March 26, 2026
Via Electronic Submission (reginfo.gov)
Acting Director Vought
Consumer Financial Protection Bureau
1700 G Street NW
Washington, DC 20552
RE: Docket No. CFPB-2026-0010, Home Mortgage Disclosure Act (Regulation C), OMB Control Number 3170-0008
Dear Acting Director Vought:
The National Community Reinvestment Coalition (NCRC) submits this comment in response to the Bureau’s 30-day PRA notice seeking extension of the HMDA (Regulation C) information collection. NCRC is a coalition of more than 700 member organizations dedicated to promoting fair lending, community reinvestment, and equitable access to credit. We are among the most intensive users of HMDA data in the country.
NCRC writes to strongly support the continuation of HMDA data collection and to place this PRA proceeding in the context of Congress’s repeated and deliberate affirmation and expansion of HMDA over five decades. We also address the Bureau’s burden estimate, which we believe significantly overstates actual compliance costs considering the Bureau’s own research findings and the widespread availability of commercial automation tools.
I. Congress Has Repeatedly Affirmed HMDA Data Collection.
HMDA’s data collection reflects a half-century of deliberate Congressional action. Every time Congress has revisited HMDA, it has chosen to maintain or expand the data collection, never to eliminate it.
Enacted in 1975, HMDA[1] requires lenders to disclose the geographic distribution of mortgage loans by census tract so communities could identify patterns of disinvestment. In 1989, through the Financial Institutions Reform, Recovery, and Enforcement Act (FIRREA),[2] Congress significantly expanded HMDA by adding requirements to collect borrower race, ethnicity, sex, and income data, as well as information on loan applications, denials, and pricing. This expansion was a direct Congressional response to evidence that discriminatory lending practices were widespread and that the original geographic-only data was insufficient to detect them.
In 2010, through the Dodd-Frank Wall Street Reform and Consumer Protection Act,[3] Congress again expanded HMDA, directing the CFPB to add data points on loan terms, conditions, and borrower characteristics that would improve visibility into the lending practices that contributed to the 2008 financial crisis. Congress specifically mandated collection of information on age, credit score, loan term, interest rate, property value, and other variables. This was a bipartisan legislative judgment that more data was needed to prevent another crisis and to protect consumers from predatory lending. Congress also explicitly protected core data fields including race, ethnicity, sex, age, and income from exemption, preserving the data necessary for fair lending analysis.
II. The Bureau’s Burden Estimates Are Confirmed by Industry’s Own Data.
The Bureau’s March 2023 Report on the Home Mortgage Disclosure Act Rule Voluntary Review[4] found that per-application ongoing compliance costs were modest across all institution tiers: $0.47 for high-complexity (large) institutions, $11 for moderate-complexity institutions, and $42 for low-complexity (small) institutions, all in 2018 dollars.[5] Total market-level ongoing compliance costs were estimated at approximately $67.3 million in 2018, with roughly 45% attributable to the then-new open-end line of credit reporting requirement. Variable costs passed through to consumers were estimated at just $0.02 per loan application for large institutions. The Bureau found no evidence that HMDA reporting requirements caused lenders to exit the mortgage market.[6]
Critically, these are not just the Bureau’s numbers. The report states that industry estimates of HMDA compliance costs per loan application reported to the Bureau, in response to the 2019 HMDA Proposal were similar to the Bureau’s estimates from the 2015 HMDA Final Rule.[7] The report further notes that these estimates are within a range of costs per application submitted by industry commenters that were reported in the 2020 HMDA Final Rule. For example, the mortgage lending industry submitted cost data to the Bureau on multiple occasions, through the 2019 HMDA Proposal,[8] the 2020 HMDA Final Rule,[9] and the 2023 voluntary review outreach, and each time industry’s own estimates confirmed the Bureau’s figures.[10]
III. Changing the Threshold Reporting Requirements Would Substantially Reduce the Ability of Regulators to Implement Existing Laws.
Industry has asserted that increasing the threshold value that triggers compliance will help lower compliance costs. However, even minor increases in the threshold required to report HMDA impact the veracity of the entire set of data. The vast majority of lenders make a relatively small number of loans. So even minor increases in the threshold of loans needed to trigger reporting would have an outsized impact.
As the graph above demonstrates, at 50 loans the threshold would eliminate 25% of the lenders that currently report as of 2024. A 100-loan threshold decreases the number of reporting institutions by 30%, and a 200-loan threshold means almost half of all lenders would be missing from this data. This doesn’t just hamper the industry and the public from understanding the housing market, it ties the hands of regulators charged with avoiding another housing crash like that which precipitated the Great Recession.
IV. The CFPB and the Compliance Industry Have Built an Advanced Ecosystem That Streamlines HMDA Reporting.
The Bureau’s estimated burden of 1,510,960 total annual hours should also be evaluated against the reality of the compliance technology ecosystem that now exists around HMDA reporting, which significantly decreases compliance times. The point here is not simply that compliance vendors market their products as easy to use. Vendors will always do that. What matters is the testimony of actual bankers and financial institution staff who confirm that HMDA compliance has become streamlined in practice,[11] and the underlying reason why: the CFPB and the compliance industry have together built an advanced ecosystem that automates the data uploading, validation, and verification process, reducing what was once a labor-intensive manual exercise to a largely automated workflow.
