Mortgage Lending Analysis (1996-2024)

Mortgage lending does not remain static. It responds to changes in the economy, housing market, financial system, public policy, and the communities that lenders serve. Looking across nearly three decades therefore provides a broader perspective on the mortgage market than any single year can offer.

This section traces the broad contours of mortgage lending in the Pittsburgh metropolitan area from 1996 to 2024. Using HMDA data from four benchmark years, 1996, 2006, 2016, and 2024, we examine how the region's mortgage lending landscape has evolved: who is borrowing, what kinds of loans they are receiving, which institutions are providing credit, and how lending patterns differ across communities. The analysis covers all eight counties that make up the present-day Pittsburgh MSA, providing a regional view of both long-term change and enduring patterns in access to mortgage credit.

Local institutions remain influential in an increasingly interstate mortgage market

The geographic boundaries of American banking changed substantially during the period covered by this study. The Riegle-Neal Interstate Banking and Branching Efficiency Act of 1994 removed important barriers to interstate banking and branching, making it easier for banking organizations to operate across state lines. At the same time, Section 109 sought to ensure that covered interstate branches did not simply collect deposits from host states without reasonably helping to meet their credit needs.

Against this backdrop, we might expect mortgage lending in Pittsburgh to have shifted progressively toward institutions headquartered outside Pennsylvania. The HMDA data, however, show no such steady transition. Pennsylvania-headquartered lenders accounted for 58.2% of originations in 1996, falling sharply to 37.6% in 2006 before recovering to 54.8% in 2016 and settling at 48.4% in 2024.

The fluctuations point to another force shaping Pittsburgh's mortgage market: the changing fortunes and corporate identities of the region's own large financial institutions. Mellon, historically one of Pittsburgh's largest banks, merged with the Bank of New York in 2007, shifting the headquarters of the combined institution to New York. PNC, meanwhile, remained headquartered in Pittsburgh while expanding into one of the country's largest banking organizations.

The result is a more complicated story than a gradual displacement of local lenders by national ones. Interstate banking made geography more permeable, but Pittsburgh's position as the home of major financial institutions has continued to shape who provides mortgage credit in the region. Changes involving a small number of very large institutions can therefore substantially alter both the local lending market and the concentration of deposits and branches documented elsewhere in this study.

Mortgage originations peaked in 2006, while 2024 had the lowest volume of the four years studied

Mortgage lending in the Pittsburgh region has not followed a steady long-term trajectory. Originations reached 85,084 in 2006 but fell to 46,766 in 2024, the lowest of the four benchmark years and roughly 45% below the 2006 peak. The contrast captures two very different mortgage environments. The 2006 market was near the height of an expansion in subprime and privately securitized mortgage credit; Federal Reserve research documents the subsequent collapse of that market and sharp contraction in lending after 2006. By 2024, the constraint was different: elevated rates left many existing homeowners with mortgages far below prevailing rates. Federal Reserve researchers estimate that this mortgage-rate "lock-in" explained 44% of the decline in mortgage-borrower mobility from 2021 to 2022.

The larger lesson is that mortgage lending volume is not simply a measure of whether households want homes. It reflects the financial conditions under which households can buy, sell, refinance and move. The four benchmark years therefore represent substantially different environments for accessing mortgage credit.

The dominant purpose of mortgage lending differs across Pittsburgh MSA counties

The composition of mortgage lending also varies geographically.

Allegheny and Butler: Home purchase was the most common mortgage purpose throughout the four years.
Armstrong: Refinancing was more common in 1996 and 2006, while home purchase became the dominant purpose in 2016 and 2024.
Beaver, Fayette, Lawrence, Washington and Westmoreland: Refinancing was particularly prominent in 2006, while home-purchase lending predominated in the other years.

The purpose here is not to classify counties as having "strong" or "weak" housing markets. Rather, it shows what residents are primarily using mortgage credit for. Home-purchase loans finance housing transactions, whereas refinance loans allow existing homeowners to restructure debt, obtain different interest rates or terms, and potentially extract accumulated home equity.

These differences also underscore why county-level analysis matters: mortgage activity within the Pittsburgh MSA is not uniform, and the balance between purchasing and refinancing has changed differently across counties and over time.

Access to homeownership has become more diverse, but the Black share of Allegheny County lending has barely advanced over nearly three decades

Allegheny County has become more racially diverse over the period covered by this study, and that change is reflected to some extent in its mortgage market. Asian and Hispanic borrowers account for growing shares of home-purchase originations. The trajectory for Black borrowers, however, is notably different.

Black borrowers accounted for 6.36% of Allegheny County home-purchase originations in 1996. That share declined in 2006 and again in 2016 before recovering to 6.19% in 2024, leaving it slightly below where it had been nearly three decades earlier.

