UI Says Underwriters Should Consider Rent History
Access to credit remains tight and the Urban Institute (UI) blames in part that lenders are not measuring the credit risk of borrowers appropriately. Laurie Goodman and Jun Zhu, writing in UI's Urban Wire blog say that paying rent is the most significant financial commitment of most renters. Yet, while credit reports often ding renters for missing rent payments, the performance of good tenants doesn't enter into their credit scores.
Considering a borrower's rental pay history, this could be done via bank statements, to the mortgage qualification process, they say, would make assessing renters' credit risk easier. It could also expand access to homeownership among a significant portion of the nation's population.
The authors analyzed rental payment histories to see how they might impact mortgage credit risk. Their analysis, funded by the National Fair Housing Alliance, shows that rent payment history is highly predicative of how a borrower will pay his or her mortgage.
Goodman and Zhu used loan-level credit data from all of Fannie Mae and Freddie Mac's (the GSEs') fixed-rate, fully documented and amortizing mortgages issued from 1999 through 2016 and involved in the GSEs' credit risk transfer transactions. They first sorted the loans by their payment history over two years, from Q4 of 2012 to Q3 of 2014, tallying up the number of missed payments. They then looked up the share of those mortgages that became seriously delinquent, i.e. over 90 days, over the subsequent three years, through the third quarter of 2017.
They found that a loan with a perfect payment record for 24 months had only an 0.25 percent probability of becoming seriously delinquent over the following three years. With one missed payment the probably rose to 4.36 percent, jumps to 28.2 percent with two missed payments and with three or four goes to 47.8 percent and 56 percent respectively.
Renters tend to be less affluent than homeowners, have lower credit scores, and make a smaller downpayment the authors cut the results several ways, using three categories of FICO scores, under 700, 700 to 750, and over 750, and loan-to-value (LTV) ratios under 80 percent, over 95 percent, and between those levels.
With the lowest FICO scores, the probability of a loan with no missed payments in the first assessment period becoming seriously delinquent in the second is 1.03 percent but only 0.13 for borrowers with scores over 750. Similar results were found for LTV ratios. Borrowers with ratios over 95 percent and no missed payments had a 0.53 percent probability of seriously delinquencies, for those with less than an 80 percent LTV it is only 0.22 percent.
Thus, Goodman and Zhu say, it appears that borrowers with perfect two-year records performed "extraordinarily well" over the next three years, regardless of FICO score or LTV ratio. Even if both the usual underwriting measures raised their risk, for example those with scores under 700 and LTV's between 80 and 95 percent, those who had never missed a payment had a three-year serious delinquency rate of only 1.14 percent. With one missed payment the rate for that group jumped to 10.27 percent and with three missed payments to 60 percent.
To solidify the comparison between homeowners and renters, the authors used the Census Bureau's 2016 American Community Survey (ACS), sorting homeowners with mortgages and renters by their income categories and restricting the sample to similar housing structures. While rents paid by most income groups were lower than that group's mortgage payments, to even the comparison the authors also included monthly owner costs versus monthly gross rents (including utility costs.)
After these adjustments, they found that, with two exceptions, rents and mortgage payments were comparable. The exceptions were the under $20,000 group and those with incomes over $120,000. Homeownership for those groups was generally (but not greatly) more expensive.
The authors conclude that predictive value of past mortgage payments on future performance and the comparability of monthly expenses for both renters and owners, likely make rental payment history a strong predictor of mortgage default, and thus a powerful indication for credit risk purposes. They declare this a clear evidence that rental pay history should be considered in assessing the creditworthiness of a renter attempting to buy a home.