PART 2:
RACE AND THE MORTGAGE MARKET
Gains for Hispanic Borrowers Stall, Losses for Black Borrowers Continue
October 2026
Buying a home is the main pathway towards building generational wealth for most American families. For decades, the federal government built the infrastructure that made mass homeownership possible almost exclusively for White families through lending that was affordable, tightly regulated and segregated by race. Once that project succeeded, the government restricted access to the very programs and underwriting rules just as the Civil Rights era opened the door for Black families to be able to use them.
Since then, three main racial/ethnic groups (Black, Asian and Hispanic Americans) have followed three different paths when it comes to their homeownership rates. Asian families have bought homes at rates that now outpace their share of the population. Hispanic families gained ground for years, nearly closing the gap, then stalled. Black families have not seen much progress.
Six decades after the Civil Rights era and the end of legal discrimination, the homeownership gap between Black and White families has not moved.
As the first installment of this Mortgage Market series found, home purchase lending has remained constant at a rate of about 2.9 million originations a year for three straight years. This 2nd installment of this series will explore which communities are gaining access to these home loans and which communities are still seeing declines in their home-ownership rates.
Jason Richardson, Senior Director of Research
Resources
For the complete series, including videos, interactive visualizations and localized market analysis, visit our 2026 Mortgage Market series homepage.
NCRC uses AI tools to assist with data processing, query development and report drafting. NCRC maintains full responsibility for the accuracy of all published analyses.
For customized community data requests or inquiries about membership benefits contact Ralph Cyrus (Membership Engagement Specialist) at rcyrus@ncrc.org.
Key Takeaways
1
Hispanic home purchase lending rose from 13.9% of originations in 2020 to a peak of 17.5% in 2024, nearly matching the Hispanic share of the adult population nationally (17.4%) before decreasing slightly to 17.0% in 2025.
2
Black home purchase lending peaked at 9.2% in 2022 and 2023 and then fell for two straight years to 8.2% in 2025, well below the Black share of the adult population (11.9%).
3
White home purchase originations grew by about 79,000 loans in 2025, more than the entire net increase in home purchase lending that year, even as Hispanic, Black and Asian originations combined fell by about 40,000. This concentration of growth in a single group is not typical of the trends seen in the broader 2020 to 2025 period.
4
5
Lending for Hispanic borrowers ran well below their total population share in high-cost coastal metros, such as San Jose, Los Angeles and New York, and at or above it in Miami, Orlando, Las Vegas and Phoenix, which are among the metros with the largest Hispanic populations.
The Racial Gap the Market Inherited
Homeownership rates by race have been measured for more than a century, with the distance between racial/ethnic groups widening considerably.
The chart below shows those rates over the full period of record. The Black-White gap is about the same size today as when the Census Bureau first recorded it in 1900.
Homeownership Rate and Gap by Race, 1870-2035
The most striking thing about this picture is how little has changed. While most economic measures in this country look different after a century, this one has not seen a significant positive shift. Whatever forces have had an impact on wages, education levels and household composition over that time have not closed the distance between who owns a home and who rents one.
The housing finance structure that produced that gap was built on purpose. During the Great Depression era, programs were created that would legally cement a practice of lenders avoiding lending in neighborhoods based on the race of the people who lived there. After World War II, the federal government expanded homeownership through Veterans Administration (VA) loans and Federal Housing Administration (FHA) underwriting standards that excluded Black neighborhoods by design. This resulted in White families acquiring generational wealth via homeownership while Black families continued to see the homeownership gap deepen.
While Hispanic households have made some gains in the homeownership sphere, they also are working against an inheritance gap when compared to White households. The Census Bureau did not track Hispanic origin as its own category until 1980 so the historical record here is thinner than for Black and White homeownership. What data does exist shows a pattern common among immigrant groups.
Homeownership rises with time spent in the country, with US born children of immigrants closing most of the remaining gap. By the time they reach their mid-40s, most children of immigrants have closed the wealth gap compared with their non-immigrant neighbors. This is a different pattern than what is observed among the descendants of American chattel slavery, where legal barriers to homeownership were enforced by design rather than eased by time in the country.
Same Year, Three Very Different Paths
The chart below tracks each demographic group’s share of home purchase lending for the 2020 through 2025 time period. The shares were adjusted in a way that applicants who did not report their race or ethnicity were set aside and the remaining groups’ lending rates were summed to the whole total.
Asian, Hispanic and Black home purchase lending all fell in 2025, but the trajectories of each group looked very different coming into that year. Prior to 2025, Asian and Hispanic home purchase lending was on the rise, with Asian borrowers overrepresented compared with their population.
