Applied Materials
Sells capital equipment (Semiconductor Systems ~75% of revenue) plus a high-margin recurring installed-base services/parts/subscription business (Applied Global Services ~21%); revenue tracks customer wafer-fab capex cycles, with AGS providing a recurring annuity that smooths the cycle. ~85% of system sales are leading-edge foundry/logic + DRAM/HBM + advanced packaging.
The thesis on this name
State of AI Compute
Cheapest credible WFE/packaging leader (~18-22x fwd) against a raised >30% CY26 equipment-growth guide and record 50% GM — the diversified incumbent that wins on glass-core process + backside power + GAA + hybrid bonding (incl. the 9% BESI call option, which is also the disintermediation it owns); broad-line diversification cushions single-segment air pocke…
State of AI Compute
Long AMAT over 12-18mo as the cheapest-of-the-three WFE/process leader (~32x fwd vs LRCX 38x / KLA 34x), where the analytical edge is that the just-raised >30% CY26 equipment guide plus AI-logic/GAA/backside-power/advanced-packaging mix durably holds non-GAAP gross margin near a structurally higher ~48% through-cycle…
State of AI Compute
Cheapest-of-three WFE leader but at peak-cycle margins/post-rally multiple the prob-weighted payoff is flat-to-left-skewed. Express as the AMAT-long / LRCX-short pair (the thesis's optional hedge promoted to primary) so…
State of the Memory Supercycle
The broadest WFE toll-booth — deposition + etch + epitaxy + ion implant — with combined AMAT/LRCX etch/dep share guided to expand to ~42% of WFE (up 3pts since 2023, BofA/Lapedus). Less memory-concentrated than Lam, which is the feature not the bug for a cycle-aware book: AMAT rides both the DRAM/HBM bit-growth capex and the logic side, so a memory-specific glut hits it less violently. Materials-engineering moat sits at the heart of the 1c-node DRAM transition and 3D-NAND stacking, both of which are gating the supply this cycle. A through-cycle compounder that monetizes the entire memory-maker capex wave (~$135B WFE cal-2026) without picking an HBM winner.
State of the Memory Supercycle
The broadest WFE toll-booth on the memory capex super-cycle — deposition/CMP/implant/packaging spanning DRAM, HBM and NAND; management names DRAM+HBM as the fastest-growing 2026 sub-segments.
State of the Memory Supercycle
Broadest memory-WFE exposure (dep/CMP/implant/packaging); management names DRAM+HBM as fastest-growing 2026 segments. Rent, not own.
Earnings, margins, COGS & capex
AMAT prints record revenue ($7.91B, +11% YoY) and a 25-year-high non-GAAP gross margin of 50.0% in Q2 FY26 (fact), driven by leading-edge foundry/logic and AI-memory (HBM/DRAM) demand plus a richer ASP/mix. The model is two-part: cyclical Semiconductor Systems (~75% of revenue) and a recurring, higher-margin Applied Global Services annuity (~21%, growing on a ~$200B+ installed base of >49,000 tools). Management just raised the CY2026 semi-equipment growth outlook from '>20%' to '>30%' (fact, estimate-driven guide). Cash generation is strong over the cycle (FY25 FCF $5.70B) though a single quarter can swing on working capital.
Income statement — where each revenue dollar goes
% of revenueOf every $1 of revenue, ~50¢ is cost of goods and ~18¢ operating expense, leaving ~32¢ of operating profit (~35¢ net).
Revenue trend
Margins
up — 25-yr high, +~110bps YoY from 48.9%
up — from 30.2% FY25
up — GAAP net income $2.81B > non-GAAP $2.29B this quarter
lumpy — strong full-year, weak single quarter on inventory/AR build
COGS structure
COGS is dominated by high-value bought-in components and subsystems — precision optics, RF power, vacuum/gas-delivery, robotics, and electronic subsystems sourced from MKS Instruments, Entegris, Advanced Energy, Ultra Clean Holdings, Ichor, and others — plus skilled assembly/test labor and a large field-service/installation organization (the AGS cost base). Gross margin expands when mix shifts to leading-edge/HBM tools, ASPs rise, and the high-margin services/subscription revenue grows; it is pressured by China-mix, lower-margin DRAM/ICAPS, and component cost inflation.
