2 Stocks With Monster Potential to Hold Through the Next Decade of Chaos

Rivian is presented as a top 2026 growth pick trading at ~3.2× sales versus Tesla’s >13×, but its path to EBITDA positivity has been delayed as R&D spending to build self‑driving capabilities accelerates. Oklo is pitched as a longer‑duration AI infrastructure play via small modular reactors (SMRs), with the company largely pre‑revenue and its first plant not expected online until late 2027. Both companies are framed as high‑upside, AI‑driven thematic bets rather than near‑term cash‑generative investments.

Analysis

Concentration of compute (high rack power density, campus-level load in the hundreds of MW) will reprice winners along the stack: asset owners who control dispatchable baseload or long-term bilateral PPAs will extract outsized returns relative to pure-opex data center operators. That elevates not only specialist power providers but also firms that can pair compute with guaranteed capacity — think of a spectrum from vertically integrated hyperscalers to large industrial off-takers that can contract long-dated supply.

A company that shifts R&D toward software-heavy roadmaps will likely extend a low-margin, high-capex window before any software monetization offsets hardware losses; the relevant inflection is multi-quarter to multi-year and depends on fleet utilization and subscription take-rates, not unit deliveries alone. Conversely, projects aiming to substitute grid capacity face permitting, financing and supply-chain friction that typically compress into multi-year timelines (3–7 years) — the payoff is binary and front-loaded to contract wins and regulatory milestones.

The clearest tradeable asymmetry is binary optionality: small capital in long-dated, low-cost option structures captures upside if infrastructure contracts or autonomy monetization materialize, while being capped if regulatory, financing or adoption delays occur. Meanwhile, positioning that expresses a compute-vendor consolidation (dominant accelerators vs legacy CPU vendors) is higher probability over 6–18 months because it rests on secular AI workload economics rather than one-off product cycles.

Contrarian caution: the market tends to conflate headline AI demand with immediate consumer purchasing behavior; fleet-first adoption and data center siting constraints mean retail demand re-rates can lag materially. Size exposures accordingly — treat infra/nuclear-style optionality as venture-like (low NAV weight, high scenario payoff) and auto-autonomy as operationally contingent (short-term margin pressure, long-term software capture).

Key Decisions for Investors

• Long RIVN shares (12–24 month horizon) sized 3–5% net equity risk; finance with a 12-month put protection (buy 1x 30–40% OTM put) to cap downside. Risk/reward: capped premium loss vs asymmetric upside if software monetization or fleet partnerships re-rate multiples (target +100% upside, downside limited to put strike).
• Speculative long OKLO via long-dated call spread (buy 2028 LEAP call, sell a higher-strike 2028 call) sized 1–2% portfolio risk; thesis payoff window 24–48 months tied to contracting/financing milestones. Risk/reward: limited premium at risk; potential 3–5x payout if commercial PPAs and project financing close.
• Pairs trade: long NVDA calls / short INTC stock or put (6–12 month horizon) to express continued accelerator dominance in AI; target 2:1 expected upside capture vs legacy CPU displacement. Risk: sector-wide drawdowns—size to volatility and hedge with broad-market put if beta >1.2.
• Income-funded conviction approach for RIVN: buy shares and sell near-term covered calls (30–60 day) to monetize volatility while maintaining directional upside; allocate remaining premium to buy longer-dated protective puts if downside skews increase. Risk/reward: reduces cost basis and provides income; limits upside to call strike.