The New-Home Price That Won't Move: How Builders Are Cutting 13% Without Touching the Sticker

The June 24 release from the Census Bureau and HUD landed with a contradiction baked in. New single-family home sales slid to a 580,000 seasonally adjusted annual rate in May 2026, down 7.3% from April and 6.8% from a year earlier. The supply of new homes for sale climbed to 10.3 months, up from 9.7 a year ago and the kind of double-digit overhang the market last saw in the foreclosure era. Builder confidence, measured by the NAHB/Wells Fargo Housing Market Index, sat at 35 in June, its 14th straight month below 40, deep in pessimistic territory.

And yet the headline price did not move. The median new-home pricemedian new-home priceThe middle price in the month's new-home sales, half sold for more and half for less; the figure Census reports as the headline price. was $424,900 in May, statistically identical to $424,800 a year earlier. A double-digit supply glut and the worst builder mood in years, paired with a price that looks like a flat line. That tension is the story.

Here is the narrow question this piece answers: when supply hits a 16-year-style high but the Census median holds flat, is the new-home market actually clearing through hidden price cuts that the headline never subtracts, and how large is that concealed discount today?

The method a reader can follow in one sentence: take the headline price, subtract the documented incentive loadincentive loadIndustry shorthand for total sales incentives expressed as a percentage of the sale price; a 12.9% load means incentives equal 12.9% of the price. that builders themselves report, and watch the two lines separate. The gap between them is the discount the sticker cannot show.

On June 24, 2026, Census and HUD published the May Monthly New Residential Sales report: 580,000 SAAR, the 10.3-month supply, the flat $424,900 median, and an average price up 5.0% from a year earlier to $540,600. (When the average rises while the median holds, it is a hint that the high end is still selling; hold that thought for the mix-shift debate below.) The standing inventory was about 496,000 homes for sale, only 1.4% below a year earlier even as the sales pace dropped, so each home now sits longer before it clears, which is how the supply figure pushed up to 10.3 months.

A week earlier, on June 18, the NAHB had released its June builder survey. It showed 62% of builders using sales incentives, the 15th straight month at or above 60%, with 35% reporting outright price cuts averaging 6%. And on June 11, Lennar, the second-largest US homebuilder, had reported its fiscal second-quarter results, the clearest dollar-level read on what those incentives are actually worth.

Three public sources, stacked: Census gives the headline price and the inventory, the NAHB survey gives the breadth of discounting, and the builder filings give the magnitude. Lay them on top of one another and the flat median starts to look less like resilience and more like a recording convention.

Start with the mechanism. A mortgage-rate buydownmortgage-rate buydownA builder pays cash upfront to a lender to lower a buyer's mortgage interest rate, cutting the monthly payment without lowering the home's listed price. is the lever here. Instead of dropping a home's price from $425,000 to $375,000, a builder hands cash to the lender to push the buyer's mortgage rate down, say from near 7% to something with a 4 or 5 in front. The buyer's monthly payment falls just as it would after a price cut, but the contract pricecontract priceThe gross price written into a new-home sales contract before any cash incentives or financing perks are subtracted; what Census records., the number Census records, stays at $425,000. The discount is real money. It simply travels through the financing instead of the sticker.

Now the magnitude. Lennar reported sales incentives running at 12.9% of sale price in its fiscal second quarter, against a long-run norm its leadership puts at 4 to 6%. Apply that 12.9% to Lennar's $371,000 average selling price and you get roughly $48,000 of value handed to the buyer that the contract price never nets out. That is a 13%-ish cut hiding behind a flat number.

The breadth confirms it is not one builder. With 62% of all builders using incentives for over a year straight, and a third cutting prices outright by an average 6%, the discounting is industry-wide. The flat median is sitting on top of a market that is, by the builders' own disclosures, paying less than it appears to.

The discount is real money. It simply travels through the financing instead of the sticker.

This is the intellectual center, and it has two layers. The first is the one above: the flat $424,900 is largely a measurement artifact. Census captures gross contract price, so a discount delivered as a buydown is invisible to it by construction. Read the median literally and you conclude builders have pricing power. Read the filings and you find the opposite, paid prices down by high single digits to low double digits while the headline holds.

The second layer is the one most coverage misses entirely: the hidden discount is cresting, not peaking. Lennar's incentive load did not just stay high, it fell three quarters running, from 14.5% to 14.1% to 12.9%. Management framed that narrowing as the first in three years. So the true price of a Lennar home may now be firming, very slightly, even as the headline median, still flat, registers neither the earlier steep discount nor the new firming.

Two stories could plausibly explain a flat median in a glutted market, and both deserve a fair hearing before we lean on the hidden-discount reading.

The first is mix shiftmix shiftA change in the kind of homes being sold, such as more small or stripped-down floorplans, that can move the average price without any single home changing price., a change in the kind of homes being sold. The average price rose 5.0% to $540,600 in May while the median stayed put. When the average climbs faster than the median, it usually means the top of the market is doing the selling: pricier, larger homes make up a bigger slice of closings, which can prop up the headline even if no single home holds its price. But mix shift and buydowns are not rivals; they stack. Mix shift explains why the listed median looks steady. The incentive data explains why the price actually paid is lower than any listed figure. One is about which homes sell; the other is about what buyers really hand over.

The second is genuine pricing discipline. Builders may refuse to cut list prices because land and construction costs set a floor, and because a public list-price cut would drag down the comparable sales, the comps, that appraisers use on every other home in a community. A financing incentive is surgical: it lowers one buyer's cost without resetting the whole neighborhood's paper value. That is true, but it is a reason builders discount through financing, not evidence that they are not discounting. The proof shows up in the margin. Lennar's homebuilding gross margingross marginThe share of each sales dollar left after the direct cost of building the home; a falling margin means builders are keeping less per sale. came in at a thin 15.6% in Q2, with net earnings down sharply from a year earlier, exactly what you would expect if real prices, net of those incentives, are under pressure.

For a buyer, the lesson is that in this market the sticker is the wrong place to look. The real negotiating happens at the rate lock, where a builder's cash can shave tens of thousands off the lifetime cost of the loan without ever showing up as a lower price. For inflation-watchers, the flat median quietly understates the disinflationdisinflationA slowing in the rate of price increases (prices still rise, just more slowly), distinct from outright falling prices. already running through new homes, though the effect is muted in the official numbers: shelter CPIshelter CPIThe housing piece of the Consumer Price Index, the official inflation measure; it leans on rents and owners' implied rents rather than new-home transaction prices. leans on rents and owners' implied rents, not new-home transaction prices, so this particular mismeasurement does not flow straight into headline inflation.

And the turn matters most of all. With Lennar's incentive load stepping down three quarters running, the deepest discounts may already be in the rear-view mirror. A buyer waiting for the official price to fall before jumping in could find that the best deals, the richest buydowns, have quietly come and gone, even though the Census line never showed them arriving and will not show them leaving.

In this market the sticker is the wrong place to look. The real negotiating happens at the rate lock.