Capex
~$2.26B FY25 (~8% of revenue, fact); $635M in Q2 FY26. Funds R&D cleanrooms and pilot lines — notably the EPIC Center in Sunnyville (collaborative R&D platform with customers/suppliers), capacity for advanced-packaging and ICAPS, plus IT/infrastructure. Relatively capital-light vs the fabs it sells into.
Latest earnings
Beat — non-GAAP EPS $2.86 vs ~$2.66 consensus (+~7.5%); revenue $7.91B vs ~$7.68B consensus (fact). 4th straight beat. Stock dipped post-print despite the beat + raise (high expectations / China overhang).
Q3 FY26: revenue $8.95B ±$0.5B, non-GAAP EPS $3.36 ±$0.20 (fact). Raised CY2026 semiconductor-equipment market growth outlook to '>30%' from '>20%' (fact).
- Semiconductor Systems revenue
- $5.97B (Q2 FY26), foundry/logic ~67%, DRAM ~29% of segment
- Applied Global Services revenue
- $1.67B record (Q2 FY26)
- Non-GAAP gross margin
- 50.0% — highest in ~25 years
- Capital returned
- $400M buybacks + $365M dividends in Q2 FY26
Growth drivers
- Leading-edge logic node transitions — gate-all-around (GAA/RibbonFET) transistors and backside power delivery add process steps where AMAT is over-indexed — the 'complexity tax' raises tool intensity per wafer
- AI memory: HBM and advanced DRAM drove DRAM to ~29% of Systems revenue in Q2 FY26 (fact); HBM is materials-engineering-intensive (TSV, hybrid bonding, stacking)
- Advanced packaging (heterogeneous integration, hybrid bonding, CoWoS-type flows) — AMAT serves >15 of the ~20 packaging process steps, a faster-growing SAM
- Applied Global Services recurring annuity — record $1.67B in Q2 FY26 (fact), growing on a >49,000-tool installed base with high subscription attach
- CY2026 WFE up-cycle — management's '>30%' equipment-growth guide (fact, raised from >20%) as AI capex compounds both wafer complexity and wafer volume
- ICAPS (IoT, comms, auto, power, sensors) trailing-edge demand and re-shoring (CHIPS-funded US/EU/Japan fabs)
Reported financials — SEC EDGAR
Audited GAAP figures pulled from SEC filings · latest filing 2025-12-12. The audited primary-source spine — not financial advice.
Revenue — annual (GAAP)
Margins & balance sheet — FY’25
Bull & bear
AMAT is the cheapest-of-the-three WFE/process leaders (~32-36x fwd vs LRCX ~38x / KLAC ~36x) levered to a multi-year AI-driven equipment up-cycle, with the just-raised '>30%' CY26 guide and a GAA/backside-power/HBM/packaging mix shift that durably holds non-GAAP gross margin near a 25-year-high 50%.
- The '>30%' CY26 equipment-growth guide (raised from >20%, fact) is backed by leading-edge logic + DRAM + packaging being >80% of WFE growth in 2026-27 — exactly AMAT's strongest pools
- Mix-driven margin durability: GAA, backside power, and HBM raise process steps per wafer (the complexity tax), pushing gross margin to a 25-year-high 50% (Q2 FY26) that the bull case sees holding, not reverting
- Broadest portfolio + EPIC co-development = structural share gains at the node-definition stage where switching costs are highest; AMAT sells the connected toolset peers can't
- AGS recurring annuity (~$6.4B, record quarterly $1.67B) is a higher-margin, cycle-smoothing growth engine that the market under-credits vs the cyclical systems business
- Cheapest of the trio on forward earnings with the strongest balance sheet (~$5.7B net cash) and ~$5B+/yr capital returns — a re-rate toward LRCX/KLAC multiples is the upside option
AMAT is a richly-valued (~36x fwd), deeply cyclical capex play with the heaviest China exposure of the three (~27% of revenue) into tightening export controls and a domestic-China substitution threat — the '>30%' guide is an estimate at peak optimism that de-rates hard on any AI-capex digestion or China shock.
