The complex derivatives that are PART of the blame for the economic nightmare we are entering are often overlooked by the average reader. No more.
Until today, even I had not ventured into trying to understand the risk management for CDO’s. I had always believed that the problem came from the fact that we discovered the “ability” to package more and more instruments together. This is not the case…these things are easy to package. The problem was that our risk analysis of these packaged goods produced a result saying the packages were not risky.
It sounds complicated, but a Wired article, titled “Recipe For Disaster“, explains it quite clearly and it is a must read. Here is my even more dumbed down explanation and example.
The Product: Take a bunch of mortgages from the same town. Bundle them together with a bunch of mortgages from another town. It all becomes one big product that has interest payments, like a bond. Call this a CDO.
The Pricing: What we need to figure out to price the risk is the correlation between all of these mortgages defaulting at the same time.
The Old Model: In the old days, investors would look at all of the factors that could effect the correlation between these defaulting and analyze each one independently and then try to combine it.
The New Model: Forget all the underlying factors effecting correlation of default. Instead, create a new product for each town, called a Credit Default Swap (CDS), that is a bet on the chances of that towns mortgages defaulting. Do the same for the other town. Then compare the correlation between how the price of those two bets (CDS’) move.
The Problems:
- The CDS’ have no direct tie to the underlying instrument (the town) itself
- The CDS’ move around based on traders protecting their risk instead of making investment decisions (they are an insurance instrument, not a true investment asset
- The history of correlation for CDS’ is only 10 years. All a real estate bull market!
- Lastly, the sellers of the CDS were making too much money selling too many of these…too much incentive to the insurance salesman!
To Blame: Who sold these CDS’? In the end, almost every bank did. But when it comes down to it…AIG sold the most. That was their business. Sell complicated insurance. They just did not know what they were selling.
*image source: investorbuddy.com
