Applied Materials

0 · What landed since the last dossier (provenance base)

Net: no new hard financials since 2026-06-18 (the next print is Aug 13). The refresh value is (a) upgrading Q2/guidance to primary via the transcript, (b) the memory/HBM capex read-through now grounded in management's own words, and (c) a fresh valuation + consensus snapshot (estimates have moved).

Moving Lens A · Performance / financials delta (FY2026 Q2, ended 2026-04-26) — now PRIMARY

The Q2 print is the same quarter the prior dossier covered, but figures are now confirmed from the on-disk transcript rather than web. All [primary: transcript 2026-Q2, 2026-05-14] unless noted.

• Revenue $7.91B — +13% QoQ, +11% YoY (record). Matches prior dossier.
• Non-GAAP gross margin 50.0% (+80 bps YoY) — "highest gross margin in more than 25 years." (Prior dossier's "49.9%" was the GAAP GM $3,947M/$7,910M [primary: 10-Q 2026-Q2]; the 50.0% is non-GAAP. Both true.)
• Non-GAAP operating margin 32.1% (+140 bps YoY).
• Non-GAAP EPS $2.86 (+20% YoY) — resolves a prior-dossier provenance gap. The 2026-06-18 dossier flagged FY2025/FY2026 non-GAAP EPS as "not cleanly sourced"; the Q2 non-GAAP print is now confirmed primary at $2.86 (beat consensus by ~6.5% ). GAAP diluted EPS was $3.51, flattered by $771M interest/other income [primary: 10-Q 2026-Q2] — the GAAP-vs-non-GAAP wedge the prior dossier called out.
• Segment detail — now with segment gross margins (AMAT began reporting per-segment GM this year):
• Semiconductor Systems $5.97B (+16% QoQ, +10% YoY) — record; segment GM 54.8%, "approaching 55%." Record foundry revenue on GAA + leading-edge FinFET; record ALD, epitaxy, materials-treatment revenue.
• Applied Global Services $1.67B (+17% YoY) — record; segment operating margin ~29% (highest in 2–3 yrs), on higher fab utilization. Note: the 200mm equipment business was recast OUT of AGS into Systems this year — AGS is now a clean services-only line.
• Other $280M (in line).
• DRAM revenue $1.7B, +18% YoY — the key memory line (see Lens C).
• Cash conversion: CFO $845M, capex $635M → FCF $210M for the quarter (capex elevated on the manufacturing-capacity doubling — see below). Shareholder returns $765M ($365M dividend + $400M buyback).
• China = 24% of SSG+AGS revenue (management's preferred basis) [primary: transcript]. The prior dossier cited "27% of Q2 revenue ($2,087M)" on a total-revenue basis [primary: 10-Q] — different denominator (total net revenue incl. Other), not a contradiction.

Delta vs prior dossier: none of the Q2 numbers changed; they are now primary-grounded, the non-GAAP EPS gap is closed, and we now have per-segment gross margins (SSG 54.8%, the operating engine) that weren't in the prior pass.

Moving Lens B · Guidance + catalysts (the forward story)

All guidance [primary: transcript 2026-Q2, 2026-05-14].

