The received wisdom about Applied Materials is that its shares fell after August’s earnings because the numbers disappointed. They did not. Applied reported record revenue of $9.12 billion for its fiscal third quarter ended 26 July 2026, up 25 percent year over year, with record operating income, record non-GAAP earnings per share of $3.50, and a fourth-quarter guide of $10.25 billion that was well ahead of where consensus sat going in. The stock then fell 5.1 percent in the following session, rebounded 5.6 percent on 17 August, and gave it back in a sector-wide equipment selloff on 18 August that hit every major peer. As at the close on 18 August 2026, AMAT trades at $514.33, some 28.9 percent below its closing high of $723.00 set on 30 June 2026, having still exactly doubled year to date from its $256.99 close on 31 December 2025. The gap between a record print and a negative reaction is not a sentiment glitch. It is the market repricing the quality of those earnings, and the evidence for that sits in Applied’s own cash flow statement.
Here is the number almost nobody quoted. Across the first nine months of fiscal 2026, Applied’s GAAP net income rose 44.5 percent to $7.37 billion. Over the identical nine months, its non-GAAP free cash flow fell 2.1 percent, to $3.58 billion from $3.66 billion. Three measures of the same three quarters point in three different directions: GAAP earnings per share up 46.6 percent, non-GAAP earnings per share up 20.4 percent, free cash flow down 2.1 percent. The further you travel from the accounting line toward the cash line, the worse the growth looks. Cash conversion — free cash flow as a share of net income — collapsed from 71.7 percent to 48.6 percent in a single year. That is what a 29 percent drawdown on record results is actually pricing, and it is a far more coherent explanation than “the market is being irrational about AI.”
Key facts
- Q3 FY2026 revenue $9.115 billion, up 25 percent year over year — a company record — with GAAP gross margin of 50.3 percent — Applied Materials Q3 FY2026 earnings release, 13 August 2026
- Q4 FY2026 guidance: revenue $10.25 billion ± $500 million, non-GAAP diluted EPS $4.02 ± $0.20 — Applied Materials, 13 August 2026
- Nine-month free cash flow $3.58 billion, down 2.1 percent, against nine-month net income up 44.5 percent — derived from the Q3 FY2026 release
- China revenue $2.506 billion, down 1.6 percent year over year, falling from 35 percent to 28 percent of the total — Q3 FY2026 release
- DRAM rose to 26 percent of Semiconductor Systems revenue from 22 percent — roughly $1.83 billion, with DRAM revenue up 52 percent year over year — Q3 FY2026 release
- Spot $514.33 (close, 18 Aug 2026) sits at 32.0 times the annualised run rate of the Q4 guide, against 45.0 times at the 30 June high — FinanceFeeds calculation
What actually happened in the quarter
Take the print on its own terms first, because it was genuinely excellent. Revenue of $9.115 billion compared with $7.302 billion a year earlier. GAAP operating income of $3.08 billion represented 33.7 percent of revenue, against 30.6 percent a year ago. Non-GAAP operating margin reached 34.0 percent. GAAP earnings per share of $3.17 and non-GAAP earnings per share of $3.50 were both up more than 40 percent. Cash from operations was a record $3.04 billion.
The margin story is the one management leads with, and it withstands scrutiny. I pulled Applied’s quarterly gross profit and revenue tags directly from its SEC XBRL filings and computed the year-over-year change in GAAP gross margin for every quarter back to 2019. The streak of year-over-year expansion begins with the quarter ended 30 July 2023, when margin turned up 21 basis points, and runs unbroken to the quarter just reported. That is exactly thirteen quarters. The company’s claim is not marketing; it is arithmetic, and it checks out to the quarter.
But the same dataset shows something management did not highlight. The rate of that expansion has been decaying. Year-over-year gross margin gains ran at 164 basis points in the April 2025 quarter and 150 basis points in July 2025, then fell to 20 basis points in January 2026 and 81 basis points in April 2026. The streak is intact; its derivative is not. A thirteen-quarter run of expansion sounds like a structural re-rating of the business. A run whose increments have more than halved looks more like a cycle approaching its comfortable middle. Both readings are available from the same series, and the market appears to have chosen the second.
