Some of the fastest-growing metros in the country are currently among the hardest places to own an apartment building. That sounds backwards until you look at what happened. The growth story attracted so much construction that new buildings are now competing with each other for the same renters.
That tension is the most useful thing in this month’s market data, and it has been on my mind as I work through my own market research.
Why the national rent number is misleading
Yardi Matrix’s June 2026 national report shows U.S. advertised rents up 1.0 percent through the first half of the year, with the average rent at $1,763 and national occupancy at 94.1 percent, down 60 basis points from a year ago.
A 1 percent national number sounds like a market moving sideways. It isn’t. It is two very different markets averaged together.
Gateway and Midwest metros are carrying the growth right now. New York rents are up 5.6 percent year over year and San Francisco is up 4.7 percent. Meanwhile several high-supply metros are negative. Austin is down 4.0 percent, Denver is down 3.1 percent, and Tampa is down 2.8 percent.
And it is not purely a supply story. Yardi’s data shows about 108,000 units absorbed in the first five months of 2026, down 61 percent from a year earlier, as household formation cooled. So demand has softened at the same time a historic wave of deliveries is still working through the system. Both sides of the equation are moving, which is exactly my point.
When growth attracts too much supply
Here is the thing about the metros struggling right now. Many of them still have meaningful long-term demand drivers. Austin still has real employers and real in-migration. But those markets are also going through genuine near-term multifamily weakness, because everyone saw the same growth story at the same time and developers responded with some of the largest supply pipelines in the country relative to existing inventory.
When deliveries outrun what renters can absorb, owners lose pricing power. Concessions spread, occupancy slips, and the rent growth in the original pro forma doesn’t show up on schedule. How long that lasts depends on how fast starts fall and how demand recovers, and that varies metro by metro. Nobody can hand you the exact timeline, which is itself a reason to underwrite carefully.
None of this means the Sun Belt is bad or that gateway markets are automatically better. Conditions vary by metro, by submarket, and by asset class and price point within a submarket. A workforce housing property in a corridor with no competing new product can do fine in a metro whose Class A core is drowning in lease-ups.
The combination I look for
Coming from operations, I tend to think about markets the way I used to think about staffing a region. Capacity has to match demand or the plan breaks, no matter how good the plan looks on paper.
So my market filter has two halves. The first is the demand side: job growth, population growth, income levels, household formation, and how durable the renter base actually is. The second is the supply side: existing inventory, units actually under construction, planned and permitted projects, expected delivery dates, recent absorption, occupancy trends, concession activity, and how hard it is to build there going forward.
Strong demand fundamentals with constrained supply is the combination worth pursuing. Strong fundamentals with a flooded pipeline means the market’s growth is already being handed to renters as free rent and falling asking rates. And constrained supply by itself proves nothing. A market where nobody builds because nobody wants to live there is not an opportunity. The supply picture only matters after the demand picture holds up.
How supply changes the underwriting
This is where the analysis stops being academic. When I pressure test a deal, the local supply picture directly shapes the assumptions.
Rent growth gets set against the delivery schedule, not against the metro’s five-year average. If 2,000 units are delivering within three miles over the next 18 months, aggressive rent-growth assumptions during that window become difficult to defend.
Economic vacancy and concessions get modeled off what competing lease-ups are actually offering today. Two months free at the new build down the street is a real number, and it belongs in the model. Renewal assumptions get tested too, because renters with better options renew less and negotiate harder. Stabilization timelines stretch when a renovated unit has to compete against brand new product at a similar price point.
Exit assumptions matter most of all. If the plan is to sell in year five, the question is what the supply picture looks like then, not now.
And none of that comes purely from a database. Thorough diligence can include visiting competing properties, talking with local operators and property managers who see leasing traffic every day, confirming which projects are actually under construction versus merely announced, surveying current concessions, and validating expected delivery dates rather than taking a broker package at face value.
El Paso as one market I’m studying
El Paso is one of the markets where I’ve spent real research time, and the supply side is a big part of why.
Per MMG Real Estate Advisors, El Paso has averaged roughly 500 multifamily construction starts per year over the past decade, and starts fell to 276 units in 2024, down 37 percent year over year. On the delivery side, MMG’s 2026 forecast puts the 10-year average at about 526 completions annually, with 537 units completed in 2025.
Those are different signals pointing in a useful direction. Deliveries are running near the long-run average while starts have dropped well below it — so the near-term pipeline is still working through, but the future pipeline is thinning rather than swelling. That is close to the opposite of the flooded-pipeline metros in the headlines: a market where new supply is constrained going forward, not accelerating. On its own that still proves nothing. It only becomes interesting once the demand side — the border economy, the employment base, and the durability of the renter pool — holds up under the same scrutiny. That is the second half of the test, and it is the work I am doing now.