This article explains public financial information. It does not recommend buying, selling or holding any investment.
New home sales count contracts, not completed closings
The monthly New Residential Sales report is produced jointly by the U.S. Census Bureau and the Department of Housing and Urban Development. It estimates sales of new single-family houses nationally and across four regions, together with homes for sale, months' supply, prices and the stage of construction.
A new house is treated as sold when a sales contract is signed or a deposit is accepted. The home may be completed, under construction or not yet started. The survey does not follow the transaction through closing, so the measure is closer to a timely signal of buyer-builder agreements than a count of families moving into finished homes.
What the report includes—and leaves out
The statistics cover newly built, privately owned single-family houses, including detached homes, side-by-side semidetached homes, row houses and townhouses. A unit is included only when it is separated from neighbouring units by a ground-to-roof wall and has no unit above or below it.
Existing-home resales are outside the survey, as are most apartment units in multifamily buildings. That distinction matters because new homes are only one part of the housing market. A strong new-home number can coexist with weak resale activity when builders offer incentives or have inventory available where existing owners are reluctant to list.
The headline is a seasonally adjusted annual rate
The headline sales figure is normally reported as a seasonally adjusted annual rate, or SAAR. It is not the literal number of houses sold during the month. Instead, the Census Bureau adjusts the monthly pace for recurring seasonal patterns and expresses the result as the rate that would prevail over a full year if that month's pace continued.
For example, June 2026 new-home sales were reported at a SAAR of 628,000. That does not mean 628,000 new homes changed hands in June. The release's unadjusted tables provide estimated monthly counts, while the annualised headline is designed to make differently seasonal months easier to compare.
Margins of error can be larger than the monthly move
New-home sales are sample-based estimates and can be volatile. The June 2026 headline was 1.6% above the revised May rate, but the published 90% confidence interval was plus or minus 14.8 percentage points. Census marked the change with an asterisk because it could not conclude that the movement was statistically different from zero at that confidence level.
The same discipline applies to year-over-year comparisons, regional changes and price estimates. A dramatic percentage can still be statistically inconclusive when the uncertainty range is wide. Revisions are also normal as additional survey information becomes available.
Inventory and months' supply add essential context
The report estimates how many new homes are for sale at the end of the month and converts that stock into months' supply at the current sales rate. In June 2026, the estimated inventory was 485,000 homes and months' supply was 9.3. The ratio answers a hypothetical question: how long would the listed supply last if the current annualised sales pace continued and no new homes were added?
Months' supply can rise because inventory increased, because sales slowed, or both. It can fall because buyers absorbed homes faster, because builders reduced available inventory, or both. Reading the numerator and sales pace separately prevents a ratio from hiding the mechanism.
Median price changes can reflect the mix of homes sold
The release publishes median and average sales prices for new houses sold. In June 2026, the estimated median was $398,300 and the average was $475,400. The median is the midpoint of transactions, while the average is more sensitive to very expensive sales.
Neither series is a constant-quality house-price index. If one month contains a larger share of smaller homes, lower-cost regions or builder incentives, the median can fall even when the price of an otherwise identical home did not. For price trends, compare several months and use repeat-sales or constant-quality measures where appropriate.
Mortgage rates matter, but the transmission is not mechanical
Higher mortgage rates generally raise the monthly payment associated with a given loan amount, reducing affordability for some buyers. Builders may respond through price changes, mortgage-rate buydowns, closing-cost support, smaller floor plans or a different mix of locations. Those responses can alter sales, inventory and reported prices in different ways.
Rates are not the only influence. Employment, household formation, land and construction costs, local supply constraints, credit standards and the availability of existing homes also shape demand. A same-month correlation between mortgage rates and sales does not by itself prove causation.
A six-step way to read each release
First, record the headline SAAR and both the monthly and annual change. Second, compare each change with its confidence interval. Third, note revisions to earlier months. Fourth, read homes for sale and months' supply together. Fifth, inspect median and average prices without assuming constant quality. Sixth, compare the national result with regional estimates while respecting their wider uncertainty.
Finally, place the report beside housing starts, building permits, mortgage rates and existing-home data. The goal is not to force every indicator into the same direction. Differences can reveal whether the constraint is demand, completed supply, construction timing or the mix of properties available.
Editorial boundary
This article explains the official methodology and how to interpret the New Residential Sales release. The cited June 2026 figures are the latest published observations available when this guide was prepared; they can be revised and should be checked against the current Census release.
The information is for education only. It does not predict home prices, mortgage rates, construction-company earnings or securities markets, and it is not a recommendation to buy, sell or hold any investment or property.
How the main new-home-sales measures differ
| Measure | What it tells you | Common mistake |
|---|---|---|
| Sales SAAR | Seasonally adjusted annualised contract pace | Reading it as the literal monthly count |
| Homes for sale | Estimated new-home inventory at month-end | Assuming every home is completed |
| Months' supply | Inventory relative to the current sales pace | Ignoring which side of the ratio changed |
| Median price | Midpoint of new homes sold | Treating it as a constant-quality price index |
| Monthly change | Estimated movement from the prior revised month | Ignoring its confidence interval |
Frequently asked questions
What does the new home sales report measure?
It estimates signed contracts or accepted deposits for newly built single-family houses, along with inventory, months' supply, prices and regional sales.
Is the headline new home sales number the number sold that month?
No. The headline is usually a seasonally adjusted annual rate—the monthly pace adjusted for seasonality and expressed as if it continued for a year.
When is a new house counted as sold?
The Census Bureau counts it when a sales contract is signed or a deposit accepted, even if construction has not started or the transaction has not closed.
What does months' supply mean in housing?
It estimates how many months the available new-home inventory would last at the current annualised sales pace if no additional homes entered the market.
Why are new home sales revised?
The figures come from a sample survey. Later responses and updated seasonal factors provide more information, so prior estimates can change.
Does a lower median new home price mean every home became cheaper?
No. The median can change because the mix of homes, regions and price tiers sold changed. It is not a constant-quality price index.
Official sources
Definitions and methodology were checked against these primary resources. Consult the current documents for complete details.
U.S. Census Bureau and HUD — New Residential Sales, June 2026 ↗U.S. Census Bureau — New Residential Sales Program ↗U.S. Census Bureau — Economic Indicators Definitions ↗U.S. Census Bureau — Survey of Construction Release Schedule ↗Unsplash — New Home Construction by Troy Mortier ↗.jpg?width=1800)