The chart below shows the rapid rise and retreat of both Asian and Hispanic homebuyers in the market during and after the pandemic. By contrast, the stagnation of African Americans homeownership is quite clear:
Hispanic and Black borrowers spent most of this period on separate trajectories. Hispanic homebuying rose year after year, closing much of the distance between it and the overall Hispanic share of the adult population, which is the closest any underrepresented group has come to parity in this data set. Black homebuying rates rose early in the period, flattened and then retreated back to its original level, ending the period exactly where it started.
In the final year, homebuying rates for both demographic groups trended down at the same time. Lending to Hispanic, Black and Asian borrowers fell in absolute terms and not only as a share of a shrinking market, while White lending rates rose. In earlier years that saw considerable market growth, the gains were spread across several groups rather than concentrated in only one.
For Black borrowers, it means five years of movement with nothing to show for it. For Hispanic borrowers, it means a long climb interrupted at the point where parity was within reach.
This data show the direction of the shift but not its cause. For Hispanic and Asian communities where population growth depends heavily on immigration, a sharp national falloff over the same period is a plausible contributor. Many major metro areas had also come to depend on immigration for most of their population growth before the 2025 slowdown, with recent Census data showing a population rebound in major metro areas driven by it.
The longer arc matters for anyone planning around future mortgage demand. Senior Fellow with the Brookings Institution William Frey projects that higher immigration rates would add 50 million working-age people and 17 million children under 15 by 2051, while a net zero immigration rate would shrink the working-age population as the senior population grows. Roughly half of recent arrivals identify as Hispanic and a quarter as Asian. Household formation and mortgage demand also follow these number trends as well.
Where to Loans Went
The race of a borrower and the composition of the neighborhoods they reside in are a few ways to determine what areas and regions lending actually lands. The chart below categorizes every home purchase loan according to the share of the total population it makes up in surrounding census tracts that are majority non-White:
The chart above shows that Census tract boundaries changed during the 2020 Census and are represented in HMDA data from 2022, which produces a large gap between the two years that further obscures the reasons around certain lending behavior patterns. However, the comparison worth making is within each period rather than across the break in this trend.
From 2023 onward, lending has drifted away from neighborhoods with the largest people of color populations and toward ones with higher concentrations of White individuals. While the rates of upward movement were small, they held throughout the 2023 to 2025 time period. Borrower-level lending shows a parallel pattern.
Two different measures pointing the same way is the useful signal here as a market can shift when it comes to borrower race without shifting on geography, or the reverse. When both move in the same direction, it suggests that the change is in who is able to buy a home rather than in where houses happen to be for sale.
Who is Buying in These Neighborhoods
Neighborhoods with large populations of color contain households at every income level and so do the people buying homes there. While it is not unexpected that buyers in Majority Minority Census Tracts (MMCTs) will be diverse due to historical trends, the majority of the people buying homes in MMCTs are often already relatively wealthy.
The chart below shows the race and ethnicity of buyers in each group of neighborhoods split into the four CRA income tiers:
The data shows that upper-income buyers are not disproportionately White even in the most diverse neighborhoods. In tracts that are more than three-quarters people of color, 81.3% of upper-income buyers are Black, Hispanic, Asian or of another race. The people of color share among buyers below the upper-income cutoff point in those same tracts sits at 83.4%.
The least diverse tracts show the same pattern in reverse. Upper-income and below-upper-income buyers are more than 84% White. Moving from below-upper to upper income areas shifts the White share of buyers by two points or less in either group of tracts. Ultimately, a neighborhood’s racial composition and not a buyer’s income determines who its buyers are.
However, income does still matter at the margins. Low- and moderate-income buyers accounted for 34% of purchases in tracts that are more than 75% people of color against 26% in tracts that are less than 25%. The upper-income share of purchases fell in every group of neighborhoods in 2025, from 40.4% to 36.9% in the most diverse tracts and from 48.7% to 46.8% in the least. The loan volume flowing into a community of color doesn’t necessarily translate to a community’s own residents being able to buy a home.
Lending Patterns in Metro Regions
Among the 50 largest metropolitan areas, Black borrowers reached or exceeded parity with their population share in three cities: Providence, RI, Portland, OR and Seattle. In each of those metros, the Black population is a small share of the total adult population and the share of loans to Black borrowers exceeded those low levels.
Every metro area with a large Black population is receiving loans at a rate well below their share of the adult population, including in Baltimore (76.6) and Atlanta (78.9) on a scale where 100 means that the rate of lending matched the overall population.