Management’s framing was unambiguously confident. “As the rapid global adoption of AI drives unprecedented demand for our materials engineering solutions, we are further raising our Semiconductor Systems revenue expectations for calendar 2026 and are confident we will grow faster than the market this year. Based on the increased demand visibility we are receiving from our customers, we expect another strong growth year for Applied Materials in 2027,” said Gary Dickerson, President and CEO, in the earnings release filed with the SEC.
Where the growth actually came from
The geographic table on page eleven of the release is the most underrated disclosure in the entire document, and it reframes the China debate completely.
Applied’s revenue grew by $1.813 billion year over year. China contributed negative $42 million of that: $2.506 billion against $2.548 billion a year earlier, down in absolute dollars, and down from 35 percent of the total to 28 percent purely because everything else grew around it. United States revenue doubled to $1.367 billion, Europe tripled to $483 million, Korea rose 31 percent and Southeast Asia grew 92 percent.
This matters for how you weigh the standing China bear case. The conventional worry is that export controls will one day remove a large slice of Applied’s revenue. The data says that adjustment has substantially already happened: China is no longer growing, and the company has replaced it with domestic and European fab construction. FinanceFeeds has tracked how Chinese fabs have worked around equipment export curbs, and the rise of domestic Chinese memory makers documented in the CXMT listing prospectus shows why the mainland’s tool demand has not vanished so much as changed hands. The residual risk is real but it is now a 28 percent exposure that has already stopped contributing growth, not a 35 percent exposure about to be surprised. Management is more optimistic than that still: Hill told analysts China was 26 percent of Semiconductor Systems plus services revenue — a narrower denominator than the 28 percent of group revenue in the filed table — and added that Applied now expects its China revenue to increase this calendar year, led by investment in 28-nanometre foundry-logic.
The concentration that grew instead is DRAM. Within Semiconductor Systems — $7.04 billion of the quarter’s revenue — DRAM rose to 26 percent from 22 percent, while foundry and logic slipped to 67 percent from 69 percent and flash fell to 7 percent from 9 percent. In dollars that puts DRAM at roughly $1.83 billion for the quarter. Hill was more precise on the call: “DRAM revenue, which includes HBM packaging, grew by 52% year-over-year to record levels.” And he named DRAM first among the drivers of the coming half.
“Applied Materials achieved its 13th consecutive quarter of year-over-year gross margin expansion, demonstrating the increasing value we create by enabling better chips, systems and fab returns,” said Brice Hill, Senior Vice President and CFO. “We expect continued strong revenue growth in the second half of the calendar year, particularly in DRAM as well as leading-edge foundry-logic and advanced packaging. Looking further ahead, we are making additional manufacturing capacity investments to support projected demand through the end of the decade.”
That is the bear case stated by the company itself, without meaning to be. The fastest-growing slice of the mix is the most notoriously cyclical product in semiconductors. Memory capital expenditure has historically swung far harder than foundry-logic in both directions, which is precisely the dynamic behind the peak-cycle anxiety we covered around SanDisk’s investor day and in our bull-bear work on Micron.
What the sell-side actually objected to
It is worth being precise about the complaint, because it was narrower than the price action suggested. Morgan Stanley analysts wrote that “following peer earnings that pointed to system shipment growth of more than 40%, we thought AMAT could match this trajectory and guide OctQ revenue to $10.5bn,” before concluding: “We were wrong: AMAT guided to $10.25bn.” Summit Insights Group made the same point to Reuters, noting that Applied’s topline performance “has lagged peers such as ASML and Lam Research.” The objection was not that the guide was weak. It was that it was $250 million short of a whisper number set by rivals who had reported first.
That comparison is the context most coverage missed. Lam Research guided its September quarter to $8.10 billion against $7.13 billion of consensus and rose about 18 percent. KLA beat, guided broadly in line, and fell 10.5 percent. ASML raised full-year guidance for the second time this year and closed lower. Applied was not singled out: it landed in a sector where the bar had been reset by whoever printed first, and where two of four major names sold off on good numbers.
Notably, no analyst downgraded the stock. Price targets moved in both directions — JPMorgan raised from $515 to $660, Bernstein from $675 to $700 and RBC from $520 to $600, while BofA cut from $720 to $650, B. Riley from $790 to $700 and UBS from $705 to $675 — but every one of those firms kept a Buy or Outperform rating. The consensus target sits near $641. The professional community did not change its mind about the business; it changed its mind about the price.
