Narrower mortgage spreads barely improved housing affordability
Mortgage spreads delivered most of the decline forecasters expected. A rising 10-year Treasury yield canceled the gain before borrowers received much relief.
See the accompanying note for methodology, base-year sensitivity and data provenance.
For two years, many housing forecasters expected narrower mortgage spreads to lower mortgage rates and improve housing affordability. The mortgage spread narrowed. By July 2026, it had closed four-fifths of the gap between its June 2023 peak and historical median. Yet the median home remained almost as unaffordable as when the spread peaked.
The 10-year Treasury yield explains why: it rose while the mortgage spread fell, leaving the 30-year mortgage rate only 0.20 percentage points lower. Borrowers received little of the expected relief. Most of the excess spread above the historical median has now disappeared, so further relief depends mainly on a lower Treasury yield. The forecast has moved beyond mortgage-market conditions to the forces that set long-term yields: expected short rates, inflation and growth, Treasury supply, and the term premium.
Median-home costs exceed 40% of income
The Atlanta Fed's Home Ownership Affordability Monitor asks whether a median-income household can afford a median-priced home. To estimate the annual cost, it assumes a 10% down payment and a 30-year mortgage, then adds principal, interest, property taxes, homeowners insurance and private mortgage insurance. A 40% reading means that estimated home costs consume 40 cents of each pretax dollar the household earns. The measure uses 30% as its affordability threshold, so 40% marks an extreme burden.
The measure was near 30% in 2019, then moved above 40% during 2022 as home prices remained high and mortgage rates rose. Its latest uninterrupted run began in March 2023. Every monthly reading from then through the latest data, March 2026, exceeded 40%. That makes 37 consecutive months, compared with 22 during the 2000s housing bubble.
The measure peaked at 45.3% in October 2023. In the first quarter of 2026, it still averaged 41.2%: 11.2 cents of each pretax dollar above the affordability threshold.
The measure describes access to the median home, not the finances of households that complete purchases. Redfin found that 32.6% of 2024 purchases across 40 large metros were all-cash, while existing owners can use accumulated equity. The mortgage lock-in effect also changes which owners sell and which households transact. The measure therefore captures exclusion pressure on marginal buyers more directly than payment stress among completed transactions.
High prices now coincide with high mortgage rates
Earlier affordability extremes had different dominant drivers. Rates dominated in 1987: the 30-year mortgage averaged above 10%, but home prices were low relative to income. Prices dominated in 2006: home prices had outpaced income while the mortgage rate averaged 6.4%. Today combines both.
The latest complete annual data end in 2024. By then, the Case-Shiller home-price index had reached 4.85 times its 1987 level, while median household income had reached only 3.21 times its 1987 level. The mortgage rate averaged 6.7%.
A consistent payment comparison assumes a 20% down payment and includes only principal and interest. Under that calculation, the median-home payment consumed 30.4% of median income in 2024, compared with 29.2% in 2006.
Although the payment burden exceeded 2006 by only 1.2 points, no earlier year combined a record Case-Shiller index level with a mortgage rate above 6%.
Mortgage spreads can narrow without a rate cut
That 6.7% mortgage rate was not a single price. It combined the 10-year Treasury yield, the baseline cost of long-term money, with a mortgage spread that compensates lenders and investors for additional costs and risks. In shorthand:
From 2019 to 2024, the mortgage rate rose 2.79 percentage points, with the Treasury yield driving roughly three-quarters of the increase and a wider mortgage spread supplying the rest. That smaller share mattered because it could reverse without a policy-rate cut.
Mortgage investors face a particular risk: when rates fall, homeowners can refinance and end a stream of higher interest payments. More volatility makes that choice more valuable to homeowners and more costly to investors, who respond by demanding additional yield. The MBS-Treasury option-adjusted spread measures the extra yield after models account for that embedded option.
During 2022 and 2023, rate volatility remained high, Federal Reserve balance-sheet runoff reduced Fed purchases of agency mortgage securities, and banks reduced their demand after SVB. Private buyers required more yield to hold the securities. The option-adjusted spread widened even as originator margins narrowed, making investor pricing the source of the wider mortgage spread.
Private demand could recover and rate volatility could decline without a policy-rate cut. Historically, the option-adjusted spread returns toward its median faster than the Treasury yield.
If the mortgage spread narrowed while the Treasury yield remained high, would housing affordability improve?
Spreads narrowed, but mortgage rates barely fell
The answer is now clear. The mortgage spread narrowed from 2.97 percentage points in June 2023 to 1.93 points in July 2026, closing four-fifths of its gap to the historical median. But the Treasury yield moved in the opposite direction. Its 0.83-point rise offset most of the 1.03-point spread decline, so the mortgage rate fell only 0.20 points.
For borrowers, that offset matters more than either rate component. Holding 2024 home prices and income constant, the small mortgage-rate decline reduced the modeled principal-and-interest burden from 30.4% to 29.75% of income.
The broader Atlanta Fed measure moved in the same direction, though home prices, income, taxes, insurance and private mortgage insurance also affect it. The measure fell from an average of 43.3% in 2024 to 41.2% in the first quarter of 2026, reversing only about 15% of the 14.3-point deterioration from 2019 to 2024.
The new-home market should have been the exception. Builders subsidized mortgage rates below the headline rate, giving buyers relief that the national average did not capture. Yet new single-family sales declined roughly 9% while the mortgage spread narrowed.
Housing starts remained flat, and existing-home sales stayed between roughly 4.0 million and 4.3 million annualized without a sustained increase. These figures do not prove that affordability caused weak activity, but they show that narrower mortgage spreads did not coincide with a broad housing recovery.
The Treasury offset consumed most of the spread decline: borrowers received little relief, and housing activity showed no recovery.
Further relief needs a lower Treasury yield
Only 28 basis points now separate the mortgage spread from its historical median. The spread remains at the 71st percentile, so the correction is incomplete but the remaining gap is small. If that gap closed while the Treasury yield remained unchanged, the mortgage rate would fall to 6.23% and the fixed principal-and-interest burden to 28.9% of median income. That small final move would help more than the previous 1.03-point decline because this calculation keeps the Treasury yield steady. Even so, it would reverse only about one-fifteenth of the 12.7-point deterioration since 2019.
Further mortgage-rate relief now depends mainly on a lower 10-year Treasury yield. The question therefore shifts from mortgage-market pricing to the forces that set long-term yields: expected short rates, inflation and growth, Treasury supply, and the term premium.
Affordability can still improve through lower home prices, higher incomes, or a credit event that reprices both homes and rates. Each path requires separate evidence, and this analysis predicts none of them. The remaining mortgage-spread correction cannot restore affordability by itself.
Most of the mortgage-spread correction has occurred. Most of the affordability loss remains.