Made-Up Definitions

By James Kwak
Many commentators who want to blame Fannie and Freddie for the financial crisis base their arguments on analysis done by Edward Pinto. (Peter Wallison bases some of his dissent from the FCIC report on Pinto; even Raghuram Rajan cites Pinto on this point.) According to Pinto’s numbers, about half of all mortgages in the U.S. were “subprime” or “high risk,” and about two-thirds of those were owned by Fannie or Freddie. Last year I pointed out that Pinto’s definition of “subprime” was one he made up himself and that most of the “subprime” loans held by Fannie/Freddie were really prime loans to borrowers with low FICO scores. Unfortunately, I made that point in an update to a post on the somewhat obscure 13 Bankers blog that was mainly explaining what went wrong with a footnote in that book.
Fortunately, there’s a much more comprehensive treatment of the issue by David Min. One issue I was agnostic about was whether prime loans to people with low (<660) FICO scores should have been called “subprime,” following Pinto, or not, following the common definition. Min shows (p. 8) that prime loans to <660 borrowers had a delinquency rate of 10 percent, compared to 7 percent for conforming loans and 28 percent for subprime loans, implying that calling them the moral equivalent of subprime is a bit of a stretch. Min also shows that most of the Fannie/Freddie loans that Pinto classifies as subprime or high-risk didn’t meet the Fannie/Freddie affordable housing goals anyway — so to the extent that Fannie/Freddie were investing in riskier mortgages, it was because of the profit motive, not because of the affordable housing mandate imposed by the government.
Min also analyzes Pinto’s claim that the Community Reinvestment Act led to 2.2 million risky mortgages and points out that, as with “subprime” loans, this number includes loans made by institutions that were not subject to the CRA in the first place. Of course, the CRA claim is ridiculous on its face (compared to the Fannie/Freddie claim, which I would say is not ridiculous on its face) for a number of reasons, including the facts that only banks are subject to the CRA (not nonbank mortgages originators) and most risky loans were made in middle-income areas where the CRA is essentially irrelevant.
Mainly, though, I’m just glad that someone has dug into this in more detail than I did.