PANW Palo Alto Networks, Inc. · NASDAQ

What this is. A technical and fundamental reading of one company. The technical half describes where the price has been and which levels it has respected; the fundamental half describes what the business earned and what the price is paying for it. They answer different questions and neither settles the other.

Where it stands. Palo Alto Networks closed Friday at $371.59 — 6.84% below the 52-week high it set on August 13, and 166.24% above the 52-week low it made on February 24. It sits at 89.5% of its own twelve-month range, above every moving average on this page, and 61.86% above its 200-day. It is up 94.53% over the year and 148.72% over six months. Almost the entire move is six months old.

What the price is paying for. Behind it sits a business that took $10.61B over the last twelve months and kept $842.8M — a net margin of 7.95% — for which the market is asking 323.12× trailing earnings, on a market capitalisation of $302.85B. Revenue over that window grew 19.51%. Net income over the same comparison fell 31.93%.

The number that explains the multiple. The most recent quarter Palo Alto has filed — Q3 2026, to April 30 — recorded revenue of $3.002B, up 31.15% year on year, and a net loss of $177M. The same quarter appears in the earnings calendar as $0.85 of EPS, 7.19% ahead of the consensus. Both figures are real, they are not the same measurement, and the section near the foot of this page explains which one is used where and why the P/E reads as it does.

The calendar. The next report is Tuesday, September 1, 2026 — two sessions after this page was written. The last four reports all beat the consensus estimate, and three of the four were met with a lower open the following morning.

Educational material. Not investment advice, and not a recommendation to buy or sell anything. No price target on this page is a forecast; every level is a place the market has already traded.

This chart opens on February 24, 2026 — the day of the 52-week low, an intraday $139.57 and a $141.67 close — and ends on August 28 at $371.59. That is 162.3% across six months, computed. There are not many charts on this site where the first bar of the window is also the extreme of the year, and it makes the shape unusually simple to describe: this one went up, and the interesting part is where it stopped.

It stopped on August 13 at an intraday $398.88 and a $396.00 close, the highest of the year. The nine sessions that followed took it to a $339.31 close on August 26 — 14.32% off the high in under two weeks, with the intraday low of $328.00 printed that same morning.

Then August 27 happened: a +12.83% session, from a $358.56 open to a $382.85 close. That is the single largest daily move on the entire 276-bar series — larger than anything the stock did on the way up, and larger than any of its reactions to a report. It traded 7,308,818 shares doing it, which is only ×1.28 the 30-day average. A move of that size on volume of that size is worth pausing over: it means the price travelled a very long way without an unusual number of participants.

Friday gave some of it back — −2.94% to $371.59, on ×1.03 the average. The close is 1.29% above the 20-day average, 8.20% above the 50-day and 6.84% below the high of two weeks ago, which places it in the upper half of a twenty-session range that is 21.61% wide — an unusually large box, and the direct consequence of the two sessions just described.

Level	What it is	Distance from Friday's close
$398.88	The 52-week high, traded August 13, and also the ceiling of the twenty-session range. The only level overhead on the entire chart.	+$27.29 · +7.34%
$366.86	The 20-day average. Friday closed above it, having been below it for four of the previous five sessions.	−$4.73 · −1.27%
$343.42	The 50-day average — the first line beneath the price with a quarter of history behind it.	−$28.17 · −7.58%
$328.00	The floor of the twenty-session range, traded on the morning of August 26 — the session immediately before the +12.83% day.	−$43.59 · −11.73%
$242.63	The 150-day average. The price has not traded here since early June, and there is very little between it and the range floor.	−$128.96 · −34.70%
$229.58	The 200-day average. The distance to it is the plainest statement on this page of how fast the last six months were.	−$142.01 · −38.22%

Four averages, for orientation: 20-day $366.86, 50-day $343.42, 150-day $242.63, 200-day $229.58. The close is above all four — by 1.29%, 8.20% and 61.86% against the first, second and fourth. The gap between the 50-day and the 150-day is $100.79, which is what a chart looks like when a price has moved much faster than the averages drawn through it can follow.

