Entegris
Recurring consumables + materials sold per-wafer to fabs across logic and memory; ~75% of revenue is unit-driven (tied to wafers produced, not fab capex), the remainder capex-driven. Two reporting segments since the 2024 reorg: Advanced Purity Solutions (filtration, fluid handling, contamination control) and Materials Solutions (CMP slurries/pads, advanced deposition & etch chemistries, specialty gases).
The thesis on this name
State of the Memory Supercycle
The materials/consumables toll-booth at the bottom of the stack — filtration, advanced deposition materials, fluid handling, FOUP/contamination control — sold per-wafer-processed rather than per-tool, so revenue scales with running utilization, not just capex. As DRAM moves to 1c and EUV and HBM stacking adds process steps, contamination-control and specialty-materials intensity per wafer rises. More recurring and less capex-cyclical than the equipment names, which makes it a steadier through-cycle hold; the trade-off is lower torque to the up-leg. A quiet compounder on rising memory process complexity.
State of the Memory Supercycle
The consumables/materials toll-booth that earns recurring revenue every wafer regardless of who wins HBM — ~30% of revenue is memory, and DRAM materials-intensity rises with each node and HBM TSV step.
State of the Memory Supercycle
~30% memory revenue, scales with wafer starts not bookings; recurring revenue diversifier, less cyclical than WFE.
Earnings, margins, COGS & capex
Entegris returned to growth in Q1 FY26 (+5% YoY to $812M) after a soft 2025 ($3.20B, roughly flat) as DRAM/HBM and 2nm logic re-accelerated. Gross margin is recovering toward 47% as new plants (Kaohsiung, Colorado) ramp and the Chandler closure completes; the model is ~75% unit-driven recurring consumables, which dampens the cyclicality versus pure equipment names. FCF jumped to 18% of sales in Q1 FY26, and management is deleveraging from 3.6x toward ~3.0x by year-end 2026.
Income statement — where each revenue dollar goes
% of revenueOf every $1 of revenue, ~53¢ is cost of goods and ~33¢ operating expense, leaving ~14¢ of operating profit (~7¢ net).
Revenue trend
Margins
up (toward ~47% guide; ~50bps Q1 one-time)
up as new-plant drag fades
up (interest expense ~$190M/yr is the drag)
up off 27.3% Q3 FY25 trough
up (~300bps YoY in 2025)
COGS structure
COGS is dominated by raw materials and purchased chemicals/resins (fluoropolymers, specialty chemistries, abrasive particles for slurries), high-purity manufacturing in cleanrooms (energy, ultrapure water, scrap/yield), labor, and depreciation on a heavy specialty-plant footprint. Margin swings are driven by product mix (advanced materials richer than legacy handling), factory utilization/ramp drag at new plants (Kaohsiung, Colorado dilutive into 2027), and input-chemical/energy cost. Tariffs and freight are a watched cost line given the global fab footprint.
Capex
$299M in FY2025 (~9% of sales); guided down to ~$250M for FY2026 (~7.5%). Funds capacity for advanced CMP slurries, deposition/etch materials, advanced-packaging materials, and the Kaohsiung (Taiwan) and Colorado Springs (US) greenfield plants built to localize supply near leading-edge fabs; depreciation ~$140M/yr.
Latest earnings
Beat. Non-GAAP EPS $0.86 vs ~$0.75 consensus (~15% beat); revenue $812M vs ~$808.7M est. Gross margin and EBITDA both above guidance (fact).
Q2 FY26: revenue $815-845M, gross margin 46.25-47.25%, non-GAAP EPS $0.76-0.84, opex ~$194M. FY2026: MSI growth mid- to high-single-digit, capex ~$250M, net interest <$190M, non-GAAP tax ~15%, net leverage toward ~3.0x (guidance).
- Non-GAAP EPS (Q1 FY26)
- $0.86 (beat)
- Gross margin (Q1 FY26)
- 46.9%
- FCF / sales (Q1 FY26)
- 18% ($144M)
- Net leverage
- 3.6x (target ~3.0x YE26)
Growth drivers
- Materials intensity rises every node — sub-5nm/2nm and gate-all-around add process steps and consume more filtration, slurries, and specialty chemistries per wafer (the core secular driver).
