MReport April 2021

TheMReport — News and strategies for the evolving mortgage marketplace.

Issue link:

Contents of this Issue


Page 59 of 67

58 | M REPORT O R I G I NAT I O N S E R V I C I N G DATA G O V E R N M E N T S E C O N DA R Y M A R K E T THE LATEST DATA How Automated Home Appraisal Errors Can Affect Black Homeowners Researchers recently conducted a study in an effort to understand how AVMs are working (or not) for various racial groups. A ccurate home apprais- als are key not only to attaining mortgage financing but also to establishing the level of equity in a homeowner's property. Having previously shown that in-person appraisals are subject to bias and discrimination, research associates at the Urban Institute recently conducted a study in an effort to understand how auto- mated home appraisals—which have gained in use with the COVID-19 pandemic—are work- ing for various racial groups. They say the data when analyzed shows that automated home valuation technology errors disproportionately affect residents of majority-Black neighborhoods. "Automated valuation models (AVMs) are a type of financial technology that can be an alterna- tive or supplement to in-person appraisals," explain UI fellows Michael Neal, Linna Zhu, and Caitlin Young, who authored the report. "Many lenders use AVMs as part of their lending decision process. AVMs apply mathemati- cal algorithms to a database of housing activity, including sales transactions, to calculate a home's value. Hypothetically, by limiting the human assessment element and potential racial bias, AVMs could provide a more accurate picture of a home's value in ma- jority-Black neighborhoods, which could hold promise for closing the racial wealth gap." Digging into data from Atlanta, Memphis, and Washington, D.C., the authors set out to determine if and how AVMs contribute to racial disparities in home value estimates. The researchers studied the accuracy of AVMs in major- ity-Black neighborhoods relative to majority-white neighborhoods by comparing the AVM estimate with the actual sales price. The researchers initially found the magnitude of AVM error was larger in majority-white neighbor- hoods, however, upon adjust- ing for the sales price, which is lower in majority-Black neighbor- hoods, they discovered that the percentage magnitude of error was greater in majority-Black neighborhoods. "That means the absolute dollar-value error might be less in these neighborhoods, but the pro- portional size of the error is larger because the average home price in majority-Black neighborhoods is much lower," the authors noted. "For example, the percentage dif- ference between a $100,000 AVM estimate and a $115,000 sales price (15%) is higher than the percent- age difference between a $400,000 AVM estimate and a $415,000 sales price (4%), even though the absolute difference ($15,000) is the same."

Articles in this issue

Archives of this issue

view archives of TheMReport - MReport April 2021