New home sales snapped back sharply in March, offering a rare data point of demand resilience — but bloated supply and a structural household formation deficit continue to define the underlying tension in the market.

---

New Home Sales: A Real Bounce, But Supply Stays Elevated

McBride flags a meaningful acceleration in new home sales: March came in at 682,000 SAAR, up 7.4% from February's 635,000 and 3.3% above March 2025's 660,000 pace. That sequential jump is significant enough that it can't be entirely dismissed as noise, and it brings sales back close to pre-pandemic levels — a benchmark worth noting given how far they drifted during the rate shock years.

The months-of-supply picture improved in tandem: 8.5 months in March, down from 9.1 in February. Progress, but context matters here. The normal range is 4 to 6 months — meaning the new home market is still running nearly double a healthy level. Inventory itself totaled 481,000 units, down 4.6% year-over-year from 504,000 in March 2025, suggesting builders are starting to work down the overhang.

The composition of that inventory is worth watching closely. Completed homes for sale stand at 121,000 — nearly quadruple the record low of 31,000 hit in February 2022 and well above normal. Homes under construction are at 249,000, which is elevated but roughly 22% off the July 2022 cycle peak. Most interesting: the "not started" bucket hit an all-time high of 111,000 — permits pulled or foundations broken, but no vertical construction yet. That's optionality builders are sitting on, and it signals they're watching demand carefully before committing capital to framing.

---

The Supply Shortage Debate: Erdmann's Household Formation Case

While McBride tracks the flow data, Erdmann is making a longer-run structural argument — and pushing back hard on what he calls "supply shortage truthers." The counterargument he's refuting: average household size is at all-time lows and homes are bigger than ever, so how can there be a shortage?

His answer is methodologically careful. Family household size (2+ related people) has been essentially flat for 20 years. The overall household size decline reflects a compositional shift — more childless, single-adult households — not looser housing constraints. Meanwhile, adults per family household rose post-2008 while children per family household fell by almost exactly the same magnitude. The problem: 5-year-olds don't form households. Adult children living with parents absolutely would, if units were available.

Erdmann's rough math: those 14 million "extra" adults living with family would have formed roughly 10 million additional households under a normal supply environment. That suppressed household formation is his quantification of the post-2008 shock — and it reframes the size data entirely. The fact that new single-family homes have been getting smaller since 2015 (and apartments since 2008) is, in his telling, further evidence of constraint: builders are shrinking units to hit price points, not because buyers want less space.

The implication for the next 20-30 years, which he raises but doesn't fully unpack here, is that any easing of supply constraints could unlock substantial latent household formation demand — a structural tailwind that cyclical rate moves alone don't capture.

---

Synthesis

March's sales rebound and supply compression are constructive at the margin, but the new home market is still operating well outside normal parameters. The more fundamental signal from Erdmann is that the shortage driving this market isn't a blip — it's a decade-plus of suppressed household formation that the data can obscure if you're not reading the composition carefully. For builders, the record "not started" backlog says they're not convinced the bounce is durable. For buyers and policymakers, the implication is that even a well-supplied-looking market (8.5 months!) can coexist with genuine structural undersupply at the household formation level.

---

TL;DR - New home sales rebounded to 682,000 SAAR in March, up 7.4% from February, approaching pre-pandemic levels but months of supply at 8.5 remains well above the 4–6 month normal range. - The "not started" new home inventory hit an all-time high of 111,000, suggesting builders are permitting cautiously and waiting on demand signals before breaking ground. - Erdmann makes a structural case that ~14 million displaced adults represent ~10 million units of suppressed household formation, a latent demand overhang that makes the supply shortage larger than surface-level size or headship data implies.

Compiled from 2 sources · 2 items

• Bill McBride (1)
• Kevin Erdmann (1)