Pipelines are down 40 to 50 percent, but the crews left too. How fast can new supply actually come back in 2028?

Question

The recovery case rests on absorption overtaking deliveries and staying there long enough for rents to firm. Everyone models the pipeline contraction. Almost nobody models the restart.

I came out of the trades and I still have friends running framing and mechanical crews. When multifamily starts dried up in 2024 and 2025, those crews did not sit around. They went to data centers, to industrial, to whatever public work was funded. Some went to a different state. The apprentices in the pipeline two years ago mostly didn't finish, because there was nothing to finish into. Rebuilding a framing crew that can hit schedule on a 300 unit wrap is not a hiring decision you make in a quarter.

So the question for the institutional side is how long the supply window really stays open. If it takes a developer eighteen months from the day the numbers pencil to get a shovel moving, plus twenty four months to deliver, then a 2027 rent recovery doesn't show up as competing product until 2030 or later. That extends the runway on anything bought in 2026. If instead capital floods back the moment rent growth prints positive and crews follow the money fast, the window is much shorter and the exit gets crowded.

Cost is the other half. Replacement cost is up almost 39 percent since 2020 and hasn't come down. Rents have to move a long way before new construction pencils again in most of these submarkets, regardless of labor.

How long do you think the supply gap actually lasts?

aaron_fitz10 · Answer

I voted for the metro answer, and I want to argue with the labor half of your post, solder.

Crews are mobile in a way land isn't. If starts turn on in one Sun Belt metro, the trades show up from three states over within a season, because that's exactly what happened going the other direction in 2021 and 2022. The thing that doesn't move is entitled dirt with a site plan already approved and a lender willing to fund vertical. In the metros where entitlement is fast and there's a bank of shovel ready sites sitting on developer balance sheets, the restart could be quick, because the slow part is already done and paid for. Somewhere with a two year approval cycle, you get your five years whether the labor exists or not.

So the runway an institutional buyer gets isn't a national number at all. It's a question about how many approved but unstarted units are sitting in the specific submarket, and I never see that in an OM. Broker pipeline pages show units under construction. The units that are permitted and paused are the ones that hurt you, and they can start the month your rent growth prints.

I'd rather buy in a hard-to-build submarket at a worse basis than an easy-to-build one at a great basis, and I'm aware that's the opposite of what the discount math says.

Pipelines are down 40 to 50 percent, but the crews left too. How fast can new supply actually come back in 2028?

1 reply