- Memory/HBM up-cycle — ~30% of revenue is memory; HBM TSV stacking and DRAM scaling add deposition, CMP, and contamination-control steps — Entegris gets paid per incremental step regardless of which HBM maker wins.
- Advanced packaging — >$100M annual run-rate and growing (thermal materials, carriers, advanced-packaging chemistries) as chiplets/2.5D-3D proliferate.
- 2nm logic ramp in 2026 plus near-capacity utilization at advanced nodes lifting unit-driven consumables.
- AI capex super-cycle pulling DRAM and leading-edge logic wafer starts higher.
- Geographic localization (Taiwan +18% YoY, Asia +10%) as customers qualify on-region supply near new fabs.
Reported financials — SEC EDGAR
Audited GAAP figures pulled from SEC filings · latest filing 2026-02-11. The audited primary-source spine — not financial advice.
Revenue — annual (GAAP)
Margins & balance sheet — FY’25
Bull & bear
Entegris is the picks-and-shovels consumables compounder of the AI/leading-edge era: a recurring, per-wafer materials toll-booth whose content-per-wafer rises every node and every HBM/advanced-packaging step, now re-accelerating into the 2nm + HBM cycle with margins recovering and the CMC debt steadily coming down.
- Recurring, unit-driven model (~75% of revenue) plus rising materials intensity = structural growth that doesn't depend on picking the HBM or foundry winner.
- Re-acceleration is visible: +5% YoY in Q1 FY26 with a clean EPS beat, gross margin back toward 47%, and FCF at 18% of sales.
- Memory (~30% of revenue) is a free call option on the HBM/DRAM up-cycle — more TSV/deposition/CMP steps mean more Entegris content.
- Deleveraging (3.6x -> ~3.0x by YE26) plus falling interest expense converts operating strength into faster EPS growth.
- Designed-in, qualified positions at every leading-edge fab create high switching costs and durable share.
You're paying a rich ~34x forward multiple for a still-cyclical materials company that levered up ~$3.2B for CMC, faces low-cost Chinese slurry competition and an escalating export-control regime, and whose 2025 revenue was flat-to-down — leaving little room if the AI/memory cycle disappoints.
- Valuation: ~34x forward P/E (vs ~26x semi average) and ~75x trailing prices in a lot of the recovery already; some analyst targets sit below the current ~$178 price.
- Leverage: ~$3.2B net debt and 3.6x with ~$190M/yr interest is a real drag and limits flexibility if the cycle rolls over.
- Cyclicality is not eliminated — FY2025 revenue was flat-to-down and Q4'25 was -3% YoY; a wafer-starts correction still bites.
- China/export-control overhang: China revenue modestly declined and tightening U.S. controls in 2025-2026 threaten demand and force requalification.
- Competitive/pricing pressure from low-cost regional players (Anji) and large diversified rivals (DuPont, Merck/Versum, Resonac, Shin-Etsu) in overlapping materials.
- New-plant ramp dilution (Colorado into 2027) caps near-term margin upside.
What it is worth
Forward P/E and EV/EBITDA cross-check vs semi-materials peers, anchored to FY2026 EPS power. Q2 FY26 guide midpoint ~$0.80 non-GAAP EPS implies ~$3.2-3.4 annualized; ~34x forward P/E (vs ~26x semi average) on ~$178 price; ~75x trailing.
~$115-134 (low end / older targets
cycle stalls, China headwind, multiple compresses toward semi average)
~$160-183 (consensus cluster
mid-to-high-single-digit MSI growth, margins toward 47%, leverage to ~3.0x)
~$205 (top-of-street
sustained 47%+ GM, HBM/2nm content growth, faster deleveraging re-rates the multiple)
Premium multiple reflects the recurring per-wafer model + materials-intensity thesis; the debate is whether AI/HBM growth justifies paying ~34x with 3.6x leverage and a flat 2025.
