AMAT · Applied Materials, Inc. · Stock research

« # AMAT — Applied Materials, Inc.

Current Thesis

A week after a record quarter the price is still going the other way. Q3 FY2026, reported 2026-08-13, printed revenue $9.12B (+25% YoY), non-GAAP EPS $3.50 (+41% YoY), non-GAAP gross margin 50.4% and non-GAAP operating margin 34.0%, with a Q4 guide of $10.25B ±$0.50B revenue and non-GAAP EPS $4.02 ±$0.20. The stock closed 2026-08-14 at $507.18, down 5.1% on the session. By 2026-08-21 the reference close was $492.32 — 31.8% under the $722.23 52-week high set during June's +54% month, the best since 1975 (Benzinga, 2026-06-30). The Q3 FY2026 10-Q, covering the quarter ended 2026-07-26, confirmed the call's China figure at $2,506M and 28% of revenue against $2,548M and 35% a year earlier, so the filing-detail risk carried into the document did not fire. Estimates rose, most targets rose, price fell. That divergence between an accelerating P&L and a thinning marginal bid is what dates the theme saturated, and it has now held through a full week of post-print trade rather than a single reaction session.

Bullish and bearish views on Applied Materials, Inc.

The model's bull view on Applied Materials, Inc. (AMAT), in brief: Q3 FY2026 revenue $9.12B, +25% YoY, described by the company as the highest sequential growth in its history; non-GAAP EPS $3.50, +41% YoY; GAAP EPS $3.17, +43% (press release, 2026-08-13). The bear view: The structural line already broke. The $620 shelf that carried the June breakout is gone; price is 31.8% below the $722.23 high, and the June advance has resolved as a failed breakout rather than a base. A beat-and-raise that cannot hold a bounce for five sessions is a demand… Both cases follow in full.

Bull Case

• Q3 FY2026 revenue $9.12B, +25% YoY, described by the company as the highest sequential growth in its history; non-GAAP EPS $3.50, +41% YoY; GAAP EPS $3.17, +43% (press release, 2026-08-13).
• Q4 FY2026 guide of $10.25B ±$0.50B revenue and non-GAAP EPS $4.02 ±$0.20 steps up sequentially off that record (press release, 2026-08-13); third-party coverage frames the EPS guide as ahead of a ~$3.68 consensus (stockanalysis.com wire item, 2026-08-15).
• Semiconductor Systems revenue $7.04B at a 37.7% GAAP operating margin; Applied Global Services $1.78B at 30.1% (press release, 2026-08-13). Management framed leading-edge foundry-logic, DRAM and advanced packaging as roughly 80% of 2026–2027 WFE growth on the 2026-08-13 call — the deposition, epi and CMP steps where the company's share is heaviest.
• 2026 semiconductor-equipment growth outlook raised to more than 30% on the 2026-08-13 call, with management flagging added manufacturing capacity to serve demand through the end of the decade.
• China revenue held at $2,506M in the July quarter against $2,548M a year earlier while total revenue grew 25% (Q3 FY2026 10-Q, period ended 2026-07-26) — the export-control-exposed slice shrank as a share because the rest grew, not because China collapsed.
• Record Q3 operating cash flow $3.04B and non-GAAP free cash flow $2.33B (+14% YoY); $860M returned in the quarter, split $440M repurchases and $420M dividends (press release, 2026-08-13).
• The entire published post-print target cluster sits above the market: Craig-Hallum $585, RBC $600, Argus $600 (raised from $500), Deutsche Bank $605 (cut from $680), Morgan Stanley $642, BofA $650, Lynx $650, JPMorgan $660, UBS $675, Bernstein / TD Cowen / B. Riley $700, Needham $740 — against a 2026-08-21 close of $492.32.

Bear Case

• The structural line already broke. The $620 shelf that carried the June breakout is gone; price is 31.8% below the $722.23 high, and the June advance has resolved as a failed breakout rather than a base.
• A beat-and-raise that cannot hold a bounce for five sessions is a demand read, and it points one way.
• The sell-side ceiling compressed. June's cluster reached Susquehanna $900 and Cantor $850 (both 2026-06-30); the highest post-print number is Needham $740, and cuts came in alongside the raises — UBS $705→$675, BofA $720→$650, B. Riley $790→$700, Deutsche Bank $680→$605, Morgan Stanley to $642.
• Valuation is the explicit objection in coverage: roughly 33x forward earnings after a 108% year-to-date move, with the question posed as whether the AI-equipment story is already priced (24/7 Wall St, 2026-08-14).
• Positioning is leaking at the margin. Tiger Global cut its stake 1% to 1,632,210 shares (13F coverage, 2026-08-14), and 24/7 Wall St characterised insider activity across 39 recent transactions as net selling.
• the disclosure is a datapoint on who else is positioned, not a forecast.
• Nothing company-reported resolves the argument before mid-November. The gap between a raised equipment outlook and a falling price has to be carried on third-party data alone for roughly three months.