• Q3 FY2026 guide (reports Aug 13): revenue $8.95B ±$500M (+~23% YoY), non-GAAP EPS $3.36 ±$0.20 (+~36% YoY). Segment guide: SSG ~$6.9B, AGS ~$1.75B, Other ~$300M; non-GAAP GM ~50.1%, opex ~$1.485B, non-GAAP tax rate ~11% (low — a tailwind to watch).
• Consensus for Q3: ~$3.39 non-GAAP EPS, +35.5% YoY — i.e., the Street is ~$0.03 above the guide midpoint. Low-bar setup relative to the run of beats.
• Calendar-2026 framing (the core forward calls), now management-primary:
• Semiconductor equipment business to grow >30% in CY2026 (Dickerson: "we'll grow 30% or more in the year"). Analysts (Arcuri/UBS, Rasgon/Bernstein) pushed that this implies only linear H2 growth and could prove conservative vs a possible >40%; Hill declined to raise the number but said to model Systems growing linearly from the Q3 guide through fiscal Q1 — i.e., roughly equal dollar step-ups each quarter.
• Advanced packaging to grow >50% in CY2026, strong growth again in 2027. NEXX acquisition (panel-level electroplating) + Besi hybrid-bonding JV underpin it.
• AGS long-term growth raised to mid-teens (from prior "low-double-digits"), "and higher this year" — a durable-annuity upgrade, driven by utilization + 35,000+ AIx-connected chambers.
• Mix concentration: leading-edge foundry logic + DRAM + advanced packaging = >80% of the YoY growth in total WFE spending in 2026, "similar profile in 2027" — the three areas where AMAT holds #1 process-equipment positions.
• Multi-year visibility (new texture): largest customers now provide rolling 8-quarter forecasts; management is tracking >100 factory projects globally (+10 added in the quarter); calls 2027 "another strong record year" and would not rule out 2028 growth ("secular growth… AI is the headline workload"). Clean-room/floor space (last quarter's pacing constraint) is easing as customers reallocate space; the binding constraint is now the supply chain (~2,000 direct suppliers), not AMAT's own capacity (which it has "nearly doubled," US/Europe/Singapore).

Catalyst calendar delta:

• EARNINGS — Aug 13, 2026: Q3 print + first hard FY2027 framing. This is the near-term binary the prior dossier flagged as "~Aug 2026," now date-confirmed.
• EPIC Center opens fall 2026 — 8 partners announced (TSMC founding, + Micron, Samsung, SK Hynix, Advantest, ASU, RPI, Stanford); "more soon."
• DRAM + advanced-packaging master class was held June 2026 (roadmap detail) [primary: transcript reference].
• REGULATORY (unchanged, still live): China export-control tightening remains the standing down-catalyst (see Lens D).

Moving Lens C · Memory / HBM capex read-through (the desk's core ask) — PRIMARY

This is the lens the refresh was commissioned for. Management commentary is now primary via the transcript, and it is decisively constructive on memory equipment.

AMAT's own memory print & positioning [primary: transcript 2026-Q2]:

• DRAM revenue $1.7B, +18% YoY. Customers are "aggressively adding capacity at 6F² nodes" while accelerating development of next-gen (4F², then 3D DRAM).
• AMAT is #1 process-equipment provider in memory, with cited strength in DRAM wiring, patterning, and peripheral logic; it has gained ~10 points of WFE share in DRAM over the past ~10 years and expects further DRAM share gains at upcoming transistor/architecture inflections.
• Node-transition mechanics (important for the capex-intensity read): the 1-delta DRAM node is more litho-intensive than 1-gamma, but management insists materials intensity rises going forward, especially at 3D DRAM — AMAT's served-market-per-wafer expands as DRAM goes 3D. Near-term DRAM drivers: periphery CMOS-logic upgrade (higher-perf/lower-power transistors) → "very strong growth in epi"; plus conductor etch + e-beam leadership.
• HBM / advanced packaging: packaging revenue guided >50% growth CY2026; HBM "a huge focus for everyone… stacking more chips vertically"; advanced-packaging investment is accelerating within both foundry logic and DRAM, shifting toward AMAT's 3D-stacking positions.
• NAND: management raised its NAND bit-growth forecast (a few points) but expects it satisfied by upgrades, not new wafer starts — and flags AMAT participates less in NAND upgrades than greenfield. So the NAND uptick is a weaker read-through for AMAT specifically than the DRAM/HBM strength.
• Agentic-AI kicker: management explicitly models agentic AI as incremental CPU + DRAM + NAND demand "on top of what we had been forecasting previously" — i.e., a raised memory-demand baseline vs the prior view.

