Financial regulators (e.g., OCC, Federal Reserve, NCUA) expect institutions to perform independent well-supported validation of AVMs, consistent with the Interagency Appraisal and Evaluation Guidelines (2010-42) and the Interagency AVM Quality Control Standards (2025). AVMetrics provides independent testing and supporting due diligence designed to support a clear, consistent, and defensible validation process.
Over approximately 50 pages, the report details our methodology, including test data parameters, sample size, coverage, and data sources. Our process incorporates multiple data sources and identifies about 500,000 candidate transactions per quarter in the U.S. The report presents our analytical approach, including accuracy, precision, and dispersion metrics used to evaluate AVM performance, as well as the scoring methodologies used to assess and rank models.
As part of the analysis, we map supervisory expectations to our testing framework, demonstrating how our approach supports institutions in selecting, validating, and monitoring AVMs, including the following:
- The need for appropriate expertise, including the use of independent third-party validation
- Due diligence considerations related to model design and development
- Establishment and application of minimum performance criteria
- Evaluation of model strengths and limitations across design and observed performance
- Assessment of the model-provided confidence and scoring outputs as indicators of reliability across property types and geographies
- Structured model testing, including sample sizes, geographic analyses, testing frequency, and performance measurement