Setup & Price Structure

• Reference close 2026-08-21: $492.32. Distance from the $722.23 52-week high: -31.8%. RSI(14) at 44.1 — neither washed out nor turning up.
• A three-month price change of +14.0% means the drawdown has not yet erased the summer advance on a quarterly lookback. Momentum screens measured over that window have not been forced to exit, which leaves supply available that has not yet come.
• The $620 shelf is broken and untested from below. The operative reference now sits at $480: a weekly close under it would mark the post-print decline as distribution of the whole June advance rather than a give-back of the reaction pop. Price closed roughly 2.5% above that line on 2026-08-21, so the level is live, not theoretical.
• Crowding observables, stated as observables: thirteen sell-side target actions clustered in one post-print window; retrospective retail-attention coverage ("Here's How Much $1000 Invested 10 Years Ago…", 2026-07-03; the 15-year version, 2026-08-19); options-flow coverage naming AMAT among "Whale Alerts" on 2026-08-17; and net insider selling across 39 transactions per 2026-08-14 coverage. No earnings date sits inside 30 days, so the near-term crowding risk is flow, not a print.
• What has not happened: no higher low, no volume reclaim of the pre-print range, no analyst upgrade of rating (as distinct from target). A base is a thing that can be pointed at, and there is not one yet.

Catalyst Calendar (next 30 days)

• 2026-09-08 — CFO Brice Hill at the Citi 2026 Global TMT Conference, 8:50 a.m. ET, webcast on the company IR site (press release, 2026-08-14). First live CFO commentary since the print; the place any Q4 tone change or 2027 framing would first surface.
• 2026-09-09 — CEO Gary Dickerson at the Goldman Sachs Communacopia + Technology Conference, 5:25 p.m. ET, webcast (press release, 2026-08-14). Same function at the CEO level, and the only scheduled forum for the >30% 2026 equipment growth outlook before November.
• 2026-09-10 — Quarterly dividend $0.53/share payable (record date 2026-08-20). Mechanical; resolves nothing about demand.
• ~2026-09-10 (est.) — TSMC August monthly revenue release. Highest-frequency public read on leading-edge capex and the cleanest third-party check on whether the equipment-growth outlook is tracking.
• Outside the window, for calibration: ~2026-11-12 (est.) Q4 FY2026 results, which test the $10.25B ±$0.50B and $4.02 ±$0.20 guides and carry the first FY2027 framing.

What Would Change Our Mind

The June breakout shelf at $620 is already lost, so the question is no longer whether the advance failed but whether $480 holds as the floor of a range. A weekly close below $480 would settle that against the bulls and put the whole June leg in play as distribution — the same level flagged before the print, still untested. Secondary confirmation would come from the 2026-09-08 and 2026-09-09 conference appearances passing without a reclaim of the $520–$540 area, which would mean two direct management forums produced no bid.

The read flips the other way on evidence, not on price alone: a weekly close back above $560 with the target cluster holding, or any house moving rating (not target) higher, would argue the de-rating has run its course and that the saturated label needs replacing. A TSMC monthly revenue series that keeps accelerating through September and October while AMAT drifts would widen the same divergence — informative, but it has not stopped the decline so far, and repetition of a signal that has not worked is weak evidence.

The fundamental break to watch is narrower than the tape: the >30% 2026 equipment growth outlook being trimmed, or Q4 revenue landing at or below the low end of $9.75B. Either would convert a valuation argument into an earnings argument.

Correlation Notes

• Trades as a tight correlated duplicate of the wafer-fab-equipment complex — LRCX, and to a lesser degree ASML and TEL. The three US names de-rate and re-rate together on the same capex headlines; exposure to one is largely exposure to all.
• Second-order to hyperscaler and foundry capex. TSMC, Samsung and Micron capex guidance moves the group directly; management's own framing puts leading-edge foundry-logic, DRAM and advanced packaging at roughly 80% of 2026–2027 WFE growth (2026-08-13 call).
• Sits inside the AI-capex basket that Burry named on 2026-07-01 (NVDA, TSLA, CAT, AMAT). Sessions where those four fall together with no company news are the signature of a basket unwind rather than an AMAT-specific event.
• China export-control policy is a shared shock across the complex, not an idiosyncratic AMAT risk; the 28% revenue share disclosed in the Q3 FY2026 10-Q is the number to re-measure at each filing. »