How much of the data center adjacent demand story belongs in an industrial rent assumption
The absorption rebound projected for 2026, toward roughly 220 million square feet, leans on three drivers: reshoring, manufacturing, and data-center-adjacent demand. The first two show up clearly in leasing tours. The third is where underwriting gets harder. Data-center-adjacent industrial demand is real: electrical contractors staging gear, switchgear and transformer storage, module assembly, generator service outfits. Buildings in the right submarkets are seeing LOIs from exactly that kind of user, often wanting more power than the building has and a large yard. The open question is durability. That demand is tied to a capex cycle that can compress quickly, and the tenants are often subcontractors rather than credit tenants. If a hyperscaler slows a campus, a three-year staging need can become a sublease request within a quarter. The practical choice is whether to underwrite today's power-and-yard premium into the rent assumption, or underwrite the building at generic logistics rent and treat the premium as unpaid-for upside. Underwriting conservatively at generic logistics rent protects against the capex cycle turning, but it also means losing deals to buyers pricing in the premium. Whether that is discipline or being too slow depends on how much conviction there is in the durability of that specific submarket's demand driver, and that judgment should be revisited regularly rather than fixed once.
How do you underwrite data-center-adjacent rent premiums?
17 votes