The 3.75 Million Home Sales Low May Have Further to Fall
October’s closing count predates the worst mortgage rates. Treasury yields have eased, but buyers have recovered little of the payment shock.
October’s 3.75 million sales pace looks more like a waypoint than a bottom. The National Association of Realtors reported that annualized pace for October, the weakest since August 2010. The peak in mortgage rates has yet to pass fully through the closing data, and newer contract indicators are weakening.
November and December closings will draw more heavily on contracts exposed to October’s higher mortgage rates. The latest pending-listings data suggest buyers were still retreating as those rates peaked. Four weeks of falling rates have done little to repair their budgets.
Today’s sales report describes yesterday’s mortgage market
The 3.75 million figure extrapolates a month’s sales into an annual pace after seasonal adjustment. Sales fell 5.5% from September’s revised 3.97 million pace and 8.8% from a year earlier. The Reuters poll expected 3.85 million. It was a genuine miss, with a misleading timestamp.
Reuters and AP place the contracts behind closings one to two months earlier. October transactions were mostly negotiated in August and September, when Freddie Mac’s weekly 30-year mortgage rate ranged from 6.65% to 7.03%. October’s own average was 7.33%, and its weekly rate reached 7.40% on 8 and 15 October. Those mortgage quotes will weigh more directly on November and December closings.
The contract data already point down. NAR’s September pending home sales index fell 7.3% to 66.0, and Zillow reported October newly pending listings down 9.1% from September and 16.8% from a year earlier. Those distinct contract measures both point to continued weakness before the October rate peak has fully entered closing statistics.
Zillow’s October closed-sales estimate of 304,105 was down 8.7% from a year earlier, close to NAR’s 8.8%. That confirms the current slump, not the timing of the next one.
Treasury relief is reaching traders faster than homebuyers
The 10-year Treasury yield fell from 5.24% on 1 October to 4.84% on 5 November. Freddie Mac’s mortgage rate fell from its 7.40% October peak to 7.18% on 5 November, then reached 7.11% on 12 November. On 5 November, the quoted 30-year rate exceeded the same-day 10-year yield by 2.34 percentage points, up from 1.81 on 10 September. A wider gap means less of the bond-market relief shows up in quoted mortgage rates. It compares quoted rates; it is not a measure of lender profit.
On CTOL's arithmetic, the mortgage rate rose 64 basis points between 10 September and 8 October (6.76% to 7.40%). Of that, 27 matched the rise in the 10-year yield (4.95% to 5.22%). The remaining 37 came from a wider spread. Lower Treasury yields do not automatically produce cheaper loans.
On CTOL's arithmetic, a buyer putting 20% down on NAR’s $416,700 median existing home borrows $333,360 for 30 years. Monthly principal and interest would be $2,053 at October 2025’s 6.25%, $2,292 at October 2026’s 7.33%, and $2,243 at 7.11%. Taxes and insurance are extra.
On the same home, the drop from 7.33% to 7.11% saves $49 a month. That reverses only about one-fifth of the roughly $240 increase from 6.25% to 7.33%. This is a rate-only comparison, not an observed change in what borrowers pay. A rally in bonds buys little house.
Stocks have been quicker to celebrate. In a 2 November late-day snapshot, Rocket Companies rose 9%, while Compass and Zillow gained 8% each as Treasury yields retreated. Shareholders gained. Households still face the mortgage quote.
Builders are conceding on price while resale volumes carry the strain
NAHB’s October survey put builder confidence at 29, its lowest since June 2012. Builders are responding to weak affordability with price reductions and incentives: 43% reported cutting prices, with an average reduction of 6%, and 65% offered incentives. The survey shows how common concessions are, not their average dollar cost.
The national resale median still rose 0.4% from a year earlier as completed transactions fell 8.8%. The slowdown is more visible in completed volumes than in the headline price, even though Zillow found price cuts on 26.5% of listings. Inventory stood at 1.59 million homes, or 5.1 months of supply.
The buyer mix is changing. First-time buyers accounted for 29% of sales, down from 32% a year earlier; investors and second-home buyers increased to 20% from 16%. In the West, sales dropped 14.5% from a year earlier while the $624,300 median price was flat. A stable price can conceal a market in which fewer households get through the door.
The case that 3.75 million is already the bottom
Mortgage rates have fallen four consecutive weeks.
NAR economist Lawrence Yun says “there could be sizable pent-up demand if mortgage rates decline.”
Realtor.com economist Danielle Hale says, “I don't think this is a new normal for home sales”. Zillow’s Mischa Fisher expects a larger contraction in November followed by an easing slowdown. The market can turn before the closing data notice.
Consider the winter of 2022 and 2023. From 10 November 2022 to 26 January 2023, the mortgage rate fell from 7.08% to 6.13%. The pending index rose 2.5% in December, from 75.0 to 76.9, while January closings still declined 0.7% to 4.00 million. They then surged 14.5% to 4.58 million in February. A weak closings print did not prevent a recovery.
But the rate decline then was 95 basis points; this time it is 29. Supply was 2.6 months in February 2023 against 5.1 now.
In October 2023, a 3.79 million sales pace coincided with the month mortgage rates peaked at 7.79%. The lag is no stopwatch. Goldman Sachs economists Ronnie Walker and David Mericle forecast only 3.9 million existing-home sales in 2027 even with rates near 6.4% by year-end. That allows modest improvement without a strong recovery. None of it rules out October as the actual low if buyers begin signing again soon.
Contracts will tell us when a bottom becomes believable
The decisive early test comes Wednesday, 18 November, at 10:00 ET, when NAR releases October’s pending index. A rebound from September’s 66.0 would weaken the case for another sales decline. A further fall would reinforce the view that the October pace has not yet reached its floor. A lower reading would also leave the 9 December release of November closings exposed to another reading below 3.75 million. Until then, a bond rally is not a housing bottom. Believe the contract data before anyone declares one.