Palo Alto Networks closed fiscal 2026 after the US market on 1 September with $3.410 billion of revenue and $1.02 of non-GAAP diluted earnings per share, against consensus of $3.35 billion and $0.98. Revenue beat by 1.8% and cleared the top of the company's own guided range; EPS beat by four cents, which is exactly how the chief financial officer described it — "exceeding the high end of our guided range by $0.04." Revenue grew 34.5% on the year-ago quarter and 13.6% on the April quarter. The full figure set is on the release page.
Every headline the company guides to was cleared: Next-Generation Security ARR of $9.10 billion against $8.90–8.95 billion guided, remaining performance obligations of $21.2 billion against $20.9–21.0 billion, a full-year adjusted free-cash-flow margin of 38.4% against 37.5%. Fiscal 2026 revenue of $11.480 billion landed $60 million above the top of the full-year guide.
Our preview argued that the number to watch was not the quarter but the disclosure — specifically, whether the company would keep publishing the growth rate of the business excluding CyberArk and Chronosphere. It did not. That is the single most consequential thing in this release, and everything below follows from it.
The shares had already fallen 5.24% to $362.09 during the session before the print, and traded at $356.00 after hours, down 1.68%, as of 7:56 p.m. Eastern on 1 September.
What the preview asked, and what the print answered
Whether the organic disclosure survived — it did not, and something else replaced it. In April the company published both rates itself: revenue up 31% reported and 14% excluding the two acquisitions, NGS ARR +60% against +28% organic, current RPO +34% against +17%. This release carries none of them, and no current RPO, no platformization count, no product-level ARR. What arrived instead is a different disclosure the chief financial officer flagged as new: "To provide more visibility into our growth drivers, we are introducing new revenue disclosure by platform." For fiscal 2026 that is Network & AI Security $8.35 billion (+17%), Cortex $1.92 billion (+25%), and Idira — the rebranded CyberArk — $1.26 billion on a pro forma basis (+21%). It is a genuine improvement in one direction and a loss in another: a reader can now see where the revenue sits, and can no longer see how much of its growth was bought.
Net new NGS ARR against the implied $770–820 million — $970 million, with a nine-figure asterisk. The ARR base went from $8.13 billion to $9.10 billion, so net new was $970 million, comfortably above the $770–820 million the guide implied. The subtraction is ours; both balances are the company's. But the same call disclosed what is inside it: "our Q4 net new ARR includes a nine-figure benefit from a large LLM customer migrating to Chronosphere from an incumbent vendor." Nine figures means at least $100 million, so the recurring part of the record is at most about $870 million — still a beat, and a much smaller one than the headline.
The fiscal 2026 free-cash-flow margin against 37.5% — 38.4%, and fiscal 2027 is guided to 38.0%. Adjusted free cash flow was $4.414 billion on $11.480 billion of revenue. That is the cleanest answer in the release, and the guide behind it is the least clean: the year before the stated 40% year is guided below the year that just finished. The chief financial officer's defence is that the level itself is the achievement — "Adjusted free cash flow margin has been 38% or better in each of the last four years, and we sustained the strong cash flow generation even while absorbing the impacts of large M&A."
RPO against $20.9–21.0 billion — $21.2 billion, up 34%, and guided down next quarter. Remaining performance obligations passed $20 billion for the first time. The organic half of the question is unanswered: current RPO, which the company reported both ways through April, is not in this release at all. And the first-quarter guide of $20.8–20.9 billion sits below the $21.2 billion just reported — a sequential step down of roughly $350 million at the midpoint, ours by subtraction, that still prints as +34–35% year over year.
Whether management addressed the acquisition reports — obliquely, and then announced another one. The release itself discloses the purchase of Console, an agentic-workflow platform folded into Cortex, with no price, no close date and no ARR. Asked about the pattern, the chief executive said: "As I've always maintained that M&A is not a strategy. M&A is a consequence of stuff that we do from a product development perspective." That is a position, not a capital-allocation figure. The release also contains no buyback disclosure, where the April release carried $1.0 billion repurchased and $1.0 billion remaining.
The platform guides do not add up to the revenue guide
The new platform disclosure is exhaustive for fiscal 2026. Network & AI Security's $8.35 billion plus Cortex's $1.92 billion leaves $1.21 billion of the $11.480 billion total, which is Idira's reported contribution against its $1.26 billion pro forma figure. Three numbers, one total, no residual.