The earnings-quality gap, quantified
Return to the cash flow statement, because this is where the valuation argument lives.
| Nine months to late July | FY2025 | FY2026 | Change |
|---|---|---|---|
| GAAP net income | $5,101m | $7,370m | +44.5% |
| GAAP diluted EPS | $6.29 | $9.22 | +46.6% |
| Non-GAAP diluted EPS | $7.25 | $8.73 | +20.4% |
| Capital expenditure | $1,475m | $1,988m | +34.8% |
| Non-GAAP free cash flow | $3,655m | $3,580m | −2.1% |
| Cash conversion (FCF / net income) | 71.7% | 48.6% | −23.1pp |
Three things drive the wedge. Capital expenditure is up almost 35 percent as Applied builds the capacity Hill described as supporting demand “through the end of the decade.” Working capital is absorbing cash as inventory and receivables build into the fastest sequential revenue ramp in company history. And GAAP earnings have been flattered by items that are not operations at all: the nine-month reconciliation strips out $1.01 per share of unrealised gains on strategic investments while adding back a $0.32 per share legal settlement. Strip both and the underlying trajectory is the non-GAAP figure, up 20 percent, not the GAAP figure, up 47 percent.
None of this is a scandal. Every one of those items is disclosed, and heavy capital investment ahead of a genuine demand cycle is exactly what a shareholder should want a capital-goods company to do. But it does change what the multiple should be. An investor paying for a business compounding free cash flow will pay more than one funding a capital build against a cyclical order book. The market re-rated from 45.0 times the annualised run rate of guided earnings at the June high to 32.0 times today. That is not disbelief in the earnings. It is a downgrade of their durability.
The pattern is familiar from hyperscaler results this year, where Microsoft’s capital expenditure was rewarded and Meta’s was punished on the same logic: the market pays for capex it believes converts and discounts capex it does not. Applied has provisionally joined the second group. Rising long-term yields, with the 30-year Treasury at a 19-year high, compound it, because cash flows pushed further out are worth less today.
Regulatory tension
The export-control regime cuts both ways for Applied. Restrictions on advanced tool shipments to China have capped one revenue pool — the flat $2.506 billion above is the visible result — but the same industrial policy is why United States revenue doubled and European revenue tripled. Subsidised domestic fab construction transfers demand from one geography to another, and Applied sells into both.
There is also a specific, dated catalyst that deserves more attention than it gets. Applied disclosed in an October 2025 filing that it expected the Bureau of Industry and Security Affiliates Rule to reduce fiscal 2026 net revenue by approximately $600 million. That rule was then suspended in November 2025 — and the suspension expires automatically on 10 November 2026, inside Applied’s fiscal first quarter of 2027. Unless it is extended, a quantified revenue headwind the company has never revised reappears precisely as the 2027 growth year management is promising begins. Separately, Applied took a $253 million charge in February 2026 to settle a BIS export-compliance matter covering shipments made between 2020 and 2022; that charge is the legal settlement line sitting in this year’s accounts. So the regulatory distribution is skewed modestly negative, but far less so than in 2023, when China was a third of revenue and rising. The more acute regulatory risk now sits with customers rather than with Applied itself — whether Intel’s subsidised expansion and TSMC’s overseas fabs convert announced capacity into actual tool orders on the announced timetable.
Bull case: $760
The bull case is worth $760, about 48 percent above the 18 August close of $514.33, and it requires clearing both the $723.00 closing high and the $739.67 intraday high set in June.
It rests on three conditions. Demand visibility has to convert into a 2027 wafer fab equipment upcycle broad enough that Applied grows faster than the market for a second year running. The mix shift has to keep working, and here the evidence is genuinely strong: Dickerson raised the packaging outlook from more than 50 percent growth to “more than 70% in calendar 2026” in the space of a single quarter, and told analysts that leading-edge foundry-logic, DRAM and advanced packaging together “represent around 80% of wafer fab equipment growth in 2026 and 2027.” Semiconductor Systems operating margin at 37.7 percent against 33.0 percent a year ago suggests the mix is already paying. And crucially the capital expenditure cycle has to peak, letting free cash flow snap back toward net income once the build is paid for. If conversion returns to 70 percent-plus on an earnings base 40 percent larger, the cash flow growth the market is withholding credit for arrives all at once. At $760 the shares would trade near 47 times the current guided run rate, so the bull case also needs 2027 earnings materially above that run rate to avoid multiple expansion doing all the work.
