When the Fed cut rates to near-zero in 2020, mortgage approval rates improved everywhere — but not by the same amount. Some states saw denial rates fall by more than 10 points. Others barely moved. When the Fed reversed course in 2022, the same pattern repeated: the states that had gained the most lost the most.
We tested whether a state's 2018→2020 "relief" (how much denial rates fell during the rate-cut boom) predicts its 2021→2023 "shock" (how much denial rates rose during the rate-hike bust). The result is one of the cleanest patterns in this entire dataset.
This isn't a fluke driven by one or two outlier states. We tested it three ways: removing Puerto Rico (r stays at 0.68), removing both Hawaii and California (r stays at 0.69), and running 1,000 random shuffles of the data to see how often a correlation this strong appears by chance — the answer was never. The real relationship sits far outside anything random noise produces.
| State | 2018→20 Relief | 2021→23 Shock |
|---|---|---|
| California | −11.4pp | +13.6pp |
| New Jersey | −10.7pp | +11.4pp |
| Maryland | −10.4pp | +11.2pp |
| Delaware | −10.0pp | +9.2pp |
| New Hampshire | −9.7pp | +9.7pp |
| District of Columbia | −9.5pp | +11.3pp |
| Illinois | −9.5pp | +9.1pp |
| Virginia | −9.5pp | +11.4pp |
| State | 2018→20 Relief | 2021→23 Shock |
|---|---|---|
| Wisconsin | −5.4pp | +5.3pp |
| Minnesota | −5.7pp | +6.4pp |
| Kansas | −5.8pp | +7.1pp |
| Puerto Rico | −3.9pp | −1.4pp |
| Iowa | −3.2pp | +3.8pp |
| North Dakota | −2.6pp | +3.7pp |
Puerto Rico is the one state that broke the mirror — it gained modestly during the boom (−3.9pp) and then kept improving even as rates rose (−1.4pp), the only territory in the dataset that didn't reverse. Every other top/bottom state followed the pattern almost exactly.
The mechanism is straightforward once you see the pattern: in high-cost markets like California, New Jersey, and DC, loan amounts are large relative to income, so a borrower's debt-to-income ratio is extremely sensitive to interest rate changes. A 4-point rate swing changes the math on a $700,000 loan far more than it does on a $180,000 loan. In cheaper markets — Iowa, North Dakota, Wisconsin — the same rate swing barely moves the qualifying threshold, because the loan-to-income ratio was never as stretched to begin with.
That's the real story behind the headline volume collapse: it wasn't a uniform national event. It was concentrated almost entirely in the markets that were most leveraged to begin with — and those markets reacted symmetrically in both directions.
⚡ Analyze a filePrices and rates are widely reported. Whether a lender says yes is not. In the complete 2025 federal record, denial rates across the 100 largest FHA lenders ran from 1.8% to 78.7% — same programme, same year.
And it is not simply who applies where: standardizing on state, loan amount, income, debt-to-income and loan-to-value, applicant mix explains only a 2.7× range in expected outcomes.
CFPB HMDA 2025, computed by FinanceRateCalc. Covers the highest-volume lenders published per market, not all lenders. Historical observations, not predictions. CC BY 4.0, not independently reproduced.