Palo Alto's Record $21 Billion Backlog Hides Why the Stock Fell 5%

Palo Alto Networks just had maybe its best quarter ever. Q4 revenue rose 34% to $3.41 billion, its contracted backlog crossed $20 billion for the first time at $21.2 billion, and management beat its own guidance on every line it guided. Then the stock fell about 5%, to roughly $362.

That is not a glitch. It is the market's hurdle showing up. A record result met a far higher price, and the outlook for next year says the growth the stock is paying for is about to slow.

The number the headline skips

The $21.2 billion backlog—the company calls it "remaining performance obligations" (RPO)—is real and worth understanding. It is contracted future revenue, the closest thing a software company has to a construction backlog. When it grows 34% year over year, it means customers are signing multi-year deals faster than they are being consumed. For a business like this, RPO is forward visibility, and $21.2 billion of it is a genuinely strong answer to anyone asking where the revenue is coming from.

The catch is that the level matters less than the growth rate. RPO is a stock of already-signed business; the number that predicts next year's revenue is how fast that stock is growing. And that is exactly where the stock market's disappointment lives.

For the coming fiscal year, management guides RPO to $25.2–$25.4 billion, growth of roughly 19–20%. That is down from 34% this year. Its flagship growth metric, next-generation security (NGS) annual recurring revenue, is guided to grow 22–23% in fiscal 2027, after climbing 63% to $9.1 billion in fiscal 2026. Revenue is guided to $14.1–$14.2 billion, up about 23–24%.

None of that is bad. A 22–23% growth rate at $11 billion-plus of revenue is enviable. It is the direction that moves a stock that nearly doubled in the past year.

Why 2026 looked faster than it was

Part of the deceleration is arithmetic, not weakness. Fiscal 2026's 63% NGS growth was flattered by two large acquisitions—identity security vendor CyberArk and the observability firm Chronosphere—which contributed roughly $1.6 billion of that ARR. Buying revenue makes the current-year growth rate look bigger than the organic run rate underneath it, and that comparison naturally falls out of the numbers next year.

Management also flagged a one-time effect: a nine-figure contribution from a large LLM customer migrating onto Chronosphere is expected to last only through the first quarter of fiscal 2027, which softens the near-term seasonal pattern.

The market is not worried about demand; it is worried about the price

This is the classic expectation-gap setup. The quarter itself scored high: revenue beat the $3.35 billion street estimate, non-GAAP EPS of $1.02 topped the guided $0.96–$0.98, NGS ARR beat guidance, and adjusted free cash flow hit $4.41 billion for the year at a 38.4% margin, above the company's own target. Bookings accelerated for a second straight quarter, and ~2,500 customers have now consolidated onto its platform, with that cohort renewing at over 120% net retention.

But the stock had already run roughly 100% year to date and trades near all-time highs with a market value around $295 billion. At that price it was embedding a very high growth assumption. The guidance for next year stepped that assumption down, and the stock repriced accordingly.

The valuation is the honest summary of the tension. Palo Alto trades around 26x trailing sales and roughly 85x the earnings it just guided for the coming year. That is a price that forgives almost no execution slip and requires deceleration to slow faster than the guide implies.

The margin line is the second risk

The other thing that moved the stock was costs. Q4 gross margin compressed 100 basis points to 74.8%, a mix shift toward cloud software plus rising memory and storage costs in the small hardware line. Management says cloud hosting costs will rise faster than revenue in 2027, which is why it guides operating margin to a modest 29.5%—only 30 basis points of expansion. For a business priced at 26x sales, an operating-margin story that has stopped expanding gives the multiple less support.

There is also the usually-overlooked gap between how the company reports and how it earns: the $1.02 headline EPS is non-GAAP, adjusted to strip out stock compensation and acquisition amortization. On a GAAP basis, Q4 produced a net loss of $0.35 per share. Investors are underwriting the adjusted number, so it is worth remembering what that adjusted number is smoothing.

The deeper question

The surface story is "record backlog, why did the stock drop?" The deeper question is whether the platform strategy keeps producing net new business at a price that holds. The scoreboard says the quarter was excellent, but excellent against a 70-point hurdle does not rescue a stock priced as if the hurdle were 90.

What to watch is not the revenue line—it converts from the backlog on schedule. Watch two things instead. First, net new NGS ARR in the next two quarters: after ~$970 million in Q4, whether that pace holds tells you if the booking engine is real or front-loaded. Second, gross margin, because if cloud hosting costs keep pulling it down and the 22–23% guide proves optimistic, the stock has far less cushion than its multiple requires.

A quiet confirmation that breaks the bear case: current RPO (the backlog due over the next twelve months) keeps growing at double digits even as the easier acquisition comparisons fade. That would mean the deceleration is the normalizing kind—an expensive stock returning to a sustainable growth rate—rather than the start of a slide.