Meta signs three nuclear power deals to help support its AI data centers

Meta has signed nuclear power agreements with TerraPower, Oklo and Vistra to supply its Prometheus AI data center cluster in New Albany, Ohio, a 1‑GW multi‑building facility expected online this year. The deals, financial terms undisclosed, are structured to support up to 6.6 GW of new and existing clean energy by 2035: TerraPower funding aims to develop two Natrium units (up to 690 MW deliverable as early as 2032) plus rights to energy from up to six additional Natrium units (≈2.1 GW by 2035); Vistra will supply more than 2.1 GW from two Ohio plants and planned expansions including a Pennsylvania site; Oklo will help develop a 1.2 GW campus in Pike County, Ohio. Combined with Meta’s prior 20‑year Constellation deal, these agreements secure long‑duration, firm power for AI operations and reinforce U.S. nuclear supply‑chain buildout.

Analysis

Market structure: Meta’s deals (Prometheus 1 GW online this year; commitments supporting up to 6.6 GW by 2035) shift value to firm baseload providers — direct winners are Vistra (VST), TerraPower/Oklo (developers) and Meta (reduced site power risk), while merchant gas peakers and regional capacity sellers in Ohio risk losing marginal pricing power. Capacity markets and bilateral PPA pricing will reprice toward baseload valuation in markets where these plants interconnect; uranium and HALEU supply chains see multi-year demand tailwinds.

Risk assessment: Key tail risks are regulatory/licensing delays, cost overruns and HALEU shortages that could push deliveries beyond 2032–2035 and meaningfully reduce NPV; near-term (days) impacts are sentiment, short-term (weeks–months) are re-rating of VST/META, long-term (years) are realization of cashflows and grid integration. Hidden dependencies include transmission upgrades, NRC approvals, and skilled labor — failure on any raises execution risk and forces financing dilution or contract renegotiation.

Trade implications: Prefer selective long exposure to VST (utility with Ohio nuclear footprint) and modest META exposure to de-risk AI operations; add uranium ETF (URA) for commodity upside if capex paths remain. Tactical instruments: buy 12-month VST call spread (ATM to +20% strike) to cap premium, pair long VST vs short NRG (dollar-neutral, 1–2% portfolio notional) to express baseload/gas divergence; small venture/secondary allocation (0.25–0.5%) to OKLO contingent on NRC milestones by 2028.

Contrarian angles: The market underestimates schedule risk — headlines overstate near-term flow-through to earnings, so valuation rerates may be overdone before plants generate cash. Historical parallels (hyperscaler renewables PPAs) show multi-year gestation with intermittent stock volatility; unintended consequences include higher utility leverage and regional political pushback that can widen credit spreads and impair returns.

Key Decisions for Investors

• Establish a 2–3% long position in VST (Vistra) within 1–4 weeks; set a 12-month target of +20–30% and a hard stop-loss at -12% to protect vs execution/regulatory risk.
• Allocate 2–3% long position in META now to hedge AI energy risk; if META falls >10% add incrementally, and consider selling 3–6 month covered calls to monetize near-term volatility.
• Buy a 12-month VST call spread (ATM to +20% strikes) sizing to 0.5–1% portfolio risk to express upside while limiting premium outlay; unwind on 25% realized profit or if NRC/state permits fail.
• Initiate a dollar-neutral pair trade: long VST vs short NRG (NRG) at 1:1 dollar exposure, total sizing 1–2% portfolio, expecting baseload nuclear to gain vs gas-fired merchant generation over 6–18 months.
• Place a tactical 0.5–1% position in URA (Global X Uranium ETF) for commodity upside to 2028–2035; open a 0.25–0.5% venture/secondary allocation toward OKLO or TerraPower exposure conditioned on demonstrated licensing progress by end-2028.