- ~27% China revenue (Q2 FY26) is the highest of the trio and concentrated in trailing-edge ICAPS — squeezed from both sides by US export rules and Chinese localization (Naura/AMEC/ACM climbing)
- The '>30%' CY26 number is a guide/estimate at maximum AI optimism; WFE is famously cyclical and a memory/AI-capex over-build reverses the narrative fast (Q4 FY25 already printed -3%)
- No EUV — AMAT is structurally excluded from the single highest-value, widest-moat leading-edge step (ASML's monopoly), capping its share of the most defensible margin pool
- Valuation leaves no margin for error: stock re-rated to ~$668 / ~36x fwd and dipped on a clear beat-and-raise — expectations are priced for perfection
- Customer concentration (TSMC/Samsung/Intel/Micron/SK hynix) means one capex pause or a leading-edge node slip hits revenue immediately; single-quarter FCF already showed fragility ($0.21B in Q2 FY26)
What it is worth
Forward P/E relative-value vs the WFE trio (AMAT ~32-36x vs LRCX ~38x / KLAC ~36x) cross-checked with a reverse-DCF read on the implied growth/margin.
~$430-520 (de-rate to ~24-28x on a China-control shock or AI-capex digestion that breaks the '>30%' narrative and pressures margins toward ~47-48%)
~$650-690 (current ~36x fwd holds
guide delivered, margins steady, China flat — roughly the spot price)
~$760-820 (re-rate toward LRCX's ~38x on a sustained '>30%' CY26 print, 50% margin holding, and China stabilizing — ~38x on ~$20-21 forward EPS)
At ~$668 / ~36x fwd, the price implies CY26 equipment growth lands near the '>30%' guide AND ~50% gross margin holds — the bull edge is that this mix durably sustains, leaving room to re-rate toward LRCX; the bear is that ~27% China + cyclicality make those embedded assumptions fragile.
SWOT
Strengths
- Broadest product portfolio in WFE — deposition (CVD/PVD/ECD), epitaxy, ion implant, CMP, RTP, e-beam metrology, and packaging — lets it sell a connected materials-engineering 'system' no single competitor matches
- Scale: ~$28-29B revenue, the largest pure WFE vendor; #1 or #2 share in most of its served markets and over-indexed to GAA + backside power
- Recurring AGS annuity (~$6.4B FY25, +growing) on a >49,000-tool installed base — smooths the equipment cycle and carries higher, stickier margins
- Record 50% gross / 32% operating margins (Q2 FY26) and ~$5.7B FY25 FCF fund heavy R&D ($3.2B+/yr) and large capital returns
- Deep co-development partnerships (TSMC, Samsung, Intel, Micron, SK hynix) and the EPIC Center R&D platform create switching costs at the node-definition stage
Weaknesses
- High China exposure (~27% of Q2 FY26 revenue) concentrated in trailing-edge ICAPS — exposed to both US export rules and a China-overcapacity/digestion air-pocket
- Cyclical, capex-driven demand — revenue can swing double digits quarter-to-quarter (Q4 FY25 -3%) on customer capex timing
- No EUV lithography — the highest-value, highest-moat leading-edge step belongs entirely to ASML; AMAT is excluded from that pool
- Single-quarter FCF can collapse on working-capital builds (Q2 FY26 FCF only $0.21B) even in a strong demand environment
- Lower-margin DRAM/ICAPS mix can dilute the leading-edge margin tailwind when memory/China leads spend
Opportunities
- GAA + backside power delivery + HBM/advanced packaging materially raise tool intensity per wafer — the structural 'complexity tax' AMAT is most levered to
- AI-driven WFE up-cycle — '>30%' CY26 equipment growth guide; leading-edge logic + DRAM + packaging = >80% of WFE growth in 2026-27
- Advanced-packaging SAM expansion (hybrid bonding, heterogeneous integration) where AMAT covers most process steps
- AGS subscription/recurring revenue growth and digital/AI-fab software upsell on the installed base
- Re-shoring tailwind — CHIPS Act, EU/Japan fab build-outs diversify demand away from China
Threats
- Tighter US/allied export controls (or Chinese retaliation/localization) further restricting the China market — already cut materially in FY25
- China domestic-equipment champions (Naura, AMEC, ACM Research) climbing the trailing-edge/ICAPS ladder, eroding AMAT's lower-end SAM
- Customer concentration — a TSMC/Samsung/Intel/memory capex pause hits revenue immediately
- A memory or AI-capex digestion/over-build cycle reversing the '>30%' growth narrative
- Valuation/expectations risk — stock re-rated to ~36x fwd P/E; any guide cut or China shock de-rates fast (post-Q2 dip despite beat+raise)
Moats, dependencies & bottlenecks
Moats
etch, implant, CMP, epi, metrology, packaging) No competitor sells the connected leading-edge toolset; integrated process control is a deep, multi-decade R&D moat.
