The Wrong Crisis

Robert Waldmann

What isn’t going on in financial markets ?

There are various theories of the financial crisis which suggest simple solutions and have the only defect that they are fantasies.

1) The only problem is mark to market accounting.

Market prices for mortgage backed securities (MBS) are very low because of a panic. Banks are required to list assets on their balance sheet at market prices. Banks appear to have negative equity (debt greater than assets) because of this silly rule. This caused the crisis.

This seems to be the theory of a weird left right alliance including Rep Darrel Issa (R Calif.). It is based on the idea that ideologue Congress persons can price assets better than professionals. The assets include MBS but also debt and shares of banks. This is fairly clearly a wish fulfillment fantasy. The idea is that if we resolutely say everything is fine, everything will be fine.

2) The problem is adverse selection in the MBS market and mark to market accounting.

Here there is a theory better than “I know best” for why MBS market prices are below the average hold to maturity value of MBSs. MBSs are worth less than was thought a year ago, but some of them are worse than others. For one thing, the value depends on the standards of the initiator of the underlying mortgage e.g. are they liars’ loans or normal loans. Owners of such MBSs know more about them than potential buyers. Therefore, the stuff that is for sale is the most toxic of the toxic securities. Therefore the market price is the price for the worst of the worst of MBSs. Therefore it would be irrational to sell anything which is only semi toxic. There is a separating equilibrium with low volume of trade and very low prices. Applying such prices to all MBSs understates their value.

Now without irrationality this shouldn’t matter. People who know that the stuff for sale is the worst of the worst should also know that the other stuff is better no ? Here market participants are required to be sophisticated when they buy and sell MBSs and unsophisticated when they read balance sheets.

3) The problem is mark to market, adverse selection and capital controls.

Ah now a regulation which is definitely unsophisticated. The rule is that Assets minus debt divided by debt must be over some number. If this rule is applied along with the principle that assets are marked to market it can cause two equilibria one with high asset prices and non binding capital controls and one with low asset prices and binding capital controls.

Here the problem is that all entities must be equally exposed to MBS risk and initially equally close to the capital requirement. In the real world different entities have different capital requirements. In particular capital requirements for investment banks (set by the SEC and relaxed in 2004) are much looser than capital requirements for commercial banks. Also hedge funds’ capital requirements are set by their counter-parties and are (if this is still true) about 2%. Plus Warren Buffet had $5 billion to spare. Note that the firms with the most flexible capital requirements are the ones that no longer exist. A problem for the theory.

More importantly,
binding capital requirements should cause a flow of assets from the firms with binding requirements to the ones with some slack. This flow would, among other things, make markets thick, volume high, markets unfrozen and adverse selection drowned in the flood of liquidation by firms with binding capital requirements (note the argument works fine and the metaphor isn’t even mixed). It is very hard to have a market frozen by adverse selection with market prices below the value of the average asset causing, via mark to market and regulations some firms to have to liquidate large amounts of assets. Firms that have to liquidate eliminate the separating only the worst of the worst is for sale equilibrium.

If agents are forced to sell assets at fire sale prices any agent who isn’t in the same bind should be buying at fire sale prices. It’s just not true that all entities in the world have binding capital controls. The explanation is nice in theory, but doesn’t fit the facts.

Now all these theories would imply that the original Paulson plan could work without a huge transfer from the public to banks. However, they don’t make sense.ù

Furthermore they are totally false for two more simple reasons which I put after the jump for no particular reason.

The theories assume that banks actually mark assets to market. They are supposed to, but they don’t. Page 81 of the September 20-26th Economist notes that at the end of 2007 Bank of America, Citigroup, HSBC, JP Morgan, Lehman Brothers, Morgan Stanley and Merrill Lynch marked less than half of their assets to market (which UBS Credit Suisse, Duetsche Bank and Goldman Sachs market more than half of their assets to market with Goldman Sachs over 75%). Sorry no link. I am working from uhm ink on deceased trees. Note first that uhm not quite all assets are marked to market. Note also which investment banks are still independent firms.

Finally, if markets are frozen and the market price makes a huge difference to balance sheets, it is easy and almost legal to manipulate the market price. Bank A can tell bank B “we will buy 1% of your toxic sludge for twice the current price if you buy an equal amount of our toxic sludge.” Then both can mark to that agreed price and make their balance sheets look fine (this is the favorite technique of Italian soccer teams who don’t trade players. Each buys the contract of a player for the other and they state an absurdly high price. They sometimes run out of cash with fine looking balance sheets). Now transactions which count as the market to mark too are supposed to be arms length, but no one is going to be picky in the middle of a crisis.

In fact, such a transaction would work without fraud if adverse selection were really a problem. Bank A could say “I will buy some of your MBSs but I pick which ones (at random)” that way, the expected value would be the average value of MBSs not the value of the worst of the worst. Still not quite arms length, but not obviously fiddling the price, since it would be a sensible transaction even if balance sheets didn’t matter.

Now some might think that the idea that fraudulent prices can be created when markets freeze is a fantasy. I think it has happened. The assets were long term calls options on European stock indices. Long Term Capital Management was short these assets. When it seemed fairly likely that LTCM would fail and be liquidated and when they had to liquidate to keep the balances of their REPO accounts above zero (they were excused from the 2% rule by counterparties) these assets suddenly became very risky and the market froze up. The market price became the list price chosen by brokers who knew that no one was buying or selling. The same brokers were LTCM counterparties. The list price of the options became absurdly huge. The counterparties were about to gain the right to seize the extremely valuable LTCM short positions (can seize when the mark to market value of LTCMs account at a bank is negative even if the hold for a month value is huge). It is alleged, without proof, that investment bank brokerages listed the absurd price then secretly traded at a reasonable price. Fact is that the market price was absurd and it was absurd in a direction helpful to the investment banks.

The crisis ended when the President of the New York Federal Reserve Bank, William McDonough told the investment banks to cut the crap and, in effect, divide up LTCM equally among them. It is very bad for an investment bank to anger the New York Fed. Bear Stearns was the only investment bank which refused to co-operate.