This relative stagnation is more striking when considered alongside the county's demographic change. In the 1990 Census, Allegheny County had approximately 149,600 Black residents, representing about 11.2% of its population. By 2020, the Black-alone population had grown to approximately 163,700, or about 13.1% of county residents. Thus, even as Allegheny County's overall population declined over this period, its Black population increased in both number and population share. Yet Black borrowers' share of home-purchase mortgage originations did not increase.

The lending share should not be expected to match the population share one-for-one. The population as a whole is not the same as the population seeking to purchase a home: differences in age, income, household formation, existing homeownership, wealth and application rates all affect who enters the mortgage market. Nevertheless, the divergence suggests that increasing racial diversity in Allegheny County has not translated proportionately into greater Black representation among mortgage-financed home purchases.

Application outcomes provide another reason to examine this pattern more closely. In 2024, 13.2% of Black applicants with a lender credit decision in the owner-occupied single-family home-purchase market were denied, compared with 6.1% of White applicants. The disparity cannot by itself establish why applicants received different outcomes; credit characteristics, debt-to-income ratios, loan characteristics and other underwriting factors may contribute. But when considered alongside the long-run lending trend, it identifies an important point of inquiry: Black residents represent a larger part of Allegheny County than they did three decades ago, while their representation among home-purchase originations has remained essentially unchanged and Black applicants continue to experience substantially higher denial rates.

The central question, therefore, is not simply whether mortgage lending has become more racially diverse. It is whether the opportunities created by the region's home-purchase market are reaching its changing population similarly across racial and ethnic groups.

LMI borrowers remain important participants in home-purchase lending even as the region's income geography and branch network shift

Across the Pittsburgh MSA, low- and moderate-income borrowers generally account for a larger share of owner-occupied home-purchase originations than they did in earlier benchmark years. In Allegheny County, upper-income borrowers continue to account for the largest share of home-purchase lending, but the share going to LMI borrowers has gradually increased. Similar changes are visible across much of the region.

That trend is notable because other parts of the region's financial geography have moved in a different direction as per this lending study’s branch analysis. Between 1996 and 2025, the share of the regional population living in upper-income census tracts increased from 24.2% to 32.8%, while the population shares of middle- and moderate-income tracts declined. The physical banking network has also become more concentrated in upper-income areas. Moderate-income communities experienced the largest decline in branch presence, from 4.44 to 2.53 branches per 10,000 residents, while upper-income communities experienced a considerably smaller decline, from 4.05 to 3.48. By 2025, upper-income tracts contained 38.7% of the region's branches despite accounting for 32.8% of its population.

The continued importance of LMI borrowers and communities in mortgage lending should also be considered in the context of the Community Reinvestment Act. Unlike racial composition, income is explicitly incorporated into CRA examinations. The CRA requires regulators to evaluate covered banks' records of helping meet the credit needs of their communities, including low- and moderate-income neighborhoods. For home mortgage lending, examiners consider both the distribution of loans across neighborhoods of different income levels and lending to low-, moderate-, middle-, and upper-income borrowers.

The Pittsburgh findings are therefore consistent with an important objective of the CRA: LMI borrowers have remained meaningful participants in the home-purchase mortgage market even as the region has become more affluent and its physical banking network has shifted toward upper-income areas.

Government-backed mortgages appear to be an important gateway into homeownership, particularly for Black borrowers

This portion of the analysis is restricted to owner-occupied home-purchase loans.
Government-backed mortgage lending varies considerably across the four years studied. 2016 had the highest share of government-insured loans, while 2006 had the lowest. FHA represents the largest government-backed category overall, while Armstrong and Fayette have comparatively greater use of USDA/Rural Housing loans.
The 2006 result is consistent with the national mortgage environment. FHA's national market share had fallen sharply from roughly 16% of first-lien home-purchase loans in 2000 to about 5% in 2006—as borrowers increasingly entered conventional and non-traditional mortgage products. After the private mortgage market contracted during the financial crisis, the federal government's role expanded dramatically; Federal Reserve research documents the subsequent increase in FHA, VA and other federally supported lending.
Policy choices also matter. FHA premiums, loan limits, underwriting rules, appropriations and program design have changed across presidential administrations and Congresses. Federal Reserve research, for example, found that the 2015 reduction in FHA mortgage-insurance premiums was followed by an increase in FHA's home-purchase market share from roughly 22% to 27%, illustrating how policy decisions can directly influence program utilization.
The racial dimension is especially significant locally because government-backed products accounted for a substantial share of homeownership lending to minority borrowers. In Allegheny County in 2016, approximately 52% of Black borrowers' owner-occupied home-purchase originations were FHA-insured. This suggests that FHA was not simply another mortgage product in the market: it constituted a major pathway into homeownership for Black borrowers.
Similarly, the relatively greater USDA presence in Armstrong and Fayette is consistent with USDA's focus on supporting homeownership in eligible rural areas.
Together, these findings show that government-backed mortgage programs play different roles across counties and borrower groups, making them an important part of the region's homeownership landscape.