External corroboration (directional; mixed vintage — flagged):

• DRAM sector capex ~$53.7B (2025) → $61.3B (2026), +14% YoY**; **Micron most aggressive ($13.5B, +23%), Samsung ~$20B (+11%), SK hynix strong growth.
• 2026 HBM market ~$54.6B, +58% YoY; emphasis shifting from raw capacity to process upgrades, higher-layer stacking, hybrid bonding.
• Samsung / SK hynix warn AI-driven memory shortage could persist into 2027+, customers reserving supply years ahead.
• Vintage caveat: an earlier TrendForce (Nov 2025) "memory industry maintains cautious capex in 2026" call has been overtaken by the mid-2026 supercycle framing and by AMAT's own "aggressively adding capacity" language — treat the Nov-2025 caution as stale.

Read-through summary (mechanics, no view): the memory leg of AMAT's WFE exposure is inflecting up, not digesting — DRAM +18% with customers adding 6F² capacity, HBM/packaging +50%, AGS raised on utilization, and a raised memory baseline from agentic AI. The structural bull point is rising materials-intensity at 3D DRAM (expands AMAT's per-wafer TAM). The honest counter-weight is NAND still upgrade-satisfied (muted AMAT pull) and the concentration of the memory buy in 3 customers (Micron/Samsung/SK hynix) whose capex is itself cyclical. Flag for desk: whether the DRAM buy is durable multi-year vs a 2026–27 pull-forward is the same binary as the WFE-cycle question — a deeper opus-level synthesis tying AMAT DRAM $ to the Micron/SK hynix/Samsung capex guides (and to HBM4 qualification timing) would sharpen this; it is beyond what this delta pass verified.

Moving Lens D · Red flags / earnings quality delta

Carrying the prior dossier's forensic verdict (clean accounting, exemplary cash conversion, no SEC findings; GAAP flattered by investment gains; BIS settlement + suspended denial order as the live overhang). Deltas this pass:

• No new SEC filings, no new litigation releases, no new 8-K disclosures since 2026-06-18. The BIS export-controls settlement ($253M, 56 alleged violations, suspended denial order waived after 3 yrs of clean audits) remains the standing regulatory flag — no change, still live [primary: 10-Q 2026-Q2, carried].
• Earnings-quality wedge confirmed primary: Q2 GAAP EPS $3.51 vs non-GAAP $2.86 — the gap is the $771M "interest and other income" (largely investment gains) [primary: 10-Q 2026-Q2]. Use non-GAAP/operating income to judge the core, as before.
• Low non-GAAP tax rate (~11%) guided for Q3 [primary: transcript] — a real EPS tailwind; worth watching for normalization (the FY2025 GAAP effective rate was ~24.5%). Not a red flag, but it flatters the +36% YoY EPS optics.
• Working-capital / inventory watch (unchanged direction): capex stepped up to $635M in Q2 on the manufacturing-capacity doubling, and management is "increasing build plan, inventory positions and logistics capacity" ahead of the H2 ramp [primary: transcript]. Consistent with the prior dossier's inventory-build flag — bullish if the >30% ramp lands, a working-capital drag if it slips. The Aug-13 balance sheet is the next read.
• China revenue de-risking modestly: China down to 24% of SSG+AGS (management expects China + worldwide ICAPS "flat to slightly higher" in CY2026) — continues the mix-shift-away-from-capped-China theme, quality-positive.

Net: no new forensic issues; the two standing flags (investment-gain flattering + BIS overhang) persist unchanged. The ~11% tax rate is a new, minor EPS-optics caveat.

Moving Lens E · Valuation snapshot (fresh) — with a correction to the prior dossier

Spot & multiples, mixed July-2026 vintage — directional, not precise.

