Data center construction has already dwarfed the entire institutional core real estate market
Josh Pristaw puts a number on the data center buildout that reframes the scale of what is happening. The pipeline is not merely large relative to prior data center investment. It is large relative to the whole of institutional real estate.
The number Josh Pristaw puts on the data center buildout is not one that fits easily into a conventional real estate analysis. The entire institutional core real estate index, he notes, stands at $280 billion. The data center pipeline currently under construction or in planning represents roughly three times that figure.
That comparison deserves a moment of stillness. The institutional core real estate index is the aggregate of stabilized, income-producing property that large institutions, pension funds, sovereign wealth vehicles, and endowments, hold as their baseline real asset exposure. It is the conservative, long-duration spine of institutional capital allocation. It is not a niche. It is, by design, the whole of the market.
What Pristaw’s framing establishes is that the capital being committed to data center development is not simply a large sectoral bet. It is a reordering of physical investment priorities at a scale that has no obvious precedent in commercial real estate history. Three times the institutional core index, in a single asset class, in a single construction cycle, is a different kind of fact than a percentage-point shift in office or multifamily allocations.
The whole institutional core real estate index is $280 billion so like there's 3x all of the core real estate owned by institutions is under construction or in planning in data centers. Josh Pristaw
The composition of that pipeline matters too. Pristaw’s framing covers both projects under active construction and those in planning, which means the committed capital includes activity that has not yet broken ground. That distinction is worth holding. A pipeline of this size that is partly prospective carries execution risk that a fully built stock does not. But the scale of what is already in motion, even setting aside the planning portion, represents a physical and financial commitment that will take years to resolve in either direction.
For institutional investors whose core real estate holdings are the $280 billion benchmark Pristaw cites, the implication is direct. The asset class they have treated as a stable, slow-moving store of value now sits in the shadow of a parallel construction effort that dwarfs it in absolute dollar terms. Whether that parallel effort competes for capital, competes for tenants, or simply redefines what counts as core real estate in the next decade are questions the existing frameworks were not built to answer.
Pristaw’s claim stands alone here, and it should be read as one speaker’s framing rather than a settled industry figure. But the specific comparison he draws, anchored to the $280 billion index as a known reference point, gives the claim a precision that distinguishes it from generic assertions about the AI buildout being large. The reference point is real. The ratio he places against it is sharp enough to warrant the attention of anyone whose capital is allocated against that index.
The broader question his framing raises is whether the institutions that own the $280 billion index have yet fully reckoned with what it means to have three times that figure in motion in a single competing asset class. The evidence here is one voice, not a consensus. But the arithmetic, if it holds, is the kind that does not get smaller on closer inspection.