AI boom COLLIDES with local data center opposition

Alex Epstein (Center for Industrial Progress) says opposition to data centers has blocked or delayed 75 projects totaling $130B. The pushback highlights potential constraints on future data-center buildout and associated power/infrastructure needs. While specifics on market effects aren’t quantified beyond project delays, the scale suggests a meaningful headwind for expansion plans tied to AI infrastructure.

Analysis

The market should treat this less like a demand shock and more like a supply-chain bottleneck that reallocates profit pools. If new capacity becomes politically harder to permit, the value migrates to companies that already control energized sites, interconnects, and power contracts: colo incumbents, grid equipment vendors, and utility-scale generators. By contrast, speculative greenfield developers and land banks priced off rapid build assumptions face longer cash-conversion cycles and higher financing costs, which can force multiple compression even if end-demand remains intact.

The first-order reaction is likely a modest hit to names exposed to fast capacity additions, but the bigger 1-3 month effect is regional scarcity: hyperscalers will chase friendlier jurisdictions, raising bids for power, transmission, and existing shells. Over 6-18 months, that should improve pricing power for operators with scarce MW, but only if permitting friction does not metastasize into state-level moratoria or transmission bottlenecks. The key falsifier is a policy response that accelerates interconnect approvals or a clear slowdown in hyperscaler capex guidance; either would blunt the scarcity premium.

The contrarian point is that opposition can be bullish for the best-positioned incumbents because it limits overbuild and preserves returns on existing assets. If the AI cycle remains intact, constrained supply may actually widen the moat of firms already plugged into the grid, while punishing anyone whose model depends on frictionless expansion.

Key Decisions for Investors

• Long EQIX/DLR on pullbacks over the next 1-3 months: scarcity of new supply should support occupancy, lease spreads, and renewal economics; use a 6-18 month horizon and cut if cloud capex meaningfully slows.
• Long VRT or ETN into weakness: electrical content per MW rises as projects become more complex and delayed builds still require equipment; this is a cleaner beneficiary than speculative developers.
• Pair trade: long CEG / short CORZ or IREN for 3-6 months — own existing generation and power access, short names whose valuation depends on rapid new data-center capacity conversion.
• Watchlist alert: if state or federal permitting reform emerges, cover shorts in speculative data-center developers quickly; that is the main catalyst that would reverse the scarcity thesis.