Introduction

This research investigates how informational and search frictions grant financial intermediaries market power over consumers in the United States mortgage market. The study focuses on a fee waiver programme implemented by the Government-Sponsored Enterprises (GSEs), Fannie Mae and Freddie Mac, which was designed to improve housing affordability for lower-income first-time homebuyers. While the mortgage market is traditionally assumed to be highly competitive due to low concentration and product homogeneity, the analysis uncovers a significant deviation from this benchmark. The report estimates that lenders pass through less than 40% of the value of these fee waivers to the intended beneficiaries, suggesting substantial competitive frictions.

Background and theoretical framework

To support low-income homeownership, the Federal Housing Finance Agency (FHFA) directed the GSEs to waive upfront fees, known as Loan-Level Pricing Adjustments (LLPAs), for first-time homebuyers with incomes below certain thresholds—typically 100% or 120% of the area median income (AMI). These waivers represent a significant cost reduction for lenders, with an average present value of approximately $2,500 per loan. However, lenders are not explicitly required to inform borrowers of their eligibility or the existence of the waiver. The report utilises a theoretical model of limited search to show that when a portion of consumers fails to shop around for the best rate, lenders can exert market power and extract rents. Survey data cited in the report indicates that nearly half of low-income first-time homebuyers consider only one lender and are heavily reliant on referrals from real estate agents, who may not always have aligned incentives.

Data

The researchers combined two primary datasets to conduct their analysis: confidential mortgage origination records from the Home Mortgage Disclosure Act (HMDA) and public loan-level data from the GSEs. The HMDA data provides detailed characteristics of the borrower and the mortgage, including race, income, credit score, interest rate, and loan-to-value (LTV) ratio. The GSE data allows for the precise identification of first-time homebuyers and those loans sold to Fannie Mae or Freddie Mac. The final matched dataset represents the near-universe of the US mortgage market for the years 2022 and 2023, enabling a robust evaluation of policy impacts across various demographics and lender types.

Design

The study employs a difference-in-differences identification strategy, comparing eligible borrowers to a control group of homebuyers who did not qualify for the waiver because their income was slightly above the threshold. A key component of the design is the adjustment of the outcome variable—the mortgage interest rate—for upfront fees or ‘points’ paid by the borrower. The report notes that failing to account for points would significantly underestimate the pass-through rate, reducing it from 39% to 28%. The researchers also utilised a regression discontinuity design and triple-difference estimators to ensure the results were not biased by strategic income reporting or general market trends.

Results

The analysis establishes a baseline pass-through rate of only 39% for the targeted fee waivers. This indicates that mortgage intermediaries captured over 60% of the $1.5 billion in total waivers provided during 2023–2024. The benefits were found to vary systematically across different groups. Black borrowers experienced the lowest benefits, with only approximately 20% of the waiver passing through to them. Furthermore, the report finds significant heterogeneity based on lender type: the 10 largest non-banks exhibited a pass-through rate of less than 30%, which is roughly half that of traditional banks and credit unions.

The research also examines the role of individual loan officers. By categorising loan officers by their ‘expensiveness’—based on their historical pricing independent of the lending institution—the study found that those in the most expensive quintile passed through less than 20% of the waiver savings. This suggests that loan officers retain meaningful pricing discretion and incentives to steer less-informed borrowers into higher-rate mortgages, despite post-crisis regulatory efforts to limit such practices.

Conclusion

The findings underscore that competitive forces alone are insufficient to discipline mortgage pricing in a market characterised by search frictions and information asymmetries. Subsidies delivered through intermediaries may fail to reach the intended recipients when pricing discretion is pervasive. The report suggests that policies aimed at improving housing affordability may require alternative approaches, such as greater transparency in fee structures, enhanced consumer literacy programmes, or more stringent restrictions on intermediary pricing discretion to reduce the scope for capture by mortgage lenders and loan officers.