In many markets, shopping around for the best deal is a given. Everyone expects to negotiate for a car, and nobody buys the first house they see for the listed price. But a plurality of borrowers get only one mortgage quote, often based on the recommendation of their real estate agent. And most borrowers believe they would get the same interest rate from any lender.
But borrowers with the same characteristics (e.g., credit score, down payment, and loan amount) who take out loans on the same day actually get considerably different rates.
Our new analysis shows how this trend has persisted over the past decade—despite the availability of more shopping tools and information—and that the difference in rates the same person can get on the same day has increased. The difference in rates amounts to more than $100 per month for many borrowers. Regulators can reduce the burden on consumers to find and use this information and take some of the sting out of today’s high interest rates.
Borrowers with similar profiles are offered a wide range of mortgage rates
First, we analyzed the dispersion in mortgage rates for Fannie Mae and Freddie Mac borrowers who have similar credit scores and down payments. We looked at the interest rate dispersion within each cell of the Fannie Mae and Freddie Mac loan-level price adjustment (LLPA) grid and discarded mortgages with low dollar amounts or nonstandard features. Below, we show interest rate variation for the segments of the LLPA matrix that represent superprime and near-prime conventional mortgages, as well as Federal Housing Administration (FHA) borrowers, to illustrate the wide dispersion in mortgage interest rates across the market.
For each borrower profile and each day, we focused on the interquartile range (IQR) of rates these borrowers received, or the difference between the 25th percentile rate and the 75th percentile rate. We chose the IQR as the summary statistic because it roughly indicates the average return to shopping and negotiating for the half of borrowers who received an above-median rate.
Our interpretation also assumes borrowers are unlikely to get to the lowest rate they could, and that is why we are comparing with the 25th percentile rate and not the lowest rate that day. We analyzed data from Optimal Blue, which captures close to one-third of all mortgage transactions at the time the borrower locks their rate.
The IQR for superprime and near-prime mortgages has hovered around 50 basis points on any given day since mid-2022.This is more than twice as wide as the 20 basis-point IQR on superprime mortgages from 2013 to 2021. For a $400,000 30-year mortgage, a 50 basis-point difference amounts to a little more than $100 a month. The IQR for FHA borrowers with 3.5 percent down payments and credit scores below 660 did decrease slightly from a spike in 2020, but it is still about 60 basis points going into 2024.
Variation in interest rates is wider for loans with lower credit scores. In 2023, the median daily IQR on interest rates locked on conventional loans with loan-to-value (LTV) ratios from 90 to 95 percent and FICO scores of at least 780 was 48 basis points, while the IQR on interest rates with the same LTV range and with FICO scores from 680 to 699 was 60 basis points.
Borrowers With Lower Credit Scores Experience Wider Interest Rate Variation
| FICO SCORE | LTV | ||
| 80-90 | 90-95 | >95 | |
| 780+ | 48 | 48 | 61 |
| 760-779 | 45 | 47 | 61 |
| 740-759 | 49 | 48 | 61 |
| 720-739 | 51 | 50 | 61 |
| 700-719 | 59 | 52 | 62 |
| 680-699 | 60 | 60 | 58 |
Sources: Optimal Blue interest rate lock data and Urban Institute calculations.
Note: Includes only 30-year fixed-rate purchase loans on single-family properties and conforming loan amounts above $150,000.
Our results are similar to what the Consumer Financial Protection Bureau (CFPB) found using a different dataset (regulatory Home Mortgage Disclosure Act data) and a different aggregation method, as well as Philadelphia Federal Reserve Bank research (which also uses Optimal Blue data but focuses on a somewhat different metric).
As the CFPB pointed out, a 50 basis-point difference in rates is higher than the FHA’s February 2023 30 basis-point price drop in mortgage insurance premiums and is higher than the difference in credit guarantee cost between a superprime borrower (with a credit score of at least 780) and a subprime borrower (620 to 639).
In other words, shopping around and negotiating could be as important for a borrower’s mortgage rate as a pristine credit history. As the CFPB and the Philadelphia Federal Reserve Bank research shows, higher mortgage rates do not appear to be related to customer satisfaction with the lender or with the mortgage transaction.
Moreover, this difference appears to be growing, especially in the past two years, which is consistent with results from Freddie Mac. We don’t have a good explanation for this. Theoretically, rate differences in the pandemic period could be explained by lenders rationing the historically large volume of borrowers who were all trying to get low interest rates. But the rate dispersion is actually higher after the period of low interest rates and high mortgage demand. The dispersion could also reflect large intraday moves in interest rates, but it remains higher as interest rate volatility has declined.
These results are robust. We analyzed the differences between the 90th percentile and the 10th percentile rates. That difference has a similar dynamic as the IQR plots above, yet the dispersion is wider, with 100 basis-point differences common across borrower profiles.
Transparency in interest rate distribution could bring down consumer costs
Right now, consumers can use the CFPB’s interest rate tool to determine whether they are getting a good deal on interest rates and how much they could save per month by going to a cheaper lender (though the tool does not include nondepository mortgage lenders, which compose 83 percent of agency mortgages as of May 2024).
But many consumers don’t know about this tool, and most don’t even know that a tool like this could exist since they believe rates are the same across lenders. Rather than leaving it to consumers to find it and to learn there is price dispersion, the government-sponsored enterprises, the FHA, and the CFPB could require each lender to make sure borrowers receive a rate dispersion disclosure, clearly stating how much lower the monthly mortgage bill could be if the consumer chose a more competitive lender. The CFPB could adapt its rate dispersion tool for this purpose by adding the tool’s output to the TILA-RESPA integrated disclosures, expanding it to include nondepository lenders and updating it daily.
Further, Fannie Mae, Freddie Mac, the FHA, and the US Department of Veterans Affairs already have this information on a lender-by-lender basis. They could encourage lenders whose pricing behavior is consistently out of line to reconsider their offerings.
Saving $100 in monthly costs is not insignificant for many Americans. Equipping borrowers with more information about their options could improve their financial standing and make homeownership more attainable.
Tune in and subscribe today.
The Urban Institute podcast, Evidence in Action, inspires changemakers to lead with evidence and act with equity. Cohosted by Urban President Sarah Rosen Wartell and Executive Vice President Kimberlyn Leary, every episode features in-depth discussions with experts and leaders on topics ranging from how to advance equity, to designing innovative solutions that achieve community impact, to what it means to practice evidence-based leadership.