Homeownership, subprime claim assessed

Recent commentary on the subprime market looks to removing abuses from that market — not shutting it down.

Underlying this note of caution is an assumption that, while a lot of bad things have happened in the subprime market, on balance it serves a socially useful purpose. While foreclosures are too high, the market has made homeownership possible for many who could not have achieved it otherwise. But this assumption has now been challenged. The Center for Responsible Lending, an influential consumer group, claims that the subprime market causes a net loss in homeownership. (See Subprime Lending: A Net Drain on Homeownership, available at www.responsiblelending.org.) This implies that if the subprime market were shut down, homeownership would rise, a startling claim that deserves careful scrutiny.

To determine whether the subprime market increases or decreases homeownership requires a comparison.

The first is the number of homeowners who would not be homeowners if not for the subprime market.

The second is the number of nonhomeowners who would be homeowners if not for the subprime market. If the second number is larger than the first, which the CRL claims to be the case, the market reduces homeownership.

The CRL measures the positive contribution of the market as the number of subprime loans to first-time homebuyers. It measures the negative contribution as the number of subprime foreclosures.

I will use the year 2006 as an illustration because it is the year when, according to CRL, the net loss from foreclosures peaked. Its figures show that in 2006, some 3.2 million subprime loans were made, of which 1.4 million were to purchase homes. However, only about 354,000 of those were to first-time buyers, while about 625,000 subprime loans were foreclosed. CRL subtracts 354,000 from 625,000 to get a net homeownership loss of 270,000.

Parenthetically, CRL ignores the million-plus subprime home purchasers in 2006 who where not first-time buyers. Its focus is on the homeownership rate, and these buyers already owned their homes. However, a balanced evaluation of the subprime market should not ignore its role in enabling existing homeowners to upgrade.

But returning to the main question: Was the subprime market responsible for a net loss of 270,000 homeowners in 2006? It was not; the market's contribution to homeownership was positive, not negative.

CRL's mistake is assuming that every foreclosure of a subprime loan reduces the number of homeowners by one, relative to what it would have been had the subprime market not existed. That is far from the case.

On subprime-purchase loans that foreclose, CRL implicitly assumes that the borrower could have purchased with a prime loan, which would not have foreclosed.

Of course that happens, but not often. Based on my experience, perhaps one purchaser of 10 using a subprime loan could have qualified with a prime loan.

The other 90 percent of subprime purchasers needed a subprime loan to qualify. Their foreclosure did not reduce the number of homeowners because, had they been unable to obtain subprime loans, they would not have become homeowners in the first place.

On subprime-refinance loans that foreclose, CRL implicitly assumes that the loans would not have gone to foreclosure had the borrower not refinanced into the subprime.

This is also true in some cases, but is far from the rule. Most foreclosures are triggered by job losses, illness, marital problems and similar factors that overwhelm the borrower.

Because deceptive solicitations are more common in refinance than in purchase transactions, perhaps as many as 20 percent of refinance foreclosures would not have occurred had the borrower not refinanced into a subprime loan.

The other 80 percent would have gone to foreclosure had the borrower refinanced with a prime loan or not refinanced at all.

Applying my estimates, and assuming the distribution of foreclosures among purchases and refinances is the same as on new loans, the 632,000 subprime foreclosures in 2006 accounted for a reduction of about 101,000 in the number of homeowners. That is less than a third of the 354,000 subprime loans made to first-time buyers in that year.

Of course, my numbers are only educated guesses. They may be too high or too low. I would like to see an unbiased effort to dig deeper than I have been able to do. Meanwhile, the widely held proposition that the subprime market makes a positive net contribution to homeownership still stands.

The writer is professor emeritus of finance at the Wharton School of the University of Pennsylvania: jguttentag@mtgprofessor.com.