Worth stating plainly: below $328.00 this chart is nearly empty for a long way. The price crossed the ground between the range floor and the 150-day average in a matter of weeks on the way up, which means very little was traded there, which means it offers very little in the way of a floor on the way back.

A trader measures two distances: how far the price would have to fall before the idea is wrong, and how far it could travel to the next level up. On this chart the second distance is unusually short and the first is unusually long, which is the opposite of the arrangement anybody is looking for. It is worked twice below.

Attempt one — the stop the chart suggests. Just under the floor of the twenty-session range.

The reward is smaller than the risk. That is not a subtle finding and it does not require interpretation: the only level above the price is nearer than the only level below it, so an idea measured between them risks $44.09 to make $27.29. The stop is sound — at 2.46× the fourteen-day average true range of $17.90, it is comfortably outside ordinary daily noise. It is the target that is missing.

Attempt two — a stop close enough to fix the ratio. Under Friday's low of $355.50.

The ratio now reads well, and the last line is why it should not be believed. A stop $16.59 away on a stock that travels $17.90 in an average day is inside a single ordinary session's range. It would be taken out by a normal Tuesday, and the ratio it produces is an artefact of choosing a stop too close to survive — the same arithmetic flattery this site has documented before.

And then the calendar. Palo Alto reports on Tuesday, September 1, two sessions away. The four reports in this window produced next-morning gaps of −4.10%, −8.53%, −0.89% and +6.61% — a mean absolute gap of 5.03%, computed. That is larger than the tight stop's 4.46%. A gap is not a price anybody can be filled at, so on the second version the report alone is, on average, enough to open the market on the far side of the stop.

The honest conclusion is that this chart does not currently offer a well-formed measurement in either direction. One version has no room above; the other has no room below. That is a description of a price sitting close under its own high, two days before a report — not a fault in the arithmetic.

Palo Alto has reported four times in the window this page covers, and beat the consensus EPS estimate every time. Here is what the stock did the following session — first at the open, before anybody could act, and then by the close.

^† Volume on the following session as a multiple of the 30-day average up to that point.

Four beats, three lower closes. Two of the five largest down moves on the whole series are the sessions after a report that beat — −7.42% on November 20 and −6.82% on February 18. On this company's recent record, beating the published estimate has been a poor guide to what the stock does next.

The reason is worth stating plainly, because it is the most common misreading of a page like this: the consensus estimate is what the market has already been told to expect, and it is priced in before the release is read. What moves the stock is everything else in it — the outlook, the billings, the margin, the guidance — none of which appears in the single number the headline compares. A company can beat and fall, and this one has done so three times out of four.

One asymmetry separates this stock from the other pre-earnings page published today: Palo Alto's gaps are smaller. A mean absolute gap of 5.03% is meaningful, but it is half of what the same measurement produces on Lululemon, and none of the four exceeded 8.53%.

Friday traded 5,845,766 shares against a 30-day average of 5,690,810 — ×1.03, which is as close to an ordinary day as this measurement gets. The 90-day average is 7,313,002, higher than the 30-day, so the last month has been quieter than the quarter before it even as the price has swung more widely.

The figure that deserves attention is Thursday's. The +12.83% session — the largest single move on the entire series — traded 7,308,818 shares, or ×1.28 the average. Set that against the other four largest moves on the chart: ×2.64 on May 29, ×1.82 on June 29, ×2.83 on November 20 and ×2.25 on February 5. Every one of the other big moves carried more relative volume than the biggest one did.

This page will describe that and not explain it, because the daily bars do not contain the explanation. What can be said from the array is narrow and still useful: the largest price move in a year happened on close to normal participation, two sessions before a scheduled report, and the following session gave back 2.94% of it on ×1.03. A move that large usually leaves a mark on the volume bar beneath it. This one did not.

1 · What it takes in, and what it keeps. Over the last twelve months Palo Alto booked $10.606B of revenue and kept $842.8M as net income — a net margin of 7.95%. Gross margin is 71.94%, which is a software company's number and much the more flattering of the two; operating margin is 9.65%. The distance between 71.94% and 9.65% is the cost of selling the software rather than making it.

2 · The direction those figures are moving. Revenue over the last twelve months is 19.51% higher than the twelve months before it — genuine, substantial growth. Net income over the same comparison is 31.93% lower, and earnings per share 34.29% lower: $1.15 against $1.75.

