Key takeaways
- May 2026 new home sales ran at a 580,000-unit annual rate.
- A sale is recorded at contract or deposit, not necessarily completion.
- Monthly estimates are volatile, revised, and not local demand forecasts.
Source record
12 cited sources
Last verified
2026-07-23
In this article
U.S. sales of new single-family houses were estimated at a seasonally adjusted annual rate of 580,000 in May 2026. The Census Bureau and HUD publish the New Residential Sales estimate, and FRED distributes this line as HSN1F. It measures contract activity under a specific survey definition,not 580,000 closings during May and not the number of new rental homes purchased.
U.S. new home sales statistics
| Observation month | Published annualized rate |
|---|---|
| May 2026 | 580,000 |
| April 2026 | 626,000 |
| March 2026 | 664,000 |
| February 2026 | 630,000 |
| January 2026 | 576,000 |
The May rate was below April's estimate. Monthly New Residential Sales figures are subject to substantial sampling variability and revision, so the difference should not be called a settled downturn. Values were checked July 23, 2026.
Understanding “sold” and “annual rate”
A new house is counted as sold when a sales contract is signed or a deposit is accepted. The house may be not started, under construction, or completed. This timing makes new home sales an earlier signal than a series based only on completed closings, but it also means cancellations and build timelines need separate attention.
“Seasonally adjusted annual rate,” or SAAR, describes the annual pace implied by the month's seasonally adjusted activity. It is not the actual number sold in that month. Dividing or summing annualized monthly rates without a documented conversion produces misleading totals.
The program focuses on new single-family houses. Existing-home sales, apartment-building transactions, and rental lease-ups are outside its scope. Comparisons with the National Association of Realtors' existing-home series must preserve those differences in universe and timing.
Method used here
PortfolioRental retained the five latest non-missing HSN1F observations available July 23, 2026. We preserved the published monthly dates and thousand-unit SAAR values, displaying 580 thousand as 580,000 for readability. We did not estimate unadjusted monthly transactions, smooth the line, or project June.
To audit the table, download HSN1F from FRED and compare it with the Census release tables. Save the reported sampling-error information and the release vintage. The ALFRED HSN1F archive is useful when a revised series no longer matches an earlier memo.
What can change the sales pace
Demand is only one part of the story. Mortgage conditions, prices, incentives, available lots, construction capacity, community openings, product mix, and buyer confidence can all affect contracts. A builder may preserve sales using a rate buydown or closing-cost support rather than a visible base-price reduction.
Supply matters too. A market with few completed homes may constrain sales despite willing buyers, while a large inventory of quick-move-in homes can support transactions. The national new home supply statistics place for-sale inventory relative to the sales pace, but the ratio still cannot identify conditions in one subdivision.
Cancellations complicate local interpretation because a contract can fail after it first enters the process. Teams reviewing builder competition should ask how cancellation information is defined, whether it is gross or net of resales, and what period it covers. Do not infer a national cancellation rate from HSN1F alone.
A corridor-level research plan
For an acquisition near active construction, map every competing community within the tenant and buyer search area. Record builder, opening date, floor plans, price range, lot count, construction stage, delivery window, advertised incentives, and the date each field was checked. Distinguish a planned lot from a home actively offered and a contract from a completed closing.
Next, compare county permit and completion data with deed records and builder availability. These sources answer different stages of the pipeline. Permits can be abandoned or delayed; completions can reflect contracts signed earlier; recorded deeds may lag. Reconciliation is more informative than choosing whichever count supports the transaction.
Finally, measure rental competition directly. New owner-occupied subdivisions can expand the future resale stock without adding rental inventory. Build-to-rent communities can compete immediately for tenants. Identify tenure, product, lease-up stage, advertised rent, concessions, and occupancy rather than assuming every new house affects a rental in the same way.
Portfolio decisions the series can support
At screening, HSN1F can flag whether national builder contract activity merits a closer look. During diligence, its proper role is to generate questions about incentives, new supply, and exit competition. In portfolio reviews, it may help explain why replacement choices or builder offers have changed.
It cannot set rent, absorption, or exit timing. Those assumptions require local leases, current listings, verified deliveries, household evidence, and portfolio operating records. The broader research library can organize related measures without treating correlation as causation.
Teams maintaining recurring market maps and investment records can review investment analysis support. The investor remains responsible for selecting assumptions and approving transactions.
Limits and release discipline
New home sales are sample-based estimates with confidence intervals. Initial monthly figures can be revised as additional survey responses arrive. Seasonal factors are estimates as well. A sharp monthly move may later become smaller or larger, and national totals can conceal opposite regional patterns.
Archive each release used in committee materials. Label the observation month, release date, access date, unit, and revision status. A later correction should update current analysis while the old vintage remains attached to the historical decision.
Verification checklist
- Verify HSN1F and confirm the value is a seasonally adjusted annual rate.
- Do not label SAAR as actual sales completed during the month.
- Preserve Census sampling-error and revision notes.
- Separate contract, completion, deed, release, and access dates.
- Check current and prior release vintages.
- Map local communities and define the competitive geography.
- Distinguish planned, under-construction, completed, and sold homes.
- Record incentives and cancellation definitions separately.
- Verify whether competing stock is owner-occupied or rental.
- Tie rent and exit assumptions to local evidence with expiration dates.
Frequently asked questions
What was the May 2026 new home sales estimate?
The published HSN1F value was 580,000 at a seasonally adjusted annual rate. It does not mean 580,000 houses sold during May alone.
When does Census count a new house as sold?
A sale is recorded when a contract is signed or a deposit is accepted, even if construction has not begun or finished.
Does the estimate include existing homes or apartments?
No. It covers new single-family houses under the Census survey definition. Existing-home transfers and multifamily building transactions use other sources.
Why are new home sales revised?
The estimate is based on a sample, and more complete survey information becomes available after initial release. Seasonal factors and source corrections can also affect history.
Can national sales predict local rental demand?
No. Local tenant demand, rent, concessions, competing rental supply, and property characteristics must be measured directly.
Sources and verification dates
- Federal Reserve Bank of St. Louis: New One Family Houses Sold (HSN1F), accessed July 23, 2026.
- U.S. Census Bureau: New Residential Sales, accessed July 23, 2026.
- U.S. Census Bureau and HUD: New Residential Sales Release, accessed July 23, 2026.
- U.S. Census Bureau: Survey of Construction Definitions, accessed July 23, 2026.
- U.S. Census Bureau: New Residential Sales Methodology, accessed July 23, 2026.
- U.S. Census Bureau: New Residential Sales Data, accessed July 23, 2026.
- Federal Reserve Bank of St. Louis: Monthly Supply of New Houses, accessed July 23, 2026.
- ALFRED: HSN1F Vintage Archive, accessed July 23, 2026.
- U.S. Census Bureau: New Residential Construction, accessed July 23, 2026.
- U.S. Census Bureau: Building Permits Survey, accessed July 23, 2026.
- National Association of Realtors: Existing-Home Sales, accessed July 23, 2026.
- U.S. Census Bureau: Seasonal Adjustment, accessed July 23, 2026.
Store the national release beside the community map and local records. For help designing that handoff, contact PortfolioRental.