How Did Paul Krugman Get It So Wrong?

Story Stream
recent articles

In July 2008 Nobel laureate Paul Krugman wrote that Fannie Mae and Freddie Mac (the GSEs) "didn't do any subprime lending, because they can't: the definition of a subprime loan is precisely a loan that doesn't meet the requirement, imposed by law, that Fannie and Freddie buy only mortgages issued to borrowers who made substantial down payments and carefully documented their income." (New York Times, July 18, 2008)

Earlier this month he compounded his error when he stated:

"Zombies, zombies, everywhere. One of the enduring myths of the financial crisis has been the claim that it was the result of (a) Fannie and Freddie (b) the Community Reinvestment Act, which forced poor, helpless bankers to make loans to you-know-who. It's a myth that won't go away - I get asked about it almost every time I give a public lecture - even though it has been extensively debunked." (New York Times, November 2, 2009)

Krugman could not have been more wrong in his assertions. Somehow he missed Fannie and Freddie's acquisition of $4.3 trillion in subprime, low down payment (5% or less) and Alt-A loans. How about the $2.7 trillion of CRA loans? After accounting for overlap among these groupings, he somehow missed some $5 trillion in such loans, trillions of which remain to plague the nation's economy.


Let's start with the assertion that GSEs only buy mortgages made to borrowers with substantial downpayments. While it is generally accepted that a substantial down payment would be 20% or more of a home's value, let's be charitable and call a down payment of 10% or more substantial. Over the period 1992-2007 Fannie and Freddie acquired $1.3 trillion in home purchase loans with a 5% or less, amounting to 62% of all such conventional loans originated nationwide over the same period. These loans are now defaulting at 7-8 times the level of the GSEs' traditionally underwritten loans with <=90% LTV. Fannie started buying loans with only 3% down as early as 1994 and by 2000 Fannie was buying loans with no downpayment.

How about Krugman's claim that the GSEs didn't do any subprime lending? Over the period 1997-2007 they acquired a total of $2.2 trillion in subprime loans and private securities backed by subprime loans. Conventional subprime loans came in two "flavors". The first group consisted of loans with a FICO score of less than 660 (a regulatory definition of subprime), loans which Fannie now says are similar to subprime loans in risk but have not been classified by it as subprime. The GSEs acquired $1.5 trillion of this type of subprime loan. These loans are now defaulting at 8-9 times the level of their traditionally underwritten loans with a FICO >=660. A second group consisted of private mortgage backed securities backed by subprime loans denominated as such by the originator. The GSEs acquired $700 billion of these securities, amounting to 33% of all such privately issued subprime securities. The loans backing these securities are now defaulting at 18-19 times the level of the GSEs' traditionally underwritten loans with a FICO >=660.

How about his assertion that the GSEs' loans were carefully documented? Over the period 2002-2007 they acquired $773 billion of Alt-A loans and private securities backed by Alt-A, amounting to 55% of all such loans originated nationwide over the same period. These loans are now defaulting at 9-10 times the level of the GSEs' traditionally underwritten loans.

These three loan types helped the GSEs meet their affordable housing (AH) mandates under the "Federal Housing Enterprises Financial Safety and Soundness Act of 1992" (GSE Act). They were materially assisted in this effort by big banks and thrifts originating trillions in high risk loans to meet another mandate established under the federal Community Investment Act (CRA).

Speaking of CRA, let us not forget Krugman's assertion that CRA's involvement in the financial crisis is a myth. Over the 17 year period 1992-2008, there was a total of $6 trillion in announced CRA commitments, covering all types of CRA lending. Ninety-four percent of this $6 trillion in commitments were issued by four banks or banks these four ended up purchasing by way of merger. The banks were Wells Fargo, JP Morgan Chase, Citibank, and Bank of America. As a result, CRA single family origination volume over the period 1993-2008 also exploded, totaling an estimated $2.7 trillion.

Detailed performance data for single-family CRA lending is rarely published. In a search of the top 25 banks by single family mortgage holdings, only Third Federal Savings & Loan reported on its CRA loans.

