Last week, the left-wing blogs were abuzz with renewed criticism of Ed Pinto’s data on subprime and Alt-A lending. Mike Konczal and Paul Krugman triumphantly displayed a graph from a February 2011 paper by David Min of the Center for American Progress that they claimed as proof that Pinto’s numbers—which I relied on in my dissent from the majority report of the Financial Crisis Inquiry Commission—were fraudulent. The graph is copied below.
Honestly, it’s hard to believe anyone gives these characters the time of day, let alone reads their work. The Min graph is grossly deficient in almost every way possible, and the fact that it would be cited by both Konczal and Krugman confirms their utter ignorance of this subject.
Let’s start with a basic problem. The chart is a fake. It’s mislabeled as coming from the Mortgage Bankers Association (MBA) National Delinquency Survey for the second quarter of 2010, but only the two bars labeled “Conforming” and “Actual Subprime” actually come from the MBA’s survey. The MBA survey includes only three categories—government loans, prime loans, and subprime loans. The survey does not include any of the data in the two bars to the left, labeled “Pinto high risk: Freddie>90 LTV” and “Pinto high risk: Freddie 620-659 FICO.” Accordingly, the data in those two bars had to come from somewhere else, namely Freddie Mac, but is misrepresented as coming from the MBA survey.
There are many other material deficiencies. In fact, just about everything in the graph is deceptive. If this were sales material, Min, Konczal, and Krugman would all be jailed by Elizabeth Warren. Let’s take for example the bar called “Pinto high risk: Freddie 620-659 FICO.” It shows a 10.04 percent serious delinquency rate for these mortgages. The implication is that these are all the mortgages below 660 FICO. But these are not all the mortgages in that category. Min (perhaps) neglected to mention the mortgages below 620 FICO. It turns out that 14.4 percent of those mortgages were also seriously delinquent. So Min has selected one limited group and tried to demonstrate that it is considerably smaller than the delinquency rate on the “Actual Subprime” in the chart. It’s like saying that a mortgage has an loan-to-value ratio of 80 percent, but forgetting to mention that there’s a second mortgage for the additional 20 percent.
Min’s chart is misleading for two more reasons. His argument—dutifully accepted by Konczal and Krugman—was that “Actual Subprime” (which Min never defines but are loans made and securitized by the private sector) had a much higher rate of delinquency than the portion of Freddie’s exposure that was represented by loans to borrowers with FICO scores of less than 660—which Pinto had labeled as “subprime.” This comparison is offered to demonstrate that the loans made by the private sector were worse than Fannie and Freddie’s loans.
But Min doesn’t tell us—perhaps he doesn’t know—that Fannie and Freddie were the largest buyers of these “Actual Subprime” loans, so that many fewer would have been outstanding if Fannie and Freddie hadn’t needed them to meet their affordable housing requirements. In addition, the percentages Min presents are meaningless because he doesn’t tell us the actual numbers of loans involved. For example, if there are 1000 “Actual Subprime” loans and 1 million Freddie loans to borrowers with FICO scores between 620 and 659, the latter are of course going to be more significant in terms of their effect in the financial crisis. Pinto found that what Min labels as “Actual Subprime” (and Pinto calls “self-denominated subprime”) were less than one-third of the total number of all subprime and Alt-A loans outstanding in 2008. If you want to see Pinto’s actual numbers for these low quality loans, compared to Fannie and Freddie prime loans, see Table 3, page 21 of my dissent.
Finally, even if we stipulate that the “Actual Subprime” is a significant number—enough to contribute significantly to the financial crisis—the fact that it might have been of even lower quality than the other subprime loans made by Fannie and Freddie is not relevant to Min’s claims that Pinto made Fannie and Freddie’s loans look worse than they actually were. The loans that Fannie and Freddie acquired—and Pinto identified—were of sufficiently low quality to cause these giant companies to become deeply insolvent. They have already required over $150 billion of assistance from the Treasury just to stay afloat, and their regulator has estimated that their losses may eventually total between $221 and $363 billion. Thus, they were bad enough to sink Fannie and Freddie and drive down housing prices all over the United States.
There is no end of the deceptions that the Krugmans, Konczals, and Mins will cook up in order to avoid the truth: Fannie Mae and Freddie Mac, as required by the affordable housing goals established by HUD, acquired 12 million subprime and Alt-A loans by 2008 and in the process destroyed themselves and triggered the financial crisis.