These dividend stocks could catch a tailwind from data center pushback

Community and political opposition to AI data centers—including a New York moratorium and 69% local opposition in an NBC News poll—could constrain new supply and enhance pricing power for incumbent data-center REITs. PwC forecasts annual data-center spending rising from roughly $800 billion in 2026 to $1.8 trillion by 2050, supporting demand for existing capacity. Equinix, Digital Realty and Iron Mountain have gained 36.9%, 23.3% and 42.0% year to date, respectively, after second-quarter AFFO beats and raised full-year guidance; analysts favor Equinix and Digital Realty due to scale, permitted pipelines and secured power access.

Analysis

Local opposition converts data-center development from a capital-availability problem into a scarce-power and entitled-land problem. That favors EQIX and DLR disproportionately because their existing interconnection ecosystems, utility relationships, and pre-permitted pipelines can command higher renewal spreads and lower customer churn; smaller private developers face a materially higher probability of stranded land or delayed power delivery. The near-term benefit is strongest in constrained, high-connectivity metros, where a delayed competing project can shift enterprise demand into incumbent capacity rather than eliminate it.

EQIX is the cleaner beneficiary if AI shifts from centralized model training toward latency-sensitive inference: its network-dense, retail-colocation footprint should see better cross-connect and interconnection monetization, not merely incremental megawatt leasing. DLR has greater exposure to large hyperscale deployments and therefore more upside from capacity scarcity, but also greater sensitivity to a slowdown in hyperscaler capex or delayed utility energization. IRM offers a lower-purity data-center exposure; its digital infrastructure growth can support the multiple, but its legacy storage business makes it less direct and its strong share-price performance raises execution-risk asymmetry.

Consensus appears to treat restrictions as uniformly bullish for landlords. The adverse second-order effect is that political pressure can raise construction cost, lengthen lease commencements, and make tax incentives or electricity pricing less predictable; REITs may win on stabilized-asset pricing while losing on development yield. Over the next 1-3 months, permitting headlines likely support the group, but the 6-18 month thesis requires sustained bookings, improving lease spreads, and power-delivery schedules—not just AI demand rhetoric.

Key Decisions for Investors

• Initiate a 3-6 month long EQIX / short DLR pair, sized dollar-neutral: EQIX has the more differentiated inference and interconnection optionality, while DLR is more exposed to hyperscale project timing. Reassess if EQIX booking growth or adjusted EBITDA guidance trails DLR by more than one reporting cycle.
• Maintain DLR as a standalone long only on evidence that secured-power capacity is converting into signed leases; use quarterly leasing backlog, development yield, and energized-capacity milestones as entry triggers. Falsify on a material reduction in development starts or a guidance cut attributable to power/permitting delays.
• Do not chase IRM after its outsized momentum absent confirmation that data-center leasing is accelerating faster than consolidated AFFO expectations. A better setup is to buy on a post-earnings pullback if management reaffirms development funding without raising leverage or weakening its AFFO outlook.
• Use a long EQIX and DLR basket versus a short broad REIT proxy (VNQ) over 6-12 months only if Treasury yields remain contained: scarcity-driven NOI growth should differentiate data centers from office and slower-growth property types. Exit the relative trade if long-term yields rise sharply enough to overwhelm FFO multiple support or if hyperscaler capex guidance is broadly cut.