Meta Platforms partners with Vistra, TerraPower and Oklo to power AI supercluster with nuclear energy

Meta Platforms has signed partnerships with Vistra, TerraPower and Oklo to secure up to 6.6 GW of nuclear-sourced electricity by 2035 to power its AI operations, including the Prometheus supercluster in New Albany, Ohio. Deals include purchases from existing Vistra plants, funding of TerraPower Natrium reactors in the early 2030s and support for an Oklo nuclear campus slated around 2030; the agreements follow a prior Constellation deal and aim to provide reliable, low-carbon capacity for Meta’s data centers. The announcements sent Oklo and Vistra shares up ~13% and drew positive analyst commentary (Wedbush ‘Outperform’ and $150 target on Oklo), underscoring potential upside for nuclear developers and reduced energy risk for Meta’s AI expansion.

Analysis

Market structure: Hyperscaler offtakes crystallize a new baseload demand bucket — Meta’s deals signal ~6.6 GW of incremental contracted nuclear demand by 2035, shifting bargaining power toward advanced-nuclear developers (OKLO, TerraPower) and incumbent baseload operators (VST, CEG). Regional wholesale markets (e.g., PJM/MISO) may see downward pressure on peak-terrace volatility but upward pressure on capacity prices as grid operators value dispatchable, low-carbon firm capacity.

Risk assessment: Tail risks are regulatory/political reversals, NRC licensing delays, and multi-year cost overruns (Vogtle-style) that can wipe out equity returns; probability moderate but impact high for small-cap developers (OKLO). Time horizons diverge: immediate (days) = sentiment repricing; short-term (3–12 months) = financing, permits and DOE/loan guarantee updates; long-term (3–10 years) = construction, operations and realized contracted revenues.

Trade implications: Favor asymmetric, time-limited exposure to OKLO’s optionality and defensive positions in large regulated generators. Use defined-risk option structures to capture funding/permit de-risking over 12–24 months; rotate modestly into uranium/industrial suppliers should multiple offtakes follow. Reduce exposure to merchant gas peakers in markets where capacity value is re-priced toward firm low-carbon supply.

Contrarian angles: Consensus overlooks execution risk and financing sensitivity to rates — nuclear economics are interest-rate and capex dependent; historical parallels (Vogtle, UK nuclear projects) show >2x cost overruns and multi-year delays, implying current positive moves could be overdone for small developers. If USD rates stay >3.5% or political opposition rises in key states, re-rating could be sharp and fast.

Key Decisions for Investors

• Establish a 1.5–3.0% long position in OKLO (OKLO) sized to portfolio risk tolerance with a 12–36 month horizon; hedge with a 12–18 month protective put at ~20% cost of the equity position or implement a 12–18 month call spread (buy 20–30% OTM LEAP, sell 60% OTM) to cap premium — exit/trim on +50% move or if major NRC permit milestone misses a 12‑month target.
• Add a 1–2% core utility position in Vistra (VST) to capture near-term visibility and dividends; scale in on >8% pullbacks and set a sell trigger if forward PJM power curve (Cal ’26–’28) falls >10% from current forwards or if Vistra issues a profit-warning tying to large nuclear capex overruns.
• Initiate a 0.5–1.0% strategic tech tilt in Meta (META) to own AI upside and energy-cost de‑risking; monetize with 3–6 month covered calls (~10–15% OTM) to fund cost of carry and exit on underperformance vs. Nasdaq‑100 by >8% over a rolling 60‑day window.
• Establish a 0.5–1.5% commodities/industrial exposure to uranium (e.g., URNM/URA) and specialty steel suppliers with a 6–24 month horizon — increase exposure by another 1% if three additional hyperscalers announce firm long‑term baseload offtakes within 12 months.