First commercial development check: chase the data center wave or take the boring healthcare shell?

The sector is the least useful variable in that vote. What determines whether you get your money back is the sponsor's basis in the dirt, the contingency in the GMP, and whether the lease-up assumption in the model has a signed LOI behind it. I've underwritten two ground-up medical deals in the last year where the sponsor's rent assumption was 8 percent above anything actually signed in that submarket, and both still cleared a 6.5 yield on cost on paper. Change the rent and the whole thing goes underwater before you get to the exit cap.

On data centers specifically, the deals with real tenant demand behind them are being funded by institutions writing eight and nine figure checks. If a 50k slot is being offered to you inside that space, ask what part of the capital stack you're in and who else said no first.

The two options in your poll are not the same risk shape at all once you look at how they actually get built.

A healthcare shell is a known scope. Steel package, slab, some fussy MEP for the imaging rooms if the tenant wants them. The tariff exposure on steel-intensive scopes has been running real money per square foot and it lands in the estimate where you can see it.

Data center adjacent work is a schedule problem. Switchgear and transformer lead times are the whole deal, and utility interconnect dates slide for reasons nobody in the ownership group controls. I've watched a project sit finished and dark waiting on a utility. Interest carry keeps running while it sits.

@tally the PPM usually caps calls at some multiple of your commitment, and the cap is often generous enough to be meaningless. Read that section twice.

Genuine question about the framing. Is LP development the right door for someone whose next move is their first commercial anything? I've spent a year deciding between active and passive and every time I look at ground-up as a passive position it reads like I'm buying an outcome I can't influence, on a five year horizon, with no rent roll to fall back on. I voted for none of it and I'll be curious how lonely that is.

Asking the dumb one so nobody else has to. What is a powered shell? Every time this topic comes up someone says it like it's obvious and I've been nodding along for months.

@compass a shell building with the electrical service already brought in and sized, so the tenant's fit-out doesn't wait on the utility. In data center land it means you've secured a real megawatt allocation, which is the actual scarce thing.

I'm four deals in on small retail and I voted healthcare. Reason is unglamorous. My last retail pad sat 11 months past shell completion because two of the four spaces were speculative, and I was paying carry the whole way. A single tenant outpatient build with a signed lease before you break ground removes the part of the job I'm worst at.

What's the realistic hold on a small healthcare ground-up before the sponsor can sell or refi? I've seen decks say three years and I don't believe them.

I read this whole thread twice and still don't know enough to vote, but the transformer thing @girder mentioned is the first time development risk made sense to me as a physical object rather than a spreadsheet line.

One thing the healthcare option flattens out. An urgent care operator, a dental group, and a hospital-affiliated outpatient clinic are different credits with different lease lengths, and a hospital system as guarantor is a different deal than a two-doctor practice signing personally. My list has 14 sites on it and the only ones I still like are the ones where I can name the likely tenant type before I price the shell.

@pike three years is a refi assumption, not a sale. It assumes stabilized NOI supports permanent debt at whatever rate the model guessed at.

Practical question for @sextant since you do this for a living. Given two PPMs on similar ground-up deals, what do you read first? I currently read everything front to back, which takes days and leaves me no more decided than when I started.

@canton coming from the renovation side, the thing I'd add is that ground-up scope creep looks different from ours. On a reno the surprises come out of the walls. On new construction they come out of the plan review and the tenant's change orders after permit. I've seen a medical fit-out add 40k of extra electrical because the practice bought a different imaging unit than the one the drawings were coordinated to. Nobody calls that scope creep, they call it a tenant improvement allowance overage, and the owner still writes the check.

Living the switchgear problem right now on an industrial shell, so I'll spare you the details and say the quoted lead time and the actual lead time were 14 weeks apart. Carry on that gap was about 31k. I voted industrial anyway because I understand the tenant demand better than I understand anything in the AI category.

Blunt answer, the beginner should not be in ground-up commercial at all, and I say that as someone who lives on thin margins by choice. The reason isn't difficulty, it's that development is the longest-dated bet in real estate and you learn nothing from it for three years. Do two small deals where the feedback loop is months and you'll be a better judge of a development sponsor when you eventually meet one.

@windlass the sources and uses page and the sponsor's fee schedule. Then the rent comps. Everything else is narrative.

Coming from a completely different career, the useful part of this thread for me was realizing the vote isn't really about which sector wins. @sextant and @flint are both saying the sponsor and the timeline decide it. I voted for none of it yet, which is honest about where I am.