Home / Banking Strategies / Mortgage and auto lenders grapple with disparate impact

Mortgage and auto lenders grapple with disparate impact


Mortgage and auto loan providers are caught between a rock and a hard place after a ruling by the U.S. Supreme Court upheld the theory of disparate impact, which says that lenders can be found liable for discrimination against a protected class even if said discrimination was unintended. The ambiguity of the theory, and the lack of clear regulatory guidelines, leaves them grappling for solutions.

“Lenders are in a crazy type of limbo,” says attorney Blair B. Evans, who chairs the auto finance team at Baker, Donelson in Memphis. “No one really understands what’s expected of them.”

Much of the ambiguity stems from the fact that, in many situations, federal law prohibits lenders from asking questions regarding an applicant’s age, race, color, gender and religion, even to determine if the individual is in a protected class. So, how can lenders tell if their practices are discriminatory if they can’t identify members of a protected class? Further, federal regulators haven’t provided clear guidelines for auto and mortgage lenders on what could be considered a disparate impact violation.

In the absence of such guidelines, individual lenders are taking a shot at developing their own standards which they can only hope follows the spirit of the law, Evans says. Doing that is easier for large institutions than it is for small- or medium-sized lenders because of the cost and resources needed to develop standards. However, the duplication of effort has become expensive for the industry, costing hundreds of thousands of dollars, she adds.

“It’s not like a lender can turn to an expert for help in establishing disparate impact standards,” Evans says. “It’s all a judgment call because few people out there have any experience with this.”

And the penalties for miscalculating can add up if a lender is audited by a federal regulator, she says. The Consumer Financial Protection Bureau (CFPB) estimates the cost to lenders for a compliance exam at about $28,000. However, industry trade groups have put the cost at more like $75,000 to $100,000, according to Evans, noting that these costs do not include any penalty that may be incurred.

“The cost of that kind of audit could be fatal to a small bank,” she says.

Playing Defense

One unintended consequence of the Supreme Court ruling may be to encourage financial institutions to be more cautious in their lending practices, which could actually result in fewer loans to protected classes, says compliance expert Jo Ann Barefoot, CEO of the Jo Ann Barefoot Group in Boston. “Banks may choose not to go into certain lending spaces rather than open a hornet’s nest related to disparate impact,” Barefoot says. “That is the opposite of what regulators were going for to begin with.”

Without a lot of regulatory guidance, lenders have to get good at thinking through these issues for themselves, she says. For example, lenders need to know that they can defend their practices in the case of an audit, which means they need to have a good business case for their lending practices and have carefully thought out the reasons for what they are doing.

“Lenders need to be able to show they’ve done their due diligence and thought through the implications of what they do in case they get called on the carpet by regulators,” Barefoot says. “Well thought out and reasoned documentation about lending practices is what will protect financial institutions in an audit.”

Robert Rowe, vice president and associate chief counsel, Regulatory Compliance, at the American Banking Association (ABA), says financial services companies are up to the challenge. “Our industry has always made every effort to provide fair lending practices,” Rowe says. “This is not anything lenders aren’t already committed to.”

Rowe says that the ABA supports most of the findings of the Supreme Court because the court is, in effect, mandating further clarification of disparate impact regulation, something that the association is also working with federal agencies to accomplish. Disparate impact compliance related to mortgage lending is the jurisdiction of regulators that enforce the Fair Housing Act, such as the U.S. Department of Justice and the U.S. Department of Housing and Urban Development, he adds.

The CFPB is not involved with housing compliance. However, as part of its enforcement authority over the Equal Credit Opportunity Act, the agency is attempting to eliminate discrimination in auto lending by targeting the auto finance industry. “Clarity is the issue,” Barefoot says. “The more regulators can fine tune disparate impact regulations, the better off the industry will be.”

Right now, one of the biggest challenges for lenders is determining which individuals are in a protected class without direct information from the applicant. Some banks are looking to use the massive amounts of data collected today from various sources, including the U.S. Census and Internet search providers, according to Barefoot. These lenders are using this information to glean information on who belongs in a protected class in order to ensure their lending practices don’t discriminate against them, she says.

Such searches might include looking beyond traditional criteria like credit scores, loan ratios, and how long someone has lived in a home, Barefoot adds, noting that some protected classes may have the ability to pay back loans but not the kind of clean credit histories that create high scores. For example, a person may have had a medical issue that has since been resolved, but their credit score remains low. Information from credit scoring companies doesn’t show that level of detail, she says.

Data analysis is so advanced that providers of this information say that they can determine levels of loan risk based on non-traditional methods, Barefoot adds, pointing to studies showing that a person paying for gas inside a station may be a riskier borrower than someone who pays at the pump. What college an individual went to and if they even went to college is also considered an indicator of loan risk. And although medical records are private, health issues might be extrapolated from Internet searches, Barefoot says, adding that banks need to make their own decisions as to whether this kind of data can lead to making fairer and less risky loans.

“Big data is a new frontier,” Barefoot says. “Huge amounts of data are collected but the accuracy of this as a predictive indicator hasn’t really been tested.”

Ms. Wolf is a contributing writer to BAI Banking Strategies based in Pittsboro, N.C.