Trump pushes emergency power auction to make Big Tech pay for soaring energy costs

The Trump administration has urged PJM Interconnection — which serves roughly 67 million customers across 13 states and D.C. — to run an emergency auction that would let large tech firms bid on 15-year contracts from newly built power plants, aiming to curb rising electricity prices and blackout risk tied to the data-center boom. Electricity costs are already up 6.7% year-over-year per the CPI, and academic work projects data centers and crypto could lift average bills by more than 25% in major markets by 2030; PJM also floated requiring large data centers to secure their own supply or face curtailment. The proposal could shift costs onto Big Tech, tighten future contracting and generation markets, and create earnings/expense pressure for cloud providers while benefiting developers of new generation capacity.

Analysis

Market structure: Emergency auctions and threats to curtailment shift economic rent from hyperscalers to generators and contracted suppliers. Winners are IPPs, battery/renewables developers and regulated utilities able to recover costs (e.g., NextEra, AES, Exelon); losers include data-center REITs and operators facing forced curtailment or mandated 15-year PPAs that increase marginal costs (pressure on AMZN margins by mid-single-digit percent over 12–24 months). This re-prices forward power curves and shortens liquidity on merchant capacity, increasing contract tenure and developer financing visibility.

Risk assessment: Tail risks include federal/state mandates that force immediate curtailments or retroactive cost allocation to Big Tech (low-probability, high-impact) and pre-election politicized rulings that amplify volatility. Immediate (days) — headlines/requests to PJM will move data-center names; short-term (weeks–months) — PJM rulemaking and state utility commission cases; long-term (1–3 years) — new-build capacity, higher retail bills, and capex rotation into generation/storage. Hidden dependencies: many hyperscalers already hold PPAs and on-site generation; auctions may primarily benefit new-build renewables with storage, not incumbents.

Trade implications: Favor regulated / contracted cashflows: long NEE or XLU-style positions and developers of co-located generation/storage (AES) for 6–18 months. Tactical shorts or put spreads on Equinix (EQIX) and CyrusOne (CONE) reflect curtailment and rate pass-through risk over 3–9 months. Use option spreads to cap cost: buy 3-month puts on AMZN or EQIX (10% OTM) and sell deeper OTM to finance.

Contrarian angles: The market underestimates hyperscalers’ negotiating power and existing hedges — many will internalize costs or build on-site generation, creating multi-year opportunities for EPC and turbine manufacturers (GE, CAT) rather than persistent stress on cloud margins. Reaction may be overdone for AMZN near-term; if PJM backs a voluntary auction instead of mandates, data-center names should recover quickly. Monitor PJM final rule and upcoming state rate-case dockets as binary catalysts.

Key Decisions for Investors

• Establish a 2–3% portfolio long in NextEra Energy (NEE) over 6–18 months to capture likely wins in long-term PPAs and regulated renewables revenue; trim if NEE rallies >15% or if PJM rejects emergency auction.
• Initiate a 1–1.5% short position in Equinix (EQIX) or 1% short each in EQIX and CONE for 3–9 months, targeting a 15–30% downside if forced curtailment or higher-backed-out rents hit data-center EBIT; cover if EQIX announces multi-year self-supply PPAs within 60 days.
• Buy a 3-month put spread on AMZN sized to 1% notional (buy 1x ~10% OTM put, sell 1x ~20% OTM) to hedge regulatory headline risk and potential margin pressure; roll or exit after PJM decision (~30–90 days).
• Rotate 3–5% from large-cap software/cloud exposure into utilities/renewables ETF (XLU) and a developer like AES for 6–12 months to harvest yield and contracted cashflows; exit if 10-year Treasury yield rises >75bps (tightens financing) or energy CPI prints normalize below +3% YoY.