M/I Homes Q2 2026: Orders Up, Deliveries Down

M/I Homes’ Q2 2026 orders rose 15% while deliveries fell 6% from a year earlier. Its July 29 results pair stronger contract activity with lower revenue, margin and pretax income. For competing builders, the distinction matters: more orders can replenish backlog without improving current earnings, and do not establish that future closings or profits will rise.

Orders and deliveries measure different stages

A contract records a buyer’s commitment. A delivery records a completed closing. Backlog contains contracted homes awaiting closing at the measurement date. Under its accounting policy, M/I recognizes home-sale revenue when closing, delivery and the required transfer of ownership and control have occurred.

That timing explains how orders can increase while revenue declines. Homes sold during a quarter and those delivered during it are overlapping groups, not necessarily the same homes.

Figures come from the 2026 release and 2025 release; gross-margin percentages are calculated from their income statements and rounded.

Fewer closings and a lower average closing price both weigh on housing revenue. Neither measure identifies a construction bottleneck on its own. When comparing homebuilder operating results, first ask which sales periods supplied the quarter’s deliveries before treating weaker closings as evidence that current selling efforts failed.

Backlog grew during Q2 but remained below last year

M/I’s July 31 Form 10-Q identifies the reported contract total as net contracts and reports a cancellation rate of 8.4%, down from 12.7% a year earlier.

Net-order growth can reflect both new signings and fewer cancellations. The cancellation rate does not explain why buyers withdrew or measure the probability that today’s backlog will close. Applying it again to reported net contracts would mix two different measures.

Backlog ended June at 2,426 homes, down from 2,577 a year earlier; its value fell 8% to approximately $1.31 billion. The July release therefore shows a smaller outstanding contract book despite stronger quarterly orders.

Within Q2, however, orders replenished backlog. M/I began the quarter with 2,245 homes in backlog, according to its April 22 release. The reconciliation is:

2,245 opening backlog + 2,387 net contracts − 2,206 deliveries = 2,426 ending backlog.

That calculation shows a 181-home increase during the quarter. It does not erase the gap with last year or establish a delivery schedule. For production planning, the next useful breakdown is backlog by construction stage and expected closing month. A contracted home awaiting permits creates a different workload from a completed home awaiting financing.

Orders rose with little change in community counts

Quarter-end communities held at 234, while the quarterly average edged up from 230 to 232. Using the averages in the earnings tables, net contracts per community per month increased from approximately 3.01 to 3.43—about 14%.

The calculation divides quarterly net contracts by average communities and then by three months. It shows that orders grew substantially faster than the selling footprint, but it does not track the same communities across both periods. Openings can replace sellouts, and shifts in location or price point can change the company average.

A smaller builder should compare similar products and markets, using the time each community was actually open for sales. Otherwise, a newly opened community can distort the comparison in either direction.

For land teams following new construction activity, this is a reason to examine local absorption before accelerating lot purchases. A companywide pace increase does not establish that a particular subdivision can support faster starts.

Margins show pressure; the 10-Q provides incentive detail

The consolidated gross margin in the table includes financial services. M/I’s homebuilding gross margin was lower: 19.7%, versus 22.6% a year earlier. The 10-Q attributes pressure to inventory-home mix, incentives and higher lot costs; results also included $4.2 million of inventory charges.

The filing also provides a direct incentive measure: revenue reductions for sales incentives and closing costs totaled $63.2 million, versus $47.1 million. Those disclosed amounts offer a firmer basis for assessing concessions than the margin decline alone.

Accounting treatment limits the comparison. M/I records closing-cost allowances and mortgage-rate buydowns as reductions of housing revenue, while free or discounted products and services enter land and housing costs, according to its annual report. The revenue-reduction total therefore does not capture every concession.

Nor does it reveal the incentive package attached to each new Q2 contract, the margin those contracts will eventually produce, or what a buyer can obtain today in a competing community. Faster sales may justify concessions, but the operating test is whether their cost leaves an acceptable return after carrying costs and overhead.

Compare your pipeline at the same stages

A useful competitor review connects each measure to an operating decision:

• Sales pace: Track gross signings, cancellations and net contracts separately, then calculate net contracts per active community per month. This distinguishes buyer activity from contract retention.
• Production: Divide backlog into homes awaiting starts, under construction and ready to close. Use expected delivery dates to assess staffing and starts.
• Conversion: Separate deliveries from opening backlog from homes both sold and closed during the quarter. This avoids treating all closings as evidence about older orders.
• Economics: Compare expected margin on new contracts with realized margin on deliveries, using consistent treatment of concessions, land sales and inventory charges.

Before changing prices or lot takedowns, ask the sales team which competing homes match your product, the construction team when your backlog can close, and the controller what each incentive costs. Agents need current written offers for those specific communities; a public builder’s quarterly average cannot substitute for them.