Logistics Is the Hidden AI Trade in Real Estate

What's New

Warehouses have become part of the AI infrastructure build, and most investors still treat them as a bet on online shopping. A.J. Agarwal, President and Director of Blackstone Real Estate Income Trust (BREIT), argues this in a conversation with David Levine, Global Co-Head of Blackstone Real Estate, on Inside Blackstone. About 15% of new leasing in Blackstone's US logistics business now comes from data center related uses. 3 years ago, that share was basically 0%. Allocators seeking AI exposure should add warehouse leasing data to the metrics they track.

Why It Matters

The argument places AI demand inside an asset class many portfolios already hold. It challenges the view that logistics is a mature e-commerce trade with its growth behind it. Levine ties Blackstone's edge to scale. As the largest data center developer in the world, the firm sees where projects will land and buys warehouses in those markets. Blackstone sells this view through its $60 billion flagship US real estate fund for individual investors. Warehouses and data centers make up about 50% of it.

Between The Lines

• Blackstone's data center pipeline steers where its logistics capital goes. Levine frames this as visibility. It also means both books rest on the same forecast of where compute gets built.

• The early-cycle pitch is aimed at fund flows. Agarwal says US real estate values are up about 9% from the bottom. They remain more than 10% below the recent peak. He pairs that with warehouses trading below replacement cost to argue the window is open now.

• Calling logistics the "sleeping giant" casts it as underpriced, which supports selling a portfolio built around it.

Reality Check

• The 15% leasing figure comes from Blackstone's own portfolio. So does the ratio linking e-commerce sales to warehouse space. The conversation offers no outside benchmark for either.

• The linkage cuts both ways. If warehouses and data centers share one demand base, a combined 50% allocation is closer to a single bet than two. Logistics alone is about 40% of Blackstone's global exposure, by design.

• Levine calls agentic commerce very nascent. Agarwal admits he can't say which AI applications will win. The data center case needs hyperscaler capex to stay high, and this year's is about $800 billion.

Memorable Quotes

• "Online shopping may be harmful to shopping malls, but it's actually really beneficial to warehouses." Agarwal describes the method behind the thesis: find the physical asset that wins when behavior shifts.

• "We are the largest data center developer in the world. As a result, we have really good visibility into where these projects are going to be built." Levine names the information advantage that links Blackstone's data center and warehouse investing.

• "We are intentionally not pie chart investors." Levine defends a heavy logistics concentration, which leaves the portfolio exposed if the demand link weakens.

• "The cloud is not up there someplace. The cloud is on the ground and it's a data center." Agarwal reduces the AI and e-commerce story to a claim about land and buildings.

The Wrap

Blackstone's position works as long as supply stays scarce in both sectors and data center users keep taking warehouse space. Rising construction costs support the first condition. The second needs hyperscaler spending to keep producing the productivity gains Levine says justify it. A pullback there would hit Blackstone's data center and logistics books together, since they now draw on one demand base. The first years of the multiyear recovery Levine forecasts will show which way it breaks.