GSE Critics Ignore Loan Performance

Money talks. It says the only way to measure the quality of mortgage underwriting is to track loan performance — delinquency and default rates, loss severity — in comparison with the rest of the mortgage market.  Otherwise, any analysis of the government-sponsored enterprises' role in housing finance is meaningless.

And yet, critics demanding GSE reform ignore the topic altogether.  Search through any book or article promoting the thesis that the GSEs helped cause the mortgage crisis for a passage comparing GSE loan performance with the rest of the market. Almost certainly, you will come up empty-handed.

There is no data anywhere to cast doubt on the vastly superior loan performance of the GSEs.  Year after year, decade after decade, before, during and after the housing crash, GSE loan performance has consistently been two-to-six times better than that of any other segment of the market. The numbers are irrefutable, and they show that the entire case against GSE underwriting standards, and their role in the financial crisis, is based on social stereotyping, smoke and mirrors, and little else. 

Consider Fannie Mae's historical loan performance, reported each year by the Federal Housing Finance Agency in its Annual Report to Congress.  Over a span of 37 years, from 1971 through 2007, Fannie's average annual loss rate on its mortgage book was about four basis points. Losses were disproportionately worse during the crisis years, 2008 through 2011, when Fannie's average annual loss rate was 52 basis points.  Freddie Mac's results are comparable.

By way of contrast, during the 1991–2007 period, commercial banks' average annual loss rate on single family mortgages was about 15 basis points. During the 2008-2011 period, annual losses were 184 basis points. 

Or check out the FHFA study that compares, on an apples-to-apples basis, GSEs loan originations with those for private label securitizations. The study segments loans four ways, by ARMs-versus-fixed-rate, as well as by vintage, by FICO score and by loan-to-value ratio. In almost every one of 1800 different comparisons covering years 2001 through 2008, GSE loan performance was exponentially better. On average, GSE fixed-rate loans performed four times better, and GSE ARMs performed five times better.  

Mortgage analyst Laurie Goodman estimated that private label securitizations issued during 2005-2007 incurred a loss rate of 24%, whereas the GSE loss rate for 2005-2007 vintage loans was closer to 4%.

And yet, large numbers of people remain convinced that Fannie and Freddie's underwriting standards caused the mortgage crisis.  Why is that? The only plausible answer is that people are besotted by social stereotypes.  Words like "government" and "affordable housing goals" make them jump to the unsupported conclusion that underwriting standards were compromised.

GSE critics seem disconnected from the world of business, which draws keen distinctions among words, actions and final results.  In the real world, Edward Pinto's expansive definitions of "subprime" and "Alt-A," used by the American Enterprise Institute and others, have no impact on loan performance data. Neither does the allegedly overbearing personality of former Fannie CEO Jim Johnson. Nor do any selected quotations of Barney Frank. 

GSE critics also claim that Fannie and Freddie led the rest of the market in a race to the bottom. This fanciful theory is based on a series of false equivalencies, wherein low-income borrowers are considered no different from subprime borrowers, and no different from those who took out two-year teaser rates or liar loans. 

This race-to-the-bottom narrative implies that GSE securitizations are the same as private label securitizations, as if concepts like, "nonrecourse," or "originate to distribute," versus "buy and hold," are not meaningful.

Since every GSE mortgage securitization benefits from a corporate guarantee, GSEs' retention of credit risk is diversified among huge portfolios of low risk loans booked before, during and after the bubble. By way of contrast, a private label securitization is a single static portfolio in liquidation. So it made no sense for private label deals to mimic GSE credit standards.  

And yet, Sen. Johnny Isakson, R-Ga., addressing an audience of industry professionals last month at the Zillow Forum on the Future of Housing, persisted in repeating the myth that Fannie and Freddie's underwriting standards caused the crisis.  

Soon thereafter Sen. Bob Corker, R-Tenn., greeted the nomination of Rep. Mel Watt, D-NC, to head up the FHFA by declaring, "The debate around his nomination will illuminate for all Americans why Fannie and Freddie failed so miserably." 

Let's finally have an honest debate about the GSEs. Because the only way to illuminate what happened is to examine loan performance data.

David Fiderer has previously worked in energy banking for more than 20 years. He is currently working on a book about the rating agencies.