Now run the same addition on the fiscal 2027 guides the chief financial officer gave: "Network & AI Security revenue growth of low double digits year-over-year... Cortex revenue up approximately 30% year-over-year, and... Idira revenue of approximately $1.5 billion." Read "low double digits" generously at 14% and the three sum to about $13.5 billion. Read it at 10% and they sum to about $13.2 billion. The total revenue guide is $14.10–14.20 billion.
So between $600 million and $1.0 billion of guided fiscal 2027 revenue is not accounted for by any of the three platform guides. Hold Cortex at 30% and Idira at $1.5 billion, and hitting the bottom of the revenue guide requires Network & AI Security to grow 21% — double what was guided. That arithmetic is ours. There are two readings and the release does not choose between them: either the platform guides are deliberately beatable, or the total guide carries revenue — Console, and whatever follows it — that the platform guides do not. The bars above this article show the gap.
What changed in the story
The dilution wedge is now the whole earnings story. Non-GAAP operating income guided at a 29.5% margin on $14.15 billion is $4.17 billion, up 24.4% on fiscal 2026's $3.356 billion. Non-GAAP earnings per share are guided to $4.16–4.19, up 8.3–9.1%. The gap is the share count: 764 million diluted in fiscal 2026, guided to 844–847 million, up another 10–11%. The same wedge is visible in the quarter just reported — non-GAAP net income +26.7%, non-GAAP EPS +7.4%, on a share count up 17.7% year over year. Both derivations are ours, from the company's figures.
The headline growth rate falls by two thirds in one guide. NGS ARR grew 63% in fiscal 2026 and is guided to $11.075–11.175 billion, +21.7% to +22.8%, for fiscal 2027. RPO grew 34% and is guided to +18.9% to +19.8%. Nothing has to go wrong in the business for that to happen; the acquisitions simply enter the comparative base.
And the second-half loading makes the first half look thin. The guide implies $1.98–2.08 billion of net new NGS ARR for the year. The chief financial officer put 60–61% of it in the second half, which leaves roughly $790–810 million for the first two quarters — about $400 million a quarter, against the $970 million just printed. The first-quarter ARR guide corroborates it exactly: $9.54–9.56 billion against $9.10 billion is $440–460 million of net new.
A record year that ends in a GAAP loss. Fiscal 2026 GAAP net income was $307 million against $1.134 billion, and the fourth quarter itself was a $282 million loss against a $254 million profit. The cause is not operations: a $524 million mark on the CyberArk convertible notes and related capped calls, a line that did not exist a year ago, drove $441 million of other expense. Goodwill and intangibles now stand at $29.0 billion of a $48.5 billion balance sheet — 60% of total assets, against 23% a year ago.
On our arithmetic the trailing Rule of 40 score improves to 60.3 — revenue growth of 24.5% plus a free cash flow margin of 35.8% on the reported, unadjusted figure — from 55.3 through April.
Against the model
Our Palo Alto model was published on 27 August and its base case was calibrated to the fiscal 2026 guide. All four of those inputs are now actuals, and all four came in above the guide: revenue $11.480 billion against $11.415–11.425 billion, non-GAAP EPS $3.84 against $3.77–3.79, adjusted free-cash-flow margin 38.4% against 37.5%, ending NGS ARR $9.10 billion against $8.90–8.95 billion. The largest of those beats is half a percent. The model's value sits overwhelmingly in a terminal exit at 9 times revenue five years out, so a half-percent level shift in the base quarter is noise, and the fair values are unchanged pending a revision, which is a separate decision.
Two inputs are worth more than the level. The first is the platformization count, which is the model's subscriber driver: it carries about 124 gross adds a quarter, damped to roughly 110 net at the current penetration, calibrated to the ~110 the company printed in April. The fourth quarter printed approximately 220 — "surpassing our prior record and representing more than twice the volume for when we initiated this metric two years ago." Taken with the ~2,280 the company disclosed in April, that puts the base near 2,500 (ours, by addition). Against management's own 4,000-plus by fiscal 2030, the remaining requirement is about 94 a quarter for sixteen quarters — below both the printed rate and the model's assumption. On this driver the target is now the easy part.
