CRWD — CrowdStrike Holdings, Inc.

Sentiment is anchored on the late-August 2026 Q2 FY2027 report (filed 2026-08-27), which management called the best quarter in company history: revenue +25.8% YoY to $1.471B, adjusted EPS +33% to $0.31, and record net new ARR of $332.8M (+51% YoY) pushing total ARR to $5.84B, with full-year guidance raised and the $10B ARR milestone pulled forward to FY2030. The early-September 2026 Fal.Con announcements (Falcon Guardian AIDR for autonomous agents, CrowdStrike SafeMind built with NVIDIA, Project QuiltWorks) added a new token-based revenue narrative, and sell-side targets moved up (consensus $231.58, recent average $235.70, with Morgan Stanley at $238, RBC at $260 and Scotiabank at $265). The offset: despite all of that, the stock drifted from its post-earnings high to $206.74, down 6.78% over one month as of 2026-09-13, because at ~39x TTM sales and a triple-digit forward P/E the market had already priced in a flawless quarter - a classic sell-the-news reaction rather than a change in the fundamental story.

Revenue growth is re-accelerating on a clean, monotonic six-quarter run: +19.8% YoY in Q1 FY2026 (Apr 2025), +21.3% (Jul 2025), +22.2% (Oct 2025), +23.3% (Jan 2026), +25.6% (Apr 2026), +25.8% (Jul 2026, $1.471B) - an inflection, not a blip, confirmed by record net new ARR of $332.8M, up 51% YoY, and total ARR of $5.84B. Margins are also inflecting off the FY2025 trough: GAAP operating margin went from -11.3% in Q1 FY2026 to -9.7%, -5.6%, then +1.1% in Q4 FY2026 - the only positive GAAP operating quarter in the window - before slipping back to -2.2% in Q1 FY2027 and -2.3% in Q2 FY2027 as R&D jumped 27.8% YoY to $444.2M to fund the AI/agent security push. Gross margin is stable-to-improving at 74.6% in the latest quarter versus 73.5% a year earlier (75.2% TTM). GAAP EPS remains effectively zero ($0.005 in Q2 FY2027) because of very heavy stock-based compensation, so the real earnings trend lives in cash: quarterly FCF rose from $285.0M (Jul 2025) to a record $493.3M (Apr 2026) and $405.9M (Jul 2026), taking TTM FCF to $1.573B, up 51.6% YoY on a 29.2% margin, with a healthy -4.37% accruals ratio (cash running ahead of reported net income). Important context: the FY2025 GAAP loss quarters were the one-off aftermath of the July 2024 faulty-update incident, not a structural decline - that episode has fully washed out and the trajectory since is up and to the right.

Expensive, and explicitly so: as of 2026-09-13 the deterministic valuation label is 'Rich vs. its own history.' The TTM P/E of 5,873x and PEG of 10.6 are artifacts of a $0.057 TTM GAAP EPS and are not usable. The numbers that matter: 39.0x TTM sales versus a 5-year average P/S of 21.99x; ~35.0x the $6.006B next-FY consensus revenue; ~165x the $1.254 next-FY consensus EPS (the research brief cites ~147x on a slightly different estimate); 41.2x book; EV/EBITDA of 481.6x; and a TTM FCF yield of just 0.75% (~134x FCF). Against 25.8% revenue growth, a 75.2% gross margin, a 29.2% FCF margin and $4.27B of net cash, the premium is defensible in direction but stretched in magnitude - CRWD is at roughly 1.8x its own five-year average sales multiple and there is no EPS cushion underneath it.

Do not read this stock off GAAP or even adjusted EPS - the market is capitalizing ARR and future FCF, not this year's $1.254 of non-GAAP EPS. The correct scorecard is net new ARR (record $332.8M, +51% YoY), total ARR ($5.84B against the accelerated $10B FY2030 target), and FCF margin (29.2% TTM, expanding) versus the ~35x forward sales multiple. That is what makes the setup asymmetric in both directions: if net new ARR keeps compounding at 40%+ and the new token-based SafeMind revenue becomes a second leg, the multiple is defensible and the analyst targets in the $232-$265 range are reachable; if net new ARR decelerates toward 20% - or FCF margin stalls while R&D climbs - there is no earnings floor to catch the stock, and the de-rating would be severe. Practically, treat any quarter where ARR growth slips below ~20% or FCF margin compresses as a multiple event, not just an earnings miss.

As of 2026-09-13, RSI(14) stood at 49.96 - dead neutral, neither overbought nor oversold - with price 0.37% below the 20-day SMA, 2.05% above the 50-day SMA and 43.45% above the 200-day SMA. The read is a strong primary uptrend (+87.18% over six months, +21.11% over three months) that is currently digesting sideways: -2.98% over one week and -6.78% over one month, with the 52-week range at $85.68-$233.88 and the last close at $208.86 versus $206.74 on 2026-09-13. There is no oversold entry signal here, but also no overbought blow-off - the stock is extended versus its 200-day and flat versus its short averages. Timing implication: this is a 'wait for the base to build' setup, not a chase; a pullback toward the 50-day SMA or a sustained reclaim of the 20-day would offer better entry mechanics than buying mid-consolidation at 43% above the 200-day.

