Nebius: Mispriced By Association

Nebius missed Q4 revenue estimates but beat ARR guidance and reports capacity sold out through Q1 2026 with strong contract commitments. CapEx is ~60% self-funded via cash, operations and prepayments, supported by $8.5B of convertible debt at a blended coupon below 2%. Management also highlights $4.6–$5.2B of non-core asset value; stock appears traded as a proxy for CoreWeave despite differing capital structure and self-funded growth model.

Analysis

Markets are conflating two businesses with superficially similar end-markets but materially different capital and cash-flow mechanics; that creates a dispersion trade where equity prices reflect narrative momentum more than balance-sheet optionality. The cheaper capital-cost profile on one name effectively embeds call-like upside for creditors and compresses near-term dilution risk compared with peers, which should support higher implied equity valuations once the market re-allocates on fundamentals rather than headline comparability.

Second-order winners include GPU OEMs and contract manufacturers that capture the margin tailwind from customers preferring self-funded, predictable demand — they can sell longer lead-time inventory at better prices to fewer large, stable buyers. Losers would be competitors that must raise equity to fund buildouts: they face both higher funding volatility and the risk of being out-priced for long-term contracts, which can trigger a cascade of canceled expansion projects and delayed procurement across the supply chain.

Key catalysts and risks sit on a 3–12 month horizon: contract re-pricing, conversion/credit events, and any visible change in GPU supply dynamics. A swift normalization of GPU prices or an unexpected large-cap cloud capacity ramp would reverse the re-rating quickly; conversely, visible conversion activity or asset monetization announcements are binary events that could crystallize a revaluation sooner rather than later. Hedging around those windows is essential given asymmetric outcomes — limited near-term downside if cash flows stay intact versus significant dilution or multiple compression if credit mechanics change.

Key Decisions for Investors

• Pair trade (3–9 months): Long NBIS equity (or long-dated calls) vs short CRWV equity — target asymmetric R/R of ~2:1 (30–60% upside if mispricing correct, capped ~20–30% downside). Size 1–2% NAV directional, tighten stops around any signs of convertible conversion windows opening.
• Credit-structure arbitrage (6–12 months): Buy NBIS convertible paper or first-lien bonds where available and overlay a small equity short in broader GPU/cloud basket to hedge cyclic risk — expected carry + optionality; stress test for convert conversion at 25–40% dilution scenarios.
• Event hedge: Buy puts on semiconductor capital equipment (or an index proxy) to protect the pair trade during GPU-price normalization risk windows (3–6 months). Cost should be sized to limit portfolio drawdown to <1% NAV.
• Monitor catalysts & execute: Set alerts for (a) large contract amendments, (b) convertible conversion notices, (c) material asset-sale filings. Move to take profits or add only after a confirmed market re-rating (volume + spread compression) — avoid scaling into price moves driven solely by headline momentum.