• Price ~$564.55 as of 2026-07-21 (that day +7.4%). Still near the all-time high (~$586 on 2026-06-15). A Motley Fool page-widget on 2026-07-25 indicated ~$535, i.e., a possible few-percent pullback into the print; treat the ~$564 as the sourced anchor.
• Shares ~799M [primary: 10-Q 2026-Q2] → market cap ~$451B (vs ~$480B at the prior dossier's higher ~$580–604 spot).
• Forward P/E ~47–52x on FY2026 non-GAAP. ⚠ Correction: the 2026-06-18 dossier reported AMAT forward P/E at "~36x," which is inconsistent with the math (at ~$580 and ~$11 FY2026 non-GAAP EPS the multiple is ~50x, not 36x) and with every current web source (48–52x). Treat AMAT as trading ~48–52x forward earnings, not ~36x. TTM P/E ~53x; 5-yr median ~20.4x.
• Consensus non-GAAP EPS: FY2026 ~$11.0–12.1 (Simply Wall St $11.09, raised from $9.32; MarketBeat $11.05 for FYE 2026-10-26 and $12.13 "current year"). Bridge check: Q2 actual $2.86 + Q3 guide $3.36 + a linear Q4 supports the ~$12 upper end. This is an UPWARD revision vs the prior dossier's ~$10.4 base case — the key valuation delta. FY2026 revenue consensus raised to ~$31.3B (from $29.2B).
• FY2027 non-GAAP EPS: high-end single-analyst prints of $15.9–17.0 — outliers, treat with caution; consensus not cleanly captured this pass (flag as remaining gap). Prior dossier base case was ~$11.8.
• Analyst targets: consensus PTConsensus price targetThe average of what published analysts think the share price should be. An opinion poll, not a forecast. ~$554.52 (66 analysts) — now roughly AT/just below spot (~$564); top-end "current" target ~$593.84. Delta vs prior dossier: in mid-June the Street was hiking targets into new highs (Citi $710, etc.) with the stock below them; by late July, spot has caught up to the average target — consensus upside has compressed to ~flat. Morgan Stanley upgraded AMAT to Overweight (from Equal-Weight) (the target figure surfaced in search looked stale — omitted).
• GF Value model: fair value ~$222, "significantly overvalued" — a mechanical historical-multiple model that does not price the AI-WFE step-change; noted for completeness, not endorsed.

Snapshot read (mechanics only): AMAT sits ~48–52x forward non-GAAP at an all-time high, with estimates revised UP (FY2026 EPS $10.4→~$11–12, revenue +$2B) but the average price target now level with spot. Cheapest-of-the-four-on-EV/EBITDA framing from the prior dossier is not re-verified this pass (peer multiples are June-vintage — see gaps). The prior "~36x" figure is corrected to ~48–52x.

Carried forward by reference (no material change since 2026-06-18)

• Lens 1 Overview · Lens 2 Supply Chain · Lens 3 Moats — unchanged. AMAT = broadest-portfolio WFE integrator, two segments (SSG cyclical engine / AGS annuity), Display folded into Corporate.
• Lens 7 Comps — WFE oligopoly (ASML/LRCX/KLAC). Multiples are June-2026 vintage — NOT re-pulled this pass (gap below).
• Lens 8 Catalysts (history) — the two-narrative-tension read (China controls down / AI-capex up) stands; the Aug-13 print is the next resolution point.
• Lens 9 Management — Dickerson (CEO since 2013) / Hill (CFO since 2022); ~90% of FCF returned, $14B buyback authorization, dividend doubled (15% hike March 2026, now confirmed primary via transcript). Succession still unaddressed.
• Lens 10 Forensics — see Lens D delta; BIS overhang unchanged.
• Lens 11 Projection · Lens 12 Bull/Bear · Lens 13 Devil's Advocate · Lens 14 Mgmt Questions — structure stands. Note: the projection's base-case FY2026 non-GAAP EPS (~$10.4) is now below the refreshed consensus (~$11–12) — the Street has moved toward the prior bull path; re-run the projection at next full battery.