V. Most HMDA Data Is Already Captured in the Normal Course of Mortgage Origination; Reduced Reporting Would Impose Unnecessary Costs on Industry.
The Bureau’s burden estimate implicitly treats HMDA data collection as a standalone compliance exercise, separate from the lender’s core business operations, and fails to account for the costs lenders would incur to change current systems. This framing overstates what is an incremental effort. The vast majority of HMDA-reportable data points are information that lenders already collect during the standard mortgage origination process because it is essential to underwriting, pricing, and secondary market delivery.
The Uniform Residential Loan Application[12] (URLA/Form 1003), jointly published by Fannie Mae and Freddie Mac and mandatory for all GSE-eligible loans since March 2021, directly captures most data elements reported under Regulation C. Borrower income, loan amount, property address, loan type and purpose, loan term, interest rate, occupancy type, credit score, debt-to-income ratio, and combined loan-to-value ratio are all fields on the URLA that lenders must collect to underwrite the loan and submit it to automated underwriting systems. The 2021 URLA redesign was itself explicitly designed to align with HMDA data collection requirements.
Even borrower demographic information (race, ethnicity, and sex) has been collected on the loan application since 1977 under Regulation B implementing the Equal Credit Opportunity Act.[13] Section 8 of the URLA, the Demographic Information Addendum, collects this information as an integrated part of the application, not as a separate HMDA-specific exercise. Lenders do not need to create any additional process to gather this data; it flows directly from the loan application into the loan origination system.
The Mortgage Industry Standards Maintenance Organization (MISMO) v3.4 Reference Model,[14] which underpins both the URLA and GSE automated underwriting specifications, now provides a common data framework that minimizes the gap between what lenders collect for business purposes and what they report for HMDA purposes. Simply put, lenders collect data to make loans. The software transmits that data via systems to the CFPB. The industry and the CFPB have made large strides in streamlining this process – changing it would impose costs. And the testimonials of bank staff as to the simplicity of HMDA reporting are evidence of the success.[15]
Finally, the Federal Reserve’s Consumer Compliance Outlook publication acknowledges that many banks find automated collection processes reduce the burden of compiling HMDA data.[16]
VI. Close the Purchased Loan Data Stripping Loophole to Reduce Burden and Improve Data Quality.
This loophole itself is a source of unnecessary burden. Lenders that purchase loans are currently spending staff time and resources to actively strip demographic data from records that already contain it. The originating lender collected this data as part of the application process. It exists in the loan file. The purchasing lender then dedicates compliance staff hours to removing it before HMDA submission. If the Bureau’s goal is to reduce burden, one obvious step is to close this loophole so that purchasing lenders no longer need to expend effort deleting data that was already collected. This would simultaneously reduce compliance burden and improve data quality, a rare policy change that advances both goals at once.
Current CFPB guidance allows lenders that purchase loans to remove demographic data from those records before submitting them to HMDA. This loophole has created a massive gap in the data. NCRC’s most recent mortgage market analysis found that while only 12% of loan originations lacked demographic data in 2023, over 90% of purchased loans were reported without this essential information.[17] Out of 1.4 million purchased loans reported in 2023, approximately 863,743 were sold to entities that would report them as purchased loans in HMDA, leaving over 500,000 loans that appear in the data only as “purchased” with almost no information about the borrowers they served.
VII. Certain HMDA Fields Are Essential for Fair Lending Analysis and Must Be Retained.
The HMDA dataset contains a core group of publicly reported variables derived from loan applications that are essential for understanding mortgage markets and evaluating whether lenders are meeting community credit needs. In NCRC et al. v CFPB, Chief Judge Beryl A. Howell ruled against the CFPB’s rule that raised the HMDA reporting threshold from 25 to 100 loans. She reasoned that HMDA is a disclosure and fair lending statute whose purpose is to ensure that individual lenders comply with anti-discrimination and reinvestment obligations. Such data enable community organizations and other local stakeholders to assess whether lending institutions are making responsible and sustainable loans and are refraining from abusive practices.[18]
Currently, 56 of these fields are publicly available. Together, they provide the foundation used by regulators, researchers, industry participants, and community organizations to analyze lending patterns, monitor fair lending compliance, and assess access to credit in local housing markets.
- Loan characteristicslike action taken, loan type, purpose and occupancy allow stakeholders to evaluate denial rates, the types of loan programs favored by different lenders, and the role of investors in the market.
- Borrower demographicssuch as race, ethnicity and sex allow stakeholders to evaluate which borrowers are getting loans and which are being denied, and whether costs differ across groups.
- Borrower incomeallows stakeholders to evaluate who can afford to buy, and whether, paired with area median income, lenders are reaching low- and moderate-income families.
- Property locationsuch as state, county and census tract allows stakeholders to evaluate where lending is happening, and whether lenders are serving low-income and majority-minority neighborhoods.