• The AI threat super-cycle is a structural demand tailwind: CrowdStrike's own threat-hunting data shows a surge in AI-accelerated attacks against enterprise LLMs and software supply chains, which is exactly what the new runtime agent-security products (Falcon Guardian, SafeMind) are built to sell into.
• Interest-rate and duration sensitivity: high-multiple, long-duration SaaS names compress fastest when rates rise, and CRWD's ~39x TTM sales multiple leaves it unusually exposed to any flight away from growth; cybersecurity budgets themselves are far more resilient than general enterprise software spend.
• Enterprise IT budget scrutiny and vendor consolidation: platformization helps Falcon win wallet share, but it also lengthens sales cycles and puts a premium on demonstrable ROI - a headwind to near-term net new ARR if macro conditions tighten.
• Competitive intensity from Microsoft bundling Defender into enterprise agreements and from Palo Alto's Cortex platformization and SentinelOne in endpoint - persistent pressure on endpoint pricing even as CrowdStrike defends on detection quality and platform breadth.
• Regulation and geopolitics: government/critical-infrastructure cyber mandates and AI-agent governance rules are net demand drivers, while data-localization and export controls add compliance cost and can slow international deployments.

• Q3 FY2027 earnings (expected late November / early December 2026) - the first quantitative checkpoint on whether record net new ARR momentum is sustainable and whether margin expansion resumes as R&D spend normalizes.
• Early adoption and revenue disclosure for the token-based AI portfolio (CrowdStrike SafeMind with NVIDIA, Falcon Guardian AIDR) - a genuinely new monetization vector on top of seat/workload subscriptions, and the single most important thing that could justify the multiple.
• Progress toward the pulled-forward $10B ARR target by FY2030 - each incremental ARR milestone datapoint will be a re-rating or de-rating trigger given how much terminal growth is embedded in the price.
• Continued sell-side target revisions and platform-consolidation deal wins coming out of Fal.Con 2026 (Project QuiltWorks partner ecosystem expansion).
• Potential AI-native acquisitions or deeper NVIDIA/Cyber Superintelligence Lab partnerships that expand the addressable market.

• Valuation is the dominant risk: 39.0x TTM sales versus a 5-year average of 21.99x, ~165x next-FY consensus EPS and a 0.75% TTM FCF yield mean the stock has no cushion - a single quarter of decelerating net new ARR would likely trigger violent multiple compression even with the business still growing.
• GAAP profitability remains elusive: TTM operating margin of -2.2%, ROIC of -4.2% and GAAP EPS of $0.005 in the latest quarter, with heavy stock-based compensation flattering free cash flow and diluting holders - the 'cash earnings' story is real but not the same as durable GAAP profit.
• Law of large numbers: 25.8% growth on a $5.4B TTM revenue base requires sustained, not one-off, execution; the March 2026 guidance raise to a $10B ARR FY2030 target sets a high bar the company now has to clear publicly.
• Competitive encroachment from Microsoft bundling and Palo Alto/SentinelOne - pricing pressure in endpoint can cap net new ARR even if the platform strategy is winning.
• Platform/outage concentration risk: the July 2024 faulty-update incident remains the precedent for how quickly a single-platform architecture can produce a company-wide reputational and legal event; tail risk has not been fully retired.
• Macro/rate-driven multiple compression on long-duration growth assets, which hits hardest precisely when fundamentals are still strong.

This is a genuinely elite business trading at a genuinely demanding price. The fundamentals are unambiguously good and improving - six straight quarters of accelerating revenue growth to +25.8%, record net new ARR of $332.8M (+51% YoY), $1.573B of TTM free cash flow at a 29.2% margin and growing 51.6%, $4.27B of net cash and a -4.37% accruals ratio confirming the cash is real. But at $206.74 as of 2026-09-13, you are paying ~39x TTM sales, ~35x forward consensus revenue and ~165x forward EPS, roughly 1.8x the company's own five-year average sales multiple; my own forward-multiple work puts fair value near $158, about 24% below spot. That does not make it a bad company - it makes it a stock where the multiple, not the business, is the swing factor. Own it on weakness (a base above $200 or a pullback toward the 50-day) rather than at a premium to an already-premium multiple, and let net new ARR, not the headline, be the tripwire.

Forward EV/Sales on FY2027E revenue: I applied a 26x multiple - a premium to the ~5-8x software median and roughly 1.2x CRWD's own 5-year average TTM P/S of 21.99x, justified by 25.8% re-accelerating revenue growth, a 75.2% gross margin, a 29.2% FCF margin and $4.27B of net cash - to the $6.006B next-FY consensus revenue, giving a $156.2B enterprise value; adding the $4.27B net cash and dividing by the ~1.018B diluted shares implied by the data ($210.5B market cap / $206.74) yields ~$158 per share, about 24% below the 2026-09-13 spot and well below the $231.58 analyst consensus, with anything above that level requiring the token-based AI monetization to become a genuine second growth leg rather than optionality.