Map shows growing number of data centers planned across Pennsylvania

A Silicon Valley technologist, Emilia Doda, has created and is publicly updating an open-source map that catalogs proposed and planned data centers across Pennsylvania to inform local communities. The tool highlights growing data center siting activity and the attendant demands on local land, water and electricity—issues relevant to investors and municipal planners—but contains no company-specific financial data and is unlikely to move markets in the near term.

Analysis

Market structure: Local mapping of data-center buildouts signals incremental demand for colocation and hyperscaler footprint expansion in PJM/NEPA, benefitting Digital Realty (DLR), Equinix (EQIX) and CyrusOne (CONE) as pricing power for premium metro interconnects can rise 5-15% regionally over 12–24 months. Upstream winners include utilities/merchant generators (NRG, AES) and copper/transformer suppliers; losers are local land/retail uses and municipalities that lose taxable land-use variety. Cross-asset: modest upward pressure on regional power curves (PJM day-ahead prices +2-6% over 2–3 years) and potential muni credit upside from PILOTs, while sovereign FX impact is nil.

Risk assessment: Tail risks include local moratoria or aggressive PILOT renegotiations that can delay projects 12–36 months, or grid interconnection bottlenecks forcing curtailments and stranded asset write-offs for colo REITs. Short-term (0–3 months) sensitivity centers on permitting and interconnection queue metrics; medium-term (3–12 months) on utility capex and capacity additions; long-term (1–5 years) on AI-driven demand concentration or edge-compute shifts reducing wholesale colocation growth. Hidden dependency: availability of low-cost firm power and transmission upgrades is the gating constraint — track PJM queue depth and utility rate cases.

Trade implications: Tactical 6–18 month longs in DLR/EQIX/CONE capture re-rate if regional builds proceed; hedge operational risk with short-dated puts or pair trades vs broader REIT indices. Utility/merchant generator exposure (NRG, AES) serves as a hedge to rising power prices — prefer names with PJM exposure. Use options (3–9 month call spreads on EQIX/DLR) to limit premium while targeting 15–30% upside if permit cadence accelerates.

Contrarian angles: Consensus underweights the political friction risk — public mapping increases visibility and could accelerate local pushback, creating idiosyncratic delays not priced into national REIT multiples. Markets may be underpricing grid constraint externalities that favor battery-storage and transmission contractors (BLNK? INFRA names) more than pure-play colos. Historical parallel: shale boom’s local tax disputes; expect similar PILOT renegotiation cycles that can compress IRRs by 200–500bps if widespread.

Key Decisions for Investors

• Establish a 2–3% portfolio position split equally between Digital Realty (DLR) and Equinix (EQIX) within 30 days to capture regional colocation demand; trim or take profits at +20% or if fewer than two major PA projects receive final permits within 6 months.
• Initiate a 1.5% long position in NRG Energy (NRG) or AES (AES) to play rising PJM power demand; target 12–18 month horizon and exit if realized PJM LMPs fail to rise >2% year-over-year over two consecutive quarters.
• Buy a 3–9 month call spread on EQIX (e.g., buy 9-month 5% OTM call, sell 9-month 15% OTM call) sized at 0.5–1% notional to capture a 15–30% upside re-rate while capping premium; roll or unwind if interconnection queue approvals stall for >90 days.
• Avoid/leverage short exposure to small regional RE developers or muni bonds of counties facing contested PILOTs; reduce exposure by 50% if county-level opposition motions or moratoria increase by two or more within 90 days — monitor local permitting calendars and county board minutes weekly.