Asymmetrical Information

From the desk of Mindles H. Dreck:

So Quoth the Sage

Today Warren Buffett preached on the topic of options and pension surplus, suggesting the former should be included in the P&L and the latter should be excluded. There is something distinctly nutty about arguing both these positions, as they are essentially contradictory. This may be a cunning plan, a quiet ruse by the canny sage from Omaha to distract Congress from meddling elsewhere. More on that later.

An option is a future contingent liability whose present value can be estimated with a series of probabilistic and economic assumptions. Likewise the surplus over a defined benefit pension plan liability can be estimated using probabilistic and economic assumptions. There are problems, as I have noted, with putting both types of estimates into the P&L, as real experience is virtually guaranteed to vary from these estimates. It is nonsensical to allow this practice with negative numbers (options) but disallow it with positive numbers (surplus in the pension).

Well, you ask, isn't it just conservative to take the expense and forego the income? Not really. Another amazing facet of the public debate on this topic is that no pundit I've seen has considered what happens after the options have been expensed and they expire (as they sometimes do). Let's imagine we have expensed $25 million of options and given them to our CEO. The option term has expired, the options were worthless. Now we get to take the expense back (the books don't balance if you don't!). Companies could have a lot of fun with that by issuing a series of complex, long-dated deep out-of-the money options with impossible employment contingencies in order to create a reserve during flush times. Talk about managed earnings! Indeed, because pension earnings can be put in the P&L, they are a source of earnings distortion as estimates change and experience varies from assumption. Buffett accurately points this out, but then fails to note that expensing options might produce the same result.

Let's return to my earlier point about option expense being sensitive to volatility and consider a concrete example. Take, for instance, General Electric. Rolling 60 month annualized volatility in GE common stock has ranged from a low of about 14% to a high of 30% over the last decade. Within this range, a $30 strike option with a term of 3 years (and GE trading at $27) would be worth somewhere between $2.50 and $5.37 per option (consider the full range of values presented here (55K). Now consider Tyco, where the corresponding volatility range is 18% to 55%. A strong new executive team may be necessary to revive this company. If the board grants an executive 2 million options at the current price, they could be worth anywhere from $7 million to $18 million under Black Scholes. None of that money goes out the door, but the company is forced to expense the entire amount in order to attract talent. The volatility bug bites at precisely the wrong time.

Buffett himself has conceded this point on assumptions, and also makes the well-founded observation that it is "normally unwise for Congress to meddle with accounting standards".... before making an exception just this once.

If the market uses the reported earnings figure to value the company, inclusive of both the ex post and the ex ante option accounting, we find ourselves in a truly strange situation. The market price and behavior of the stock has determined company performance instead of company performance determining the market price. This is not a situation anyone should desire. In fact, it sounds like Enron, where the rising stock price propped up off-balance sheet liabilities.

Buffett also implies that surplus pension assets never "deliver a dime to their companies". This is demonstrably false. If a company truly has pension surplus in a defined benefit plan, they can buy annuities to cover their DB plan obligations and pocket any excess. Other avenues for realizing surplus pension assets are available in a change of control, where plans are combined and/or terminated for a new plan. Of course, if they are using 11% earnings estimates to estimate the plan's funding level, they are unlikely to realize the surplus they estimated. That's the ultimate test of a plan's funding level.

The sage of Omaha appears to have been a skillful value investor, but his public politics and moralizing leave something to be desired. At least reading this diatribe was not as depressing as watching him attempt to defend large, arbitrary and redundant taxes for the express purpose of social engineering. (i.e., the estate tax his heirs won't be paying). Of all people, Buffett must be keenly aware of how much faster inherited wealth is redistributed to the worthy when it's not tied up in government or a dynastic foundation employing all the cousins.

I don't particularly like estimations in the P&L. Both items should be left out. In case you are wondering, I favor fully diluted earnings per share, which is already reported, as the best proxy for determining P/E ratios and ridding ourselves of the supposed depredations visited upon us by the option-wielding demons of the go-go '90s.

The ugly and sad fact remains that if the market ignored the cost of options, it was our own fault. Anybody with a little experience reading financial statements (let alone Warren Buffett) can take a current annual report and determine the options grants and the impact of pension earnings on a P&L. These expenses are already all there for everyone to see. Critics like to say they are "buried in the footnotes", but every financial analyst knows the cash flow statement and the footnotes are where the accounting rubber meets the road. Changing the accounting conventions won't make any difference at all, and certainly shouldn't have an impact on the market or the likelihood of future corporate scandals. In the rush for reform, option expensing is a non-event.

Hey...that's it! Buffett is creating a distraction to keep Congress from doing some real damage! And here I am going on about it.... D'oh!

The Man Without Qualities is providing excellent coverage on the logical aspects of the arguments referenced above here and here.

And you heard me right about "appears to have been a skillful investor." At some point we need to talk about the role of randomness in investment success. The fact remains that if investment success were 100% random (which I don't believe), there would still be a Warren Buffett and a Peter Lynch to worship. As Victor Niederhoffer says, everything that can be disproved should be.

Posted by Mindles H. Dreck at July 24, 2002 09:53 PM | TrackBack | Technorati inbound links