SWOT
Strengths
- ~75% of revenue is recurring, unit-driven consumables — far less capex-cyclical than equipment peers; gets paid per wafer regardless of who wins.
- #2 in semiconductor materials with broad, qualified-in product breadth across filtration, slurries, deposition/etch chemistries, and fluid handling — deep designed-in positions at every leading-edge fab.
- Secular materials-intensity tailwind — each node and each HBM/advanced-packaging step adds consumption per wafer.
- Strong, rising FCF (18% of sales Q1 FY26) and a clear deleveraging path (3.6x -> ~3.0x).
- Diversified across logic and memory (~40% advanced logic, ~30% memory) and customers, so no single fab or HBM vendor is make-or-break.
Weaknesses
- Carries ~$3.2B net debt from the CMC Materials acquisition — net leverage 3.6x with ~$190M/yr interest expense weighing on net income.
- New-plant ramp (Kaohsiung approaching breakeven, Colorado dilutive into 2027) is a near-term margin drag.
- Still exposed to the semiconductor cycle — 2025 revenue was flat-to-down; not immune, just less volatile.
- GAAP net margin only 7.4% FY2025 — heavy depreciation + interest compress the bottom line versus the gross-margin profile.
- China headwind: revenue modestly declined YoY and export-control regime is a recurring overhang.
Opportunities
- 2nm / gate-all-around ramp and HBM4 scaling materially raise materials content per wafer.
- Advanced packaging (>$100M run-rate) is an emerging multi-hundred-million growth vector.
- Cross-sell of CMC slurries with legacy filtration/handling into a single qualified supplier relationship.
- On-region capacity (Taiwan, US, Korea) wins share as customers de-risk supply chains and qualify local sources.
- Deleveraging frees cash flow for buybacks/M&A and lowers the interest drag as the balance sheet normalizes.
Threats
- U.S.-China export controls tightening through 2025-2026 could cut China-related demand or force requalification.
- Semiconductor downturn / inventory correction would pressure unit volumes despite the recurring model.
- Low-cost regional competitors (e.g., Anji Microelectronics in CMP slurries) pressuring price, especially in China.
- Large diversified rivals (DuPont, Merck/Versum, Resonac, Shin-Etsu, JSR, Fujimi) competing across overlapping materials.
- Input-cost / energy / tariff inflation and FX given the global high-purity manufacturing footprint.
Moats, dependencies & bottlenecks
Moats
products are designed-in and qualified into each fab's process recipe; swapping a filter, slurry, or chemistry risks yield and requires lengthy requalification. The core moat; sticky for years once qualified at a node.
broadest materials portfolio post-CMC (filtration + slurries + deposition/etch + fluid handling) lets it be a single qualified supplier. Breadth aids share but each line faces specialist competitors.
global high-purity plants (Taiwan, US, Korea) near leading-edge fabs that smaller rivals can't easily replicate. Capacity localization is a structural advantage as customers de-risk supply.
Materials/IP and process know-how in high-purity chemistries and CMP slurries. Real but contested by DuPont, Merck/Versum, Resonac, Fujimi, JSR, Shin-Etsu.
consumed and re-ordered every wafer, not a one-time capex sale. Underpins revenue resilience versus equipment peers.
Dependencies
Samsung, Intel, SK hynix, Micron) ~75% of revenue is unit-driven; demand tracks wafer starts, concentrated in a handful of advanced-node fabs.
Recurring model dampens but does not remove cyclicality; FY2025 was flat-to-down.
Tightening 2025-2026; China revenue already modestly declining and subject to requalification risk.
Continued node migration (2nm/GAA) and HBM/advanced-packaging adoption The materials-intensity thesis needs the roadmap to keep advancing; a stall caps content growth.
Input-cost, tariff, and energy inflation flow straight to COGS in high-purity manufacturing.
Advantages
- Recurring, per-wafer consumables model (~75% of revenue) gives revenue resilience versus capex-driven equipment peers.