Bear case: $310
The bear case is $310, roughly 40 percent below spot. It is not a worst-case invention; it is empirically anchored to what this stock has just done. Between the closing high of $723.00 on 30 June 2026 and the trough of $436.45 on 29 July 2026, AMAT fell 39.6 percent in twenty-one sessions. A repeat of a drawdown the shares experienced four weeks ago takes them to almost exactly $310.
It also lands in a defensible valuation zone. $310 sits between the 2026 opening price of $268.87 and the end-February level of $372.30, meaning the year’s re-rating substantially unwinds without a full round trip. On the annualised run rate of the Q4 guide it implies 19.3 times earnings — a mid-cycle multiple for a capital-equipment company, not a distressed one.
The path there requires nothing exotic. DRAM at 26 percent of systems revenue and growing 52 percent is a mix that historically mean-reverts hard; memory buyers cut orders faster than they place them. Realised volatility has run at 85.7 percent annualised over the last sixty sessions against 58.6 percent over the year, so a wide distribution is already priced. And the cash conversion problem does not have to worsen to hurt — it merely has to persist for another two or three quarters for the earnings-quality discount to become the consensus view rather than a minority one. The sector-wide de-rating we tracked through the summer’s AI chip selloff showed how quickly these multiples compress once the narrative turns.
What happens next
Three concrete expectations. First, the November print should hit or beat the $10.25 billion guide: the visibility management describes is a backlog, not a hope, and the ramp is already funded. Second, the line that will move the stock is free cash flow, not revenue — if Q4 does not close a meaningful part of the conversion gap, the multiple compresses further however good the top line looks. Third, the 2027 commentary is the swing factor, because the market is currently paying for roughly no growth beyond the guided run rate. Until cash conversion is answered we expect the shares to stay range-bound between the July trough and the June high, which makes November the first genuine catalyst.
Frequently asked questions
Why did AMAT stock fall after record Q3 FY2026 earnings?
Because earnings quality deteriorated even as quantity improved. Nine-month net income rose 44.5 percent while free cash flow fell 2.1 percent, as capex climbed almost 35 percent. Sell-side also flagged a guide $250m short of the whisper number peers had set. The market re-rated from 45 times the annualised guided run rate to 32 times — a judgement about durability, not about the reported numbers.
Is the 13 consecutive quarters of gross margin expansion claim accurate?
Yes. Year-over-year GAAP gross margin change computed from Applied’s SEC XBRL filings shows the streak beginning with the quarter ended 30 July 2023 and running unbroken to 26 July 2026 — exactly thirteen quarters. The caveat is that each year-over-year gain has more than halved from its 2025 peak.
How exposed is Applied Materials to China?
China was $2.506 billion, or 28 percent of Q3 FY2026 revenue, down from 35 percent a year earlier. Critically, Chinese revenue fell 1.6 percent in absolute terms, so China contributed none of the $1.813 billion of year-over-year growth. The exposure is meaningful but it is no longer a growth dependency.
What is the DRAM concentration risk?
DRAM rose to 26 percent of Semiconductor Systems revenue from 22 percent, with DRAM revenue up 52 percent year over year to about $1.83 billion, and the CFO named it first among second-half growth drivers. Memory capex is the most cyclical spending in semiconductors, so a rising share of revenue from it makes the guide less durable than the headline implies.
What are the bull and bear cases for AMAT?
The bull case is $760, requiring a 2027 wafer fab equipment upcycle, continued advanced-packaging mix gains and a peak in the capital expenditure cycle that restores cash conversion. The bear case is $310, matching in magnitude the 39.6 percent drawdown the shares suffered between 30 June and 29 July 2026 and implying 19.3 times the guided run rate.
Prices quoted are as at the close on 18 August 2026, when AMAT last traded at $514.33; the pre-market quote on 19 August was $510.00 at 07:41 EDT. The 52-week range is $154.47 to $739.67 on an intraday basis and $156.25 to $723.00 on closing prices. Bull and bear cases are analytical scenarios constructed by FinanceFeeds, not price targets, and nothing here is investment advice. Financial data is taken from Applied Materials’ fiscal Q3 2026 earnings release and its most recent Form 10-Q, both filed with the SEC; share price data is from stockanalysis.com; company statements are from Applied Materials investor relations.