Palo Alto Networks: Revenue Up 34%, ARR Up 63% - and the Bottom Line Swung From Profit to a $282 Million Loss

Palo Alto Networks reported yesterday on its fiscal fourth quarter and on fiscal 2026, which ended on 31 July. Quarterly revenue rose 34% to $3.41 billion and Next-Generation Security ARR rose 63%, but the accounting bottom line showed a loss of $282 million against a profit of $254 million in the comparable quarter. This piece decomposes the gap between GAAP and adjusted earnings, shows which single line accounts for most of it, and explains what the guidance reveals about the underlying growth rate.

By Ilan Abramov8 min read
Palo Alto Networks: Revenue Up 34%, ARR Up 63% - and the Bottom Line Swung From Profit to a $282 Million Loss
* The cover image was generated with an AI tool and is not a photograph.

Palo Alto Networks published yesterday the results of its fiscal fourth quarter and of fiscal 2026, which ended on 31 July. The growth measures are strong, the company beat its guidance, and the accounting bottom line swung to a loss.

All three of those are true at once, and the explanation sits in a single line of the report.

Errors or inaccuracies are possible. Spotted something that looks wrong? Write to me and I will correct it.

What the Headline Said

Fourth quarter20252026
Revenue2,5363,410+34%
Next-Generation Security ARR-$9.10 billion+63%
Remaining performance obligations-$21.2 billion+34%

(in millions of dollars, except where noted otherwise)

Nikesh Arora, the company's chairman and chief executive, noted in the release that it had added nearly $1 billion of Net New NGS ARR in a single quarter, and that the target is $20 billion of ARR by fiscal 2030.

And What Sits Underneath

Here the picture changes.

Fourth quarter20252026
Gross profit1,8572,304
Gross margin73.2%67.6%
Operating income497172
Other income (expense), net95(441)
Net income (loss)254(282)
Diluted earnings (loss) per share0.36(0.35)

Operating income stayed positive. What turned the quarter into a loss sits below the operating line: a $536 million swing in other income and expense, from plus 95 to minus 441.

The Single Line Responsible for Most of the Gap

The reconciliation table the company publishes shows exactly what the move from GAAP to adjusted earnings consists of.

Fourth quarter 2026$ millions
GAAP net loss(282)
Share-based compensation487
Change in fair value of convertible notes and capped calls524
Amortisation of acquired intangible assets281
Acquisition-related costs68
Litigation-related charges3
Tax adjustments(228)
Adjusted net income853

The largest item in the gap is not share-based compensation but a line of $524 million, and the footnote states plainly what it consists of: changes in the fair value of convertible senior notes acquired from CyberArk.

ניטרלי

This mechanism is worth pausing on, because it runs against intuition.

When a company carries a note convertible into its own shares and measures it at fair value, a rise in the value of the conversion right increases its liability - and is therefore recorded as an expense in the income statement.

So this is not an expense that indicates any deterioration in the business. It arises because the right to convert into those shares became worth more. No cash leaves, and operations are untouched.

And in the same breath - it is a real liability on the balance sheet. Long-term convertible senior notes now stand at $1.774 billion, against zero a year ago.

In per-share terms: of the $1.37 gap between the GAAP loss of $0.35 and the adjusted profit of $1.02, $0.64 comes from that line alone.

And What Does Recur

It is only fair to state the other side too, because not everything stripped out on the way to adjusted earnings is one-off.

Share-based compensation came to $487 million in the quarter and $1.712 billion for the full year. That is not a line that disappears next year, and the company itself writes in the release that the non-GAAP adjustments include items that are "recurring and will be reflected in the company's financial results for the foreseeable future, such as share-based compensation".

And the amortisation of acquired intangibles - $281 million in the quarter against $37 million in the comparable quarter - will run for years, because it is a direct consequence of the acquisition.

What Was Bought, and at What Price

The CyberArk deal closed on 11 February 2026, that is, in the middle of the fiscal year.

The deal
Reported valueabout $25 billion
Cash consideration$2.3 billion
Share consideration112 million shares
Per CyberArk share$45.00 in cash and 2.2005 Palo Alto shares
Regulatory clearancesUS, European Union, UK and Israel

In May 2026 the CyberArk product portfolio was consolidated under a single brand named Idira, Palo Alto's identity security platform. Chief financial officer Dipak Golechha named it in the release alongside Network & AI Security and Cortex as one of the three platforms that drove the quarter.

Goodwill on the balance sheet rose from $4.567 billion to $22.010 billion - an addition of about $17.4 billion.

The Dilution, and What It Cost Per Share

דובי

The basic share count for the quarter rose from 669 million to 817 million - about 22%.

And this is what that did to adjusted earnings:

Fourth quarter
Adjusted net incomefrom $673m to $853m, +27%
Adjusted earnings per sharefrom $0.95 to $1.02, +7.4%

Earnings grew 27%, earnings per share grew 7.4%. The difference is the shares issued to buy CyberArk.

For the full year the picture is milder, because the deal closed only in February: adjusted earnings +25%, adjusted earnings per share +15%.

The Guidance, and What It Reveals

This is the part I find most interesting in the report.

Fiscal 2027 guidance
Revenue$14.10 to $14.20 billion, growth of 23% to 24%
Next-Generation Security ARR$11.075 to $11.175 billion, growth of 22% to 23%
Remaining performance obligations$25.2 to $25.4 billion, growth of 19% to 20%
Adjusted operating margin29.5%
Adjusted earnings per share$4.16 to $4.19
Adjusted free cash flow38.0% of revenue

ARR is growing 63% today, and the company guides to 22% to 23% next year.

That is not a forecast of a dramatic slowdown in the business. It is what happens when an acquisition laps. CyberArk entered the books in February 2026, so it inflates the growth rate in every quarter where the comparison base still excludes it. Once it sits on both sides of the comparison, the underlying growth rate is what remains.

So 63% is a correct number that cannot be compared with last year, and 22% to 23% is the company's own estimate of the pace without the acquisition effect.

And What the Stock Did

The shares fell about 5% during the trading session of 1 September - a day when the broad market fell and yields rose - and slipped about 2% more in extended trading, after the results. The report was published after the close.

A pattern is worth noting: this is the fourth consecutive quarter in which the company has beaten guidance, and in three of them the stock fell in response.

הזווית שלי

דעה אישית של אילן אברמוב - לא ייעוץ ולא המלצה

This report is a particularly good example of one working rule: when the gap between GAAP and adjusted earnings jumps, it is worth checking what it consists of before deciding how to treat it.

Here the gap is made of three quite different kinds of item, and in my view they deserve different treatment.

The first is the $524 million on the convertible notes. This is an accounting expense arising from a rise in the value of a conversion right, with no cash leaving and no bearing on how the business performed. I do not count it as operational damage.

The second is the amortisation of acquired intangibles. It is also non-cash, but it does represent something real - the company paid $25 billion, and amortisation is how that price is spread across the years. Anyone who always strips it out ends up with a picture in which acquisitions are free.

And the third is share-based compensation, $1.712 billion for the year. That is a recurring expense, the company says so itself, and it is well explained by the share-count column - which grew.

And what I take from all of it is not a conclusion about the company but a distinction that is easy to miss: the interesting figure in this report is not 63% but 22%. The first is an accurate description of what happened, and it includes an acquisition. The second is management's own estimate of the pace once the acquisition stops inflating the comparison.

When two numbers like that appear in the same release, the second is the one that deserves the attention. Not because the first is wrong, but because it is not comparable.

(It is important to stress: this is my personal opinion only, and nothing herein constitutes a recommendation to take any action.)