I’m a data center CEO. Communities deserve better partners, not blanket bans

New York imposed what it calls the country’s first statewide moratorium on new large-scale data center construction, with additional freeze/restriction bills appearing in states such as Maine and Georgia. The piece argues blanket bans won’t solve underlying concerns (e.g., grid/transmission costs to ratepayers, water usage, and community disruption) and instead calls for tougher, measurable standards for transparency and infrastructure funding. Overall, the policy direction suggests heightened regulatory risk and added development friction for data center operators, even as demand for capacity remains.

Analysis

This is less a demand shock than a friction shock: the industry does not disappear, but the hurdle rate rises because permitting risk, interconnect timing, and community-exaction costs move from nuisance to valuation input. That tends to compress returns on new builds first, then widen the spread between incumbents with powered land, operating campuses, and local relationships versus developers whose model depends on fast entitlement and flip economics.

The near-term market issue is not revenue, it is schedule slippage and capex inflation. If more states follow this path, hyperscalers and colo tenants will likely re-route projects toward friendlier jurisdictions, which could lengthen lease-up cycles in constrained metros but also tighten supply and support pricing for existing capacity. That is structurally supportive for the best-positioned REITs, while punishing developers whose equity stories assume smooth scaling.

The contrarian miss is that community pushback can be bullish for the wrong names: scarcity often improves pricing power for installed capacity and raises the value of every megawatt already entitled and energized. The bigger risk is political contagion; if this evolves from a local permitting issue into a broader cost-shift debate around grids and water, the timing on AI infrastructure spend could slip by quarters, not weeks. Reverse the thesis if states pivot from moratoriums to standardized fee-and-commitment frameworks, or if utility commissions allow fast-track interconnect recovery without ratepayer backlash.

Key Decisions for Investors

• Long EQIX / short DLR for 1-3 months as a relative-value expression of entitlement scarcity. EQIX should be less exposed to new-development friction, while DLR has more sensitivity to pipeline timing; stop if DLR guidance shows no delay in development starts or if moratorium risk fades from state legislation.
• Add PWR and ETN on weakness as a multi-quarter beneficiary basket. If developers must fund substations, transmission, and onsite power more often, grid-build and electrical equipment demand becomes more private-capex-driven; thesis breaks if cost recovery shifts back to utilities or if project cancellations dominate over infrastructure spend.
• Avoid chasing broad AI-infrastructure beta until the policy path is clearer. The right alert is on hyperscaler capex guidance from AMZN, MSFT, GOOGL, and META: if any of them cite site-permitting delays or power availability in the next 1-2 quarters, that is the point to short high-multiple AI infrastructure names.
• If state moratorium headlines spread beyond New York over the next 30-60 days, consider a tactical short XLC vs long XLRE only if data-center supply constraints show up in leasing data. Without evidence of delayed capacity, this is more of a regulatory overhang than a sector-wide earnings cut.