Ensuring transparency and trust in home valuation

A new era of scrutiny for AI-powered valuation tools began on July 17, 2024, when six federal regulatory agencies (the Consumer Financial Protection Bureau, Federal Deposit Insurance Corp., Federal Housing Finance Agency, Federal Reserve Board, National Credit Union Administration, and the Office of the Comptroller of the Currency) issued the Interagency Automated Valuation Models (AVM) Final Rule. Pursuant to the Dodd-Frank Wall Street Reform and Consumer Protection Act, this rule implements mandatory quality control standards for AVMs used in credit decisions for a consumer’s principal dwelling, directly impacting how financial institutions and their partners develop, test, and deploy these models.

Effective October 1, 2025, the rule requires financial institutions to adopt and maintain policies, practices, procedures, and control systems to ensure AVMs adhere to quality control standards designed to meet five key objectives:

1. Ensure a high level of confidence in the estimates produced.
2. Protect against the manipulation of data.
3. Seek to avoid conflicts of interest.
4. Require random sample testing and reviews.
5. Comply with applicable nondiscrimination laws.

This federal mandate serves as a clear signal: the era of "black box" algorithms in essential consumer finance is ending. It compels the industry to prioritize model validity and governance and necessitates a commitment to transparency and robust model governance. Since housing is the largest asset class, it needs to be valued accurately. By mandating rigorous quality control standards, the U.S. government is effectively setting a foundation to underwrite the validity of the AVMs. This regulatory framework is crucial for enabling the government to then use these models, which have a high level of confidence in their estimates, to craft policy.

Furthermore, the requirement to comply with applicable nondiscrimination laws directly addresses the challenge of algorithmic bias in artificial intelligence (AI). The federal ruling's explicit focus on non-discrimination requires companies to employ sophisticated fair lending counsel and Model Governance review processes to audit models for any prohibited bias input variables. This extends beyond obvious factors like race or religion to include variables like income, credit, or public assistance status if they are shown to have a disparate impact.

‍

Cotality’s approach to validation and confidence

Accurate and reliable valuations are the cornerstone of successful risk management and regulatory compliance. Cotality's robust testing and ongoing monitoring of its AVM products, specifically the Total Home Valueˣ (THVx) estimates, are designed to not only meet but exceed the mandates of the new AVM Final Rule, with a focus on delivering high-confidence estimates and unparalleled model integrity; including:

• ‍Blind testing: A key differentiator in ensuring confidence is our methodology for "blind" testing. Cotality's testing is designed to keep the AVM completely unaware of the actual sales activity on the subject property during the valuation process.
• Scores and metrics: To ensure clients have a high level of confidence in the estimates produced, Cotality provides two crucial, quantified confidence scores: Confidence Score and Forecast Standard Deviation (FSD). Clients can also configure customized confidence score thresholds to ensure only values meeting their specific regulatory standards are delivered.

‍

Meeting the mandate for random sample testing

The Final Rule explicitly mandates financial institutions to require random sample testing and reviews. Cotality supports clients in meeting this objective in three comprehensive ways, backed by massive datasets:

1. Internal blind testing results: We provide clients with our internal blind testing results on a quarterly basis, allowing institutions to integrate our proven performance data into their own compliance reviews.‍
2. Quarterly due diligence reports: These reports deliver thorough, quantifiable metrics essential for robust monitoring. By consolidating sales data nationwide, Cotality achieves a sample size between 500,000 and 1 million records each quarter. The metrics reported cover essential performance indicators across various geographies (National, State, County) and property types/price tiers‍
3. Client-submitted file testing: At no charge and on a predetermined schedule, clients are invited to submit test files based on their internal sampling.

Furthermore, Cotality collaborates with independent third-party testers whom clients may engage for professional, unbiased testing and monitoring services, adding an additional layer of external accountability and confidence.

‍

Don’t navigate the new regulatory landscape alone

The 2024 Federal AVM Ruling marks a pivotal shift, elevating AI governance and model integrity from best practice to a regulatory mandate across the housing industry. At Cotality, this change reaffirms the necessity of our core methodologies, especially blind, out-of-sample testing, ensuring the valuations underpinning America's largest asset class are accurate, reliable, and fully compliant with non-discrimination laws. Using our network, scale, and technology, we deliver faster, smarter, more human-centered experiences to strengthen your business; fostering true trust in the technology we build. Don't navigate this new landscape alone, learn more about Total Home Valueˣ (THVx).