At the end of September 2025, the average 30-year fixed mortgage rate touched 6.22%, its lowest point in three years. Six weeks later it was back above 6.5%. That round trip explains why forecasts keep missing: rates do fall, but they fall in lurches, and they give ground back almost as easily as they take it.
So when will mortgage rates drop? Not on a date you can circle on a calendar. On a set of conditions that have to line up and then hold for a few months. Here is what those conditions actually are, which numbers to watch, and how to make a housing decision in the meantime rather than waiting for a moment that keeps sliding to the right.
The Fed doesn’t set mortgage rates, but it bends them
Your 30-year fixed rate isn’t set by the Federal Reserve. It tracks the yield on the 10-year Treasury, plus a spread that reflects what investors demand to hold mortgage-backed securities instead of government debt.
The spread is the part people overlook. Between 2010 and 2019 it averaged roughly 1.7 percentage points. In 2023 and 2024 it ran closer to 2.5, sometimes 3. That’s why mortgage rates stayed near 7% even during stretches when Treasury yields were easing. The bond market was charging extra for uncertainty.
For rates to genuinely drop, two things have to move at once: Treasury yields need to come down, and the spread needs to narrow. A Fed cut helps only indirectly, by lowering expectations for short-term rates over the next couple of years. If inflation data keeps surprising to the upside, a cut can even push mortgage rates higher, because investors start pricing in hotter inflation later.
The three numbers that decide the timing
Core inflation, month over month
Watch core PCE and core CPI, the versions that strip out food and energy. The year-over-year figure gets the headlines, but the monthly change is what moves bond desks. Three or four consecutive months of core inflation running at an annualised rate near 2% is the most reliable signal that mortgage rates have room to fall. We had a stretch like that in late 2025, and rates responded within days.
The 10-year Treasury yield
Mortgage rates follow this more closely than anything else. A 10-year yield that settles below 4% for a few weeks usually pulls the 30-year fixed toward the low 6s. Below 3.7% and you are looking at high 5s. You don’t need a trading terminal for this. Any finance site shows the yield, and it leads the weekly Freddie Mac survey rather than trailing it.
Payrolls and the unemployment rate
A softening labour market cools wage growth, which cools services inflation, which pulls yields down. The trigger isn’t one weak jobs report. It’s a trend: unemployment drifting up three or four tenths over several months, job openings falling, hiring slowing. That backdrop is what let rates fall in late 2024 and again in autumn 2025. What actually has to happen before mortgage rates go down is mostly this, in sequence.
A realistic timeline instead of a hopeful one
Most 2026 forecasts cluster around a 30-year fixed rate in the high 5s to low 6s by year end, with plenty of zig-zag on the way. Anyone promising 4% mortgage rates is describing a recession scenario, not a base case. What the 2026 market really points to is modest relief, not a return to 2021.
The pattern worth internalising: mortgage rates are forward-looking. They tend to fall before the Fed cuts and stall or reverse once the cuts arrive, because the cut is already priced in. Wait for a Fed announcement and you are usually acting after the move has happened.
Why the drop feels slower than the headlines suggest
Two frictions. Lenders price in the future, so a big improvement in the inflation data can show up in rates within days and unwind just as fast when the next report disappoints. The spread stays sticky too. Servicing costs, capital requirements and prepayment risk don’t vanish when the economy improves, and that keeps a floor under how far mortgage rates can fall relative to Treasuries.
There’s also a demand effect. When rates dip, applications jump, and lenders have less reason to sharpen pricing. Refinance demand swinging sharply in either direction is a decent real-time read on how attractive pricing has actually become.
If you’re buying anyway: a decision framework
Waiting has a cost, and it isn’t only the rent you keep paying.
- Price a one-point drop. On a $400,000 loan, falling from 6.75% to 5.75% saves about $265 a month. Falling to 6.25% saves about $130. Real money, but often smaller than a bidding-war premium on the right house.
- Compare payments, not rates. A 6.5% rate on a $350,000 house can cost less each month than a 6.0% rate on a $390,000 one. Rising inventory in your market may matter more than 50 basis points.
- Ask about a float-down. Many lenders let you lock now and renegotiate once, usually for a fee, if rates improve before closing. What happens if mortgage rates drop after you’ve locked explains the mechanics and the price of each option.
- Use a temporary buydown or ARM only with a plan. Either can bridge you to a refinance, but only if your income and timeline are steady.
The refinance math most people get wrong
Break-even is arithmetic: total closing costs divided by the monthly saving. If refinancing costs $5,200 and saves $190 a month, you need 27 months to get whole. Plan to move in two years and that isn’t a deal, it’s a donation.
Two refinements matter. Closing costs on a refi often run 2% to 3% of the loan, sometimes more if you roll them into the balance. And a no-cost refi isn’t free; you pay through a higher rate, which stretches the break-even in a different way. A practical target is a rate improvement of at least 0.75 to 1 percentage point, or a break-even under 30 months.
How to handle the rate lock without gambling
A 60-day lock protects you from a spike and blinds you to a decline. That’s a fine trade if you close in six weeks and a poor one if you close in five months. Floating is a bet on data you can’t control, and the downside, losing your payment, usually outweighs the upside.
A full walkthrough for buyers and refinancers, including when to lock and when to float, sits in this practical step-by-step guide for buyers and refinancers. The short version: lock when you can comfortably afford the payment at that rate, and buy the float-down option instead of banking on a drop you can’t predict.
What to watch in the next thirty days
Two releases will do more to answer the question than any forecast: the next core PCE print and the next payrolls report. If core comes in at 0.2% month over month and payrolls land under 100,000 with unemployment ticking higher, expect the 30-year to drift toward the low 6s, maybe below. If core runs 0.3% or more and hiring holds above 150,000, the current range holds steady.
Check the 10-year yield once a week, same day each time. Under 4% for two consecutive weeks means lenders are starting to compete and the spread is narrowing, which is the moment to collect quotes rather than read about rates. Pushing 4.5% means waiting costs you nothing and an early lock makes little sense. Write your target down in dollars per month, not in basis points. Rates will do what they do; the payment is the number you actually have to live with.