- Loan pricingsuch as interest rate, points, fees and lender credits allows stakeholders to evaluate what borrowers pay, and whether those costs vary by race or ethnicity.
- Lender identifierssuch as Legal Entity Identifier and institution type allow stakeholders to evaluate which lenders are active in a market, how they compare to peers and whether banks, mortgage companies or credit unions serve communities differently.
- Loan and property valuessuch as loan amount, property value and combined loan-to-value ratio allow stakeholders to evaluate what kinds of homes different groups can access and how affordability is changing over time.
VIII. Conclusion
Taken together, these data points enable regulators, researchers, community organizations, and the public to identify disparities in lending outcomes across racial, ethnic, and income groups. Without these data, discriminatory lending patterns would be significantly more difficult to detect and address.
Homeownership remains the primary wealth-building vehicle for most American families. Ensuring that access to mortgage credit is offered widely and equitably is not only a matter of civil rights but of economic policy. Any weakening of data collection requirements would reduce the sizable benefit the data currently provides without providing much if any cost savings. Weakening HMDA requirements would also disproportionately harm the very communities the statute was designed to protect.
Respectfully submitted,
Jason Richardson
Senior Director of Research
National Community Reinvestment Coalition
[1] Home Mortgage Disclosure Act of 1975, Pub. L. 94-200, title III, 89 Stat. 1125, codified at 12 U.S.C. § 2801 et seq. Implemented by Regulation C, 12 C.F.R. Part 1003.
[2] Financial Institutions Reform, Recovery, and Enforcement Act of 1989, Pub. L. 101-73, § 1211, 103 Stat. 183 (1989).
[3] Dodd-Frank Wall Street Reform and Consumer Protection Act, Pub. L. 111-203, § 1094, 124 Stat. 1376, 2097 (2010).
[4] https://files.consumerfinance.gov/f/documents/cfpb_hmda-voluntary-review_2023-03.pdf
[5] Id. at page 7-9 and Section 6 beginning at page 113.
[6] Id.
[7] Id. at 6.
[8] Id. at 128.
[9] Id.
[10] Id.
[11] See footnote 13, infra.
[12] Fannie Mae Selling Guide B3-2; Freddie Mac Seller/Servicer Guide. The redesigned Uniform Residential Loan Application (URLA/Freddie Mac Form 65/Fannie Mae Form 1003) became mandatory for all GSE-eligible loans effective March 1, 2021. See FHFA, Redesigned Uniform Residential Loan Application (2019).
[13] Equal Credit Opportunity Act, 15 U.S.C. § 1691 et seq.; Regulation B, 12 C.F.R. § 1002.5(a)(2) (requiring creditors to collect demographic information for certain dwelling-secured loans).
[14] Mortgage Industry Standards Maintenance Organization (MISMO), MISMO v3.4 Reference Model (2017). MISMO is a subsidiary of the Mortgage Bankers Association.
[15] Judy A. Edwards of Greer State Bank stated that she was able to accomplish her HMDA submission in a short period of time, calling the process a testament to the software’s user friendliness. https://www.rataassociates.com/testimonials/; TruHome Solutions reported completing HMDA submission in record time, with doubled reporting volume taking less than half the time of the prior year. https://www.rataassociates.com/testimonials/; Banker’s Bank of Kansas reported reducing compliance work hours from approximately 160 hours to 30 hours using Ncontracts’ platform, an 81% reduction. https://www.ncontracts.com/case-studies; Mark Hutchinson, VP of ERM and Internal Audit at Langley Federal Credit Union, stated that without compliance automation, his team would still be doing a lot of manual compliance work. https://www.ncontracts.com/case-studies; Wolters Kluwer, whose HMDA Wiz product promises automated workflows, precision geocoding, and examiner-ready summaries in minutes. https://www.wolterskluwer.com/en/solutions/wiz/hmda-wiz; Ncontracts, whose Nlending HMDA platform advertises reducing quarterly analytics workload from 96 hours to as little as 3 hours. https://www.ncontracts.com/products/hmda-software; Coviance, whose HMDA Compliance Validation Tool advertises saving lenders hours if not days or weeks in collecting and submitting data, with automatic updates to reflect current CFPB requirements. https://www.coviance.com/news/mastering-hmda-reporting-how-coviances-automation-tool-simplifies-the-process
ComplianceTech, whose HMDA Ready product offers quick import functionality from loan origination systems. https://compliancetech.com/hmda-compliance-software-hmda-ready/.
[16] Consumer Compliance Outlook, HMDA Data Collection and Reporting: Keys to an Effective Program (Fourth Issue 2020) https://www.consumercomplianceoutlook.org/2020/fourth-issue/hmda-data-collection-and-reporting-keys-to-an-effective-program/.
[17] NCRC Mortgage Market Analysis (October 2024) https://ncrc.org/ncrc-mortgage-market-analysis/
[18] Chief Judge Beryl A. Howell, Civil Action No. 20-2074 (BAH), United States District Court for the District of Columbia, NCRC et al. v CFPB, p. 38, https://law.justia.com/cases/federal/district-courts/district-of-columbia/dcdce/1:2020cv02074/220484/26/
