Monday, July 6, 2009

A Rather Egregious Case of Confusing Cause and Effect

You have heard all the talk about correlation not being causation. About the perils of research: researchers finding what they (unconsciously) planted in the evidence; about being factually accurate about the observations and yet totally wrong on the conclusion; about the epistemology of science. Etc.

All those abstract issues were beautifully rendered in a real life and easily comprehensible example thanks to the research of Professor Mark Garmaise of UCLA’s business school. The Financial Times reported the results of the professor’s iconoclastic research in which he showed that “before crisis, US mortgage brokers fed loans of deteriorating quality to the banks they did most business with.” This, the professor concluded, proved that brokers abused the trust of the banks and in doing so, planted the seeds of the crisis which, everyone remembers, began in the subprime mortgage area:
”At the beginning of a relationship, the bank’s natural intuition is to avoid fly-by-night brokers they barely know,” said Professor Garmaise, who likened bank-broker dynamics to a marriage. “The broker knows this, so they are on their best behavior, but over time the broker gains credibility and each additional mortgage matters less.”

Such breached of trust accounted for a staggering 22 per cent of late mortgage payments and 28 per cent of foreclosures in the nearly five years covered by the study.
That brokers initiated progressively lower quality mortgages is a matter of record. Very little by way of research is needed to confirm this well-documented race to the bottom. But just about everything else in the research is wrong. Far from being hapless victims, banks were the instigators of the problem. They pressured the brokers to keep the mortgage supply chain going no matter what the quality.

In the two-part series on the destruction of Fannie Mae and Freddie Mac I documented this pressure and explained the reasons for it. In Part 2, I quoted the following telling passage from a New York Times story:
[William D. Dallas, the founder of a mortgage brokerage] recalls being asked to make more “stated income” loans, in which lender do not verify the information provided by borrowers and brokers with tax returns, pay stubs or other documentations. The message, he said, was simple: You are leaving money on the table – do more of them.
Rewriting history is a favorite sport among the powerful who stand exposed by the light of the past. (A Soviet era politician once quipped that “nothing is more unpredictable than the past.”)

That is where the role of true scholarship, i.e., disinterested scholarship, comes in: to filter the noise injected into past events by the self-serving spin of interested parties.

But there are no disinterested scholars in the U.S. universities. A finance professor in a major business school simply cannot function without constantly bringing in grant money. And when you take the king’s shilling, you play his tune. More: you dance to his tune.

This is the academic cadre that is expected to find a way out of the crisis.