Freddie Mac put the 30-year fixed-rate mortgage at 6.95% for the week ending September 17, its highest point since January 2025. The rate a borrower was actually being quoted that week was closer to 7.2%. Both numbers are correct — they measure different things, and underwriting off the wrong one understates your cost by roughly 30 basis points.
What happened
The Fed raised rates the day before. On September 16 the FOMC voted 12-0 to raise the target range by 1/4 percentage point to 3-3/4 to 4 percent.
Mortgage rates did not rise because of that vote. Thirty-year mortgages track the 10-year Treasury plus a spread, not the federal funds rate. Bisnow reports the 10-year broke 5% for its first sustained period since 2007, and the spread on top of it widened to 1.97% from 1.92% the week before. A 10-year near 5% plus a spread near 2% puts you in the neighborhood of 7% — which is where the 30-year now sits.
Mike Fratantoni, chief economist at the Mortgage Bankers Association, made the point that long rates had already "baked in the expectation of hikes" before the Fed acted.
The survey detail: the 30-year jumped 19 basis points from 6.76% the prior week and stands 69 basis points above the 6.26% posted a year earlier. The 15-year averaged 6.26%, up from 6.09%. Freddie Mac chief economist Sam Khater said the 30-year continues to "fluctuate as markets assess economic data".
The headline rate lags your actual quote
Freddie Mac's survey is weekly, and the company describes the reading as an average of loan rates offered from the prior Thursday through the Wednesday before release. In a week when rates are moving, that lag is not a rounding error.
Redfin published both measures side by side: a daily average 30-year rate of 7.24% on September 16, against a weekly average of 6.76% for the week ending September 10. HousingWire's rate tracker carried the 30-year at 7.23% alongside its September 19 analysis.
So the number in the headline is an average of rates offered over the week that ended the day before it was published. It is not what your lender will quote you on Tuesday.
Why it matters for property managers
Your underwriting is off by about 30 basis points if you use the headline. Thirty basis points on a $2 million loan is $6,000 a year of debt service that is not in your model. On a portfolio with several refinancings queued up, that compounds into a real hole in next year's distributions.
The maturity calendar does not care what the Fed does next. Bisnow reports roughly $875 billion — about 17% of the $5 trillion in outstanding commercial mortgages — scheduled to mature in 2026, with 13% of multifamily mortgages among them. Loans written in the low-rate years are being refinanced into a 7% market. That gap is the single biggest line item most owners will face this cycle.
Coverage, not rate, is what kills a refinancing. Art Rendak, president of Inland Mortgage Capital, who heads the company's bridge lending program, described borrowers arriving at a 0.9 debt service coverage ratio — below the 1.0 line where the property covers its own debt. That is one lender describing the would-be borrowers who show up at its door for a refinance, not a market-wide statistic, but it is the shape of the problem: at 7%, a property that penciled at origination may no longer cover the new payment.
The demand side cuts the other way, and it helps you. Redfin put pending home sales at 299,126 for the four weeks ending September 13, down 5.4% year over year and the lowest level in nearly three years. Households that cannot buy at 7% do not disappear. They renew. Soft asking rents and firm occupancy can coexist, and right now they are.
What to do this week
This is general information, not legal or financial advice. Loan terms, and the rules that govern them, vary — confirm anything here with your lender and your own advisors before you act on it.
- Re-underwrite off a live quote. Ask each lender for today's indicative rate on the specific loan, and use that number. Retire the weekly headline from your models.
- Stress your coverage at 7.25%, not 6.95%. If DSCR lands under 1.20 at the higher number, you have a conversation to start now rather than 60 days before maturity.
- Pull the next 24 months of maturities into one page. With 13% of multifamily mortgages maturing in 2026, the question is which of yours, and in what month.
- Write scenarios with the conditions attached. Logan Mohtashami's base case of 6.50% to 6.75% with the 10-year back near 4.48% holds only if the conflict ends and oil moves lower; his 8% case requires the conflict to get worse, mortgage spreads to widen a bit further, economic data to stay solid, and the Federal Reserve to stay silent as the long bond heads higher. Neither is a forecast you can bank on — they are bookends for your budget.
- Do not plan on the Fed pulling mortgage rates down. It raised rates and long rates were already positioned for it. Mortgage pricing will follow the 10-year and the spread.
- Separate the debt problem from the leasing problem. Renter demand is being reinforced by the same rates that are hurting your refinancing. Price renewals to the occupancy you are actually holding.
What we're watching
- Next Thursday's Freddie Mac print, and whether 6.95% was a spike or a step.
- The spread. At 1.97% it is doing as much damage as the Treasury move, and it can narrow without any help from the Fed.
- Oil and the conflict. They are the hinge in both of Mohtashami's scenarios — every rate path this quarter runs through them.
- Whether the 10-year holds above 5%. Bisnow reports it had not sustained that level since 2007.
- Pending sales. If they keep sliding, the renter pool keeps growing, which is the one part of this that works in an operator's favor.