Anonymous said...

"Warren Buffett has never identified the Gap as an investment he made."

My response:
"Billionaire investor Warren Buffet's investment holding company Berkshire Hathaway (nyse: BRKA - news - people) raised its stake in clothing retailer Gap (nyse: GPS - news - people) last year to 8 million shares, according to a Schedule 13F that was denied confidential status on March 23 and released yesterday by the Securities and Exchange Commission. Buffett had moved Berkshire's holdings in Gap, US Bankcorp (nyse: USB - news - people), Nike (nyse: NKE - news - people) and First Data (nyse: FDC - news - people) to a confidential 13F from a public 13F in a Feb. 15 filing with the SEC. The SEC can grant such a request for up to one year to limit severe market reaction. Buffett previously owned 2.4 million shares of Gap as of June 30, 2000, and raised the stake to 8 million three months later."

http://www.forbes.com/2001/04/03/0403facesscan.html

"And stock options used for any other purpose besides compensation are, and have been, expensed. It is a bizarre loophole and abortion of logic that the exact same instrument would be expensed for one use, and not expenses for another."

My response: Maybe they shouldn't be expensed for the other purpose either.

Lastly, just because Intel does it doesn't mean its the right thing to do. Intel is a semiconductor manufacturer, not a financial theory bellweather.

What I didn't explain is that I'd like more disclosure of options so that investors could understand just how much are being granted to employees. The factor that I mentioned, but didn't fully illustrate would be one way to help investors understand the true impact of excessive option issuance. The current method of trying to value a highly speculative and variable amount is craziness and is why most professional investors exclude the number when looking at core earnings.

 

Post a Comment

<< Home

Subscribe in NewsGator Online Blogroll Me! Subscribe with Bloglines Who Links Here

Email Me