The second is the ARR path, where it is not. $20 billion of NGS ARR in fiscal 2030 from $9.10 billion today is a 21.8% compound rate for four years, and fiscal 2027 is guided at 21.7–22.8%. The target is therefore not a stretch above the guide — it is the guide, repeated three more times without a single decelerating year, in a base that has just stopped being flattered by acquisitions. That is the assumption the model's bull and chief-executive cases are paying for, and the quarter neither proved nor broke it.
One correction to our own framing is owed. The model's bear case is anchored on the fiscal 2026 free-cash-flow margin being guided backwards to 37.5%. It was not: 38.4% was delivered. The fact survives the correction anyway, moved forward a year — fiscal 2027 is guided to 38.0%, again below the year before it, again in front of a stated 40% target.
What is coming
The first quarter of fiscal 2027, ending 31 October, is guided to $3.300–3.310 billion of revenue (3.1% below the quarter just reported), non-GAAP EPS of $0.96–0.98 on 837–844 million shares, NGS ARR of $9.54–9.56 billion and RPO of $20.8–20.9 billion. Three of those four are sequentially lower than what was just delivered.
For the year: revenue $14.10–14.20 billion, a 29.5% non-GAAP operating margin, EPS $4.16–4.19, RPO $25.2–25.4 billion and an adjusted free-cash-flow margin of 38.0%, with fiscal 2028 still framed as the 40% year.
Three things settle over the next two prints. Whether the platform disclosure is repeated quarterly rather than annually, which is what would make the fiscal 2027 guide checkable at all. Whether first-half net new ARR lands near $400 million a quarter as the second-half loading implies, or the shape is being managed. And whether the reconciliation between the platform guides and the total revenue guide closes through the platforms or through the next acquisition.
What we learned
- The organic disclosure is gone; a platform split replaced it. No ex-acquisition revenue, ARR or current RPO growth this quarter. In its place, fiscal 2026 revenue by platform: Network & AI Security $8.35B, Cortex $1.92B, Idira $1.26B pro forma.
- Nearly $1bn of net new ARR includes a nine-figure one-off. NGS ARR went $8.13B to $9.10B — $970M of net new against the $770–820M the guide implied — and the CFO said it carries a nine-figure benefit from one migrating customer.
- Fiscal 2027 needs $2.0bn of net new ARR, 40% of it in the first half. The $11.075–11.175B guide implies $1.98–2.08B; with 60–61% guided into the second half, the first two quarters average about $400M against the $970M just printed.
- The platform guides do not add up to the revenue guide. Network & AI up low double digits, Cortex about 30% and Idira about $1.5B sum to roughly $13.2–13.5B against a $14.10–14.20B total — $0.6–1.0B unaccounted for.
- Fiscal 2027 operating income grows about 24%; earnings per share grow 8–9%. The wedge is dilution: non-GAAP diluted shares were 764M in fiscal 2026 and are guided to 844–847M, up another 10–11%.
Palo Alto Networks reported the quarter and fiscal year ended 31 July 2026 after the US close on 1 September 2026. Revenue, the product and subscription split, the GAAP and non-GAAP results and their bridge, cash flow, the balance sheet, NGS ARR, remaining performance obligations and all first-quarter and fiscal 2027 guidance are the company's own, from its results release; the full figure set is on the Q4 FY2026 earnings page. Quotations, the revenue-by-platform figures, the platform guides, the nine-figure ARR benefit, the second-half ARR loading and the platformization count are from the earnings call held the same afternoon and sit on the call page. Consensus of $0.98 and $3.35 billion is press-reported from third-party estimate feeds on a non-GAAP basis and is not a series this site stores or verifies. Ours rather than the company's: every net new ARR figure, which is one balance subtracted from another; the implied per-quarter first-half rate; the platform-guide addition and the growth rate it implies for Network & AI Security; the implied fiscal 2027 operating income and the dilution comparison; the trailing Rule of 40; the ~2,500 platformization base; and the fiscal 2030 compound rates. The model comparison uses our Palo Alto Networks model of 27 August 2026, whose fair values and 9x exit multiple are assumptions, not company forecasts. The $362.09 close and $356.00 after-hours quote are 1 September 2026; a live quote will differ. All guidance is forward-looking.
