Should You Forget NuScale Power and Buy Oklo Stock Instead?

Bank of America says nuclear power could become a $10 trillion opportunity as U.S. electricity demand rises from AI data centers, industrial load growth, and electrification. The article highlights two SMR plays, NuScale Power and Oklo, with different go-to-market strategies: NuScale is targeting a 6 GW utility-scale system for TVA, while Oklo has a 1.2 GW deal with Meta for AI data centers. The piece is generally constructive on nuclear energy demand, but it is largely commentary rather than new company-specific operating data.

Analysis

The market is beginning to price nuclear as a power-supply solution to AI load growth, but the first-order trade is less about broad utility exposure and more about who captures the premium for speed, site control, and customer specificity. That favors OKLO structurally in the near term because data-center buyers can justify paying up for dedicated behind-the-meter reliability, while SMR is a longer-dated utility procurement story that is more exposed to regulator, rate-base, and permitting cadence. In other words, the economic value of nuclear is shifting from commodity electrons to contracted uptime, and that changes the mix of winners across developers, EPCs, grid equipment, and industrial gas/thermal management vendors.

The second-order risk is that the SMR narrative can get ahead of deployability. Even if enthusiasm remains high for months, actual value creation depends on financing, licensing, and supply-chain execution over years; any delay pushes these names back into a capital-market dependence cycle where dilution becomes the hidden cost of optionality. A more subtle competitive dynamic is that the biggest beneficiaries may be non-obvious incumbents with nuclear-adjacent capabilities—large industrials and engineering firms that can sell design, safety systems, fuel handling, switchgear, and long-lead components without taking full technology risk.

Consensus is probably underestimating how binary this setup is: either a handful of flagship projects create an investable template, or the market spends another cycle rewarding story over throughput. The near-term catalyst path is not broad utility adoption but customer announcements, power-purchase structure, and evidence of project de-risking; absence of those updates over the next 3-6 months would compress multiples quickly. Conversely, a single credible financing/FOAK milestone could re-rate the group sharply because the market is paying for scarcity of validated commercialization, not current cash flow.

Key Decisions for Investors

• Long OKLO / short SMR as a 3-6 month relative-value pair: OKLO has a clearer path to monetizing AI-driven willingness to pay; SMR is more exposed to utility-cycle delays and capital dilution risk.
• Buy OKLO on pullbacks only after project-validation headlines; use a starter position with a 20-25% downside stop because sentiment can outrun execution, but upside can extend 40-60% on any credible customer/financing milestone.
• Avoid outright long SMR ahead of visible licensing progress; if you want nuclear exposure, prefer a basket with smaller sizing until commercial FOAK risk is reduced over the next 6-12 months.
• Pair long META with a small hedge in OKLO if looking for AI infrastructure beta: META has direct upside from secured power, while OKLO captures the supply-chain value of that spend; this is a cleaner expression than owning generic power utilities.
• Watch for a sympathy rally in nuclear-adjacent industrials and grid equipment suppliers over the next 1-2 quarters; these names may offer better risk/reward than the SMR developers because they monetize project activity without binary reactor execution risk.