How Fair Lending Examinations Work
Federal regulators conduct fair lending examinations of mortgage lenders through a structured analytical process. The examination typically begins with a review of the lender’s HMDA data to identify statistical disparities in denial rates, pricing, or geographic lending patterns between demographic groups. If HMDA data reveals disparities, the examiner conducts a more detailed analysis that may include comparative file reviews (matching applicants from different demographic groups with similar financial profiles and comparing their outcomes), policy and procedure reviews (evaluating whether lending policies or underwriting overlays have potential disparate impact), and interviews with lending personnel.
The comparative file review is a central component of fair lending examinations. Examiners identify pairs or groups of applicants from different racial or demographic backgrounds who have similar credit scores, income levels, LTV ratios, and other underwriting characteristics. If the minority applicant was denied while the similarly situated non-minority applicant was approved, or if the minority applicant received a higher interest rate, the disparity must be explained by legitimate, non-discriminatory underwriting factors that differ between the files. Unexplained disparities across multiple file comparisons may support a finding of discrimination.
If a fair lending examination identifies violations, the regulator may pursue formal enforcement action, enter into a consent order requiring corrective measures, refer the case to the DOJ for litigation, or require the lender to provide restitution to affected borrowers. Examination findings are confidential, but enforcement actions and consent orders are public.
How Adverse Action Notices Are Generated
When a mortgage application is denied, the lender’s underwriting system or underwriter identifies the specific reasons for denial. These reasons are drawn from a standardized list of denial codes (such as those published by the FFIEC) and must correspond to the actual factors that caused the denial. The lender selects up to four principal reasons, ranked by significance, and includes them in the adverse action notice sent to the applicant.
The notice must be delivered within 30 days of the denial decision. It includes the applicant’s name, the date of the denial, the specific reasons (e.g., “Debt-to-income ratio too high,” “Insufficient length of employment,” “Derogatory credit history”), the name and contact information of the lender, information about the applicant’s rights under ECOA and the Fair Housing Act, and the name of the credit reporting agency that provided the credit report. If the denial was based in part on information from a credit report, the applicant is entitled to a free copy of the report within 60 days.
Automated underwriting system denials (such as a DU or LPA “refer” finding) must still be supported by specific adverse action reasons. The AUS may generate reason codes, but the lender must ensure those codes accurately reflect the factors driving the denial and are not generic placeholders. Inconsistent or vague adverse action reasons are a common finding in fair lending examinations and can suggest that the lender is not adequately documenting the bases for its decisions.
How HMDA Data Is Used for Fair Lending Analysis
HMDA data is submitted annually by covered lenders to their federal regulator and is made publicly available through the FFIEC and CFPB. Analysts use HMDA data to calculate denial rate disparities (comparing denial rates for minority applicants vs. non-minority applicants), pricing disparities (comparing average APR or rate spread for originated loans across demographic groups), and geographic distribution analysis (comparing the percentage of lending activity in majority-minority census tracts versus the overall market).
HMDA data includes robust geographic detail (census tract level), which enables analysis of whether lenders are adequately serving all communities within their assessment areas. A lender whose HMDA data shows very few applications or originations in majority-minority census tracts compared to peer lenders may be flagged for a potential redlining examination. Similarly, a lender whose HMDA data shows a statistically significant denial rate disparity between Black and white applicants, after controlling for available variables, may be flagged for a disparate treatment examination.
It is important to note that HMDA data does not include several key underwriting variables (credit score, LTV ratio, reserves, property type details). This limitation means that raw HMDA disparities do not prove discrimination; they identify patterns that warrant further investigation using complete loan file data. Regulators use HMDA as a screening and prioritization tool, not as a standalone basis for enforcement action .
Related topics include qualified mortgage (qm) rules explained, respa explained: real estate settlement procedures act, role of fannie mae and freddie mac in mortgage lending, fha program structure and guidelines overview, va loan entitlement and eligibility framework, and mortgage regulations: a borrower’s guide.