30-Year Mortgage Average Reaches 7.03% as September Forecasts Break
The 30-year fixed-rate mortgage averaged 7.03% in the United States for the week ending Thursday, September 24, 2026, according to Federal Reserve Economic Data.
That print caps a steady climb through September. The Wall Street Journal reported 6.91% on September 8. Freddie Mac's survey series put the average at 6.76% for the week ending September 10 and 6.95% for the week ending September 17, according to Freddie Mac and its archive. On a week-to-week basis, that is 19 basis points higher into the September 17 week, then another 8 basis points to the 7.03% print for the September 24 week. The prints come from different series. The direction does not.
The benchmark behind the move rose in parallel. The 10-year Treasury yield briefly climbed back above 4.8% on September 8, according to CNBC. It hit 5% around September 14, according to CNN. The mortgage move followed the Treasury move by days, not weeks.
September forecasts did not capture it. Real estate and mortgage expert Judi Kutner at Gorilla Movers predicted in MarketWatch's September outlook that the average 30-year fixed would stay between 6.5% and 6.8%. That outlook was published September 4. An earlier MarketWatch compilation from September 5, 2025 quoted analyst McBride with the same 6.5% to 6.8% range. Both ranges now sit below realized September averages.
The broader context here is pricing, not just levels. For agency MBS desks, a 20-basis-point-plus backup in the primary rate over roughly two weeks forces duration extension, pipeline mark-to-market, and wider primary-secondary spreads to compensate for volatility and fallout risk. Lenders reprice rate sheets intraday. Borrowers in process face lock expiration arithmetic. Rate sheets tell the story first. Volumes confirm it later.
Looking at what this means for origination, the speed matters more than any single print. A gradual drift allows lock decisions and seller concessions to adjust. A 27-basis-point gap between the September 10 survey week and the September 24 FRED week compresses that adjustment window. It raises monthly payment sensitivity on new purchases, impairs refinance incentive almost entirely above 7%, and keeps existing low-coupon inventory locked in. That is a quantity effect as much as a price effect. For portfolio managers, higher coupons change prepayment assumptions and extend expected cash flows. That repricing feeds back into secondary execution.
In my view, the forecast miss is instructive. Anchoring September expectations to 6.5% to 6.8% assumed range-bound Treasuries and mean-reverting spreads. Once the 10-year broke through 4.8% and then 5%, that anchor failed. Desk quants will focus less on whether 7.03% holds and more on convexity hedging flows and MBS liquidity if volatility persists. The error was directional, and direction drives hedging.
