[co-author: Matthew Tikhonovsky]
- On June 24, 2024, six federal agencies finalized a rule to create safeguards for automated valuation models (AVMs) in the real estate industry.
- AVMs are an increasingly popular tool, often powered by AI, used by mortgage originators to estimate property values for mortgages and lending services.
- The new rule requires companies that utilize AVMS to implement five quality control factors to ensure that their models uphold data accuracy and security and avoid discriminatory impacts.
- The rule will go into effect a year after its publication in the Federal Register.
On June 24, 2024, six federal agencies — the Department of the Treasury (Treasury), the Federal Reserve System (FRS), the Federal Deposit Insurance Corporation (FDIC), the National Credit Union Administration (NCUA), the Consumer Protection Financial Bureau (CPFB), and the Federal Housing Finance Agency (FHFA) — finalized a rule on Quality Control Standards for Automated Valuation Models (AVMs), an increasingly popular tool, often powered by AI, used for real estate valuations and mortgage lending processes.
The finalized rule creates five quality control factors for companies that utilize AVMs to follow to ensure that their models — and the property valuations and credit decisions that they serve as the basis of — are neither inaccurate nor discriminatory.
Below we provide an overview of AVMs and the proposed rule to regulate AVMs, before turning to final AVM rule.
The Boom in AVMs
In recent years, AI has impacted nearly every industry, and the real estate industry is no exception. AVMs, which often rely on artificial intelligence and machine learning, estimate a property’s value for mortgage and lending purposes. Widely considered to be more reliable than human appraisals, AVM’s valuations and predictions often serve as the basis for credit decisions and securitization determinations, providing access to credit for first-time homeowners and businesses alike.
The growing use of AVMs has been accelerated by advancements in AI technologies, coupled with the COVID-19 pandemic and the resulting shortage of mortgage appraisers, according to a report by the Brookings Institution. These models are now widely used in the real estate industry, including by the Federal National Mortgage Association (Fannie Mae) and the Federal Home Loan Mortgage Corporation (Freddie Mac), because they speed up and reduce costs in the home valuation and mortgage lending processes., Nonetheless, concerns exist about the potential discriminatory impact of AVMs on housing valuations and credit decisions. The Fair Housing Act (FHA) prohibits discrimination in the purchasing and renting of homes, which covers mortgage lenders that utilize AVMs. Research has detailed that AVMs may perpetuate the same biases against certain protected classes of homeowners that often exist in human-performed appraisals because AVMs ultimately rely on human-created data that may contain the same biases, as well as historical data that may reflect past discrimination and disparities in housing. While AVMs may not completely eliminate biases in property valuations, they have still been shown to reduce biases in property valuations relative to human appraisals.
Proposed Rule
On June 1, 2023, the six federal agencies issued a proposed rule on quality control standards for AVMs. The proposed rule was promulgated in response to Section 1473(q) of the Dodd-Frank Act of 2010, which added a new section to the Financial Institutions Reform, Recovery, and Enforcement Act of 1989 and directed the six agencies to release guidance requiring companies that use AVMs to comply with four quality control factors. In addition to the four quality control factors outlined in the Dodd-Frank Act, which focused on data accuracy and security, the June 2023 proposed rule also introduced a fifth quality control factor, requiring AVMs to comply with applicable nondiscrimination laws.
The proposed rule foreshadowed the Biden administration’s October 2023 Executive Order (EO) on AI, which “encouraged” the directors of the FHFA and CFPB to “consider using their authorities” to “evaluate automated collateral-valuation and appraisal processes in ways that minimize bias,” with the goal of “[addressing] discrimination and biases against protected groups in housing markets and consumer financial markets.”
Finalized Rule
With relatively few changes to the proposed rule, the final rule applies to mortgage originators and secondary market issuers that use AVMs for “certain credit decisions or securitization determination.” The rule broadly defines AVMs to include “any computerized model used by mortgage originators and secondary market issuers to determine the value of a consumer’s principal dwelling collateralizing a mortgage.” The rule requires mortgage originators and secondary market issuers to “adopt policies, practices, procedures, and control systems to ensure” that their use of AVMs meets the following five quality control factors:
- Ensure a high level of confidence in the estimates produced;
- Protect against the manipulation of data;
- Seek to avoid conflicts of interest;
- Require random sample testing and reviews; and
- Comply with applicable nondiscrimination laws.
Importantly, the rule does not set specific “policies, practices, procedures, or control systems” that companies that use AVMs must put in place to ensure compliance with the rule. Rather, the rule leaves it to these companies themselves to develop their own policies and practices to ensure compliance with the rule’s five quality control factors. As modeling technology and AI continues to evolve, “this flexible approach will allow institutions to refine their implementation of the rule as appropriate,” according to the rule. Furthermore, the rule provides mortgage originators and secondary market issuers “the flexibility to set their quality control standards for covered AVMs as appropriate based on the size, complexity, and risk profile of their institution.”
The rule will go into effect a year after its publication in the Federal Register.
[View source.]