Guide
New home sales explained
A homebuyer signs a contract on a spec house still under roof framing. That signature counts toward new home sales the month it happens — even if the closing is six weeks away and the buyer still needs a mortgage commitment. The Census Bureau’s New Residential Sales series tracks those contract signings on newly built units, seasonally adjusted and annualized as a SAAR rate. It is smaller in unit volume than existing home sales but far more important for homebuilders, lumber and appliance suppliers, and GDP residential investment. Builders watch new orders to set production schedules; equity investors trade XHB and ITB on the print; macro strategists pair it with housing starts and construction spending to map the housing pipeline. This guide covers Census methodology, median price and months’ supply, the starts-to-sales-to-spending sequence, mortgage-rate sensitivity, a Harbor Development release-day worked example, an indicator decision table, common pitfalls, and an investor checklist.
What new home sales measure
The Census Bureau and HUD jointly publish New Residential Sales each month. The headline number is the seasonally adjusted annual rate (SAAR) of new single-family houses sold — meaning contracts signed or deposits accepted, not necessarily closed or occupied. A SAAR of 700,000 implies that if the current month’s pace continued for twelve months, roughly 700,000 new homes would be sold.
Key distinctions every release-day reader must internalize:
- Contracts signed, not closings — unlike NAR existing home sales, which count MLS closings, Census new sales count when a buyer commits to purchase from a builder.
- New construction only — resales, conversions, and mobile homes outside the survey scope are excluded.
- Sample-based with revision — Census surveys a panel of builders and extrapolates; the first print is often revised next month as more builders report.
- Nominal median price — the published median sales price is not inflation-adjusted and can move with mix (larger homes in the South vs condos in the Northeast).
New home sales typically represent 10–15% of total U.S. home transactions in normal cycles but account for a disproportionate share of cyclical volatility in construction employment and building-materials demand.
Release timing
The report usually lands around the last week of the month at 10:00 a.m. ET for the prior month — often the same morning as durable goods or GDP-related releases. Confirm the exact date on the economic calendar because holiday weeks shift publication.
Median price, inventory and months’ supply
Census publishes the median sales price of new houses sold — the midpoint of transaction prices that month. Builders also report houses for sale at month-end (completed units plus spec homes under construction offered for sale). From those two series analysts compute months’ supply of new homes:
Months’ supply = homes for sale ÷ monthly sales rate
Interpretation mirrors the resale market but with builder-specific nuance:
- Below 4 months — tight new-home inventory; builders have pricing power; incentives (rate buydowns, free upgrades) shrink.
- 4–6 months — balanced production; builders match starts to sales without heavy discounting.
- Above 6 months — oversupply risk; builders cut starts, increase incentives, or write down land options.
Unlike the resale market’s lock-in effect, builders can adjust supply mechanically: cancel projects, slow permit pulls, or convert specs to rentals. Months’ supply above seven during a rate spike often precedes negative earnings revisions for public homebuilders within one to two quarters.
Regional and size breakdown
Census releases sales by census region (Northeast, Midwest, South, West). The South dominates unit volume; the West often leads median price. Subscribers to detailed tables can track average square footage and price per square foot trends — useful for separating true price inflation from buyers shifting toward smaller floor plans when mortgage rates rise.
New home sales vs existing sales vs housing starts
Housing macro dashboards stack three Census/NAR series that answer different questions at different pipeline stages:
| Series | Publisher | What it counts | Timing vs cycle |
|---|---|---|---|
| Housing starts | Census | Ground broken on new units | Leading — future supply |
| New home sales | Census | Contracts on new construction | Coincident — builder order book |
| Construction spending (residential) | Census C30 | Dollars put in place | Lagging — cash flow to contractors |
| Existing home sales | NAR | Closings on resales | Coincident/lagging — turnover, not new supply |
In a typical upswing, starts lead sales by several months as builders break ground on communities before every lot is under contract. In a downturn, sales fall first as buyers balk at payments; starts lag as builders work through backlog and land commitments. Watching starts stay elevated while new sales collapse is a classic late-cycle warning — finished inventory piles up until months’ supply forces production cuts.
Mortgage rates, incentives and buyer mix
New-home buyers are overwhelmingly payment-sensitive financed purchasers. When the 30-year fixed mortgage rate jumps, builders respond faster than resale sellers because they control pricing and can buy down rates through lender partnerships.
Builder incentives that move the headline
- Mortgage rate buydowns — builder pays points so buyer gets 5.5% instead of 6.8%; boosts contract signings without cutting list price in the median series immediately.
- Closing-cost credits — reduces upfront cash; helps first-time buyers who are down-payment constrained.
- Free upgrades — appliances, flooring; keeps headline price stable while improving value.
- Spec-to-rent conversion — if sales stall, builders sell blocks to single-family rental REITs; removes units from new home sales but clears inventory.
During 2023–2025 many public builders reported record gross margins while unit volumes fell — they preferred margin over volume until months’ supply threatened land writedowns. Tracking net orders in builder earnings calls (contracts minus cancellations) often gives a cleaner read than the Census headline alone.
