As the debate to overhaul the financial system heats up, discussion about what to do with mark-to-market (MTM) accounting has also increased. James Chanos' op-ed in the Wall Street Journal this morning defends the fair value accounting practice as "a compass for investors to figure out what an asset would be worth in today's market if it were sold in an orderly fashion to a willing buyer." However, MTM during the September meltdown acted less like a compass and more like an anvil when it came to valuing assets.
Chanos and others are right that there is a lot of good thinking in MTM accounting. The principle is that the value of an asset is whatever someone will give you for it. Logically, that would be marked to today's market price.
The problem with this, and something that Chanos does not answer in his WSJ commentary is, that value is subjective both short-term and long term. Chanos argues "we should work to make MTM accounting more precise," but the point is that you can not be precise with varied market actors looking at asset values differently.
Opportunity cost is a huge mitigating consideration not accounted for with mark-to-market. For example, in an emergency or natural disaster, luxury goods (paintings, jewelry, etc) lose their short term value compared to food, shelter, or means of escape. A can of beans was much more valuable during Katrina than a pair of diamond earrings, unless you were in Seattle or New York, away from the storm. And once the storm passed and the flood waters drew back, the value of beans and diamonds reverted to their normal worth in the New Orleans area.
If there was an universal price on diamonds set by MTM rules, how would it account for the subjective value assessment of individuals in New Orleans or New York during Katrina?
But we can take it a step further. Assume that all of America is hit by a hurricane. The value should be the same in New York and New Orleans. But its possible that there is someone who already has a lot of food, more than enough to ride out the storm which will end at some point. He could be willing to trade supplies for a luxury good, thinking ahead to when the crisis will pass. Opportunity cost for him plays into the subjectively varying price of diamonds or anything else.
Mark-to-market tries to take the principal of free markets but sticks it in a box and controls it. There is rarely one fixed price for an asset in a fluctuating market. In a free market, baselines emerge, but they are flexible. Varied opportunity cost and subjective value mean that you can't effectively mark-to-market an asset in an economic downturn. The market is made up of varying actors with varying perspectives that can't meaningfully be lumped together.
So what is the best policy? The desire for fair value accounting standards is significant. I understand the goal is to prevent fraud, and I stand by regulations that do prevent fraud--but not if they create a host of unintended consequences. The diagnosis may be right, but the prescription wrong. The challenge we face is information asymmetry that might allow banks "to substitute their own wishful-thinking judgments of value for market prices," as Chanos critiques. But this means we need to look at ways of increasing asset transparency, not rigid accounting methods.