Opinion · Positioning · August 31, 2026
CrowdStrike Beat and Rose 22%. Marvell Beat and Fell 14%. Guidance Was the Difference.
CrowdStrike and Marvell both reported last week, and both beat Wall Street’s revenue estimates by a wide margin: CrowdStrike grew revenue 25.8% year over year, Marvell grew 37%. Since each company’s own report, CrowdStrike is up roughly 22% and Marvell is down roughly 14%. Neither move is about the quarter that already happened — both companies delivered a genuinely strong one. The entire gap traces to what each management team said about the quarter ahead: CrowdStrike raised its outlook and followed up Monday with a new product launch; Marvell guided margins down and left open questions about the timing of its largest new customer relationship. This is not a new pattern on this site — it is the same one I wrote about August 24, when the market sorted names by whether they had current earnings at all. Today’s version is narrower and, I think, more useful: both of these companies clearly have earnings. The market still only wanted to pay for one of their forecasts.
1. The pattern, shown directly
Index both stocks to 100 at the same starting point and the divergence is immediate and durable, not a one-day blip. Both names sat within a few points of each other twelve sessions ago. Marvell spiked into its own earnings date on anticipation, then gave essentially all of it back the moment guidance landed — a 10.28% single-session drop the day of the print, followed by Monday’s further 2.29% decline on continued coverage of the same margin and Google-timing questions. CrowdStrike moved the opposite way: a roughly 20.5% single-session jump on its own earnings date, a modest give-back the next session, then Monday’s 5.77% rally on a new product launch layered on top of the raised outlook. Both lines cross the same “earnings window” on the chart; they exit it moving in opposite directions.
Both series indexed to 100 at the session twelve trading days before August 31. The shaded band marks the two-session window spanning each company’s own earnings reaction and Monday’s follow-through.
| Session | CRWD index | MRVL index | Read |
|---|---|---|---|
| Pre-earnings (12 sessions ago = 100) | 100.00 | 100.00 | Common starting point |
| Session before each report | 87.20 | 110.40 | Diverging anticipation — MRVL bid up, CRWD sold off |
| Earnings-reaction session | 105.07 | 97.57 | CRWD +20.5% on the print; MRVL −10.28% on the same-day guide |
| Aug 28 close | 100.67 | 97.57 | CRWD gave back some of the pop; MRVL held flat, still absorbing Friday |
| Aug 31 close (today) | 106.48 | 95.33 | CRWD extended on a product launch; MRVL extended on the same guidance concern |
2. Why guidance dominates when the multiple is already rich
Neither stock is cheap, which is exactly why guidance carries so much weight. Friday’s News coverage put Marvell at roughly 53x forward earnings going into its report; a same-week Zacks piece put CrowdStrike at 32.97x forward price-to-sales against an 18.51x industry average. At those multiples, a company isn’t being paid for the quarter it just delivered — both already priced in a beat. It is being paid for the next several years of growth the market believes management can still deliver, and that belief moves on a single sentence in a guidance range. Marvell said its highest-margin business is being diluted by a faster-growing but lower-margin one; the market treated that as evidence the growth story is starting to cost something. CrowdStrike said its AI Detection and Response business “nearly tripled its annual recurring revenue sequentially,” per Zacks, and followed it Monday with a concrete new product; the market treated that as evidence the growth story still has room. This is the same mechanism I described August 24 in a piece about the market sorting names by whether they had current earnings at all — except both of this week’s subjects clear that bar easily. The sorting variable has simply moved one level up the stack, from “does this company earn money” to “is the rate at which it earns money about to accelerate or decelerate.”
It is also worth noting how fast the story can flip. This site’s August 20 coverage logged CrowdStrike falling 5.60% with no company-specific catalyst attached, as part of a broader security-software selloff. Eleven trading sessions later, the same stock is up more than 20% since its own report and leading a sector-wide rally. Nothing about CrowdStrike’s business changed in either direction on August 20 — the stock was simply caught in a sector mood that had nothing to do with it, the same way Marvell’s slide this week is arguably more about its guidance sentence than its actual growth rate. Sentiment and fundamentals share a road, but they do not always drive at the same speed.
3. Sizing what the market is actually paying for
Zacks’ own earnings-revision estimates, cited in a Monday piece on CrowdStrike, project 33% EPS growth for fiscal 2027 and 26% for fiscal 2028. Those are analyst consensus figures, not this site’s forecast, but they are useful to size: compounded, they imply CrowdStrike’s EPS would be roughly two-thirds higher at the end of fiscal 2028 than where it stands today, if both estimates land exactly.
| If Zacks' consensus EPS growth estimates hold… | Cumulative effect on EPS |
|---|---|
| FY2027 alone (+33%, consensus) | +33.0% |
| FY2028 alone, applied after FY2027 (+26%, consensus) | +26.0% on the new base |
| FY2027 + FY2028 combined, compounded | +67.6% |
Assumptions, stated plainly. The 33% and 26% figures are Zacks’ own consensus EPS-growth estimates for CrowdStrike’s fiscal 2027 and 2028, as reported in Zacks coverage published September 1, 2026 — not AEA’s independent forecast. The 67.6% combined figure is simple compounding (1.33 × 1.26 − 1), which assumes both estimates land exactly as projected two separate times in a row; consensus estimates on high-growth software names are revised frequently and have not always held, including at CrowdStrike itself, whose own revenue-growth rate the same Zacks coverage described as “decelerating” from “35%+ to projected 21.6% by fiscal 2028.” This table does not predict CrowdStrike’s stock price; a 32.97x forward P/S multiple already implies the market expects strong growth, and whether that growth materializes at the compounded rate above, faster, or slower is precisely the open question Monday’s rally was a bet on.
What would change this read
WatchTwo things that would tell me the gap is closing, not widening
1. Marvell’s next dated update on the Google relationship. The stated reason for the stock’s continued weakness is a lack of disclosed scope and timing on that deal; a concrete update, in either direction, would likely move the stock more than another quarter of pure revenue growth. 2. Whether CrowdStrike’s AIDR growth rate holds. “Nearly tripled sequentially” from a small base is an easy comparison to beat once; whether it can repeat anywhere close to that pace next quarter is the real test of whether Monday’s rally was pricing a durable new growth driver or a one-quarter number.
I hold both names — Marvell only on the watchlist, CrowdStrike as a 2-share position in the book — and I made no trades in either this week. The lesson I am taking from placing them side by side isn’t about either company specifically; it is a reminder that at the multiples the market is currently paying for AI-linked growth stories, the trailing quarter is close to a formality. The report that actually moves the stock is the one about the quarter that hasn’t happened yet.