An Ex-OpenAI Researcher With $20 Billion Just Built a Neocloud Trifecta: Here Are the 3 Stocks He's Buying.

Nebius, CoreWeave, and IREN are highlighted as rapidly growing AI-focused neocloud stocks, with projected 2026 revenue growth of 550%, 147%, and 46%, respectively. The article argues their valuation multiples of 6x, 3x, and 7x next-year sales look inexpensive relative to long-term growth potential, despite concerns about infrastructure spending. It also notes Nvidia’s strategic stakes/options in the group and cites a potential 58% CAGR for the neocloud market from 2025 to 2031.

Analysis

The market is still pricing these as financing-and-capex stories, but the second-order winner is the entire GPU supply chain: every incremental neocloud dollar effectively tightens capacity for power equipment, liquid cooling, networking, and land entitlement names long before it shows up in revenue. If these platforms keep converting demand into contracted capacity, the scarce input shifts from chips to megawatts, which is structurally supportive for firms that can secure interconnects and cheap power, and negative for slower hyperscale allocators that need broader, lower-IRR workloads to justify buildouts.

The more interesting read-through is that valuation compression may already be close to peak fear if revenue ramps remain intact into the next 2-4 quarters. At 3-7x forward sales, the market is implicitly betting that one of two things happens: either growth decelerates sharply after the initial build cycle, or equity dilution eats the upside before operating leverage appears. The setup is asymmetrical because the first credible proof of utilization stability can rerate these names multiple turns higher, while the failure mode is mostly execution/financing, not demand destruction.

The contrarian view is that the consensus may be underestimating how fragile the economics are if power prices rise, GPU lease rates reset, or customer concentration narrows. These businesses can look cheap on sales while still being value traps if gross margin expansion lags capex by even 6-12 months. The highest-probability catalyst for a reversal is not an AI demand slowdown, but a capital markets window that lets one of these firms de-risk its balance sheet; absent that, the group remains hostage to narrative and funding terms.

Key Decisions for Investors

• Long NBIS / CRWV / IREN basket for 3-6 months as a relative-value basket trade against broader AI infra, with the thesis that any one company proving sustained utilization can re-rate the group; size modestly because balance-sheet risk dominates.
• Prefer CRWV as the cleanest direct AI workload proxy, but hedge with a paired short in a hyperscale-adjacent cloud beneficiary if you want to isolate neocloud beta; upside is multiple expansion on contract durability, downside is dilution or lower-than-expected margin conversion.
• Buy 6-9 month call spreads on NVDA as a supplier-beneficiary hedge to long neocloud exposure; if neocloud capex persists, NVIDIA remains the toll collector even if one platform stumbles.
• Avoid unhedged single-name longs into earnings or financing events unless near-term liquidity is visible; the biggest drawdown risk is not miss on growth, but a capital raise at a weak valuation multiple.
• Watch for any announcement of long-term power or land contracts as a catalyst to add; those are the closest things to moat-expansion evidence in this model and can justify a 1-2 turn forward-sales rerating.