2 Stocks With Monster Potential to Hold Through the Next Decade

The article highlights two speculative beneficiaries of AI infrastructure spending: TMC The Metals Company, which could supply nickel, cobalt, copper and manganese from the Clarion-Clipperton Zone, and Nano Nuclear Energy, which is developing a portable small reactor for on-site data center power. TMC estimates operations could begin in late 2027, while Nano still lacks commercial regulatory approval but is in government programs that may accelerate deployment. Overall, the piece is bullish on long-term optionality but emphasizes high execution and regulatory risk.

Analysis

The market is starting to price a second-order AI bottleneck: not just compute, but the physical inputs and power behind compute. That creates a reflexive setup where any credible path to more copper, nickel, cobalt, or behind-the-meter generation gets bid as an “AI infrastructure” proxy, even if the core business remains pre-commercial. The winners in the near term are likely the financing chain and option market, not the operating businesses themselves, because the equity value is being driven by narrative optionality ahead of cash flow.

For TMC, the real catalyst is regulatory de-risking, but the more important issue is whether the project economics can survive a world where deep-sea mining faces sustained ESG litigation, permitting drag, and a likely multi-year capex ramp. If approvals advance, the upside can be abrupt because supply is highly levered to a single jurisdictional decision; if approvals stall, the downside is slow bleed through dilution and capital scarcity. The market is underweight how far out the first meaningful volumes likely are, which makes the stock more of a 12-24 month binary than a durable compounder.

NNE has a different risk profile: it benefits from the scarcity value of firm power for data centers, but the path to commercialization is longer and more regulation-intensive than the market is likely discounting. The key second-order effect is that hyperscalers may prefer quicker solutions first—gas turbines, PPAs, grid interconnects, and modular battery storage—before adopting novel nuclear, which pushes out revenue recognition and increases financing risk. In that sense, the “AI power” trade is real, but the cleaner expressions may still be incumbent utilities, gas infrastructure, and select uranium exposure rather than the most speculative reactor developers.

Consensus is probably overestimating near-term monetization and underestimating policy optionality. If the U.S. explicitly leans into strategic mineral and domestic power security, both names can re-rate sharply; if that support softens, both trade back toward pure cash-burn narratives. The asymmetry is attractive only for small sizing and with the understanding that catalysts are measured in quarters to years, not weeks.

Key Decisions for Investors

• Trade TMC as a binary regulatory optionality position only: small long via calls or a tight equity tranche for 6-12 months, with expectation of high gamma into permit/news flow and a hard stop if approvals slip or financing risk widens.
• Prefer a pair trade: long NNE or TMC call spread vs short the basket of mature industrial names exposed to AI capex skepticism; the thesis is that narrative premium expands faster than fundamentals, but cap the downside with defined-risk options.
• For investors wanting the cleaner AI-power exposure, rotate toward OKLO/SMR on weakness rather than chasing NNE strength; they are closer to commercialization and should outperform speculative pre-revenue names if the market shifts from story to execution over the next 6-18 months.
• Hedge speculative long exposure in TMC/NNE with a short in an established renewable or utility proxy if capital rotates away from unproven energy solutions; this protects against a rerating compression once the market re-prices timeline risk.
• Do not size these as core longs: target position sizes <1% NAV each, because the distribution is dominated by policy and permitting outcomes; use event-driven add points only after confirmed regulatory milestones.