Cancellations and the contract vs closing gap
Census counts signed contracts; builders later report cancellation rates when buyers fail appraisal, lose jobs, or walk away from deposits. Rising cancellations with firm headline new sales means the Census print will revise down. Pair the release with the latest D.R. Horton, Lennar, or Pulte order commentary for a reality check.
GDP, employment and materials demand
New residential construction feeds directly into private residential fixed investment in GDP — structures, improvements, and brokers’ fees on new sales. Weak new home sales eventually translate into lower starts, fewer construction jobs, and softer orders for lumber, HVAC, and kitchen appliances.
Cross-asset links investors monitor on release day:
- Homebuilder ETFs (XHB, ITB) — most rate-sensitive equity sector on the print; single-family-heavy builders move more than diversified firms.
- Lumber futures (LBS) — correlated with starts with a lag; new sales inflection is an early sentiment read.
- Building-products equities — roofing, insulation, fixtures; order books follow builder starts by one to two quarters.
- Treasuries — very weak new sales plus rising unemployment supports bonds; surprisingly hot sales with tight supply can worry inflation hawks if shelter costs stay sticky.
New home sales do not directly enter CPI shelter the way rents do, but sustained builder pricing power eventually flows into owners’ equivalent rent models and new-tenant lease comps.
Worked example: Harbor Development release-day read
Harbor Development is a regional homebuilder with 2,400 lots across four Sun Belt metros. The CFO runs a twenty-minute Census release review before the weekly production meeting:
- Read SAAR new sales vs three-month average — if the print is 675k SAAR and falling while the three-month average is below 700k, Harbor holds Q3 starts flat rather than accelerating land development.
- Check months’ supply nationally and in the South — national supply at 8.2 months triggers a review of Harbor’s spec inventory; any community above nine months gets a 3% base-price reduction or enhanced buydown package.
- Compare median price YoY to Harbor’s average selling price — if Census median is down 2% YoY but Harbor ASP is flat, the mix may be holding; if both fall, cut land-option takedowns.
- Cross-check housing starts and builder earnings — if starts are still rising while new sales fall two months in a row, Harbor delays new community openings by 90 days.
- Log one paragraph for the board — e.g. “NHS 668k SAAR (-5.4% m/m). Months’ supply 8.0. Median $428k (-1.2% YoY). Starts still elevated — expect incentive wars in Phoenix. Maintain 340 quarterly closings target; defer Phase 2 of Ridgefield.”
The discipline is to tie one production decision to each release rather than treating the headline as noise.
Indicator decision table
| Your question | Best series | Why |
|---|---|---|
| Are builders signing contracts today? | New home sales (Census) | Contract-based order book for new supply |
| How much resale turnover is happening? | Existing home sales (NAR) | Closings on previously owned homes |
| How much construction will happen next year? | Housing starts and permits | Leading physical activity |
| When do dollars hit subcontractors? | Construction spending (C30) | Put-in-place cash flows, lags starts |
| Are buyers getting pre-approved? | MBA purchase applications | Weekly, leads Census by weeks |
| Public builder order trends | Builder earnings (net orders) | Cancellations and regional detail |
Common pitfalls
- Confusing new and existing home sales — different agencies, methodologies, and economic meaning; never add the two series together.
- Treating the first print as final — Census revises; large surprises often shrink next month.
- Ignoring months’ supply — sales can rise on incentives while inventory builds, setting up the next downturn.
- Comparing nominal median prices across decades without adjusting for size — today’s new homes are larger; price per square foot is cleaner.
- Equating new sales with starts one-for-one — multi-year communities and backlog decouple the series.
- Missing cancellation waves — contracts signed in January may cancel by March if rates spike.
- Overweighting one region — national SAAR masks Texas strength vs California weakness.
- Expecting instant Fed reaction — housing is slow-moving; the Fed watches shelter CPI and payrolls more than any single NHS print.
Investor checklist
- Know release date and time (typically last week of month, 10:00 a.m. ET).
- Compare SAAR sales to consensus and three-month moving average.
- Read months’ supply and median price YoY alongside the headline.
- Check regional tables if your equity exposure is geographically concentrated.
- Pair with latest housing starts and MBA purchase applications.
- Scan major builder earnings for net orders and cancellation rates.
- Note revision to prior month before trading the surprise.
- Map signal to GDP residential investment, not total consumption.
- Track incentive intensity in builder press releases for margin risk.
- Log your read in one paragraph tied to a concrete position or sector view.
Key takeaways
- New home sales count Census-tracked contracts on new construction, annualized as SAAR.
- Months’ supply tells you whether builders are ahead or behind demand.
- Starts lead, spending lags — place new sales in the middle of the pipeline.
- Homebuilder equities react most; resales matter for sentiment, not this series directly.
- Always distinguish new sales from NAR existing closings on release day.
Related reading
- Existing home sales explained — NAR closings, resale inventory, and lock-in dynamics
- Housing starts explained — permits, single-family vs multifamily, leading supply signal
- Construction spending explained — Census C30 put-in-place dollars and GDP fixed investment
- Mortgages explained — rates, payments, and affordability math for buyers