The news that major investment banks were trying to bring back synthetic collateralized debt obligations last week was met with a lot of head-shaking and eye-rolling. Less noted is what the Financial Times reported over the weekend—it's not working. And the reason why it's not working turns out to be kind of inspiring. Apparently people learned their lessons from the last go-round after all!
The basic way the product was supposed to work is this. You have a bunch of loans that you can aggregate together into a bundle. Then instead of simply slicing up homogenous elements of the bundle, you create different "tranches." There's a "senior" tranche that's meant to be very safe since it gets paid off first. Then there's a "mezzanine" tranch with more risk and higher yield, and a "equity" tranche with the most risk and the highest yield. By tranching in this manner you can create arbitrary risk/return profiles regardless of the riskiness of the underlying assets. At least in theory this is a good way of matching borrowers' needs for money with investors' desire to manage risk. In practice, it turned out that the senior tranches weren't nearly as low-risk as advertised and blah blah blah, you know the rest.
So why did the comeback fail? Well, for the best possible reason—nobody wanted to buy the senior tranches.
Mark Hale from Prytania Investment Advisors explained it to the FT: "Many of the traditional buyers no longer exist, and those that survive have very bad memories." In other words, the system worked. Lots of buyers of these senior tranches got so badly misled that they ended up going out of business. Those who remain are both suspicious of the products and suspicious of the investment banks that are trying to sell the products. Lessons have been learned, and reputational and financial penalties are being paid by institutions whose past products worked out poorly for customers. A very nice contrast to a lot of the continuing bad news we hear out of the financial sector.