Took a shelved 96 unit apart and rebuilt it until it penciled. Broke ground last month
A sponsor I underwrite for had a 96 unit deal that died in early 2024. Land was tied up, plans were 60% done, and the yield on cost was 5.4% against exit cap assumptions in the low fives. Dead. They were going to let the contract lapse and eat about $300k of predevelopment.
I got asked to re-underwrite before they killed it. Here's what came out.
Original budget: $25.7M total, $268k a unit. Land $4.2M. Podium parking under four stories of wood. Unit mix heavy on two bedrooms because the sponsor liked them.
What we changed.
Land. The seller had been at $4.2M for two years with no other buyer. We came back at $2.6M cash or $2.9M with $1.1M of it carried on a subordinate note for five years, interest only. He took $2.9M. That carry is subordinate to the construction loan and the lender needed a standstill agreement to allow it, which took six weeks and was the single most annoying part.
Parking. Podium out, surface and tuck-under in. That cost us 14 units, so the project is 82 units now. Hard cost per unit dropped from about $212k to $174k. Losing 14 units of revenue against saving $38k a unit on 96 units of cost was the whole trade and it was close enough that I ran it four ways.
Mix. Shifted from 55% twos to 62% ones and studios. Rent per unit falls, rent per square foot rises about 11% in this submarket, and the unit count per building envelope goes up.
Where it landed: $19.4M total, $237k a unit, stabilized yield on cost 6.5%. Against the same exit cap assumption that's 125 basis points of spread instead of 15. That's a deal.
The part that nearly broke it: the parking ratio. Surface and tuck-under got us to 1.15 spaces per unit against a code requirement of 1.5. We needed a variance and the neighborhood association came out against it. It passed 4 to 3 after we added a transit contribution and a bike room nobody asked for. If that vote goes the other way the whole restructure collapses, because the podium coming out was the load-bearing change.
What I'd keep: going back to a seller who has had no other offer in two years. The $1.3M price reduction was worth more than every other change combined and it cost one uncomfortable phone call. The sponsor did not want to make it.
What I'd watch: the delivery date is late 2027. Everything about that timing depends on the supply pipeline staying thin, and I'm one developer among many making that same bet.