This CoreWeave Rival Could Double in 2026

Nebius, a GPU-cluster provider to AI hyperscalers spun out of Yandex, reported Q3 revenue growth of 355% year-over-year and an annual run rate of $551 million, and management projects ARR of $7–9 billion by the end of 2026 while increasing contracted power from 1 GW to 2.5 GW. The company — like competitor CoreWeave — remains deeply unprofitable and currently trades at roughly 64x sales, but the piece argues that if AI spending and market appetite for growth-at-all-costs persist, Nebius could see substantial upside (potentially doubling), with the primary risk being delayed path to profitability.

Analysis

Market structure: Hyperscalers (AMZN, GOOGL) and GPU vendors (NVDA) are primary winners as external GPU-cloud suppliers like NBIS/CRWV monetize overflow AI demand; expect NVDA revenue upside of +10–30% if third‑party demand persists. Small‑cap GPU cloud providers gain pricing power short term (Q1–Q4 2026) due to constrained supply and large capex commitments (NBIS guidance implies ~13x–16x revenue growth from $0.55B ARR to $7–9B). Incumbent cloud providers could see margin tailwinds or share pressure depending on who secures long‑term capacity contracts.

Risk assessment: Tail risks include renewed export controls on GPUs (probability ~10–20% in 12 months) and power/permitting delays that can push NBIS/CRWV deployments out 6–18 months, crushing forward valuations. Immediate risk (days) is headline-driven IV spikes; short term (weeks–months) earnings/contract updates will reprice; long term (2026+) the binary is profitability or growth narrative fatigue. Hidden dependency: near‑total reliance on Nvidia GPU roadmap and third‑party power contracts; an NVDA inventory ramp or hyperscaler in‑house buildout would be devastating.

Trade implications: Direct plays: establish a tactical 1–2% long in NBIS (ticker NBIS) as a high‑beta growth punt with 100% upside target by end‑2026 and a 30% hard stop; size to 2–3% max household risk. Pair trade: long NBIS (1–2%) / short CRWV (1%) if you believe faster expansion is underpriced for NBIS; alternatively, long NVDA (2–4%) as a safer play on GPU pricing. Options: buy 9–12 month NBIS call spreads (delta ~0.35) or sell 6–9 month OTM NVDA puts for income if comfortable owning at strike ~10–15% below spot.

Contrarian angles: Consensus overweights growth at the expense of capitalization — market may underprice capex burn and dilution risk; if NBIS fails to convert contracts into revenue by H1 2026, downside could exceed 50%. Historical parallel: early cloud/CSP proliferation (2010s) saw winners consolidate — expect M&A risk (acquisition or margin squeeze) within 12–36 months. Watch for unintended friction: grid/power shortages and ESG permitting that can delay revenue recognition and flip the story from growth to cash‑flow risk.

Key Decisions for Investors

• Establish a tactical 1–2% long position in NBIS (high‑risk growth exposure). Set target of +100% by 31‑Dec‑2026 and stop‑loss at −30% from entry; trim to half at +50%.
• Add a 2–4% long position in NVDA as a structural hedge on GPU demand; consider selling 6–9 month OTM puts at ~10–15% below current spot to collect premium if willing to own on pullback.
• Implement a pair trade: long NBIS 1–2% / short CRWV 1% to express relative conviction in Nebius' higher near‑term growth; rebalance if NBIS underperforms guidance by 20% or CRWV reports stronger enterprise wins.
• Buy 9–12 month NBIS call spreads (delta ~0.35) sized to 0.5–1% notional to play asymmetric upside while limiting premium risk; if IV doubles around earnings, roll to later expiries.
• Monitor three specific catalysts in next 60–120 days before scaling: (1) NVDA supply commentary, (2) NBIS contract power deployment updates vs. 2.5 GW target, (3) any export/regulatory notices — pause new adds if any miss guidance by >15%.