In a recent post, Patrick mentioned some proposals that have been floating around that supposedly would help the banks. One of them is to suspend “mark to market” accounting, which comes from the idea that a lot of the assets at the banks only look so bad because of overreaction. This meme has actually been around for about two years now (it first cropped up when the banks were valuing things at 90% because surely that was impossible) but most of the commentators I respect have never given it any credence. They argued that if housing prices went to historical values then default rates would be so high that the market prices were realistic.

The idea has been hard to test because the banks have refused to sell them to prevent price discovery. As long as they don’t sell then they can pretend that they are worth what they want and indeed, while the market bid price for a lot of this stuff is very low, most banks are still carrying it at rather high valuations. A few events have happened (like Lehman collapse etc) that have forced sales, but the mark to market rules allowed banks to ignore those as “distressed sales.”

Of course many of these assets have complex rules and trigger events that force liquidation once a certain percentage of the pool goes into default. The banks have no choice about this, and looking at how many pools have entered into default and the recovery rates gives a look into what they are actually worth. Apparently, not much. In fact, it’s possible that they are worth even less than the market bids, but it’s hard to tell because each one is so different.

I’ll let Hilzoy summarize how bad it is:

It’s actually hard to understand how the banks managed to do this badly: you’d think they could have done better hiring people off the street and paying them to put all those nice little loan documents into piles at random, or tossing mortgages down the stairs and bundling them based on how they landed. They certainly didn’t need to hire people with advanced math degrees and pay them seven- or eight-figure salaries to get these kinds of results.

And how about those ratings agencies? They would have done a better job using a Magic 8-Ball to rate the CDOs. (“Signs point to junk!”)

I have been hearing for years and years about how the financial services sector pays such exorbitant wages because the people who work there are so immensely talented that they are cheap at $50 million a year. I never particularly bought that line before. But I never imagined that all those Masters of the Universe would do quite this badly. If we had paid them $50 million a year to go far, far away and leave our financial system alone, it would have been a bargain.