Monday, February 23, 2009

All About The Gaussian Copula Function

Our friend Felix Salmon has written a very insightful article for Wired magazine on the implications of the Gaussian copula function for the capital markets. We recommend that anyone, who is curious about why correlation and index arbs desks are doomed in this high vol environment, should read it.

Sphere: Related Content
Print this post

3 comments:

Anonymous said...

that was excellent thank you.

methinks its companion is equally good, if not better, only because it speaks of the possible future, instead of probable past.

http://www.wired.com/techbiz/it/magazine/17-03/wp_reboot

Anonymous said...

There are a quite a few factual problems with this article, which massively overstates the importance of the Gaussian copula formula with respect to the failure of the CDO and securitization markets.

The Gaussian Copula formula (and its variants) is a foundation used to price and risk manage primarily synthetic CDOs backed by corporate CDS.

Full capital structure cash flow CDOs backed by corporate bonds, corporate loans and mortgages (both residential and commercial) are not structured, priced or rated using the Gaussian copula approach.

In fact, most mortgage and loan CDO desks did not start looking at "correlation" until 2006, a full 15+ years after these types of CDOs were created. And thinking about these CDOs in a "correlation" framework did not change how they were created, valued or traded

The rating agencies, although aware of CDS levels (i.e. market implied default probabilities) and the Gaussian copula approach, did not use them to assign ratings to mortgage and loan cdos, or even corporate cds backed CDOs. Instead the rating agencies took an historical default probability approach.

Now, that's not to say that corporate synthetic CDOs have not blow up (they have), and the framework does have significant limitations (as demonstrated by these blow ups). But the impact the Gaussian formula has had on the mortgage and loan CDOs is minimal. In fact most would disagree that it has had any impact at all, as it just was not used in these markets. Mortgage and loan CDOs blew up for other entirely other reasons.

Unknown said...

What was the reason though?