The building products market spent 2020–2022 in an anomalous expansion window driven by pandemic-era conditions that are now fully unwound. Five charts tell the structural story — and what it means for manufacturers planning the next 18 months.
U.S. homeownership peaked at 69.0% in 2004 and has not recovered. The 2020–2022 bump — driven by pandemic-era mobility, historically low rates, and pull-forward demand — created the illusion of structural recovery. It wasn't.
By 2024, the homeownership rate had returned to 65.6%. If the 2000 tenure structure had held, there would be approximately 2.3 million more owner-occupied households today. Instead, those households exist as renters — in the segment of the market that does not drive discretionary replacement spend.
This decline occurred across multiple economic cycles, including strong labor markets, rising incomes, and historically favorable financing conditions. It is not cyclical. It is structural.
Manufacturers planning against 2020–2022 run rates are planning against a distortion window, not a baseline. The replacement-capable base is capped — and drifting lower.
Total household formation has been resilient — but the composition of that growth is the issue. Since 2000, renter-occupied households have grown 36%. Owner-occupied households have grown only 20%.
This matters because replacement decisions, timing, and scope differ fundamentally between owners and renters. Owners control the timing and scope of replacement — accessing home equity, financing upgrades, making discretionary choices about product quality. Rental replacement is episodic, budget-driven, and typically lower-intensity per household.
The U.S. economy can generate household growth without generating replacement-capable household growth. That's what has been happening since 2004 — and the divergence has been widening.
Sustaining historical unit volumes from a capped owner base requires higher replacement intensity per owner household. The arithmetic becomes more demanding every year renter growth outpaces owner growth.
Building permits are the most predictive single leading indicator for new construction product demand. A permit issued today becomes a product order in 12–16 weeks. By the time that demand shows up in your sell-through data, the supply chain needed to respond three weeks ago.
The chart shows FRED:PERMIT (blue, solid) indexed against a sell-through demand proxy (amber, dashed) shifted forward by 12 weeks. The gap between when the permit signal moves and when the demand confirmation arrives is your planning window — and most teams are operating without it.
Teams that read sell-through as their primary demand signal are operating reactively, by definition. The permit data has already told the story.
Permit trends visible right now are the demand signal for Q2–Q3 2026. If you're waiting for sell-through to confirm, you're 12 weeks late — in both directions. The permit softening visible in H2 2025 will show in orders before summer.
The NAHB Housing Market Index (HMI) measures builder sentiment. At 42 — below the 50 threshold that separates expansion from contraction — builders have now been pessimistic for six consecutive periods. Permit volume has held elevated over the same stretch.
These two series are now moving in opposite directions. That divergence is the signal. When permit activity holds high while builder confidence falls, it historically reflects a lag between builder commitments made in more optimistic conditions and the emerging reality of slower demand. The correction in permits typically follows within two quarters.
This is exactly the kind of cross-series tension that doesn't show up in any single dataset — and that most teams miss because they're reading one series at a time.
The current HMI-permit divergence is a leading indicator of a permit correction in Q2–Q3 2026. Manufacturers who adjust production planning now will have a 6–8 week advantage over those waiting for the permit data to confirm the softening.
Multi-family starts are structurally volatile at the monthly level — not because the underlying demand is volatile, but because of how multi-family construction is reported. A single large project in a Census region can move the monthly number by 100K+ units.
The chart shows raw monthly multi-family starts (red, jagged) alongside the 3-month rolling average (blue, smooth). The two lines are tracking the same underlying demand. Only one of them is readable.
The more important signal for multi-family isn't monthly starts at all — it's the pipeline on the boards. The projects permitted and under construction represent 24–36 months of forward demand. Watching monthly starts change is a poor proxy for that pipeline.
Multi-family monthly starts should never drive a production planning or inventory decision. The 3-month rolling average is the minimum read. The permit pipeline — projects on the boards with approvals in place — is the real signal for manufacturers planning 12–24 months out.
Four decisions every building products manufacturer should be revisiting right now.
The structural signals in this report look different for windows than for HVAC, for new construction-focused products than for R&R. We do this analysis category-specifically — with the data to back it up.