CrowdStrike upgraded by Wolfe on Anthropic AI Catalyst

Wolfe Research upgraded CrowdStrike to "outperform" from "peer perform" with a $450 PT, implying 21.8% upside from $369.58 and sending shares up ~3% in pre-market trade. The firm models ARR accelerating to 25.5% in FY27 (up 1.7 ppt), net new ARR of 32% in FY27 (vs company guidance 22.5%), and revenue of $5.99B in FY27, $7.42B in FY28 and $9.04B in FY29. Key risks include a potential near-term negative reaction to the expected Anthropic Mythos model announcement and longer-term competition from AI model providers encroaching on cybersecurity.

Analysis

The market is treating the impending large-model release as a near-term binary news event, but the real commercial impact will be decided by two slower, structural forces: procurement cadence and telemetry depth. Expect a sharp volatility window in the first 48–72 hours after the announcement as quant and momentum flows reprice perceived competitive risk, while measurable enterprise spend changes will emerge only over the next 6–18 months as renewal cycles and vendor evaluations complete.

Second-order winners are vendors that convert new telemetry into recurring ARR — that favors platform players that bundle endpoint, identity and log telemetry into higher-retention deals, and infrastructure suppliers that get incremental billings from increased telemetry ingestion and on-prem inference (an underappreciated capital-intensity tail). Conversely, pure-play EDR/feature vendors without strong cloud telemetry or identity capabilities are vulnerable to margin compression and customer churn if customers consolidate to one integrator.

Key risks that could reverse the current constructive view include rapid model-driven feature parity (fast commoditization of detection), aggressive price competition that pressures NRR, or an enterprise pause in procurement while security teams re-architect controls around the new model — each would show up in bookings within 3–12 months and compress multiples. The higher-conviction outcome is asymmetric: if CrowdStrike (and peers) translate incremental AI telemetry into stickier multi-product deals, upside is realized over 12–36 months; if AI vendors disintermediate core detections, downside is front-loaded and severe for single-product EDR vendors.

Key Decisions for Investors

• Long CRWD via 12–18 month call spread (buy nearer-term ITM or ATM calls, sell further OTM) to capture platform re‑rating if FY+1 bookings accelerate; target 30–40% upside over 12 months, max loss = premium paid (limited by sold calls).
• Pair trade: long CRWD / short S (SentinelOne) on a 3–12 month horizon to express platform vs endpoint-only exposure — aim for CRWD to outperform by ~15–25%; size short to keep net delta neutral and cut if both names move >20% intraday.
• Buy SMCI on a 6–18 month thesis that increased telemetry and on-prem inference demand lifts server/accelerator orders; asymmetric payoff if enterprises pursue hybrid inference (target +30–50%), risk is further cloud migration compressing on-prem demand.
• Event-specific tactical rule: if the Anthropic/model release sparks a >8–12% one-day gap move in CRWD, scale into the long call spread at 50–70% of planned size on the pullback and use the remainder to average in over next 4–8 weeks as renewal data arrives.
• Risk management: set stop-loss triggers for pair and directional trades at 15% adverse move, and monitor three datapoints quarterly (net new ARR growth, NRR trends, and new multi-product deal count) — if any fall >150bps vs trend, reassess positions.