CHICAGO – It is universally recognized that a key factor underlying the 2007-2008 financial crisis was the diffusion of collateralized debt obligations (CDOs), the infamous special-purpose vehicles that transformed lower-rated debt into highly rated debt.

As these structures lost popularity on Wall Street, however, they gained popularity on the other side of the Atlantic.

After all, the European Financial Stability Facility (EFSF), created by the eurozone countries last May, is the largest CDO ever created. As with CDOs, the EFSF was marketed as a way to reduce risk.

Unfortunately, the outcome could be similar: the entire banking system sent into a tailspin.

CDOs are a form of financial alchemy: special-purpose vehicles that buy the financial equivalent of lead (low-rated mortgaged-backed securities) and finance themselves mostly with the financial equivalent of gold (highly sought-after AAA bonds).

This transformation is based on one sound principle and two shaky ones.
The sound principle is excess collateral. If there is $120 of collateral guaranteeing a $100 bond, the bond is safer, no doubt.

How much safer, however, depends upon the returns on the pool of bonds that compose the CDO.

The first shaky principle is that if the return on these bonds is highly correlated, so that they all default at the same time, overcollateralization is not much help.

By contrast, if the returns are uncorrelated, it is extremely unlikely that they will all default at the same time, making overcollateralization sufficient to guarantee a safer return.

Unfortunately, no precise mathematical model can determine correlation across securities, which is always an educated guess based mostly (sometimes entirely) on past behavior.

Thus, the explosion of CDOs in the United States during the housing boom was predicated upon the shaky assumption that house prices never fall nationwide.

The second shaky principle is that, in order to validate these instruments, issuers of CDOs relied on credit-rating agencies. Historically, these agencies had been reliable in predicting the risk of corporate defaults.

Much of their credibility, however, depended on a fragile balance of power. Since each issuer represented a small fraction of their revenues, rating agencies were unwilling to compromise their reputation for the sake of any single issuer.

The CDO market, however, was concentrated: six or seven issuers controlled most of the market, and this market ended up representing 50% of the entire revenues to be gained through ratings.

All of a sudden, issuers had much more influence on the rating agencies, which, like any good seller, were ready to bend a little not to alienate important customers.

As a result, the CDO market did not so much spread risk as it shifted and hid it. When the US housing market started to turn south, the biggest underwriters (such as Countrywide) did not go bankrupt right away, because they had sold the vast majority of their loans to the CDO market.

Eventually, however, the uncertainty created by these CDOs nearly brought down the entire US banking system.
Europe is following a similar path.

The EFSF, created to assist countries facing "illiquidity,” is designed exactly like a CDO. The EFSF buys the bonds of the countries which find it difficult to finance themselves in the marketplace (for example, Ireland) and issues bonds that are AAA rated.

How is this alchemy possible? Once again, it relies on overcollateralization, an assumption on the joint distribution of possible outcomes, and the inevitable seal of approval of the three major credit rating agencies.

With the EFSF, the overcollateralization takes the form of guarantees by other eurozone countries. Among the major countries, however, only France and Germany have an AAA rating.

How can a bond guaranteed in large part by countries such as Italy and Spain (likely candidates for a fiscal crisis) provide AAA status to Irish bonds? According to Standard and Poor’s, for example, "The rating on EFSF reflects our view that guarantees by ‘AAA’ rated sovereigns and freely available liquidity reserves invested in ‘AAA’ securities will, between them, cover all of EFSF’s liabilities.”

But the value of the guarantees depends on the situation. As long as the only country to be rescued is Ireland, there is no problem. But if the EFSF has to guarantee Spain, would Germany really be willing to step in and use its taxpayers’ money to cover Spanish banks’ losses?

And how compromised might French and German banks become – and thus how much more fiscal strain is in store for France and Germany?

Here, too, no mathematical formula can help, because what we need to test is the reasonableness of our assumptions. This is why rating agencies’ opinions are so valuable. Unfortunately, one must wonder to what extent these ratings are distorted by the eurozone countries’ political power.

Since the crisis, rating agencies have been under attack, and major regulation, which could severely affect their business, has been discussed on both sides of the Atlantic.

Given this, how free are credit rating agencies to express their opinion on the very institutions that will regulate them? We will have to await the pricing of the credit-default swaps on EFSF debt, which will be issued in January, to see whether markets believe the ratings.

Independent of this test, however, Germany seems to have squared the circle: it helped the countries in trouble, without coughing up (for the moment) any euros. As with CDOs, however, this might be a Pyrrhic victory.

The EFSF has bought short-term gain in exchange for much larger losses if the situation deteriorates: a fiscal crisis in Spain could bring down the entire edifice.

After the sub-prime mortgage crisis, politicians alleged that the market was short-sighted and irrational, and rushed to propose new regulations. While some of the criticism might have merit, what gives politicians the moral authority to criticize?

After all, as the EFSF shows, their orientation can be more short term and irrational than the market’s, repeating the same mistakes because they seem not to have learned from them.

The market’s verdict is likely to be uncompromising. As Oscar Wilde said: "Fool me once, shame on you; fool me twice, shame on me.”

Copyright: Project Syndicate, 2010.