Fast pay-off, low down payment loans perform well
By Laurie Goodman and Ellen Seidman Jun Zhu :: January 30th, 2014
(This post is the first in a series on “Housing the next generation.”)
We’ve heard a good deal recently about how upcoming potential homebuyers will be locked out because of low wealth, high debt, and uncertain, lower incomes. While some potential new homeowners might have all of these characteristics, many will almost certainly fall into just one or two categories.
In particular, we would not be surprised to find potential homeowners who are wealth- and/or credit-constrained, but who have relatively good and steady incomes. Can they be served in a way that meets their needs and is safe and sound for lenders and investors?
We can get some hint of the answer by looking at the performance of 15-year fixed rate loans over the last 15 years. Sure, a disproportionate share of 15-year loans are refinances, but we have a large enough dataset to be able to look at 15-year purchase loans, and also to separate out high loan-to-value (LTV), low FICO score loans—that is, loans to the lower-wealth, poorer-credit borrowers of the past.
The good news is that the faster equity appreciation of a 15-year loan, in comparison to a 30-year loan, makes a huge difference in loan performance. As shown in the chart below, for our entire database of loans (CoreLogic Prime Servicing Data, which includes about 62 percent of outstanding loans), only 1.8 percent of purchase money 15-year fixed rate loans ever went 90 days delinquent, compared with 6.9 percent of purchase money 30-year fixed rate loans. And that’s over the entire boom-and-bust cycle, from before 2000 through 2013.
As illustrated above, significant differences are apparent for purchase loans to borrowers with FICO scores below 700 (2.6 percent versus 10.7 percent) and for loans with LTVs above 90 (4.0 percent versus 8.2 percent). Even more impressive: the 7.5 percent default rate for greater than 90 LTV, less than 700 FICO 15-year mortgages is similar to the 6.9 percent default rate for greater than 90 LTV, 700-750 FICO 30-year borrowers. These differences held when we used a logit model to hold other factors (occupancy, state, issue year) constant.
While there may be multiple reasons for this similarity in performance, the faster equity buildup of a 15- year loan is an important contributing factor. Moreover, and not considered in the chart (which only measures delinquencies), faster equity growth means that when a 15-year loan defaults, the loss severity is likely to be lower.
As shown by comparing the first red bar to the second blue bar, low down payment 15-year purchase loans outperform 30-year purchase loans of all LTVs. Taking that into account could open the door to homeownership to many more low-wealth families, even those without pristine credit, if their incomes allow them to afford a higher monthly payment.
Should everyone be required to have a 15-year loan? No. The advantage of the 30-year mortgage has always been lower monthly payments, and for many borrowers, those are key to being able to purchase a home. Moreover, good underwriting, which often was not the case during part of the period at issue but that will be required by the ability to pay rules under the Dodd-Frank Act, will enhance the quality of all loans. But this analysis provides one illustration that a higher down payment isn’t the only road to good performance.
Follow Ellen Seidman on Twitter at @esseidman.