- Format-agnostic: paid on incremental process steps whether the winner is HBM, DRAM, or leading-edge logic.
- Broadest materials portfolio post-CMC enabling single-supplier qualification and cross-sell.
- Secular materials-intensity tailwind — content per wafer rises every node and every HBM/packaging step.
- Strong and rising FCF (18% of sales Q1 FY26) funding deleveraging and reinvestment.
- Diversified end-market and customer base (~40% advanced logic, ~30% memory) limits single-name concentration.
Weaknesses
- ~$3.2B net debt and 3.6x leverage from the CMC deal; interest expense compresses GAAP earnings.
- Still semiconductor-cyclical — 2025 revenue flat-to-down, not immune to corrections.
- New-plant dilution weighs on margins into 2027.
- China/export-control exposure is a recurring demand and compliance overhang.
- Price pressure from low-cost regional slurry competitors and large diversified materials rivals.
- Rich valuation (~34x forward) leaves little margin for execution slips.
Bottlenecks
- Net leverage of 3.6x and ~$190M/yr interest expense constrain capital allocation until deleveraging reaches ~3.0x.
- New-plant ramp drag — Kaohsiung approaching breakeven and Colorado dilutive into 2027 — caps near-term gross-margin expansion.
- Customer qualification cycles are long — winning content at a new node takes time and front-loads cost before revenue.
- China export-control exposure can strand or delay demand and force costly requalification of supply.
- High-purity manufacturing is capital- and energy-intensive, tying margin to utilization and input costs.
Top signals & trends
Top signals
The single biggest swing factor; HBM step-count growth is the per-wafer content lever.
Q1 FY26 hit 46.9% (incl. ~50bps one-time); sustained 47%+ confirms mix + ramp leverage.
On-track deleveraging lowers interest drag and frees capital allocation.
Already modestly declining; further tightening could cut demand or force requalification.
Watch for the dilution to fade; Colorado stays dilutive into 2027.
Confirms the structural content-per-wafer thesis is intact.
Trends
Lifts wafer starts and per-wafer materials content simultaneously.
Core secular driver — more steps, more filtration/slurry/chemistry per wafer.
>$100M run-rate and growing; new thermal/carrier/packaging materials demand.
Threatens China demand and adds compliance/requalification friction.
Wins share via local capacity but requires capex and ramp dilution.
Pricing pressure from players like Anji, particularly in China.
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.
Specialty chemical / fluoropolymer producers (e.g., Chemours, Daikin, 3M-class fluoromaterials) Supply fluoropolymers, resins, and specialty chemistries that go into filters, membranes, and fluid-handling products.
Abrasive / nanoparticle and base-chemical suppliers Silica/ceria abrasives and reagents for CMP slurries and cleaning chemistries.
Industrial gas and ultrapure-water / energy providers High-purity manufacturing is energy- and utility-intensive; gases (Air Products/Linde) and power feed COGS.
High-purity packaging and global freight for contamination-sensitive shipments.
Largest leading-edge foundry; major consumer of filtration, slurries, and advanced chemistries (Taiwan +18% YoY).
Korea-listed; logic foundry + DRAM/HBM/NAND — large memory + logic materials buyer.
Korea-listed; HBM leader — direct beneficiary link to Entegris' memory/HBM content.
US DRAM/HBM/NAND maker — core memory customer.
Leading-edge logic + foundry (incl. Arizona/Ohio fabs) consuming advanced materials.
Mainstream + specialty logic adds to the unit-driven consumables base.
Took #1 in semiconductor consumables in 2024; overlaps in advanced cleans, CMP slurries, deposition materials.
German-listed; deposition materials and specialty gases — direct overlap in advanced materials.
Japan-listed; CMP slurries, packaging materials — top-5 in slurry market.
Japan-listed; high-purity materials and photoresists; broad electronic-materials scale.
Japan-listed; CMP slurries (Fujimi) and resists/materials (JSR) — top-5 slurry share.
Pall competes in liquid filtration; Anji (China) disrupts CMP-slurry pricing in China with low-cost offerings.