Some of the advances in structured product technology over the past decade have truly tested the limits of comprehension of both man and machine. As soon as one leveraged instrument was created, somebody else was figuring out a way to take leverage on that leverage. But investment banks and hedge funds had some of the most high-powered computers and statistical minds on the case at each step of the way. So how do you know when you've gone too far? Where do you draw the line in the sand between useful investment tool and mathematical video game? John Thain has an idea.
In a recent Q&A session at Wharton, JT recalled a time before he and Ken Lewis became partners in crime. When the world was a different place.
These instruments were so complicated. One of these -- I'm talking now about ABS CDOs, and actually the CDOs-squared are even worse. Merrill created one. I picked one particular one. To actually model out the things that are inside an ABS CDO [is difficult] because derivatives are already inside the CDOs. They're already derivatives created out of derivatives. So you have to go way underneath to get to the actual pools of mortgages. To model correctly one traunch of one CDO took about three hours on one of the fastest computers in the United States. There is no chance that pretty much anybody understood what they were doing with these securities.
Creating things that you don't understand -- that the buyer doesn't understand, that the rating agency doesn't understand, that the regulator doesn't understand -- is really not a good idea no matter who owns it. I think that the degree of complexity that was created in the securities, and the lack of anybody's ability to really understand how they were going to perform, was simply an error and a bad thing. The fact that the firms that created them were stupid enough to own them doesn't make me feel any better.
John Thain Comes Clean [Seeking Alpha]