CoreWeave: Deeply Undervalued Against Peers

CRWV has a $66.8B backlog, with $28.0B slated for recognition over the next 24 months, underpinning a projected 145% revenue surge in FY2026. An analyst rates CoreWeave Strong Buy, noting forward P/S is expected to fall below 2 by FY2027 and the stock appears undervalued versus peers despite uncertainty on the timing of profitability.

Analysis

CoreWeave’s strength is a demand-side story that ripples across the AI hardware and power ecosystem: GPU vendors, high-throughput networking vendors, and regional utilities are the natural upstream beneficiaries while legacy colocation and CPU-focused cloud offerings look structurally disadvantaged. The most important second-order dynamic is pricing of GPU-hours — if GPU supply tightness persists, specialized providers keep pricing power; if supply normalizes, realized billing rates can compress faster than headline backlog converts to cash.

Key near-term catalysts are utilization, effective price per GPU-hour, and customer concentration disclosures; these will move the stock within earnings windows (days–quarters). Medium-term (6–24 months) risks center on capex cadence, milestone-based contract writebacks, and the potential for hyperscalers to internalize capacity — any of which can flip forward visibility into dilution and force multiple contraction.

Consensus is likely underweight execution and margin risk: backlog visibility is not a free option unless cost-to-serve and churn remain stable. We prefer a staged exposure that monetizes directional conviction while protecting against outcomes where revenue recognition proves lumpy or margins slip; monitor GPU-hour realized pricing and top-5 customer share as the primary real-time diagnostics.

Key Decisions for Investors

• Buy CRWV Jan 2028 LEAP calls (12–36 month horizon) sized 2–4% of equity allocation; scale on two catalysts: sustained sequential utilization improvement and expanding realized GPU-hour pricing. Hedge with a 10–20% OTM put or sell short-dated calls against 25–50% of position to finance premium. Risk/reward: asymmetric — limited premium loss vs multi-bagger on execution and multiple re-rating.
• Pair trade: Long CRWV / Short EQIX (or DLR) equal notional for 6–12 months to express GPU-workload migration; target 400–600bps relative outperformance. Exit on (a) CRWV utilization miss, (b) hyperscaler disclosure of major internal build, or (c) sector-wide multiple compression.
• Construct a conservative income entry: buy CRWV shares and sell 3–6 month covered calls to harvest premium while maintaining directional upside; roll calls out on each quarter if utilization metrics improve. Use proceeds to buy 6–12 month protective puts if guidance flags concentration risk.
• Event hedge: establish a small short position in CRWV sized to cover potential post-earnings gap risk (days–weeks) if customer concentration or milestone cancellations are disclosed; convert to longer hedge if management signals increased capex or financing need. Target P&L defense rather than alpha generation.