Home Mortgage Disclosure Act (HMDA)

The Home Mortgage Disclosure Act is the reason we know whether banks are lending fairly in your neighborhood. Enacted in 1975 (12 U.S.C. §§ 2801–2810) and implemented by the CFPB's Regulation C, HMDA requires most mortgage lenders to collect and publicly disclose detailed data about every home loan application they receive — including the applicant's race, ethnicity, sex, income, the loan amount, the property location, and whether the application was approved or denied. This data is the foundation of fair lending enforcement in America. When researchers discover that Black applicants are denied mortgages at twice the rate of white applicants with similar credit profiles, they're using HMDA data. When regulators identify a bank that's systematically avoiding lending in minority neighborhoods, they're using HMDA data.

Current Law (2026)

Parameter	Value
Governing law	12 U.S.C. §§ 2801–2810; Regulation C (12 CFR Part 1003)
Regulator	Consumer Financial Protection Bureau (CFPB)
Who must report	Banks, credit unions, and mortgage companies meeting volume thresholds
Reporting threshold	Originated 25+ closed-end mortgage loans or 200+ open-end lines of credit in each of the 2 preceding years
Data points collected	~100+ data fields per application (expanded under 2015 rule)
Key data fields	Race, ethnicity, sex, income, loan amount, interest rate, property type, census tract, action taken
Public disclosure	Annual — data available via CFPB's online HMDA data tools
Enforcement	CFPB, Federal Reserve, OCC, FDIC, NCUA, HUD, DOJ
Penalty for noncompliance	Civil money penalties up to $1M+/day under applicable banking laws

Legal Authority

12 U.S.C. § 2801 — Congressional findings and declaration of purpose (Congress found that some depository institutions failed to provide adequate home financing on reasonable terms, contributing to income disparities and urban decline)
12 U.S.C. § 2802 — Definitions (defines "mortgage loan," "depository institution," and key terms; includes banks, savings associations, credit unions, and other mortgage lending institutions)
12 U.S.C. § 2803 — Maintenance of records and public disclosure (requires covered institutions to compile and make available to the public, by March 31 of each year, data about each mortgage application — including loan type, purpose, amount, action taken, location, and applicant demographics)
12 U.S.C. § 2804 — Enforcement (CFPB prescribes rules; enforcement delegated to appropriate federal banking agencies for their regulated institutions)
12 U.S.C. § 2805 — Relation to state laws (HMDA does not preempt state mortgage disclosure laws that are consistent with federal requirements)
12 U.S.C. § 2806 — Compliance improvement (authorizes CFPB to provide compliance assistance)
12 U.S.C. § 2809 — Institutional reporting requirements (requires reporting by census tract and county to show geographic distribution of lending activity)
12 U.S.C. § 2810 — Disclosure by the CFPB (CFPB must make aggregate data available by census tract, enabling community-level analysis of lending patterns)

How It Works

Every covered lender must report detailed information about each home loan application it receives during the calendar year. The 2015 HMDA rule (effective 2018) dramatically expanded the required data points to over 100 fields per application under Regulation C (12 CFR Part 1003), including loan amount, interest rate, loan term, property value, combined loan-to-value ratio, debt-to-income ratio, credit score, automated underwriting results, applicant demographics (race, ethnicity, sex, age), and the reason for denial if applicable — an expansion designed to let regulators and the public detect discriminatory lending patterns that older, simpler data couldn't reveal. HMDA applies to depository institutions (banks, savings associations, credit unions) and non-depository mortgage lenders that meet minimum origination thresholds: currently 25 closed-end mortgage loans or 200 open-end lines of credit originated in each of the two preceding calendar years, with smaller lenders exempt. The data serves three purposes: it's the backbone of fair lending enforcement (regulators use it to identify lenders whose approval rates or lending patterns suggest racial or ethnic discrimination); it helps communities assess whether local lenders are serving local housing credit needs; and it supports public research by academics, journalists, and advocacy organizations studying lending disparities and housing market trends.

