Nuclear Power Is the Energy Story of the Decade. This Stock Is Built to Last.

The article argues that AI-driven data center growth could create a large demand opportunity for NuScale Power’s small modular reactors, with global data center spending estimated to reach $7 trillion by 2030. NuScale is highlighted as the first SMR approved by the NRC and a potential faster, lower-carbon power source, though its first system may not be online until 2030 or later. The piece is largely promotional commentary rather than a new operational update, so near-term market impact should be limited.

Analysis

The market is treating SMR as a clean expression of AI-driven power scarcity, but the more important second-order effect is valuation dispersion across the power stack. If hyperscalers move from “net-zero narrative” to “energy availability first,” the winners are not just reactor developers; they are the firms that can monetize site-level baseload, grid interconnects, and long-duration power contracts. That creates a larger opportunity set in utilities, gas infrastructure, EPC, and even power-equipment suppliers than in a single pre-revenue SMR name.

SMR is still a long-duration call option on regulatory execution, not an operating business with visible cash conversion. The key catalyst path is binary and slow: permitting, first deployment, then proof of uptime and unit economics over multiple years. In the meantime, the stock can rerate on headlines, but the fundamental gap between “approved design” and “bankable fleet rollout” remains the core risk; any schedule slip into the post-2030 window would compress the multiple sharply.

The underappreciated loser is likely flexible fossil generation that depends on an AI demand surge to justify new buildout without carbon penalties. If SMRs or other low-carbon solutions become credible at scale, the upside for new gas peaker capacity gets capped, and the market may overpay for near-term dispatchability that tech buyers increasingly cannot justify under ESG and policy constraints. Conversely, the near-term beneficiaries may be firms selling turbines, switchgear, cooling, and transmission gear needed regardless of which electrons win.

Consensus appears to be overestimating how quickly “AI electricity demand” converts into SMR revenue and underestimating how many interim winners can monetize the same thesis sooner. The right way to express the view is not a full-size outright long in SMR, but a basket around the constraint: power infrastructure with current earnings and optionality on nuclear buildout. If nuclear adoption accelerates, SMR benefits; if it stalls, the support winners still have the backlog and cash flow.

Key Decisions for Investors

• Initiate a small starter long in SMR only as a long-dated optionality position; size at 25-50 bps and use a 12-24 month horizon, because the upside is asymmetric but dilution/schedule risk is high before first deployment.
• Prefer a basket long in grid/power infrastructure beneficiaries over SMR beta: add exposure to electrical equipment, transmission, and cooling names over the next 1-2 quarters, as they monetize AI capex immediately regardless of reactor timing.
• Avoid chasing gas-peaker developers here; use any rally in conventional generation names to trim, since low-carbon baseload credibility can cap medium-term rerating if SMR or alternative solutions gain traction.
• Pair trade idea: long power-infrastructure beneficiaries / short speculative pre-revenue clean-energy names that depend on multi-year commercialization, to isolate AI power demand without paying for execution risk.
• If SMR gaps up on nuclear policy headlines, consider selling upside call spreads rather than chasing equity; the stock is more likely to trade on milestone disappointment than on near-term fundamentals.