- Third Federal Savings & Loan reported that its "Home Today" community development program constituted just 3.2% of its owned mortgage loan portfolio ($299.3 million), yet these loans represented 32% of its 90+ delinquencies. Its Home Today delinquency rate is 33% vs. 2% on its non-Home Today portfolio.

Other "sightings" of CRA loan performance include:

- Bank of America, one of the nation's largest CRA lenders, noted on its Q3:08 earnings call with equity analysts that while its CRA loans constituted 7% or $18 billion of its owned residential mortgage portfolio, they represented 29% of net losses, with an annualized loss rate of 1.26%.

- The Shorebank (Chicago) has a 19% combined delinquency and non-accrual rate for its single-family first mortgage loan portfolio The Shorebank is the nation's first community development bank. In addition to its 19% rate on single-family first mortgages, it has a 12% rate on its multi-family lending, a 9% rate on its commercial real estate, a 13% rate on its commercial and industrial lending, and a 31% rate on its construction and development lending. All rates are as of 6.30.09. These loan categories account for 98% of its total lending portfolio;

- On a more general note, a recent Fed study of CRA loans, as reported by then Fed Governor Kroszner (, identified CRA loans as a type of subprime loan and noted that "CRA-related subprime loans performed in a comparable manner to other subprime loans."

There exists a proxy for national CRA performance since approximately 50% of CRA originations since early this decade were acquired by the GSEs to help them meet HUD-mandated AH goals. CRA created the supply and the GSEs created the demand. We do know both the quantity and performance of the GSEs' loans that were AH goals rich. There were two types of AH loans that have special bearing on CRA lending - loans with down payments of 5% or less and loans to borrowers with impaired credit (generally represented by borrowers with FICOs below 660).

Fannie's traditionally underwritten loans have a serious delinquency rate of 1.8%.
How are Fannie's $650 billion in loans with the above high risk characteristics doing?

- Down payment equal to or less than 5% - 11.56% serious delinquency rate
- FICO < 620 - 16.08% serious delinquency rate
- FICO >= 620 and < 660 - 11.32% serious delinquency rate

On average this rate is 7x the rate on traditionally underwritten loans. This result should come as no surprise since Freddie Mac published its estimated default rates by loan-to-value (LTV) in the late 1990s and reported that its 95% LTV loans had about 6 times the default rate of 80% loans.

CRA and GSE Act promoted "innovative or flexible" lending practices such as downpayments of 5% or less, acceptance of impaired credit, higher debt ratios and creative definitions of income. This loosened underwriting resulted in total CRA originations and non-overlapping GSE AH acquisitions by the GSEs of $7 trillion over the period 1993-2007. This tsunami of high risk lending spawned and sustained a housing bubble unlike any this country has ever seen.

Why did GSE Act and CRA's mandated lending have such a huge impact? Historically home prices were determined by supply and demand at a local level. These two acts changed this local dynamic. Both operated nationally, due to the fact the Fannie and Freddie, along with the big banks responsible for the overwhelming majority of announced CRA commitments, were all national in scope. They were not only largely independent of local supply and demand pressures; their loosened credit standards created demand.

It was a short leap from getting the facts surrounding the GSEs' and CRA's role in the financial crisis wrong, to being wrong about the causes of the financial collapse. On September 2, 2009 Krugman wrote an opinion piece entitled "How Did Economists Get It So Wrong?" In it he chastised the economics profession for its "blindness to the very possibility of catastrophic failures in a market economy...." (New York Times, September 2, 2009)

Of course, Krugman just assumed that the failure related to "a market economy". Because Krugman is wrong about Fannie, Freddie and CRA, he missed the fact that government policies, not market failure, caused the losses suffered by homeowners, investors, taxpayers, commercial banks and investment banks.

Edward Pinto is the chief risk officer and co-director of the Internal Center on Housing Risk at the American Enterprise Institute.    

Show commentsHide Comments

Related Articles