This blog reproduces an e-mail exchange concerning the credit default swap (CDS). The gist of the e-mail was that the credit crunch is being driven by the CDS, that there had been wild speculation in the CDS instruments, that the credit crunch had been set up so that speculators could profit on these swaps, and that they should be regulated. More or less, a conspiracy theory. There are lots of these going around. Most are fantasies.

The rest of this entry places the CDS into an insurance framework.

It then goes on to my brief interpetation of how this credit crunch came to be.The CDS is credit insurance, or insurance written to insure against the default of a bond or other credit instrument. The amounts of CDS written are very, very large.

Credit insurance is now being tested as a viable kind of insurance. It may or may not survive. Unlike life and fire insurance, it has two challenges. The distribution of failure may have enormous variance. Imagine that in a given year, there was a severe epidemic and many people died. Many life insurance companies would probably fail. They count on a normal distribution, such as such severe events happening once every 700 years. They build up reserves against it, but they could be insufficent if a Black Plague event occurred.

But credit insurance may face these severe events once every 70 years or even less.

Second, not everyone dies in an epidemic, and the insurance company may have a big pool of insured across the country or world. The risks are independent to some helpful extent. With the insuring of mortgages, what happens is that if one mortgage fails, the odds are many, many will fail because they all depend on one factor, which is home prices. When home prices fall, the mortgages go under as the equity is less than what is owed. If the people in the insured pool behind the mortgage instrument all have similar characteristics or live in similar areas, then the risks are not independent from those sources too. All of this undermines the CDS as an insurance nstrument.

So far, we have not actually heard specifics of how much the CDS contributed to AIG, Lehman, Bear Stearns failures. We hear other things, like that the actual mortgage-backed securities have lost value (not the CDS that insure them.) We hear that many people fear the repercussions of CDS holdings causing failure, but we lack hard information on this.

If the CDS risks are spread over many institutions and holders, then the losses will be dispersed and failures won’t be caused by these CDS. If some institutions wrote a lot of CDS insurance and now have to pay off, then they have a risk exposure and may fail. The monoline insurers like MBIA, we know focused on credit insurance, and they are goners. Their business model was no good, for the reasons I just gave. In the good years, underwriting these risks was like coining money. You can ask, how short-sighted can companies be? Quite a lot.

The problem here is that risk is not really as quantifiable for some risks as people might think or hope for. And yet there is a big demand for insurance against all sorts of risks.

The result is that insurance companies sometimes loosen their underwriting standards and go too far in underwriting risks.

A contributor to this problem is my field of finance and financial education. Insurance has been neglected as a field of study as compared with capital markets and corporate finance. Finance professors got all happy when they discovered a means of understanding risk in the “beta” coefficient. But they failed to heed the warnings of Mandelbrot and Black that the distributions of returns had infinite variance, which makes very unlikely events occur more often than a normal distribution suggests. Black wrote a note called “The Holes in Black-Scholes,” pointing out problems in his own option pricing model that underlies the pricing of CDS. But teachers and students went their merry way, happy to have any kind of model. The adage is that some model is better than no model.

Professors happily accepted jobs on Wall Street and joined investment bankers at very, very high pay in inventing new derivatives and finding pricing models for them. This intensive development of new instruments (the CMO, CDO, MBS, ABS, CDS, etc.) interacted with the huge demand produced by banks who could market them to Fannie and Freddie and other institutions. This interacted with the housing bubble. The breaking of that bubble then precipitated the waterfall of defaults.

Real world accidents often involve a confluence of several unlikely events that interact and were hard to foresee. The credit crunch is something like that except that there were many warnings and nothing was done to stop it. I think history will show that the banks were strongly encouraged by the authorities and Congressional incentives to make sub-par loans. This coincided with the Fed’s money supply expansion and the finance profession’s giddy success and failure to understand fully the risks they were dealing with. The result of this convergence is a big accident showing up in widespread insolvency of financial institutions.

As the credit crunch unfolds further, we may yet see the CDS take center stage. As matters now stand, home mortgage defaults themselves have driven the losses. Other categories of loans that have been packaged up into tranches may run into defaults next.