CoreWeave shares plunge after Q4 losses, revenue outlook misses estimates

CoreWeave reported Q4 revenue of $1.572 billion (vs. $1.55B est.) but swung to an EPS loss of $0.89, well below the estimated loss of $0.21, driving a 17.6% drop in the stock. Adjusted EBITDA was $898 million while GAAP net loss was $452 million (adjusted net loss $284 million); backlog totaled $66.8 billion and the company added ~260 MW of active power, bringing capacity to >850 MW with 3.1 GW contracted. Management guided Q1 revenue of $1.9–$2.0 billion (consensus $2.2B) and expects $17–$19 billion in annualized run-rate revenue exiting 2026 and >$30 billion exiting 2027, signaling strong long-term growth ambitions but near-term softness that rattled investors.

Analysis

Market structure: CoreWeave’s miss reweights near-term winners toward GPU OEMs (NVDA) and diversified data‑center landlords (EQIX, DLR) while penalizing pure‑play GPU cloud operators (CRWV, smaller competitors) because CoreWeave’s +260 MW add (now >850 MW) and 3.1 GW contracted power show supply ramp may outpace near‑term monetization. The $66.8B backlog and $898M adjusted EBITDA imply strong long‑run demand but the Q1 guide ($1.9–2.0B vs $2.2B est.) signals a mismatch in sales conversion and potential pricing pressure for spot GPU compute.

Risk assessment: Tail risks include a liquidity/covenant shock if capital markets tighten or capex overruns force equity dilution (high impact, low prob over 12 months), tighter export controls on AI chips, or a large customer churn event given probable concentration. Immediate horizon (days) = higher IV and potential follow‑through selling; short term (weeks–months) = Q1 execution and utilization metrics; long term (quarters–years) = whether adjusted EBITDA converts to GAAP profitability as capex intensity moderates.

Trade implications: Near term favor defined‑risk bearish positions on CRWV to capture guide‑driven repricing while positioning long NVDA exposure to play GPU demand; rotate 2–4% portfolio weight from small GPU cloud names into EQIX/DLR for defensive, cash‑generative infrastructure exposure. Use options to control risk around specific catalyst dates (Q1 print, NVDA supply updates).

Contrarian angles: Consensus underweights CoreWeave’s backlog and positive adjusted EBITDA, so a durable 12–18 month recovery trade is plausible if utilization improves or M&A emerges; however the market may have rationally discounted execution/liquidity risk — downside to ~30–50% near term is credible. Watch for insider buying, large contract renewals, or Nvidia supply commitments in the next 3–6 months as reversal triggers.

Key Decisions for Investors

• Establish a defined‑risk short on CRWV: buy 3‑month (Jun 2026) put spread sized to 1–1.5% of portfolio (target 30–45% downside, stop if spread loses 20%); rationale: Q1 guide miss + widened losses should pressure price within 3 months.
• Pair trade: go long NVDA equal‑dollar 1% portfolio vs short CRWV 1.5% for 3–6 months to capture asymmetric upside in GPU OEMs versus execution risk in a GPU cloud operator; rebalance if NVDA outperforms by >25%.
• Hedge existing CRWV exposure with a 12‑month collar: buy Jan 2027 puts ~25% OTM and sell calls ~15% OTM to finance protection; adjust if CRWV reports two consecutive quarters of utilization >70%.
• Rotate 2–4% portfolio from small GPU/cloud names into EQIX and DLR (add 1–2% each) over 30–60 days to reduce capital‑intensity risk and lock in more predictable cash flows while monitoring CoreWeave’s Q1 conversion and Nvidia supply updates.