The Housing Supply Cliff Is Farther Off Than It Looks

Supply is fading. That matters. But a decline in construction does not mean every market has reached an inflection point. The better signal is whether new household demand is beginning to outrun the units already expected to arrive.

Apartment deliveries are slowing, and rent growth has started to firm. That does not mean the housing market has turned everywhere. The next phase of this cycle will be local. Markets tighten when household formation runs ahead of new supply. They stay soft when expected completions continue to outrun demand.

To see where that balance could shift, we compared projected household formation with expected completions across major U.S. metros, measuring single-family homes and 5+ apartment buildings separately.

Two findings stand out:

• The consensus that a supply cliff makes 2026 the multifamily inflection looks early. Across nearly the entire screen, expected new 5+ apartment supply still outruns projected new 5+ household demand through 2027.
• The single-family forward balance is tightest in a small set of supply-constrained coastal and Northeastern metros and still loose throughout most of the Sun Belt.

Why the flow balance matters

Rent growth and concessions tell us what has already happened. The forward balance between new demand and new supply tells us where conditions may be heading next.

On the supply side, we translate each metro’s building permits into expected completions. Permits are observed through mid-2026 and projected at each metro’s recent run rate into 2027.

On the demand side, we project population forward, adjust for changes in age structure, and translate that growth into new households — then estimate whether those households are more likely to occupy a single-family home or a 5+ apartment.

The gap between projected demand and expected completions is scaled by each metro’s existing housing stock, so that a 5,000-unit shortfall in New York is not treated the same as a 5,000-unit shortfall in a much smaller market.

A positive balance means demand is running ahead of new supply. A negative balance means new supply is running ahead of demand.

One note on language: this is a forward snapshot, not a year-over-year change. When we call a market soft or tight, we mean the level of that balance now — where expected supply stands against expected demand.

This is a flow measure. It does not capture the vacancy or concession overhang already sitting in a market. That makes it a useful early signal, but not a complete indicator of pricing power.

Why demand is expected to cool

The demand side of this balance is softening for a straightforward reason. Population growth surged during the pandemic and has since pulled back, in part because immigration has come down from its recent highs. Fewer new residents means slower household formation.

In our forecast, net new household formation slows from 2026 into 2027 and keeps drifting lower — a gradual cooling is expected, not a collapse, but enough that demand is no longer racing ahead of supply the way it did during the boom.

That is the backdrop for everything below: expected new supply is being measured against a demand stream that is easing, not accelerating.

Where we disagree with the consensus

The prevailing view is that a coming supply cliff will be a tailwind for multifamily, and that 2026 could mark the inflection point. Our model disagrees.

The cliff in completed apartments is real. Read the multifamily pipeline back to front, though, and it gets stronger at every earlier stage (year-to-date through July, versus a year earlier, from the Census Bureau):

• Completions: down 14%. This is the number everyone points to.
• Starts: up 12%. The units right behind completions are not slowing; they are increasing.
• Permits: up 6%. The units behind the starts are growing too.

A decline in completions is not a decline in supply. Fewer apartments are being finished right now, but more are being ‘started and authorized’ behind them — into a vacancy rate that is elevated but has held relatively stable. The pipeline is not emptying. It is refilling.

At the same time, the demand side is losing momentum. Population growth has cooled and household formation is slowing in our forecast. So the supply cliff is arriving just as the demand that would absorb it softens.

Absorption has held up so far. The real question is whether it can keep doing so as population growth slows, household formation cools, and near term macroeconomic headwinds keep people in place. If it can’t, the multifamily inflection arrives later than investors expect — because supply is still coming while demand is losing steam.

Apartments: still more supply to absorb

Start with apartments, because this is where the “the boom is fading, so the market has turned” story is most tempting — and most premature.

