How to Wreck a Train

A paper called "Anatomy of a Train Wreck" by Stan Liebowitz tries to explain the US housing market crash. Here's the abstract (my emphases):

Why did the mortgage market melt down so badly? Why were there so many defaults when the economy was not particularly weak? Why were the securities based upon these mortgages not considered anywhere as risky as they actually turned out to be? It is the thesis of this chapter that, in an attempt to increase homeownership, particularly by minorities and the less affluent, an attack on underwriting standards was undertaken by virtually every branch of the government since the early 1990s. The decline in mortgage underwriting standards was universally praised as an ‘innovation’ in mortgage lending by regulators, academic specialists, GSEs, and housing activists. This weakening of underwriting standards succeeded in increasing home ownership and also the price of housing, helping to lead to a housing price bubble. The bubble increased the number of housing speculators with estimates indicating that one quarter of all home sales were speculative sales prior to the bubble bursting. The recent rise in foreclosures is not related to the subprime/prime distinction since both markets had similar size increases in foreclosures that occurred at exactly the same time. Instead, the adjustable-rate/fixed-rate distinction is the key to understanding the rise in foreclosures. This is consistent with speculators turning and running when housing prices stopped rising. It is not consistent with the nasty-subprime-lender hypothesis currently considered to be the cause of the mortgage meltdown.

The "attack on underwriting standards", according to the author, was largely based on a study that suggested widespread discrimination in lending against minorities, but which turned out to be based on faulty data; a reanalysis removing cases with obviously incorrect data produced no evidence for discrimination. All in all, the analysis is similar to the one presented by Steve Sailer, scattered throughout many posts at his blog.

The weakest link in his narrative is pointed out by Liebowitz himself:

We know where the idea of flexible underwriting standards came from and we know how relentlessly it was pushed by almost every government organization or quasi-government organization associated with the industry. But how did investors, who are supposed to be cool and rational, misperceive the risk so badly? One of the questions about the current crisis is why purchasers of mortgages (i.e., mortgage backed securities) were willing to treat them as AAA and perhaps more surprisingly, why the rating agencies were willing to give them AAA ratings.

Although it is not clear that any answer to this question can be completely satisfactory, I believe that if it is understood how universal the idea of ‘flexible underwriting standards’ had become, how dangerous it was to suggest anything else (and risk being labeled a racist) and how strong this force is, even now, it becomes possible to understand how investors, who, just like other human beings, are prone to mistakes (the dot com bubble is another recent example), might be led by the same arguments that were being repeated by so many others.

He goes on to present some limited evidence for this.

Lending is one of the many markets in which widespread discrimination on the basis of race (or sex or whatever) seems unlikely to pose a major problem. Even if all lenders are racist, it seems likely there is one who likes her profits better than her discrimination and will have a go at raking in the profits her competitors leave lying on the street. And if you're a government pressuring businesses into practices that seem obviously bonkers (such as giving six-figure loans to people who produce no proof of income), you should know you're doing something wrong.

The paper does not address the question how the bursting of a bubble in one market in one country could cause a worldwide financial crisis.