The last filed quarter states it in its strongest form. Q3 2026, to April 30 and filed June 3: revenue $3.002B, up 31.15% on the same quarter a year earlier — and an operating loss of $183M, a net loss of $177M, and diluted EPS of −$0.28. Revenue grew by nearly a third and the quarter did not make money.

3 · What the market is paying. 323.12× trailing earnings, on a market capitalisation of $302.85B across 744M shares. The company's sales over the last twelve months are $10.61B, so the business is valued at roughly 28.6× what it sells in a year.

A P/E of 323× is not a number that can be compared to much, and this page will not dress it up as one. What it means arithmetically is that the trailing earnings are very small relative to the price — and the section below shows exactly why they are small, which is that one quarter of the four in the window recorded a loss. What it does not mean is that the market expects the next twelve months to look like the last twelve; a multiple is a ratio between today's price and the past, and it carries no forecast inside it. Whether the price is reasonable depends entirely on figures that have not been filed yet, and this page has none of them.

4 · Three years, for shape.

Revenue has risen every year and by a lot — $6.89B to $8.03B to $9.22B, and the trailing twelve months at $10.61B continues it. The profit line does not follow the same path. The 2024 net income of $2.578B is roughly 3.8× that year's operating income of $684M — computed — which means most of it arrived from below the operating line rather than from selling more software. This page reports that shape and does not name its cause, because the income statement in hand shows the amount and not the reason. The practical consequence is that 2024's $3.64 of EPS is not comparable with the years on either side of it, and any growth rate measured from it will mislead.

Two figures for completeness: beta is 0.893, meaning the stock has historically moved slightly less than the wider market — which, on a stock that rose 94.53% in a year and 12.83% in a day, is a reminder that beta is a long-run average and not a description of any particular session. Palo Alto pays no dividend; the entire return comes from the price.

The last quarter appears as two very different numbers depending on where you look, and this page uses both, so it owes you the difference. The filed income statement gives diluted earnings per share of −$0.28 — a loss. The earnings calendar records the reported figure as $0.85, and a beat of 7.19%. Neither is wrong, and on this company the gap is about as wide as it gets.

$1.13 a share across roughly 744M shares is about $841M in a single quarter — computed. The filed statement records a net loss of $177M for the same three months. The difference is made up of charges that a company's own presentation typically sets aside — share-based compensation, acquisition-related amortisation and similar items — and whether they should be set aside is a real argument rather than a trick. They are genuine costs by any accounting standard, and they are also not cash leaving the business this quarter. Serious people hold both positions.

Which is used where. The P/E, the margins and every TTM figure on this page come from the filed statements — the stricter, fully-audited view, and the one that ties to the balance sheet. That is the entire reason the P/E here reads 323.12×: the trailing twelve months include a quarter that lost money, which pushes trailing EPS down to $1.15 and the ratio up. On the other basis it would be dramatically lower. The earnings table further up compares reported EPS against the consensus that was set for reported EPS, because comparing a figure to an estimate of a different figure would manufacture a beat that never happened.

Mixing the two is the most common arithmetic error made with a page like this, and on this company it is not a small one: it is the difference between a quarter that earned $0.85 and a quarter that lost $0.28.

The business described above grew revenue 31.15% in its last filed quarter and lost $177M doing it. Both halves of that sentence were equally true on June 9, when the stock closed at $260.52, on August 13, when it closed at $396.00, and on Friday, at $371.59. The quarter did not change by 52%. The price did.

That gap is the reason both halves of this page exist. The fundamentals describe what you would be buying. The chart describes what other people have recently been willing to pay for it. Neither answers the other's question — and on this stock the distance between the two questions is unusually visible, because a 323× multiple and a +94.53% year are the same twelve months seen from two different measurements.

The August 27 session makes the same point inside a single day: +12.83%, the largest move on the chart, with no filing, no report and nothing in the fundamentals changing between Wednesday's close and Thursday's. Whatever moved the price that day is not in the income statement, because the income statement was identical on both sides of it.