Across nearly the entire screen, expected new 5+ supply still outpaces projected new 5+ demand through 2027. But if you want to know where the apartment inflection shows up first, look at the markets closest to flipping positive — the ones where the forward pipeline has thinned the most:

• San Antonio, TX: +0.1% of stock — the only apartment market in the screen where expected demand could run ahead of expected supply.
• Little Rock, AR: −0.04%
• Winston-Salem, NC: −0.1%
• Deltona–Daytona Beach, FL: −0.1%
• Houston, TX: −0.1%

These are the markets to watch — closest to balance — the first that could tip into deficit if absorption holds. San Antonio is the clearest example of the mechanism: it built heavily in 2022, is still absorbing that wave, but has since cut its apartment permitting by roughly 90% from the peak. Heavy supply now, almost nothing new behind it — which is exactly what a market on the cusp of a deficit looks like. But note how thin the signal is: only one metro is actually positive, and the next four are still negative. Even the leading edge of the apartment recovery hasn’t clearly turned.

At the other end, the largest expected-supply-over-demand gaps are:

• Madison, WI: −2.2% of stock
• Fayetteville, AR: −2.0%
• Columbus, OH: −1.5%
• North Port–Sarasota, FL: −1.5%
• Cape Coral–Fort Myers, FL: −1.3%

Deliveries can roll over, but fewer deliveries do not automatically mean a tighter market. Part of the reason we are not at an inflection point yet is that the builder pullback slowed this year — and in multifamily, reversed.

As the pipeline figures above show, 5+ starts and permits are both running higher than a year ago, even as completions fall. On the other side, projected new apartment demand is expected to ease rather than accelerate: in our forecast, net household formation slows by roughly 3% from 2026 to 2027 and continues to drift lower after that. Softening deliveries meeting softening demand is not an inflection. That combination is exactly what keeps the vacancy rate from falling.

A full pricing-power view also needs to account for current vacancy and concessions. While asking-rent growth is firming, concessions remain elevated in most markets — so the rent a landlord actually collects has recovered less than the headline suggests.

Single-family tightness is narrow

The single-family markets with the largest positive forward balances are concentrated in places where it has long been difficult to add new homes.

These eight are the single-family markets where projected demand is ahead of expected new supply. Just eight out of the roughly one hundred markets we screened.

The coastal and Northeastern markets on this list are also persistently supply-constrained. That means part of the signal is structural rather than a fresh cyclical turn.

The softest single-family metros are boom-era growth markets

The largest negative single-family balances are concentrated in several boom-era growth markets across the Sun Belt and Mountain West. These regions are not uniform — Miami is the second-tightest single-family market in the screen, and San Antonio is positive in both housing types.

• North Port–Sarasota, FL: −3.4% of stock
• Boise, ID: −3.2%
• Cape Coral–Fort Myers, FL: −2.7%
• Fayetteville, AR: −2.4%
• McAllen, TX: −2.3%

In these metros, permit-implied expected completions still exceed the number of households projected to form. The forward flow balance does not yet point to a bottom.

Single-family construction is actually pulling back more consistently than multifamily. Year-to-date through July, single-family completions are down 10%, starts down 7%, and permits down 4% (Census) — a steady retreat across the whole pipeline, not the mixed picture multifamily shows. But here is the problem: the pullback isn’t much larger than the expected pullback in demand. Single-family permits are down about 4%, while projected single-family household formation slows by roughly 3% from 2026 to 2027. When supply and demand ease at nearly the same rate, the balance doesn’t change. A builder pullback only tightens a market if it outruns the slowdown in demand — and here, it essentially matches it. That is why even a real, broad-based construction pullback is not yet expected to shift the balance in most single-family markets.

Nationally, the inventory of new homes for sale has been rising — the picture you expect when more homes are still being completed than the market is absorbing. The current decline in homebuilding will not clear that standing supply overnight.

What this means

Single-family conditions are expected to remain tight in a short list of supply-constrained coastal and Northeastern metros. Several boom-era growth markets are still working through more expected single-family supply than projected household demand can absorb. Apartments face an even broader imbalance, with new 5+ supply still running ahead of new demand across nearly the whole screen.

Supply is fading — but so is demand, and two things are keeping most markets short of an inflection. Household formation is cooling as population growth slows, and the builder pullback itself decelerated this year, so the pipeline is still being fed. In most markets the units already expected to arrive still outrun the households expected to form.

How we estimate this. We translate each metro’s building permits, observed through mid-2026 and projected at the recent run rate into 2027, into expected single-family and 5+ apartment completions using the historical timing between authorization and completion. We compare those expected completions with projected household demand by structure type and scale the gap by existing housing stock. The 2027 estimates depend on recent permit issuance continuing near its current pace.