The CFPB publishes HMDA data annually through online tools that allow anyone to search, filter, and download loan-level records under 12 U.S.C. § 2810 — individual borrower names are not published, but the combination of census tract, loan amount, and demographic data can be highly granular, with privacy protections through data binning for certain sensitive fields. Violations of HMDA reporting requirements can result in civil money penalties and supervisory actions. More significantly, HMDA data frequently serves as the statistical foundation for fair lending enforcement actions under the Equal Credit Opportunity Act and Fair Housing Act, enforced by the CFPB — meaning inaccurate HMDA reporting can both trigger its own penalties and mask violations of other civil rights laws. See also Fair Credit Reporting Act for related consumer data protections.

How It Affects You

If you're a homebuyer researching lenders: Before you apply, use HMDA data to check whether your shortlisted lenders actually lend in your neighborhood and to people with your demographic profile. The CFPB's HMDA Data Browser at ffiec.cfpb.gov/data-browser lets you filter by lender, geography (state, county, or census tract), loan type, and applicant demographics — all for free. Look at approval rates, denial rates, and median interest rates for conventional purchase loans by lender in your county. A lender approving 85% of applications from white applicants and 55% from Black applicants in the same county is a statistical red flag. HMDA data doesn't prove discrimination by itself — credit scores, DTI ratios, and other factors affect approval — but it tells you whether a lender's track record warrants a closer look before you commit.

If you're denied: under the Equal Credit Opportunity Act (ECOA), you have the right to a written explanation of the reasons for denial within 30 days. HMDA data shows your lender's overall pattern; your denial notice tells you the specific reason. If the explanation doesn't match your financial profile — or if you received worse terms than similarly situated borrowers — file a complaint with the CFPB at consumerfinance.gov/complaint or with the HUD Fair Housing hotline at 1-800-669-9777. Both agencies use HMDA data as statistical context when investigating individual complaints. The intersection of your denial and the lender's aggregate HMDA pattern is the evidence regulators start with.

If you're a community organization monitoring fair lending: HMDA data is the foundation of every serious fair lending advocacy campaign. The CFPB HMDA Explore tool at ffiec.cfpb.gov/explore.html provides aggregate statistics by lender, census tract, and demographic group without requiring data download or technical expertise. For deeper analysis, download the HMDA LAR (Loan Application Register) data for specific lenders or geographies and run your own comparisons. The 2018 data expansion added credit scores, DTI ratios, interest rates, loan-to-value ratios, and automated underwriting results — which means you can now control for creditworthiness when identifying disparities, not just compare raw approval rates.

Effective advocacy strategies: (1) compare a lender's denial rates for minority vs. white applicants in the same census tract for the same loan type; (2) compare median interest rates offered by the same lender to minority vs. white borrowers with similar income and loan amounts; (3) map lending volume by census tract to identify geographic redlining (low volume in high-minority areas despite apparent demand). With this data, you can file complaints with your state attorney general (California, New York, Illinois, and Massachusetts have active state-level fair lending enforcement programs that have stepped up as the CFPB pulled back) or petition the OCC, FDIC, or Federal Reserve to take enforcement action against a specific lender during their CRA examination cycle. Submitting HMDA-based evidence during a bank's Community Reinvestment Act public comment period is one of the most direct levers available.

If you're a mortgage lender managing HMDA compliance: HMDA reporting is mandatory if you originated 25+ closed-end mortgage loans (or 200+ open-end lines of credit, like HELOCs) in each of the two preceding calendar years. If you're near the threshold, check annually before assuming you're exempt — crossing the threshold mid-year still requires full-year reporting. Coverage also extends to branches and subsidiaries; assess your full enterprise footprint.

The annual LAR submission deadline is March 31 for the preceding calendar year's data. The CFPB's HMDA Filing Platform at ffiec.cfpb.gov/filing accepts electronic submissions and provides validation checks — run your data through the platform's edit checks before your final submission. Common error categories: missing or invalid census tract codes (use the FFIEC geocoder tool); stale or incorrect denial reason codes; inconsistent ethnicity and race collection (you must ask and record in separate fields as of 2018); incorrect loan purpose codes for refinances vs. purchase.

Accuracy matters beyond compliance penalties. Systemic HMDA data errors are a primary trigger for ECOA/FHA fair lending examinations — regulators who spot anomalies (unusual denial rate patterns, unexplained demographic disparities) in your HMDA data open supervisory examinations that go far beyond the data quality question. Build an internal HMDA audit process: before submission, run statistical comparisons of your own denial rates across demographic groups for each loan product and geography, flagged against industry benchmarks. If your data shows a 2× or greater denial disparity for minority applicants relative to white applicants in the same product/geography/income tier — even after adjusting for credit factors — that's what an examiner will find, and you need a documented business justification ready. Contact your primary federal regulator (OCC, FDIC, Federal Reserve, or CFPB depending on charter) for pre-examination HMDA compliance guidance.

