I can build the 12-unit myself for about 1.4M all in, or I can buy a similar one two streets over asking 1.85M with a 7.1 cap on trailing rents.

The assumption doing the most work on the build side is that stabilized value. You're projecting 1.75 to 1.8M, which implies a cap rate somewhere around 7.5 to 8.5 percent on your delivered rents. The 7.1 cap on the existing asset is the market's current read on a proven income stream in the same submarket. If your delivered rents come in light, or if cap rates tick up another 25 to 50 basis points by the time you're stabilized and trying to refi, that spread between 1.54M all-in and stabilized value narrows fast. The GC margin you're capturing is real, but it's the first thing that gets absorbed if lease-up takes four months longer than modeled.

The risk you did not name is construction loan extension. If you're targeting a refi in late 2026 and lease-up drags into Q1 2027, you may be negotiating an extension on a construction facility at the same moment the refi market is repricing. Construction lenders and permanent lenders are reading the same rate environment. Confirm your construction loan terms, specifically extension options and their cost, in writing with your lender before you close on the land.

On the buy side, your read on 1.85 as 2022 pricing is the real question. The 7.1 cap on trailing rents is what matters here, and trailing is doing a lot of work in that phrase. If current market rents in that submarket are above what the seller has been collecting, there's value-add left in the asset that the cap rate does not reflect. If trailing rents are at or above market, you're buying a stabilized asset at full price with no operational upside, which is a different conversation entirely.

What are current market rents per unit in that submarket versus what the existing 12-unit is collecting, and how much of the 13,200 monthly is at-market versus below?

The one thing I have never seen work is underwriting a lease-up in a new 12-unit like it's going to fill linearly. What actually happens is you get four or five units absorbed fast, then you hit a wall around month three or four and carry two or three vacant units for another six to eight months while your pro forma is screaming stabilized. That gap alone can eat 40 to 60k you didn't model, and it's almost never a demand problem, it's a sequencing problem with finishes and certificate of occupancy on the back half of the building.

The 1.85 ask doesn't bother me as much as the trailing rent figure does. Thirteen-two collected on a 12-unit in 2025 is 1,100 a unit, and I want to know how old those leases are and what concessions got them signed. If half that book renewed in 2022 and 2023 at rates that made sense then, you're buying a cap rate that's going to look different in 18 months when those leases roll and the market either holds or it doesn't. The stabilization work the seller did might be the problem, not the asset.

The 7.1 cap at 1.85 is only clean if those trailing rents hold and the next lease cycle doesn't give anything back. Two streets over matters a lot here, and if that submarket has been absorbing new supply, the existing tenants might already be at or above where new leases will clear in 2026. I'd want to see the actual lease expiration schedule before I got comfortable with that 13,200 number, because a 12-unit can look fully stabilized and still have four units rolling in the same quarter.

On the build side the 140k carry assumption feels tight if rates stay where they are through the back half of 2025. My last ground-up was smaller and I underestimated soft costs and carry by about 60k over what I had modeled, and that was on an 11-month timeline not 18. The refi risk in late 2026 is the part I'd be losing sleep over, not the lease-up.