The live alternative to CBRE / JLL H2 reports — sample it now.

Transaction volume exploded in June, with 293 deals totaling $748.9 billion across 359,137 MW—a 1,027% month-over-month surge from May's 26 deals. Blackstone's $105 billion acquisition of Anthropic's data-center assets on June 10th anchored the month, but the real story is breadth: deal count increased 14,550% year-over-year, signaling a fundamental shift in how hyperscalers are acquiring capacity. Geographically, Richland Parish (Louisiana) and the unnamed "One" market each claimed 5,000 MW of the month's transacted volume, with Las Cruces (New Mexico) at 4,500 MW and Ashburn (Virginia) at 4,296 MW across 161 facilities. These four markets concentrated 18,796 MW—52% of monthly volume—underscoring continued preference for power-dense, low-cost jurisdictions and legacy fiber hubs.

This velocity reframes the next two quarters as a capacity-crunch inflection. The market has shifted from on-balance-sheet development to aggressive portfolio acquisition, a posture typically adopted when grid-queue depth and new-build lead times exceed hyperscaler deployment windows. With global installed capacity at 849.7 GW and only 40 facilities added in June, net capacity growth remains constrained relative to transaction appetite, suggesting buyers are cannibalizing existing supply rather than waiting for greenfield projects to reach online. Expect M&A multiples to compress, secondary-market (non-hyperscaler) divestiture activity to accelerate through Q3, and power procurement hedges to tighten in Ashburn and ERCOT-adjacent markets where demand concentration is highest.

The single biggest shift in Q2 2026 was not capacity deployment or headline deals—it was a decisive market verdict against Western Europe and a corresponding capital reallocation toward inland North America. Across 313 markets scored, 175 received CAUTION verdicts and 120 received AVOID ratings, with Dublin, London, Amsterdam, Frankfurt, and Manchester forming a clear exclusion zone for institutional capital. Simultaneously, Blackstone deployed $165 million across four separate transactions in the 90-day window, while AirTrunk added $30 million, signaling that PE dry powder is flowing away from the traditional gateway hubs and toward frontier power-constrained regions. The operative pipeline tells the story: 11,063 facilities remain operational (91 GW), but only 154 are under active construction (33.7 GW), indicating that available capital is not driving incremental brownfield expansion but rather acquiring and consolidating existing stock in defensible markets.

This window reveals a structural inversion in capital allocation that will define the next two to four quarters: hyperscaler and megafund appetite have bifurcated. AWS (504 facilities, 14.1 GW) and Microsoft (301 facilities, 11.1 GW) continue to anchor the operational base, but their new investment posture has shifted from gateway acquisition to power-secured inland development—a position now confirmed by the 377 M&A transactions totaling $807.3 billion across the 90 days. The verdict distribution (56% CAUTION, 38% AVOID, 6% BUILD) is not a market weakness signal; it is precision. Those 18 BUILD markets—anchored by Williston, ND (MISO), Cheyenne, WY (WECC), and Rural SPP—are power-abundant, interconnect-accessible, and unencumbered by the regulatory friction that has strangled Northern Europe. PE capital, no longer competing for trophy assets in saturated coastal markets, is bidding aggressively for scale in secondary clusters where power is the binding constraint solved.

Monitor three signals through Q3 and Q4: first, the velocity of Blackstone and Brookfield M&A activity in Williston and Upper Peninsula Michigan—any slowdown signals PE capital fatigue and a return to coastal consolidation; second, whether AWS or Microsoft announce >500 MW anchors in Rural SPP or Cheyenne, which would confirm the inland shift is not speculative but structural and multi-year; and third, European facility disposition announcements—London and Amsterdam assets are not vacant, and we will see whether hyperscalers divest or indefinitely hold, a choice that determines whether European CapEx restarts or contracts into 2027. The power gap is real, visible, and now capital-constrained. Position accordingly.