New-home buyers had 8.4 months of supply to choose from in September, Census and HUD reported on October 27. That is a national, seasonally adjusted figure for new single-family houses, set at the pace they were selling that month. A month earlier the supply was 8.5 months, and a year earlier it was 8.2.
How 476,000 houses for sale turns into 8.4 months
Months’ supply, Census’s definitions page says, is the ratio of houses for sale to houses sold. It shows how long current inventory would last at the current sales rate if no new houses were built. In September that meant 476,000 houses for sale set against sales running at an annual rate of 676,000.
The release puts the result as “a supply of 8.4 months at the current sales rate.” Both inputs are seasonally adjusted, and the annual rate is the seasonally adjusted monthly value multiplied by 12. The supply figure therefore describes a pace, not a forecast of how long the inventory will really last.
For a buyer planning a winter move, the question is whether 8.4 months is a lot or a little, and the release supplies no benchmark. It does supply the direction: 8.5 months in August and 8.2 months in September 2025. Anyone reading a trend into those three numbers is reading past what the margins allow.
Retirement deadlines, in your inbox: Miss an enrollment or claim deadline and the money can be gone. Join the Retirement Money Brief and we’ll send you our Property Tax & Utility Relief Finder as a welcome gift.
Get the Brief + the relief finder →
We may earn a commission from links on this page.
Why supply barely moved while both inputs did
Houses for sale fell to 476,000 from 481,000 in August, a drop of 1.0 percent with a margin of plus or minus 0.9 percent. That is the clearest move in the release, because the change is larger than its margin. Sales slipped 0.6 percent in the same month, with a margin of plus or minus 13.8 percent.
Inventory fell a little faster than sales, so the ratio eased from 8.5 to 8.4. Over the year, houses for sale fell 2.3 percent from 487,000 while the sales rate fell 5.3 percent from 714,000. Sales dropped faster than inventory, which is why supply edged up from 8.2 to 8.4 months.
Those yearly comparisons carry wide margins. Census gives plus or minus 4.0 percent on the inventory change and plus or minus 12.5 percent on the sales change, both larger than the changes themselves. Nothing in the pairing says that buyers gained or lost bargaining room in September.
A 2.4 percent rise with a 14.1 percent margin of error
The supply figure rose 2.4 percent against September 2025, and Census attaches a margin of plus or minus 14.1 percent. The monthly dip of 1.2 percent carries plus or minus 14.7 percent. Both margins dwarf the changes, so the 8.2, 8.5 and 8.4 are best read as roughly the same number.
Census’s methodology page explains why estimates carry such ranges. Sampling error reflects the fact that only a particular sample was surveyed rather than the entire population, and Census builds 90 percent intervals around each estimate.
Nearly all of the doubt in the supply figure comes from the sales side. The monthly change in houses for sale carries a margin of plus or minus 0.9 percent, while the sales rate carries 13.8 percent. Because the annual rate is the monthly value times 12, 676,000 works out to roughly 56,300 sales a month, and it is that sales estimate, not the count of houses for sale, that leaves the 8.4 open to question.
A ratio divides one estimate by another, so error in either input flows into the result. That fits the supply change’s 14.7 percent margin, which is wider than the 13.8 percent on sales alone. Census also marks all of it preliminary: three months of data are revised with each month’s release.
National new-house supply is not the supply on any one street
The release is a national total of new single-family houses. It says nothing about a metro area, a price band or a particular builder’s lots, and a buyer in one county can face far fewer or far more houses than 8.4 months implies.
New houses are also only part of what a buyer can choose from. The report counts houses for sale in new construction, not resale houses, so the 8.4 months says nothing about existing-home inventory.
The 476,000 figure is the stock and the 676,000 pace is the flow. Months’ supply is the first divided by the second, and a move in either one shifts it. A buyer comparing winter against spring would be watching two estimates that stay open to revision for three months after they first appear.
Whether a buyer’s winter is in the numbers yet
The September release says new-house inventory held near eight and a half months of sales at a slightly slower pace. It supports no claim that buyers gained or lost leverage during the month, and it names no cause for the small shifts.
The open question is the one the margins leave open: whether 8.4 months is a pause in a slow rise from 8.2 or noise around a flat level. Later monthly releases will revise September and show which.
Is this a buyer’s winter?
Eight months of national supply says little about the cost of any one house, owned or for sale. A winter move still comes down to what the current house costs to hold and what borrowing against it would run.
Housing statistics have no product of their own, so the turn here is to the owner’s side. Cut Your House Costs covers home equity options compared, from a HELOC to a reverse mortgage, and the home-cost audit worksheet.
The guide runs 11 pages. Its Compared section sets the equity options side by side, and its Worksheet section is the audit.
Get the home equity comparison and the cost audit before a winter move →
This article was produced with AI assistance and edited for accuracy against the sources linked above.
