WASHINGTON – An automatic appraisal sounds fair, as if taking the human element out of lending decisions can erase bias by lenders who may not even know they’re making biased decisions. However, the U.S. Consumer Financial Protection Bureau (CFPB) says it doesn’t work that way.

CFPB announced an initiative to make sure computer models that help determine home valuations are fair, the Small Business Advisory Review Panel for Automated Valuation Model (AVM) Rulemaking. The options will now be reviewed to determine their potential impact on small businesses.

“It is tempting to think that machines crunching numbers can take bias out of the equation, but they can’t,” says CFPB Director Rohit Chopra. “This initiative is one of many steps we are taking to ensure that in-person and algorithmic appraisals are fairer and more accurate.”

An algorithm is a complicated math formula loaded into a computer. It can take a wide range of data, weigh the importance of each item and generate a decision based on the criteria programmed within the algorithm.

When underwriting a mortgage, lenders typically require an in-person appraisal – an estimate of the home’s value – but many now also use algorithmic computer models. The technical term for them is automated valuation models (AVM).

However, both in-person and algorithmic appraisals appear susceptible to bias and inaccuracy, absent appropriate safeguards, CFPB says. Given automated valuation models’ crucial role, the Dodd-Frank Wall Street Reform and Consumer Protection Act required CFPB and other regulators to implement rules on the models because both too-high valuations and too-low valuations cause problems.

Overvaluing homes can lead to higher rates of foreclosure. During the run-up to the 2008 collapse, for example, housing prices became inflated, and one reason for that was lenders extending mortgage credit – often with toxic or predatory terms – without regard to borrowers’ ability to repay because overvaluations suggested that lenders’ risk was lower than believed.

Low valuations can jeopardize sales and prevent homeowners from refinancing. In some cases, systematically low valuations can exacerbate and extend already existing disparities in the housing market, CFPB says. The Federal Housing Finance Agency recently identified discriminatory statements in some home appraisals, and both Fannie Mae and Freddie Mac have found appraisal disparities for communities and borrowers of color.

CFPB says it’s “particularly concerned that without proper safeguards, flawed versions of these models could digitally redline certain neighborhoods and further embed and perpetuate historical lending, wealth and home value disparities.”

CFPB says its goal, along with federal partners, is to:

• Ensure a high level of confidence in the estimates produced by automated valuation models
• Protect against the manipulation of data
• Seek to avoid conflicts of interest
• Require random sample testing and reviews
• Account for any other such factor that the agencies determine to be appropriate

© 2022 Florida Realtors®