2 Nuclear Stocks Worth Buying After the SpaceX IPO

SpaceX’s IPO prospectus says more than 90% of its $28.5 trillion addressable market is tied to AI, making the company’s long-term growth dependent on scaling its AI division and securing energy for data centers. The article highlights constrained global power supply and suggests SMRs from NuScale Power and Oklo as potential solutions, though both remain pre-commercial and economically uncertain. The piece is largely thematic and speculative, with modest implications for AI infrastructure and nuclear SMR stocks.

Analysis

The market is likely underpricing how capital allocation by AI developers changes the adjacent power trade. If AI infrastructure remains constrained by baseload availability, the first-order winners are not the reactor developers themselves but the enabling stack: uranium fuel, grid interconnectors, switchgear, turbine/OEM suppliers, and regulated utilities with large load corridors. That means the economic surplus from AI demand may be captured earlier and with less execution risk by incumbents tied to power delivery than by pure-play SMR names that still face multi-year commercialization uncertainty.

The key second-order effect is sequencing. Even if SMRs ultimately win a share of AI load, their near-term role is mostly a signaling function that validates long-duration power scarcity, which can compress the valuation gap between renewables-plus-storage and nuclear-adjacent infrastructure. In the next 6-18 months, the biggest catalyst is not a reactor milestone but evidence of hyperscaler capex shifting into physical power assets, which would benefit utility-scale transmission and thermal equipment vendors before it benefits reactor developers. Conversely, any easing in data-center growth or a faster-than-expected drop in AI compute intensity would weaken the urgency premium in this theme.

On the risk side, both SMR equities remain classic “optionality stocks”: most of the value depends on commercialization timelines that can slip by years, and financing conditions matter more than headlines. A higher-rate regime or any cost-overrun at pilot sites would likely re-rate the whole basket lower, because the bear case is no longer technology feasibility but capital intensity and customer patience. The contrarian view is that the consensus is over-focused on the novelty of nuclear and underfocused on the utility of time-to-power; the market may be paying up for distant perfection while ignoring cheaper, faster substitutes already being deployed.

The better expression is to own the bottlenecks around AI power rather than the unproven end-state. If large AI operators continue to pursue dedicated energy solutions, that supports a multi-year capex cycle in grid hardware and regulated generation, while pure-play SMR names remain event-driven and binary. The asymmetric trade is to stay constructive on the broader power-enablement complex and treat SMRs as venture-style exposure, not core infrastructure.

Key Decisions for Investors

• Long GOOGL / long NVDA / short SMR basket via options or equity: favor the hyperscaler and compute beneficiaries over the highest-execution-risk nuclear optionality for the next 6-12 months; target 2:1 upside/downside if power scarcity keeps AI capex in focus while SMR timelines slip.
• Long BAC as a financing proxy only if project finance/utility capex inflects: use as a relative-value expression on capital deployment into infrastructure, but size modestly because the cleaner alpha is in power equipment and regulated utilities, not banks.
• Avoid initiating new long OKLO or SMR outright before a first commercial proof point; if already long, use call spreads or size caps to keep theta decay from dominating over a 12-24 month horizon.
• Pair trade: long grid/infrastructure enablers versus SMR developers for 3-9 months; the thesis is that time-to-revenue and balance-sheet quality matter more than reactor upside until a customer project is actually under construction.
• Set a catalyst watchlist for hyperscaler power contracts and utility interconnect announcements; if those accelerate, rotate into the power supply chain immediately, because that is the earliest monetizable leg of the AI energy theme.