3 Things Smart Investors Know About the Nuclear Power Comeback

Nuclear power is gaining long-term support as more than 20 countries committed at COP28 to triple capacity by 2050, 14 major financial institutions backed that objective in September 2024, and the World Bank lifted its longstanding nuclear-financing ban in June 2025. AI hyperscalers are driving demand for 24/7 power: Microsoft signed a 20-year PPA with Constellation tied to restarting Three Mile Island Unit 1, while Meta entered 20-year nuclear PPAs with Constellation and Vistra. Small modular reactors could reduce deployment timelines to roughly 24-36 months, with GE Hitachi BWRX-300 projects targeted for Ontario in 2029-30 and under consideration by TVA.

Analysis

The investable bottleneck is not nuclear demand but deliverable firm power: licensed sites, interconnection rights, fuel conversion/enrichment, and construction execution. CEG and VST monetize scarcity now through contracted and merchant power optionality, whereas SMR developers remain duration assets whose valuations require serial deployment rather than announcements. Hyperscaler procurement can also raise regional power-price curves, benefiting incumbent generators and transmission owners more reliably than the technology vendors.

Over the next 1-3 months, data-center load forecasts, utility integrated-resource plans, and new long-duration PPAs are the relevant catalysts; broad policy endorsements have limited incremental valuation impact. The key risk to CEG/VST is that hyperscalers accept gas-backed capacity or curtailable load structures, reducing the premium for 24/7 clean power. A sharp decline in forward power prices, weaker data-center capex guidance from MSFT/META, or adverse capacity-market outcomes would falsify the near-term scarcity thesis.

The consensus error is treating all nuclear equities as a homogeneous AI-power trade. GEV has a more balanced risk/reward because grid equipment, electrification and gas-turbine demand can generate earnings before SMR economics are proven; SMR and OKLO carry financing, fuel-supply, licensing, and first-of-a-kind construction risk that can force dilution well before meaningful revenue. Structural upside over 6-18 months is strongest in existing nuclear fleets and grid suppliers, while pure-play SMR upside is better expressed only after firm customer deposits, construction notices, and funded fuel arrangements.

Key Decisions for Investors

• Maintain a 6-12 month long CEG / short SMR pair: own contracted existing-fleet cash flows and scarcity pricing while hedging nuclear-theme beta. Reassess if CEG forward power capture weakens materially or SMR secures a fully financed, binding construction order with customer-backed deposits.
• Accumulate GEV on 5-10% pullbacks for a 12-18 month horizon; grid and generation equipment offer multiple demand pathways beyond SMRs. Thesis breaks on sustained backlog erosion, margin-guide cuts, or evidence that utility capex is being deferred.
• Treat OKLO as an event-driven watch rather than a core long until it discloses sufficient committed project funding, fuel sourcing, and a credible path through licensing. Any position should be sized as venture-style optionality with a 50%+ drawdown tolerance.
• Monitor MSFT and META quarterly capex and regional load disclosures over the next two earnings cycles. Upward power-procurement commitments support CEG/VST; a shift toward gas, demand response, or slower AI infrastructure deployment is the signal to reduce exposure.