Rental data may improve credit model predictability by 10% or more

Combining information about a consumer's rental payments with traditional credit data made it easier to predict their ability to repay than standard performance indicators alone, TransUnion found in a new study.

The study, which analyzed how well 1 million consumers with rent data furnished to TransUnion at the end of 2018 paid debts in 2019, found that the inclusion of rental data can make credit models at least 10% more predictive of new 90-plus-day delinquencies for various types of consumer finance products, including mortgages, credit cards, auto and personal loans. That’s significant because outside of the experience during the 2020 pandemic, when long-term forbearance has been permitted, servicers commonly consider consumers with payments late by 90-days-plus to be in default.

The findings in the new TransUnion research could lend momentum to two major government-related mortgage investors’ recent efforts to encourage the reporting and use of rental-payment information in lending. One of the two government-sponsored enterprises that buys a significant number of the home loans made in the United States, Fannie Mae, has begun experimenting with rental data’s use in single-family underwriting. The other, Freddie Mac, has been working with a business partner to encourage multifamily borrowers’ and tenants’ involvement in reporting.

While other studies have found incorporating rental data can give newer versions of consumer credit scores a lift, potentially increasing borrowers’ ability to obtain loans at lower rates, the most recent research focuses more on what the outcomes are for mortgage servicers and lenders.

“Historically, the analysis we have done has been from the consumer’s vantage point, or what having rental data on a credit report does to a score. What’s groundbreaking about this analysis is the fact that it examines the lender’s perspective,” said Maitri Johnson, vice president of tenant and employment screening at TransUnion, in an interview.

The divergence in predictability instances where rent payment histories were included came into play primarily for borrowers who lacked a traditional credit history or had lower scores, Johnson said. This is in line with Fannie’s goal to broaden the range of lower-income borrowers eligible for homeowner loans through its rental data initiative.

“TransUnion’s findings support Fannie Mae’s belief that if someone is paying rent consistently, it’s likely they could pay their mortgage consistently too,” said Steve Holden, vice president of single-family analysis at Fannie Mae, in a press release. “This...demonstrates the potential of using technology and data to responsibly remove long-standing barriers to credit access, while helping to ensure consumers receive a more fair and inclusive credit eligibility assessment.”