My gap note paid out at 22% annualized on a 14 MW shell and I still can't tell if the power arbitrage was the trade or the lease structure was

The land basis is the trade, full stop, Gainesville corridor doesn't care what's sitting on it.

Interconnection rights on tenant exit, that's the number I'd need in writing before I slept on this.

The one thing I never see work is underwriting the NNN income as the primary return and then being surprised when the real exit is a land play. Sponsors know this going in and the income story is just the part that makes the preferred return look covered on paper.

Gainesville 2019 basis at $180k is almost unfair, that corridor move to $900k happened faster than most 10-year models assumed because NoVA power constraints pulled demand south faster than anyone priced in back then. Your 2031 contracted power is probably worth more to a competing operator at lease roll than the tenant is paying now.

The question I'd be sitting with is what the sponsor's promote looks like if they sell the land pre-roll versus waiting for the hyperscaler to exercise any expansion option, because those two outcomes hit your $400k very differently.

@ehernandez74 here, so genuinely curious what your exit assumption is on the land if the hyperscaler exercises any purchase option buried in that NNN.

The NNN framing is doing a lot of work to keep LPs calm and I'm not sure that's a good thing long term. If the real bet is $900k-per-acre land in a power corridor, you want that underwritten explicitly, not buried under lease comps and pref waterfalls. I'd want to see what the sponsor's exit assumes per acre in 2031 versus what the market is already pricing in adjacent parcels today, because if those two numbers are close, the upside is already in the rearview.

The 2019 land basis at $180k is the part that's hard to replicate now, but I'd be careful about how you're marking the current $900k corridor comp. Prince William County parcels with confirmed power capacity are trading at that number, but raw Gainesville acreage without an executed interconnection queue position is still clearing closer to $500-600k in deals I've seen paperwork on through Q1 this year. That gap matters a lot if the hyperscaler doesn't renew and you're selling land, not a powered shell. Your gap note returning 22% annualized is real money, but if the exit model is quietly assuming $900k land value, I'd stress that number pretty hard before you feel good about the structure.

The thing I have never seen work is the assumption that a hyperscaler stays quiet through a lease roll and just renews on similar economics. What actually happens is they either expand aggressively, which triggers a whole capital conversation you are not positioned for as a gap lender, or they go dark and the asset reverts to a land play with a concrete slab and stranded power infrastructure that a new tenant has to engineer around. Your 38 acres in that Gainesville corridor is real, the $900k comps are real, but the slab and the cooling layout are not neutral, they constrain the next buyer's imagination and that shows up in the cap rate conversation two or three years before the lease actually expires. I put a second behind a similar structure in Ashburn in 2021, smaller deal, about 6 MW, and by the time I started asking hard questions about what the land looked like unencumbered by that specific build-out, nobody had a clean answer. The 22% annualized sounds right for the risk you were actually carrying, which was not income risk, you are correct about that, it was optionality risk on a very specific piece of ground that may or may not pencil the same way once someone has to price the demolition.