If you're a fair lending researcher or journalist: HMDA's 2018 data expansion was a step-change for research. You now have, for each application: credit score range, debt-to-income ratio, loan-to-value ratio, automated underwriting result, interest rate, loan term, negative amortization flag, balloon payment flag, interest-only flag, prepayment penalty flag, and more. This allows statistical controls for creditworthiness when analyzing racial disparities — separating the discrimination question (same-credit-profile borrowers treated differently) from the wealth-gap question (demographic groups have different average credit profiles). The persistent finding: even after controlling for credit score, income, DTI, and LTV, Black and Hispanic applicants face denial rates and interest rate premiums that can't be fully explained by financial risk factors alone. The 2023 data shows Black applicants denied conventional purchase mortgages at approximately 2× the rate of white applicants at comparable income levels.

Key analysis resources: the CFPB Research and Reports page at consumerfinance.gov/data-research publishes annual HMDA data summaries. The Urban Institute Housing Finance Policy Center (urban.org/housing) publishes quarterly HMDA analyses. The National Community Reinvestment Coalition (ncrc.org) maintains a HMDA analysis tool and publishes fair lending maps. For AI underwriting disparate impact: use HMDA's automated underwriting result field (APPROVE/REFER from DU/LP) combined with denial rate data to identify lenders where algorithmic referrals convert to denials at different rates by demographic — that's the disparate impact evidence pattern under ECOA's Regulation B. The Trump CFPB's reduced enforcement posture (2025-present) has shifted the fair lending enforcement burden to state AGs and private litigation; tracking state enforcement actions gives you the current enforcement landscape more accurately than federal enforcement statistics alone.

State Variations

HMDA does not preempt state mortgage disclosure laws, and several states have enacted additional requirements:

Some states require additional data collection beyond federal HMDA requirements
State fair lending laws may use HMDA data as evidence in enforcement actions
State Community Reinvestment Act equivalents may incorporate HMDA data into assessments
State attorneys general frequently use HMDA data in fair lending investigations

Implementing Regulations

12 CFR Part 1003 — CFPB Home Mortgage Disclosure (Regulation C): the CFPB's implementing regulation for HMDA, prescribing who must report, what data must be collected, how data is submitted, and what public disclosures are required. The 2015 HMDA rule (effective January 1, 2018) dramatically expanded both the universe of covered institutions and the number of required data fields. Key provisions:
- § 1003.2 — Definitions: "application" includes both written applications and oral requests for covered loans; preapproval requests are applications if the institution's approval program meets specific criteria (written commitment, limited conditions); "covered loan" includes closed-end mortgage loans and open-end lines of credit secured by a dwelling (including HELOCs)
- § 1003.3 — Exempt institutions and excluded transactions: an institution is exempt if it originated fewer than 25 closed-end mortgages (or fewer than 200 open-end lines of credit) in each of the two preceding calendar years; smaller community banks, credit unions, and non-bank lenders below these thresholds do not report; the partial exemption for smaller reporters (fewer than 500 closed-end or 500 open-end covered loans) allows simplified reporting on some data fields
- § 1003.4 — Compilation of reportable data: covered institutions must collect 48 data fields per loan/application for closed-end mortgages (with additional fields for open-end credit); expanded fields from the 2015 rule include: interest rate, loan term, origination charges, points, discount points, lender credits, property value, combined loan-to-value ratio, debt-to-income ratio, credit score model and value, automated underwriting system and result, and submission channel; the applicant's race, ethnicity, sex, and age remain required; denial reasons are required for denied applications
- § 1003.5 — Disclosure and reporting: financial institutions must submit their annual loan/application register (LAR) in electronic format by March 1 following the calendar year; submission is to the institution's primary federal regulator (CFPB for large non-banks, OCC for national banks, FDIC for state non-members, Federal Reserve for state members, NCUA for credit unions); the CFPB makes modified LARs (with personally identifiable information removed) publicly available through the FFIEC HMDA Platform
- § 1003.6 — Enforcement: HMDA violations subject institutions to civil money penalties under HMDA § 2804; the "bona fide errors" defense allows institutions to avoid penalties for unintentional data errors if they have reasonable procedures to avoid errors and correct them promptly after discovery
Regulation C's 2015 expansion (effective 2018) transformed HMDA from a relatively simple geographic lending disclosure into a comprehensive loan-level database covering pricing, underwriting, and demographic data. This expansion enables regulators and researchers to identify redlining (geographic discrimination), disparate pricing (charging higher rates to protected-class borrowers), and discriminatory underwriting (denying qualified protected-class applicants). CFPB and DOJ use HMDA data as the statistical foundation for fair lending investigations under the Equal Credit Opportunity Act and Fair Housing Act. Public HMDA data is freely accessible at ffiec.cfpb.gov/hmda and used by journalists, academics, advocacy groups, and state regulators to track lending patterns by race, income, and neighborhood. Recent rulemakings: 88 FR 7464 (2023) — CFPB adjusted the HMDA partial exemption thresholds; see the CFPB HMDA resources for current reporting thresholds as they are adjusted periodically.
12 CFR Part 226 — Truth in Lending (legacy Regulation Z) (HMDA cross-references for mortgage lending disclosures)

