The index and its scale
The NAHB/Wells Fargo Housing Market Index registered 35 in the June 2026 release. The index is a diffusion measure: any reading below 50 means more builders rate current single-family sales conditions poor than good. It is derived from a monthly survey panel of NAHB members, not from transaction records.
NAHB later described the June figure as upwardly revised to 36 in its July publication. Estate Wire cites the figure as originally published for June and notes the revision rather than blending the two, because the association does not publish a revision history for the series.
The concession data is the substantive part
Alongside the headline, NAHB reported that about 35 percent of builders cut prices in June, with an average reduction of roughly 6 percent, and that about 62 percent used sales incentives of some kind. Incentives in this market most commonly take the form of mortgage-rate buydowns, closing-cost credits and option upgrades rather than list-price changes.
This matters for how published price data should be read. A rate buydown funded by the builder reduces the buyer's monthly cost without reducing the recorded sale price, so it does not appear in the Census median new-home price or in repeat-sales indices. The effective discount in the new-home market through 2026 was therefore larger than headline price series show.
Where sentiment went next
July's index slipped to 34 and the share of builders cutting prices rose to about 37 percent with incentive use near 63 percent. By the September release the index had fallen further, and Census reported starts and permits both easing in August.
Analysis: the sequence through the summer is internally consistent — sentiment softening, discounting widening, and groundbreaking slowing — but the HMI is an opinion survey and cannot by itself establish the cause. It is most valuable as a near-real-time read on builder behaviour that the official statistics confirm only months later.
How the June reading was revised
The June index was initially published at 35 and subsequently revised up to 36. That one-point revision is small, but it matters for anyone comparing months: the July decline was reported as two points from a revised 36, not one point from the originally published 35.
The concession data moved in a straight line across the spring. The share of builders cutting prices was 32 percent in May, 35 percent in June and 37 percent in July, while the average reduction held at 6 percent in both June and July. Sales incentives were used by 62 percent of builders in June and 63 percent in July.
Analysis: a stable average discount alongside a rising share of discounters describes widening participation rather than deepening cuts. More builders reached for the same-sized lever. That is a different market condition from one in which a fixed group of builders is cutting progressively harder, and it is the distinction the headline sentiment number hides.