U.S. September Housing Starts Fall to Four-Month Low, Single-Family Permits Hit Three-Year Low
nashnova research
U.S. housing starts dropped 5.2% in September to an annualized 1.22 million units, missing every economist's forecast; with the 30-year mortgage rate near 7.5%, sky-high borrowing costs are choking new-home supply and dragging on GDP for a fifth quarter out of six.
How bad are the numbers?
Total starts fell to an annualized 1.22 million, down 5.2% month-on-month and 7.4% year-on-year — residential construction is shrinking on an annual basis.
Single-family starts dropped to a one-year low of 859,000; single-family permits slid to 845,000, the lowest since early 2023. This means → permits are "future starts," so a permit trough signals supply will tighten further in the months ahead.
Multi-family projects (apartments and condos) fell nearly 7% month-on-month. Volatile by nature, but the direction matches: the entire housing-construction sector is cooling.
Why are builders pulling back?
The core issue is affordability. The 30-year fixed mortgage rate reached 7.40%–7.5% last week, a near-three-year high. In plain terms = on a $1 million loan, moving from 5% to 7.5% adds roughly $1,600 a month — enough to price out a large share of potential buyers.
Builder confidence has collapsed: the NAHB (National Association of Home Builders) sentiment index fell in October to its lowest since June 2012. The association blamed four pressures — rising material costs, high mortgage rates, surging diesel prices, and tighter immigration policy squeezing the labor pool.
Weak sales have left new-home inventory piling up. Builders sitting on unsold houses have little incentive to break ground on more. This reflects a vicious loop: high rates → weak demand → rising inventory → fewer starts → even less supply down the road.
What is the completions data signaling?
Single-family completions fell last month to a six-year-plus low. This means → it is not just that builders are reluctant to start new projects; even projects already under way are finishing more slowly. The entire construction pipeline is decelerating.
Fewer completions → slower drawdown of existing inventory → even less motivation for builders to launch new projects — a second negative-feedback layer stacked on top of the first.
What does this mean for the broader economy?
Residential construction has dragged on GDP in five of the past six quarters. Before this release, the Atlanta Fed's GDPNow model projected Q3 residential investment falling at an annualized 1.3% — and this data only reinforces that downside.
In plain terms = "residential investment" is a GDP line item covering homebuilding, renovation, and related spending. When it runs negative quarter after quarter, housing has flipped from an economic engine to a brake.
Is the weakness nationwide, and how reliable is the data?
Single-family starts in the South — the country's largest housing market — fell to their lowest since January. The West and Midwest declined in tandem. This is not a regional pocket; it is a national cool-down.
A caveat: the government report flags a 90% confidence interval of −17% to +6.6% for this month's change. Any single month may be revised, but the directional trend — sustained decline — has been confirmed across multiple months.
Where is the structural contradiction?
Even as newly built single-family homes face a glut, the U.S. housing market overall remains undersupplied. This means → unsold new homes and a national housing shortage coexist — the problem is not total volume but price: high rates shrink the pool of qualified buyers while structural scarcity keeps home prices elevated.
The key thing to watch: whether builders can clear their current inventory backlog and restart the construction cycle before mortgage rates drop meaningfully — that will determine when residential investment stops being a drag.
市场有风险，内容仅供研究参考，不构成投资建议。