Sunday, October 5, 2008

Mark-to-Market vs Mark-to-History

By now it should be no question that the Fair Value Accounting rule has driven some firms into a death spiral, and been an important amplifier in the financial nuclear chain reaction we've been witnessing.

To illustrate the point, as if it needs more illustration, let's consider a sudden cut-off of gasoline supply. The cause of the cut-off is such that we can be reasonably sure of its revival, but can't be sure when. The Lambo you just bought last year is worth little on the spot market.

"That's fine," you say to yourself, "I'll just park it for now."

Unfortunately, it's not fine. You used it as part of the collateral on your new Starbucks shop. WaMu texts you at 4pm asking for more collateral. "Or else c u in court," it says, "Sorry, must do MTM." But you happen to be tight on cash. You look around the house and come to the conclusion that the only option is to give up. You let your 10 Starbucks employees go. The next month five of them go delinquent on their mortgages. And these happen to be the last straw that pushes WaMu under the 6% capital ratio stop.

WaMu's insistance on Fair Value Accounting ultimately caused its own demise in this fictional scenario. It's stupid. It's unfair to everybody involved. It's self-inflicted damage, mutually-ensured destruction.

However, supporters of Fair Value Accounting also have a point. Imagine the same scenario above, with one of the following variations:

  1. The cause of the gasoline cut-off is such that we can't be sure it could ever revive.
  2. WaMu thinks that you will default soon regardless because it knows your Starbucks is in trouble. It'd have to sell the Lambo on the market afterwards since it doesn't intend to park it at its own expense and take the risk on it.

It's simply not the regulators' job to judge the market outlook or intention of the company. And, if the company management, when given the option of suspending mark-to-market accounting, comes out saying "we don't intend to sell these mortgages anyway", how can anybody trust them? This is not even a swipe at honesty of corporate manangement. Unexpected events may force companies to change strategy with little warning.

But the solution is so simple it's puzzling why it hasn't been discussed widely and considered sincerely. Just ask companies to report both mark-to-market and mark-to-X on financial assets. The market will have to decide how much of which to use to evaluate the company, and price its stock, bonds, CDS, etc accordingly.

Take Lehman as an example. Its rapid collapse caught most people by surprise, at least before 9/11, the Thursday before its demise. While some hedge funds had been shorting Lehman for months, I'm not sure even they believed Lehman would go down so quickly, especially after the Fed window opened up after the Bear Stearns bailout. If Lehman had the option of dual reporting, people would've valued it somewhere between the two numbers, perhaps even closer to the mark-to-X number.

Trust and confidence can be a self-fulfilling prophecy. If people hadn't been afraid of Lehman's capital solvency, to a large degree due to artificial, nominal paper losses imposed by Fair Value Accounting, they would have continued doing business with it, albeit with somewhat heightened prudence. And Lehman could've survived -- if it had survived till the credit market stabilizes, it would've survived, looking pretty sitting on massive gains as most of the previous Fair Value Accouting write-downs are eventually recouped. The AIG (AIG) bailout may not have been necessary as a result. And tax-payers could've saved a cool $700B (probably more).

The X in mark-to-X could be the moving average market price of the asset, or assets of similar nature, over the last, say, one year after forward discounting. Many details must be carefully considered, of course. But I'll not dwell in them here.

This dual reporting approach has an interesting self-correcting effect if capital requirement is based on mark-to-history. In a bull market, the effectively higher capital requirement, due to time lag and lower-than-market price basis, limits the risk growth while in a bear market, effectively lower capital requirement offers relief and time to adjust. The latter is especially critical in a disaster scenario.

This dampening effect contrasts with the destabilizing effect of positive feedback of mark-to-market: companies are allowed to take on more risk based on paper gains in a bull market, reaching the height of risk just as correction is due, and forced to delever in a bear market, just when they can least afford to do so.

Even the most fervent defenders of mark-to-market admit its destabilizing effect under severely adverse market conditions. Mark-to-history capital requirement coupled with dual reporting can substantially eliminate the negative effects while retaining the positive ones.

No comments: