Dive Deep into Mortgage Data
Mortgage lenders are more likely to turn down buyers of color than white applicants with similar debt-to-income and loan-to-value ratios, and algorithms are likely to blame, according to an in-depth analysis of 2.4 million loan applications to the purchase by The Markup, a nonprofit newsroom that “oversees big tech.”
Nationally, black applicants were 80% more likely to be denied conventional mortgages eligible for support from Fannie Mae and Freddie Mac compared to white applicants with similar qualifications, according to the analysis.
The markup 2019 loan data analysis found that Latino homebuyers were 40 percent more likely to be turned down, and that Asian / Pacific islanders were turned down 50 percent more often than whites. The odds of Native American home buyers being turned down were 70% higher than for whites.
“In all cases, potential borrowers of color looked almost exactly the same on paper as white applicants except for their race,” the study authors said.
In addition to the nationwide findings, The Markup reported racial disparities in 89 metropolitan areas, including New York, Los Angeles, Chicago, Houston, Atlanta, Dallas, Phoenix, Minneapolis, and Washington, DC.
The markup also singled out seven lenders with the greatest disparities for applicants of color: DHI Mortgage Company, Lennar Mortgage, Pulte Mortgage, Freedom Mortgage Corporation, Movement Mortgage Corporation, Fairway Independent Mortgage Corporation and Navy Federal Credit Union.
All of these lenders have told The Markup that they comply with fair loan laws, and several have challenged the publisher’s analysis.
Details of The Markup’s analysis of local markets and individual lenders, as well as the methodology employed, have been published separately (How we investigated racial disparities in federal mortgage data).
Credit industry groups question findings
Lending industry groups criticized The Markup’s methodology, saying its analysis did not take into account borrower credit scores or analyze applications for government-backed FHA, VA and USDA mortgages.
“An individual’s credit history can help explain why seemingly comparable applicants don’t always end up with the same loan outcome,” the American Bankers Association told The Markup in a written statement. âMarkup’s analysis not only takes into account credit history, it also does not include the millions of mortgages resulting from the Federal Housing Administration and other government loan programs. These programs are specially designed to serve low to moderate income families most at risk of being denied a mortgage. This glaring omission paints an incomplete picture of the mortgage market.
The Mortgage Bankers Association provided similar comments to The Markup before publication. After the analysis was published on Wednesday, the MBA denounced him as “not only deeply flawed, but clearly biased in its premises” and distorting “the problems and solutions needed to address the very serious problems that lead to uneven outcomes related to black home ownership and wealth creation “.
Although the Home Mortgage Disclosure Act (HMDA) requires lenders to provide loan-level credit score data to regulators, the Consumer Financial Protection Bureau removes this information from public data, “in part because of lobbying from the mortgage industry. to remove them, citing borrower confidentiality. Said the authors of The Markup’s analysis.
Authors Emmanuel Martinez and Lauren Kirchner – who both have extensive experience in data-driven investigative reporting – said they were also unable to analyze decisions made by Fannie Mae and Freddie Mac’s underwriting algorithms. Like credit scores, lenders report these decisions in their HMDA records, but they are not available in public files.
Credit scores and algorithms to blame?
Nonetheless, Martinez and Kirchner concluded that credit scores, along with the algorithms used by individual lenders and Fannie and Freddie, are likely the culprits for the disparities they say their analysis uncovered.
The classic FICO score that Fannie and Freddie ask lenders to use, they said, is “widely viewed as detrimental to people of color because it rewards traditional credit, to which white Americans have more access.” It doesn’t factor in on-time payments for rent, utilities and cell phone bills, among other things, but will lower people’s scores if they fall behind and get sent to debt collectors. Unlike newer models, it penalizes people for past medical debts even if they have since been paid. “
Some of the data taken into account by Fannie and Freddie’s underwriting algorithms – including assets, employment status, and debt – can also put people of color at a disadvantage, Martinez and Kirchner noted, citing other research.
âThis is a relatively new world of automated underwriting engines that intentionally may not discriminate, but probably do so by effect,â said David Stevens, former president and CEO of Mortgage. Bankers Association at The Markup.
Stevens and others told The Markup that outsiders have little visibility into the inner workings of the algorithms used not only by Fannie Mae’s Desktop Underwriter and Freddie Mac’s Loan Prospector systems, but also by similar software used by the FHA, USDA, and individual banks and lenders.
Fannie Mae – who recently announced plans to start factoring in on-time rent payments when assessing borrowers with thin credit records – provided a written statement to The Markup saying Desktop Underwriter is analyzing claims loan “regardless of race”.
Fannie and Freddie “said their algorithms are regularly evaluated for compliance with fair loan laws, internally and by the FHFA and the Department of Housing and Urban Development,” reported The Markup. “HUD said in an email to The Markup that it had asked the couple to change the underwriting criteria as a result of these reviews, but would not release details.”
Potential for new regulations
In a written statement provided to The Markup, the MBA said regulators have access to a richer data set and have not been troubled by HMDA data in the past.
“The Federal Reserve, CFPB and other regulators have made it clear that denial disparities in HMDA data are not in and of themselves critical to the valuation of a fair loan,” said the MBA. âFair loan reviews include a much richer body of information about loans and borrowers. “
In a new 2020 HMDA data analysis, however, the CFPB found that black and Hispanic borrowers “continued to have fewer loans, were more likely to be turned down than non-Hispanic white and Asian borrowers, and paid median interest rates and total costs of higher loan. It is clear from this data that our economic recovery from the COVID-19 pandemic will not be robust if it remains uneven for mortgage borrowers of color. “
And The Markup noted that the new housing secretary in the Biden administration, Marcia Fudge, recently told Axios that she thinks one of the reasons black homeownership rates have declined is that “we have never fully enforced the Fair Housing Act.”
In June, the HUD announced its intention to reinstate the rules put in place by the Obama administration in 2013 to tackle âdisparate impactsâ – unintentional, but nonetheless unjustified, discriminatory practices. This could create legal liability for lenders who use artificial intelligence and algorithms to assess borrowers.
Fannie and Freddie’s federal regulator, the Federal Housing Finance Agency (FHFA), wants mortgage giants to meet new goals that support lending in minority census tracts. The FHFA, which wants at least 35% of the purchase mortgages guaranteed by Fannie and Freddie to be taken out by low- and very-low-income borrowers, is seeking comments on the proposed targets for 2022 to 2024.
Email Matt Carter