If a data center is camouflaged in the woods, will anyone hate it?

San Antonio's District 5 has more than a dozen data centers across roughly 55 square miles, with the count projected to reach about 20 in coming years if proposed hyperscale AI facilities proceed. The concentration along two main roads is increasing neighborhood concerns over traffic and other local disruptions, highlighting potential community and infrastructure constraints on data-center expansion.

Analysis

The relevant market signal is not incremental AI demand, but a rising local-permitting and social-license constraint on converting announced capacity into energized capacity. For Digital Realty (DLR), Equinix (EQIX), and private hyperscalers, the binding constraint increasingly shifts from land and shell construction to power delivery, noise mitigation, water use, and community concessions; this raises project cost and lengthens revenue commencement dates. In Texas, concentrated opposition can also make municipally controlled utilities and local zoning more important than state-level pro-growth policy.

The second-order beneficiary is the power-delivery stack rather than the data-center landlord: Vertiv (VRT), Eaton (ETN), Quanta Services (PWR), and selected generation owners such as Vistra (VST) and Constellation Energy (CEG) gain if customers must procure dedicated generation, substations, backup systems, transmission upgrades, and on-site cooling solutions. Over 6-18 months, stricter siting conditions may favor incumbents with already-energized campuses and developers able to absorb mitigation capex, while smaller speculative projects face financing and utilization risk. This is marginally negative for AI-capacity timelines, but not necessarily AI infrastructure spend; delayed grid access can increase spend per commissioned megawatt.

Consensus appears too focused on aggregate data-center pipeline announcements and too little on the gap between contracted land, utility interconnection, and operational load. The near-term equity effect is likely limited unless local opposition produces an actual permit denial, interconnection deferral, or incremental utility-rate mechanism; absent those events, this is a watch item rather than a standalone short. A reversal would be rapid if utilities secure new generation/transmission commitments or if developers demonstrate low-water, low-noise designs that neutralize neighborhood resistance.

Key Decisions for Investors

• Maintain a 6-12 month preference for VRT and ETN over DLR/EQIX: power-quality, cooling, and electrical-distribution content rises when deployment is constrained by local infrastructure. Reassess if backlog growth or book-to-bill decelerates for two consecutive quarters.
• Watch-list a long PWR / short DLR pair over 3-6 months only after evidence of permitting or energization delays at major Texas campuses. The thesis is capex shifting from data-center shell construction toward grid and substation work; exit if DLR lease commencements and development yields remain on or above guidance.
• For Texas power exposure, retain VST and CEG as conditional beneficiaries of high-load interconnection demand, but do not add solely on this report. Trigger additions only on disclosed long-term load contracts, approved capacity additions, or sustained ERCOT forward-price strength; downside is policy intervention, subsidized new supply, or lower AI utilization.
• Set alerts for municipal hearings, utility interconnection-queue updates, water-use restrictions, and new local noise/setback rules in major Texas data-center corridors. A documented delay or mitigation requirement is the catalyst to reduce exposure to speculative data-center development and increase exposure to electrical equipment contractors.