TTM	Trailing twelve months — the last four reported quarters added together, rather than the last full financial year. On this company one of those four is a loss, which is why the trailing figures read as they do.
EPS (diluted)	Profit divided by the share count, counting shares that could exist if every option and convertible were exercised. The stricter of the two usual figures.
GAAP vs reported ("adjusted")	The filed, audited statement versus the company's own presentation with certain charges set aside. Both are published; they are not interchangeable, and mixing them is the error the section above exists to prevent.
P/E	Price divided by earnings per share. What the market pays for one dollar of last year's profit. It contains no forecast.
Net margin	Profit as a percentage of revenue. How much of each dollar taken in is kept.
Gross margin	Revenue less the direct cost of delivering the product, as a percentage of revenue. What survives before sales, research, wages and tax.
Operating income	Profit from running the business, before interest, tax and one-off items. When net income is far larger than operating income, most of the profit came from somewhere other than operations.
Gap	The difference between one day's close and the next day's open — price movement that happened while the market was shut, and which no order placed inside market hours can be filled at.
ATR (average true range)	The average distance a price travels in a day, including any overnight gap, over the last N days. Used here to ask whether a stop is inside or outside ordinary daily noise.
Beta	How much this moves when the wider market moves, averaged over a long window. Below 1 means historically less than the market — which says nothing about any single session.
52-week high / low	The highest and lowest price traded in the last year, measured here from the daily bars so each has a date attached.
Moving average (SMA)	The average close of the last N days, redrawn daily. A shorter one hugs the price; a longer one describes the trend. When the price is far above the long one, the trend has moved faster than the line measuring it.
Reward-to-risk	Distance to the next level up divided by distance to the stop. Arithmetic on levels; it says nothing about probability — and, as the second attempt above shows, it can be flattered by choosing a stop too close to survive an ordinary day.
Consensus estimate	The average of the forecasts analysts have published for a coming report. It is what the market has already been told to expect — which is why a company can beat it and the stock can still fall, as this one has after three of its last four reports.

Next earnings: Tuesday, September 1, 2026 — two sessions after this page was written — with consensus at $0.977 EPS on $3.350B of revenue. That revenue figure would be 11.60% above the $3.002B of the last reported quarter, computed. The EPS figure is compared against reported EPS, not the filed statement, so the meaningful comparison is with the $0.85 the calendar records for last quarter — 14.9% higher, computed — rather than with the −$0.28 that was filed.

It is the one scheduled event on the calendar, and it arrives with the price 6.84% below a 52-week high set sixteen days earlier and two sessions after the largest single-day move on the chart. The last four reports produced overnight gaps averaging 5.03% in absolute terms, and all four beat the estimate the market had been given.

Palo Alto pays no dividend, so there is no ex-date on the calendar and no income to offset a price decline.

The technical half. 276 daily bars from the platform's own price pipeline, pulled August 29, 2026, covering 2025-07-25 to 2026-08-28. The OHLC, the 52-week extremes and their dates, every return, the 30- and 90-day average volumes, the average true range, the five largest moves, the twenty-session range and all four moving averages are computed from that array. The four legs in the timeline are close-to-close between bars located by an eleven-bar pivot scan rather than chosen by eye. Nothing was read off a chart by sight, and the chart above is generated from the same array by tools/analysis-chart.mjs, so a level in the prose and a line on the picture cannot disagree.

The fundamental half. Revenue, gross profit, operating income, net income, diluted EPS and the share count come from Palo Alto's own filed income statements through the same provider — four quarters for the trailing-twelve-month window, four more for the year-on-year comparison, and three financial years for the table. Reported and expected EPS, and every gap and next-session move beside them, come from the earnings calendar and the daily bars respectively. The derived figures are marked computed where they appear: the margins, the P/E, the sales-to-market-cap ratio, the average absolute gap, the leg percentages, the net-to-operating-income ratio for 2024, and the reported-to-filed EPS difference in the section that explains the two figures.

No external sources, no models, no price targets. Nothing here came from an analyst note, a forecast or another site. Where the chart has no level to measure to — above $398.88, and between $328.00 and $242.63 — this page says so rather than supplying a number. Where the filings show an amount but not its cause, as with the 2024 net income, the amount is reported and the cause is not guessed at.