Pending Legislation

HR 166 (Rep. Green, D-TX) — Fair Lending for All Act: add sexual orientation, gender identity, and ZIP/census tract to ECOA protections; create CFPB fair-lending testing office; expand HMDA data collection. Status: Introduced.

Recent Developments

The 2015 HMDA rule's expanded data collection (effective 2018) was the most significant change to mortgage data reporting in decades, adding interest rates, credit scores, loan terms, and dozens of other fields. Subsequent regulatory actions partially rolled back some requirements — raising the reporting threshold from 25 to 100 closed-end loans in 2020 (later partially reversed) and exempting some data points for smaller lenders. The CFPB's annual HMDA data releases continue to reveal persistent disparities in mortgage lending by race and ethnicity, fueling ongoing policy debate about fair lending enforcement, algorithmic bias in automated underwriting, and the adequacy of existing anti-discrimination tools.

CFPB HMDA enforcement paused under Trump (2025): The Trump CFPB reduced HMDA supervisory examinations and fair lending enforcement actions. The CFPB's Office of Fair Lending and Equal Opportunity — which uses HMDA data to identify discriminatory lending patterns — has seen reduced staffing and lower priority under Acting Director Vought and subsequent leadership. State attorneys general (particularly California, New York, Massachusetts) have increased state-level fair lending enforcement using HMDA data to fill the federal gap.
HMDA data and AI mortgage underwriting: HMDA data reveals that algorithmic underwriting systems, while facially race-neutral, produce denial rate disparities that closely track historic redlining patterns. Black applicants are denied conventional mortgages at approximately 2x the rate of white applicants with similar financial profiles. The CFPB's 2023 AI examination guidance — which the Trump CFPB has reviewed but not formally rescinded — addressed how fair lending laws apply to algorithmic decision-making. Disparate impact theory under ECOA and FHA requires lenders to justify algorithmic disparities with legitimate business necessity.
HMDA threshold adjustments: A 2020 CFPB rule had raised the closed-end mortgage reporting threshold to 100 loans (from 25), exempting many community banks and credit unions. The U.S. District Court for D.C. vacated that threshold change on September 23, 2022 (NCRC v. CFPB), restoring the original 25-loan threshold from the 2015 rule. CFPB and prudential regulators have stated they will not pursue enforcement against institutions that originated 25–99 closed-end loans in 2020–2022 in reliance on the vacated rule, but the 25-loan threshold has been the operative law since the 2022 ruling.
HMDA and commercial real estate: HMDA covers residential mortgage lending but not commercial mortgages. The CRA Section 1071 rule (small business data collection, finalized 2023) extends analogous data collection to small business loans including commercial real estate — providing HMDA-like transparency for business lending. The Trump CFPB has delayed and proposed rescinding the Section 1071 rule, preventing this expansion of lending data transparency to the commercial sector.

In March 2026, President Trump signed an Executive Order titled "Promoting Access to Mortgage Credit," directing the Federal Reserve Vice Chairman for Supervision, CFPB Director, NCUA, FDIC, and OCC to reduce regulatory burdens that have driven up the cost of homeownership and restricted access to mortgage credit.