Palo Alto Networks just posted its best quarter in years. Revenue up 15%. Earnings crushed estimates by 10%. Net income up 62%. And the stock dropped 6% after hours.
The culprit? Q3 EPS guidance of $0.78 to $0.80 — a mile short of the $0.92 Wall Street wanted. But here's the thing: that miss is almost entirely acquisition math, not business deterioration. And if you look past the EPS noise, the actual business is accelerating.
The Numbers
| Metric | Q2 FY2026 | Q2 FY2025 | YoY Change |
|---|---|---|---|
| Revenue | $2.59B | $2.26B | +15% |
| GAAP Net Income | $432M | $267M | +62% |
| Non-GAAP EPS | $1.03 | $0.81 | +27% |
| NGS ARR | $6.3B | $4.7B | +33% |
| RPO | $16.0B | $13.0B | +23% |
| Non-GAAP Op. Margin | 30.3% | 30%+ | Stable |
Every single line beat consensus. Revenue topped estimates. EPS blew past by 10 cents. RPO came in above the $15.78B StreetAccount number. Three straight quarters of 30%-plus operating margins.
So why the selloff?
The Acquisition Hangover
Nikesh Arora has been on a shopping spree that would make Bezos blush. In the past two months alone:
- CyberArk — $25B. The biggest deal in Palo Alto's history. Identity security.
- Chronosphere — $3B+. Cloud observability.
- Koi — announced today. Secures AI agents.
That's north of $28 billion in acquisitions. The CyberArk deal closed earlier this month. Chronosphere closed in January. When you bolt on $28B in acquired companies, your share count balloons and your near-term EPS gets crushed — even if the underlying business is printing money.
Q3 EPS guidance of $0.78 to $0.80 reflects 812 to 817 million diluted shares. A year ago, Palo Alto had roughly 700 million. That's 16% more shares eating into earnings. The business didn't slow down. The denominator got bigger.
Full-year non-GAAP EPS guidance was also trimmed slightly to ~$3.65-$3.70, largely reflecting acquisition-related dilution costs rather than organic weakness. The headline looks soft. The reason behind it is mechanical.
The Number That Actually Matters
Forget the EPS guidance for a second. Look at NGS ARR guidance for Q3: $7.94 to $7.96 billion. That's 56% year-over-year growth. Up from 33% this quarter.
Read that again. Next-generation security ARR is accelerating from 33% to 56% growth. In a $6B+ ARR business. That doesn't happen by accident — it happens because CyberArk and Chronosphere are getting bundled into the platform and customers are consolidating.
Q3 revenue guidance? $2.94 to $2.95 billion. That's 28-29% growth and actually above the $2.60B estimate. So Palo Alto is guiding for revenue acceleration too. The only "miss" is on the one metric most distorted by acquisition dilution.
Full-year revenue guidance tightened to $11.28 to $11.31 billion, with 22-23% growth and a 37% adjusted free cash flow margin. For a company that just swallowed $28B in acquisitions, that's execution.
The Platformization Thesis
Arora has been talking about "platformization" for two years now. The pitch: instead of buying 30 different security tools from 30 vendors, buy everything from Palo Alto. Network security, cloud security, identity (CyberArk), observability (Chronosphere), AI agent security (Koi).
The Q2 numbers suggest it's working. RPO hit $16B — that's contracts customers have already committed to but haven't been delivered yet. It grew 23% and beat estimates. When enterprises sign up for the platform, they sign multi-year deals. That backlog is the real signal.
And platformization is accelerating "due to AI," per Arora. As companies deploy AI agents across their infrastructure, they need unified security. You can't secure an AI agent with five different point solutions from five different vendors. That's exactly Palo Alto's pitch — and exactly why they bought Koi today.
The Bear Case
The skeptics have a point. $25B for CyberArk is expensive — and paying a premium for identity security when CyberArk was already richly valued raises integration risk. More than 20 acquisitions under Arora's watch means a lot of moving parts. Non-GAAP operating margins at 28.5-29% for the full year are solid but not expanding.
Then there's the valuation question. $PANW trades at a P/E north of 130x — steep for any software company, but especially steep right now. The broader SaaS sector is under a cloud of AI displacement fears, with investors questioning whether legacy software vendors can maintain pricing power as AI agents automate workflows. Palo Alto's cybersecurity niche is more defensible than most (you can't automate away the need for security), but that premium multiple still needs sustained execution to justify.
CHART_PLACEHOLDER_0
There's also a question of how much acquisition-driven ARR growth counts. When you buy $25B of CyberArk, you inherit their ARR. The organic growth rate matters more than the headline number, and Palo Alto doesn't break that out cleanly.
Bottom Line
Wall Street is trading the EPS miss. But the EPS miss is acquisition math, not business fundamentals. Revenue is accelerating. ARR is accelerating. The backlog is growing faster than expected. And Palo Alto is assembling the most complete cybersecurity platform in the industry at exactly the moment AI is forcing enterprises to consolidate their security stack.
The stock dropped 6% on a quarter where every actual business metric beat estimates. At 133x earnings with a software sector under pressure, you're paying a real premium for that thesis. But the underlying business justifies confidence — this is a disconnect driven by dilution mechanics, not deterioration. If the platformization flywheel keeps spinning, the multiple compresses on its own.
PANW