>49,000 tools, ~$6.4B FY25 AGS; high subscription attach makes the base sticky and margin-accretive across the cycle.
Tools are qualified into customer nodes years ahead (EPIC Center, TSMC/Samsung/Micron/SK hynix partnerships); re-qualifying a competitor is costly and slow.
Largest WFE budget funds the next node, but peers (LRCX in etch/dep, ASML in litho, TEL) out-moat AMAT within specific steps.
IP / process know-how in advanced packaging & GAA/backside-power steps Over-indexed to the fastest-growing leading-edge steps, but trailing-edge IP is increasingly replicable by Chinese entrants.
Dependencies
Samsung, Intel, Micron, SK hynix) Top customers concentrate revenue; a capex pause or node slip hits the order book immediately. Foundry/logic ~67% of Systems.
Subject to US/allied export controls; already cut materially in FY25 and exposed to further tightening or Chinese retaliation/substitution.
Demand is capex-timing-driven and swings double digits; the '>30%' CY26 guide is an up-cycle estimate that can reverse.
Entegris, Advanced Energy, Ultra Clean, Ichor) RF power, optics, gas/vacuum delivery, robotics; supply tightness or cost inflation pressures lead times and gross margin.
Customer adoption of GAA / backside power / HBM / advanced packaging Margin/growth thesis depends on these transitions proceeding on schedule; AMAT is over-indexed but not in control of the roadmap.
Advantages
- Largest, broadest WFE portfolio — sells a connected leading-edge process toolset (deposition/etch/implant/CMP/epi/metrology/packaging) no rival fully matches
- Over-indexed to the fastest-growing leading-edge steps — GAA, backside power, HBM/DRAM, and advanced packaging — the structural complexity-tax winners
- Recurring AGS annuity (~$6.4B FY25) on a >49,000-tool base smooths the cycle and lifts blended margins
- Best-in-class profitability and cash generation — 50% gross / 32% operating margin (Q2 FY26), ~$5.7B FY25 FCF, ~$5B+/yr capital returns
- Strongest balance sheet of the trio (~$5.7B net cash) plus the largest R&D budget to fund the next node
- Cheapest forward multiple among AMAT/LRCX/KLAC despite comparable AI leverage — relative-value optionality
Weaknesses
- Highest China exposure of the three (~27% of Q2 FY26 revenue), concentrated in trailing-edge ICAPS most exposed to export rules and Chinese substitution
- No EUV — structurally absent from the single highest-moat leading-edge step
- Cyclical revenue with double-digit quarterly swings (Q4 FY25 -3%) and concentrated customer base
- Lumpy near-term FCF (Q2 FY26 only $0.21B) on working-capital builds despite record revenue
- Lower-margin DRAM/ICAPS/China mix can dilute the leading-edge margin tailwind
- Valuation re-rated to ~36x fwd with little error margin — stock dipped on a clear beat-and-raise
Bottlenecks
- China revenue ceiling — export controls cap the addressable China ICAPS market regardless of demand, and rules can tighten further
- Excluded from EUV lithography — the highest-value leading-edge step (and its margin pool) belongs to ASML, capping AMAT's leading-edge share
- Customer capex timing — revenue cannot grow faster than TSMC/Samsung/Intel/memory makers choose to spend; concentration amplifies any pause
- Working-capital intensity in an up-cycle — inventory/AR build can throttle near-term FCF even when demand is strong (Q2 FY26 FCF $0.21B)
- Component/subsystem supply (RF power, precision optics, robotics) — single-source subsystems can gate tool build rates and lead times
Top signals & trends
Top signals
The core thesis variable; any reaffirm/raise validates the up-cycle, any walk-back de-rates the stock fast.