The chart below ranks the top 50 metro areas by the gap between Black home purchase lending and the Black share of the adult population from most underrepresented to most overrepresented:
The data suggests that Hispanic borrowers split along cost lines rather than their actual population size. In the highest-cost coastal metros, Hispanic lending runs well below the Hispanic share of adults — 44.3 in San Jose, CA, 63.1 in Los Angeles and 67.2 in New York. Several metros with the largest Hispanic populations sit at or above parity, including Miami at 110.4 and Orlando at 125.2. Lending rates also run above the population share across a set of Midwestern and Southeastern metros with much smaller Hispanic populations.
Metro areas with small population bases show the widest swings in the parity index, with the raw shares being worth examining closer. For instance, in Louisville, KY, Hispanic borrowers received 17.1% of home purchase loans (13,689 loans in total) while Hispanic adults make up 6.1% of the population.
Two factors this analysis does not account for is that the Hispanic population in the US is not uniform, with first-generation and multi-generation Hispanic families buying homes at different rates in proportions that vary by metro. Also, one’s household income relative to a particular metro’s local market also matters. For example, San Jose, CA is among the highest-income metros in the country, with Hispanic households earning below their metro’s median income would show lower rates of homebuying for that reason alone.
The deliberate and systemic effort to enforce de-facto segregation by restricting access to housing has survived its legal demise. Housing finance has proven even more efficient at denying access to minority borrowers. The intransigent gap in White and Black homeownership has survived every legal effort to curtail it and it seems poised to continue to expand.
A more detailed analysis for all 50 metro areas can be found in this companion data workbook. NCRC members interested in an analysis of their own metro area can contact NCRC’s research team directly.
Resources
For the complete series, including videos, interactive visualizations and localized market analysis, visit our 2026 Mortgage Market series homepage. NCRC members can request customized community data requests as well. Those interested in membership benefits should contact Ralph Cyrus (Membership Engagement Specialist) at rcyrus@ncrc.org.
Series Introduction
Methods and Definitions
Data Sources
Home Mortgage Disclosure Act (HMDA) Data: Primary data source covering national mortgage lending patterns from 2018-2024, representing approximately 88% of all mortgage applications processed annually
US Census Bureau Data: Used for demographic information, population statistics, and income data including the American Community Survey and Decennial Census data
Consumer Financial Protection Bureau (CFPB) Data: Source for HMDA data collection and release
Brookings Institution Research: Referenced for demographic projections and population growth analysis
Federal Financial Institutions Examination Council: Source for HMDA data products
Analysis Period and Scope
Time Frame: 2018-2024
Loan Types Analyzed: Focus on home purchase loans for owner-occupied, site-built, 1-4 unit properties except as noted
Data Processing Methods
Race/Ethnicity Calculation: Detailed subgroup identification method that prioritizes specific ethnic codes (11-14 for Hispanic subgroups, 21-27 for Asian subgroups, 41-44 for Pacific Islander subgroups) over broader categories
Missing Data Treatment: “No Data” loans excluded from demographic calculations rather than treated as a separate racial category
Year-over-Year Comparisons: Multi-year data compared using identical calculation methods across the 2018-2024 period
Key Metrics and Definitions
Low- and Moderate-Income Borrower (LMIB): Borrowers with household income below 80% of area median income
Low- and Moderate-Income Census Tract (LMICT): Geographic areas where median family income is at or below 80% of metro area median family income
Majority-Minority Census Tract (MMCT): Census tracts where racial/ethnic minorities comprise more than 50% of residents
Cost Per Dollar: Calculated as (Total Payments Over 30 Years + Closing Costs) ÷ Original Loan Amount
Market Share: Percentage of total loans in a market originated by a specific lender
Calculation Formulas
Percentage Calculations:
- Low and moderate-income borrower percentages: (LMIB/Total Loans) × 100
- LMI Tract percentages: (LMICT/Total Loans) × 100
- Majority-minority tract percentages: (MMCT/Total Loans) × 100
- Race and ethnicity percentages: (Race group/(Total Loans-No Data)) × 100
Data Quality and Limitations
Coverage Limitations: Analysis limited to loans with reported demographic data (approximately 4.7 million of 5.3 million total loans in 2024)
Census Boundary Changes: 2020 Census redrew neighborhood boundaries, affecting historical comparisons for majority-minority tract analysis starting in 2022
Missing Data Impact: Growing number of loans without demographic data affects trend analysis accuracy
Multiracial Identity Challenges: Difficulty measuring lending equity for people identifying as multiple races
Terms
