Housing and construction market · October 22, 2026

PulteGroup orders ease while its home backlog grows

By Workvado / Litequo LLC · Published October 27, 2026 · Updated October 27, 2026

The builder reported 6,208 net new orders and a 10,298-home backlog at September 30, offering a company-level view of residential demand and production flow.

PulteGroup’s October 22 third-quarter report showed a homebuilder delivering more houses than it sold into backlog, while new orders declined from a year earlier. The Atlanta-based builder reported 6,876 closings for the three months ended September 30, home-sale revenue of USD 3.859 billion, and 6,208 net new orders. Orders were down 6 percent year over year, but the company ended the quarter with 10,298 homes in backlog, up 4 percent, valued at USD 6.4 billion. Those figures describe PulteGroup’s own operating pipeline; they are not a national housing-start measure or a forecast for the wider market.

The order and closing counts give a useful but limited picture of the company’s near-term production flow. Closings are homes transferred during the quarter, while net new orders are contracts booked net of cancellations. Backlog is the reported stock of homes under contract but not yet closed at quarter end. The gap between 6,208 orders and 6,876 closings is not by itself evidence that production is falling: prior-period orders can close now, and the company’s community count, cancellation activity, construction cycle, and buyer mix affect the relationship. The report gives a snapshot through September 30; readers should not treat it as a count of homes currently under construction at every stage. Pulte’s average community count rose 7 percent to 1,077, while net new orders declined 6 percent. That combination implies fewer net orders per average community during the quarter than a year earlier, although the two reported measures are not a same-store cohort calculation and should not be presented as one. It points analysts toward absorption and sales pace as questions to monitor: did a broader footprint offset slower demand, and were new communities still filling their planned production schedules? The published release does not provide a project-level inventory or a full local absorption breakdown, so companywide figures cannot establish which markets gained or lost momentum.

Backlog value increased 3 percent to USD 6.4 billion, slightly slower than the 4 percent rise in unit backlog to 10,298. That pattern is consistent with modestly lower average value per backlog home, but calculation from rounded totals is only an approximation. It may reflect geography, product and buyer mix, options, or pricing; it does not tell a contractor what a specific trade package will cost or when a neighborhood will be built. Pulte’s segment tables offer geographic detail, but aggregate backlog remains a sales measure with timing uncertainty. Builders can cancel or renegotiate contracts, and the home must still pass through permitting, construction, financing, and closing.

Margins also illustrate the distinction between a sales pipeline and a construction budget. Pulte reported a 24.8 percent home-sale gross margin, including USD 26 million of land impairments that lowered the measure by about 70 basis points. Selling, general, and administrative expense was 10.3 percent of home-sale revenue, compared with 9.4 percent a year earlier. These are company accounting measures, not gross margins on a contractor’s work or a direct measure of building costs. Land write-downs reflect an assessment of expected recoverability; they should not be mistaken for a cash payment to suppliers during the quarter.

The demand context was weak across broader builder sentiment as well. NAHB’s October Housing Market Index put single-family builder confidence at 29 on its 0–100 scale, with current sales at 33, expected sales at 32, and prospective-buyer traffic at 18. NAHB said 43 percent of surveyed builders cut prices in October and 65 percent used sales incentives. Freddie Mac separately reported an average 30-year fixed mortgage rate of 7.32 percent on October 22, down from 7.40 percent the prior week but above its 6.19 percent reading a year earlier. These sources measure sentiment and mortgage quotes, not Pulte’s customer-specific loan rates or national home sales.

The quarter therefore contains both a builder-specific backlog expansion and evidence of demand friction. Pulte’s backlog count rose even as orders fell, while the national builder survey showed weak confidence and low buyer traffic. The figures do not conflict: existing contracts may be converting into backlog as new sales slow, and national sentiment is not a substitute for one builder’s results. For residential contractors and suppliers, the useful implication is to avoid projecting a single demand signal into a production plan. A builder can still have substantial contracted work while adjusting incentives, land purchases, construction starts, or community schedules in response to current buyer conditions. A project-by-project dashboard can separate homes sold from homes started and units under construction, reducing the risk of treating every backlog contract as immediate labor demand. Teams can compare projected starts with inspection calendars, material release dates, and local trade capacity. If incentives are concentrated in particular subdivisions, that may affect buyer conversion and option choices differently from a general price change; the published companywide results do not reveal such neighborhood-level decisions, so contractor forecasts need direct coordination with the builder’s local team.

Procurement teams can use the report to frame questions rather than predict work volume. A subcontractor watching Pulte can track the next quarter’s orders, cancellation rate, backlog conversion, average community count, and regional starts, then compare those measures with local permit filings and publicly available bid opportunities. Suppliers can ask whether purchasing schedules align with backlog homes under construction or with speculative starts, since those carry different commitment levels. The release does not break down quantities of lumber, concrete, HVAC equipment, or trade labor required by quarter; translating backlog units into materials demand would require project schedules and house specifications that are not disclosed. Any decision to increase inventory should also account for the difference between a sales contract and a final closing. Financing approvals, buyer cancellation rights, appraisal outcomes, and construction timing can affect whether a home transfers as planned. The company’s report gives a quarter-end snapshot, not a guarantee that every backlog home will close in a particular month.

The October report is useful because it separates a builder’s completed sales, incoming contracts, and unclosed pipeline, allowing residential firms to ask whether business is moving from one stage to the next. It cannot establish a national construction recovery or contraction by itself. The next verification points are Pulte’s subsequent quarterly data, local permits and starts, and whether the backlog converts into completions without a further decline in orders. This distinction helps housing professionals separate a quarter’s accounting and sales data from the work that will actually reach a jobsite. For suppliers, matching purchasing capacity to local starts and builder forecasts is more defensible than multiplying a company’s backlog by a generic material allowance. Where the company has not disclosed its neighborhood schedule or specifications, the local project team remains the right source for a demand forecast.

Report date

October 22, 2026

Quarter ended

September 30, 2026

Closings

6,876 homes

Net new orders

6,208 homes, down 6% year over year

Backlog

10,298 homes, up 4%; value USD 6.4 billion

Sources

1. PulteGroup Reports Third Quarter 2026 Financial Results, PulteGroup, Inc., read October 27, 2026
2. PulteGroup 2026 Third Quarter Form 10-Q, U.S. Securities and Exchange Commission, read October 27, 2026
3. NAHB/Wells Fargo Housing Market Index: October 2026, National Association of Home Builders, read October 27, 2026
4. Mortgage Market Survey Archive, Freddie Mac, read October 27, 2026