Read “The Paradox of Securitization” on the Carnegie Council’s Policy Innovations website.

Read the full article on the Carnegie Council website.

Sylvain and Ann specifically address the subprime crisis, how it happened, who is responsible, and whether economic calamities of this kind could be avoided. To this last question, their answer is “yes” — with the establishment of effective standards and accountability to address the dynamic nature of structured finance.

“Because I’m an aerospace engineer by training, I was asked to make an analogy between structured finance transactions and keeping an airplane from crashing,” Sylvain says. “There’s a widespread assumption that financial markets are ruled by self-interest and therefore beyond the control of standards and scientific laws. However, we have to remember that there was also a time when there was no real science of flight, either. Eventually people decided that it was unacceptable for airplanes to keep crashing all the time because of the loss of life, and then the appropriate science, standards and accountability were developed to minimize the danger. Airplanes still crash today, but rarely. The same could be true of structured securities if we cared enough,” he says.

“No one has developed a theory of how to do deals yet, because the consequences are not considered severe enough. There still seems to be an assumption that it’s all right for markets to crash, because it doesn’t kill anyone — at least not directly. But market crashes cause huge damage. Why should they be accepted? Why shouldn’t a science be developed to reduce risk, and standards and accountability be imposed to penalize those who contribute to financial failure? We need to change the philosophy of the standard approach to finance — that self-interest dominates, and that loss and ruin are acceptable and even inevitable. At the same time, we have to become willing to engage risk. Using another analogy, when passing through a green traffic light, we have to take the precaution of still looking both ways for someone who may be running a red light, instead of treating the signal as some kind of protective barrier that will keep the other cars from killing us.”

R&R has been forging relationships in Japan through the Japan Society in New York City, where Ann is a member of the Business Advisory Committee of the U.S.-Japan Innovator’s Network. The Network is an initiative that brings together innovation leaders from Japan and the United States to explore fresh approaches to collaborative problem-solving. For more information, visit the U.S. Japan Innovator’s Network Web site.

For further information on viewing this episode of “NHK Special” in the U.S., see programming and subscription information on the TV Japan Web site.

First, the team.

The client (let’s call it Equilox) was a company formed by a disgruntled managing partner of a well-branded equipment lessor. Equilox had adopted their same systems and same basic underwriting approach. Equilox understood securitization. They knew the static pool, not the portfolio, was the unit of measurement. They knew that credit scores could be used to discern relative value between deals with similar partial loss curves. Equilox had a contract with a credit scoring firm (let’s call it Scorex) for the equipment lease sector. Scorex had a modeler from a well-branded credit scoring firm. Equilox had also lined up a monoline insurance company to wrap the senior notes, and an investment bank to underwrite the transaction. What the monoline’s endgame was, is now topic du jour. The investment bank had no stake in any outcome other than a sale. R&R’s role, as usual, was to build the transaction model to find the cheapest cost of capital consistent with robustness using Scorex inputs to modulate cash flows, and then offer that analysis to the bank and the monoline as a negotiating tool, for a fee.

Now the dilemma.

Credit scores cannot be used in a securitization as raw scores. They first must be mapped to the distribution. It may have changed. If it has changed, they must be revalidated. This mapping exercise showed something much worse. Scorex had no predictive value. Furthermore, the modeling exercise using pure static pool data showed the target ratings were wrong, given the structure. It is amply evident today that ratings (like credit scores) must also be validated by relating them to a distribution of risk. But since, at the time, the monoline did not understand the relationship between the rating and the risk distribution, our analysis was rejected in favor of the investment bank’s counsel.

Now comes the punch line. How many Equilox tranches do you think are out there in today’s market. Some are much better than you think. Some are much worse than you think. You won’t know which to trust.