AAPI – Asian American and Pacific Islander: Demographic label that groups together Asian and Pacific Islander communities
AHO – Access to Home Ownership: Office of Hawaiian Affairs program that guarantees portions of home loans for Native Hawaiian first-time homebuyers
CDFI – Community Development Financial Institution: Specialized lenders focused on serving underserved communities
CFPB – Consumer Financial Protection Bureau: Federal agency that oversees mortgage lending and consumer financial protection
CRA – Community Reinvestment Act: Federal law requiring banks to meet credit needs of their entire communities, especially low-income areas
FHA – Federal Housing Administration: Government agency that insures mortgages
GSE – Government-Sponsored Enterprise: Companies like Fannie Mae and Freddie Mac that buy mortgages from lenders
HMDA – Home Mortgage Disclosure Act: Federal law requiring lenders to report detailed mortgage lending data
HoPI – Hawaiian or Pacific Islander: Demographic category for Native Hawaiian and Pacific Islander populations
HUD – US Department of Housing and Urban Development: Federal agency that oversees housing programs
IHBG – Indian Housing Block Grant: Federal program funding housing development on tribal lands
LEI – Legal Entity Identifier: Unique identification code for financial institutions
LMI – Low- and Moderate-Income: People or areas with incomes at or below 80% of area median income
LMIB – Low and Moderate-Income Borrower: Borrowers with incomes below 80% of area median income
LMICT – Low- and Moderate-Income Census Tract: Geographic areas where median incomes fall below 80% of regional average
MIP – Mortgage Insurance Premium: Monthly fee paid by FHA borrowers to protect lenders against default
MMCT – Majority-Minority Census Tract: Neighborhoods where racial/ethnic minorities make up more than 50% of residents
RHS – Rural Housing Service: USDA program providing housing assistance in rural areas
VA – Veterans Affairs: Government department that provides benefits to military veterans, including mortgage guarantees
YoY – Year-over-Year: Comparison between the same period in consecutive years
The Home Mortgage Disclosure Act (HMDA) data is collected and released each year by the Consumer Financial Protection Bureau (CFPB). This dataset offers unparalleled details about 88% of the mortgage applications processed each year. This information is critical for any regulator, advocate or lender that wants to understand the market. Data of this kind promotes fair and efficient markets.
This series of research briefs will offer a deep analysis of this data and help policymakers, the general public and National Community Reinvestment Coalition (NCRC) members understand current mortgage market trends at the local level. There are a great number of topics that this data will help us explore via a series of episodes with easy to understand reports, policy suggestions, videos, data visualizations and maps. These insights can help various organizations and market actors to utilize this data to support fair lending programs and initiatives in their communities.
There were several key takeaways and findings in the 2024 HMDA data that we will discuss in future episodes. This introduction and summary will be updated as new episodes in this series are published.
Key Takeaways
Key Findings
- Declining Low-Income Access: Lending to low- and moderate-income borrowers fell to 14.2% in 2024 (the lowest level since 2018), reflecting severe affordability challenges.
- Hispanic Market Growth: Hispanic borrowers now exceed their population share in mortgage lending, reaching 17.7% of home purchase loans in 2024.
- Persistent Black Homeownership Gap: Black borrower participation remains stagnant at 8.9% (well below their 11.7% share of the adult population), with declining shares in major metro areas.
- Less-Regulated Lenders Displacing Banks: Mortgage companies and credit unions – whose lending activity is not covered by key economic opportunity laws like the Community Reinvestment Act – have greatly expanded their share of lending. Mortgage companies are making ⅔ of home purchase loans in 2024. Credit unions are now making more cash out refinance loans than banks and hold nearly the same share of the home equity market that banks do, without the oversight offered by the CRA..
Access and Affordability
Low- and moderate-income (LMI) home purchase lending continues its long decline, now at just 25.8% of all home purchases on owner occupied, 1-4 unit site built homes. Upper income borrowers dominate homebuying, even in LMI and majority minority census tracts.
Demographic Shifts
Hispanic borrowers continue to expand their market presence, and in 2024 for the first time on record were slightly over-represented in home purchase lending relative to overall population share. 17.7% of loan originations in 2024 went to a Hispanic borrower, exceeding the 16.8% percent of the overall adult population who identify as Hispanic. In contrast, the Black borrower share of the market remains well below their population representation (8.9%), with declines in key markets like Atlanta, Houston and Washington, DC.
Mortgage Market Series
PART 2: Race and the mortgage market
Subscribe
Get NCRC news and
alerts by email.