Margin durability is the analytical edge; holding ~50% confirms the GAA/HBM/packaging mix thesis, slippage breaks it.
Largest swing factor on the downside; ~27% mix means a rule change moves the model materially.
Rising memory mix signals AI-capex breadth; a memory digestion would show up here first.
The recurring annuity; sustained growth de-risks the cyclical systems business and supports the multiple.
Watch for a re-rate toward peers (upside) or a sector-wide de-rate on a cycle scare (downside).
Trends
Directly lifts AMAT's strongest served markets; underpins the '>30%' CY26 guide.
Adds process steps per wafer where AMAT is over-indexed — structural tool-intensity tailwind.
Materials-engineering-intensive; AMAT serves most packaging process steps, a faster-growing SAM.
Caps and erodes ~27%-of-revenue China business; Naura/AMEC/ACM climbing the trailing-edge ladder.
Diversifies demand geography and adds a multi-year capex base outside China.
Up-cycle today, but the same cyclicality that drives the beat-and-raise can reverse into an over-build correction.
Ecosystem & competitor graph
Suppliers feed the company; customers pull from it. Line thickness shows the strength of each tie (supply-chain dependency, customer earnings contribution). Hover to isolate a tie.
RF/microwave power, vacuum, gas/flow and motion subsystems — critical process-tool components.
Advanced materials, filtration, gas/liquid purity and specialty chemicals for the tools and the fabs they feed.
Precision power conversion / plasma power supplies for deposition and etch chambers.
Outsourced critical subsystems and gas-delivery modules assembled into AMAT tools.
Fluid-delivery subsystems and gas-panel assemblies for semiconductor capital equipment.
Lasers, optics and photonics components used in process and metrology systems.
Largest leading-edge foundry customer; GAA/N2/backside-power and CoWoS packaging buildout drives core demand.
Foundry + DRAM/HBM/NAND — buys across logic and memory; co-development partner.
18A/14A leading-edge logic + foundry (IFS); US re-shoring capex customer.
DRAM/HBM/NAND; signed long-term AI-memory R&D collaboration with AMAT (HBM/DRAM).
HBM leader; co-development partner with AMAT on next-gen DRAM/HBM/NAND (US$5B AI-memory R&D hub).
Trailing-edge/ICAPS foundry customer (specialty/auto/IoT nodes).
Etch + deposition leader, over-indexed to memory/NAND 3D scaling and HBM; trades ~38x fwd, the most direct deposition/etch overlap. Stock +~114% YTD 2026.
Process-control / metrology-inspection near-monopoly (~50%+ share); highest-margin of the trio, ~36x fwd. Adjacent rather than head-to-head but takes share of the same WFE wallet.
Lithography monopoly (sole EUV supplier) — the leading-edge step AMAT cannot serve. Complementary more than competitive, but owns the highest-value WFE pool.
Japanese broad-line WFE peer overlapping in deposition, etch, coat/develop, and cleans; AMAT's closest global breadth competitor.
Largest China domestic WFE champion; climbing the trailing-edge/ICAPS ladder where AMAT has high China exposure. Named as analysis only, not a recommendation.
China-focused cleaning/ECP/advanced-packaging equipment maker gaining share in trailing-edge and packaging steps.