Meta's massive nuclear power deals will help US 'win' AI race against China, executive says

Meta has signed 20-year power purchase agreements to buy nuclear power from Vistra’s Perry and Davis-Besse plants in Ohio and the Beaver Valley plant in Pennsylvania and will support small modular reactor development with Oklo and TerraPower; the deals could supply up to 6.6 GW of capacity by 2035. Combined with a prior Illinois nuclear deal, this positions Meta as one of the largest corporate buyers of U.S. nuclear power, materially undergirding its data-center energy needs, supporting local skilled jobs, and strengthening the company’s argument that reliable, large-scale baseload power is strategic to U.S. competitiveness in the AI race with China.

Analysis

Meta’s 20‑year PPAs and stated 6.6 GW by 2035 materially tilt corporate demand toward firm low‑carbon baseload; immediate winners are Vistra (VST) for near‑term contracted cashflows and SMR developers (OKLO/TerraPower) for optional long‑dated upside. Merchant gas peaker operators and regional spark‑spread dependent generators are at risk of lower utilization and volatile margins as baseload supply increases regionally, which should cap wholesale price spikes and reduce power volatility over multi‑year horizons.

Primary tail risks are NRC/licensing delays, construction cost overruns, HALEU fuel bottlenecks and political/regulatory shifts that can push delivery past 2035 and destroy developer equity; expect headline moves in days, financing/permitting battles over 3–12 months, and realization (or failure) of projects across 3–10 years. Hidden dependencies include transmission upgrades, interconnection queue positions and supply‑chain inflation; key catalysts to watch are NRC combined licenses, DOE cost‑share awards and first SMR commissioning dates.

Actionable trades: favor long VST for secured PPA exposure and small, option‑capped exposure to OKLO (binary development upside). Rotate modest exposure from broad energy (XLE) into regulated/contracted utility exposure (XLU) and AI infrastructure beneficiaries (META, NVDA) because energy certainty lowers operating risk for data center expansion; use pair trades and LEAPs to control risk and capital.

Contrarian view: the market underprices execution and grid bottleneck risk — rallies in small SMR names may be overdone absent near‑term permitting wins. Historical PPA waves (pre‑2010 renewables) produced durable asset build but also extended interconnection delays; expect similar two‑to three‑year bottlenecks that create tactical opportunities in grid services and capacity markets.

Key Decisions for Investors

• Establish a 2–3% long position in VST (Vistra) common stock targeting 12‑month upside of 25–35%; place a 12% stop‑loss and add 1:1 exposure with 12–18 month LEAP calls (strike ~170–190) to lever upside while controlling cash exposure.
• Allocate 0.5–1% speculative position to OKLO via 2026–2028 call spreads (buy lower strike, sell higher strike) to cap premium; set a hard review at 12–18 months and exit/roll if NRC/DOE milestones are not achieved within that window.
• Rotate ~2% of portfolio from XLE into XLU (1%) and META (1%) over the next 3 months; implement a 1–2% pair trade long VST / short XLE to express secular shift from thermal peakers to contracted nuclear baseload over 12 months.
• Buy Vistra corporate or project bonds (3–7 year maturities) sized 3–5% of FI sleeve if yield‑to‑worst >5.5% or credit spread >200bp vs Treasuries, as long‑dated PPAs materially improve cashflow visibility and should compress spreads on financing milestones.