Entegris, Inc. (NASDAQ: ENTG) — A Real Moat and a Real Deleveraging Story, Priced as if Both Are Already Finished

⚡ Claude’s Take

This is the author’s own independent opinion and general information, not investment advice. The analysis that follows (sections 1–15) takes no position and carries no price target; the single directional view in this report is confined to this clearly-fenced block.

Verdict: HOLD / AVOID at $178.77 — a high-quality business at the wrong price. Not a short (the moat and the deleveraging are real and the earnings are not at a cycle peak), but no margin of safety here. The accumulate-on-weakness zone is the mid-$90s to ~$120 (≈ 20–24x adjusted EBITDA / ~6–7x EV/Sales), where you are paying a fair price for the self-help and getting the AI-content optionality for free. Conviction: medium.

Entegris is a genuinely good business — the consumables-and-purity layer of the semiconductor stack, where ~75% of revenue is consumed as wafers are processed (not when fabs buy equipment), gross margins held above 42% even at the 2023 trough, and the moat is a real one: materials qualified “copy-exact” into a customer’s process flow, where switching means risking yield. That is demand-side captivity in Greenwald’s taxonomy, and it shows up in the financials. The problem is not the business; it is what management paid to make it bigger and what the market is now paying for the repair. The 2022 CMC Materials acquisition (~$6.5B, debt-funded, at the cycle peak) doubled revenue but cut normalized ROIC from 14–18% to roughly 5–7% and ROE from ~60% to 16%, and saddled the company with $3.95B of goodwill (47% of assets) and a 2029 refinancing wall. The last three years have been a competent, creditable cleanup — net leverage ~6–7x → ~3.8x, capex $466M → ~$250M, free cash flow −$103M → +$396M. That self-help is the actual thesis, and it is working.

But the tape has already paid for a finished job. After a +39% move in six sessions — a 2.4-beta amplification of a sector-wide AI-semiconductor rally, not anything Entegris-specific — the stock sits at its all-time high, at the 97.5th percentile of its own decade of P/S history (~9.6x EV/Sales, ~35x adjusted EBITDA, ~69x free cash flow). A reverse-DCF says you are underwriting ~24% free-cash-flow CAGR for a decade. That is a heroic ask for a company whose own 2025 incentive plan paid 0% on “revenue growth in excess of markets” — i.e., the content-per-wafer outgrowth engine that the whole bull case depends on stalled last year. My framing: this is a crowded, high-beta, momentum/sector trade at a record multiple, not a falling knife and not a value setup. The quality is in the price twice over. The single piece of evidence that would flip me bullish: two or three quarters of clear MSI outperformance with adjusted-EBITDA margin recovering toward 30%+ and a buyback restart as leverage crosses ~2x. The single piece that would flip me more bearish: another year tracking the market with no outgrowth, or a sector air-pocket — into which that 2.4 beta cuts both ways. Tag: “You don’t get paid for being right about a great business if you pay the all-time-high price for it.”

📈 Stock Price Action — Five-Year Event Map

Entegris has made a full five-year round-trip and then some. From a pre-downturn high near $157 (January 2022) it fell to a five-year low of ~$61.84 (28 December 2022), recovered fitfully through 2023–2024, slid again to a 52-week low of ~$68.70 (November 2025), and has since rocketed to an all-time high of $178.77 (18 June 2026 — the current price and the close of this report). The 52-week range is roughly $68.70 → $178.77; the stock is at its high (0% off) and ~+160% off the November 2025 low. With a beta of ~2.4, Entegris is a high-amplitude expression of the semiconductor cycle — it does not move, it lurches. (Price moves are FACT; attributed causes are INTERPRETATION.)

The defining feature today: Entegris is at its all-time high on a sector-driven spike, with valuation at the ~97th percentile of its own history. The opportunity/mispricing judgment belongs to Claude’s Take above; this map only records what happened and why.

1. Executive Summary

Entegris is a ~$3.2B-revenue supplier of the consumable materials and contamination-control products that semiconductor fabs consume as they manufacture chips — chemical-mechanical-planarization (CMP) slurries and pads, deposition and ion-implant precursors, specialty etch and clean chemistries, liquid and gas filtration and purification, and the FOUPs, carriers and fluid-handling systems that protect ultra-pure materials in transit. It is not an equipment maker. That distinction is the whole point of the business: roughly 75% of revenue is “unit-driven” — consumed per wafer processed and therefore tied to production volumes (wafer starts) and GDP-like end demand, rather than to the famously cyclical capital-equipment budget. The remaining ~25% is capex-driven (fab construction and, to a lesser degree, wafer-fab equipment).

The business carries a real, identifiable moat. Entegris materials are qualified “copy-exact” into a customer’s process flow; switching suppliers risks manufacturing yield — the single most expensive failure a fab can suffer. That demand-side captivity shows up where a moat should: gross margins held in a 42–46% band through a deep 2023 downturn (42.5% at the trough), and management targets 3–6 points of revenue growth above the market via rising “content per wafer” as nodes shrink and stacks deepen. The closest comparable, MKS Instruments, is structurally weaker — MKS occupies the “squeezed middle” of subsystems, while Entegris sits in the “consumables/materials layer with genuine monopolistic niches.”

The complication is capital allocation and price. In July 2022, at the cycle peak, Entegris closed the ~$6.5B (incl. assumed debt) debt-funded acquisition of CMC Materials at ~14–15x EBITDA. The deal roughly doubled revenue but cut normalized return on invested capital from 14–18% to ~5–7%, cut ROE from ~60% to 16%, and layered on $3.95B of goodwill (47% of total assets) plus ~$3.75B of debt. The three years since have been a creditable deleveraging campaign: management divested non-core pieces (QED, Electronic Chemicals to Fujifilm, PIM, the MacDermid alliance), some of which were parts of CMC sold back at lower multiples than paid, took net leverage from ~6–7x to ~3.8x, cut interest expense from $312M to $200M, and let free cash flow inflect from −$103M (FY22) to +$396M (FY25) as the capex super-cycle ($466M peak) normalizes toward ~$250M. The investable thesis is therefore a balance-sheet-repair and free-cash-flow-recovery special situation, not a scale-driven margin compounder — and even the self-help comes with fresh management instability (a new CEO in August 2025 plus three CFOs in fourteen months) and a 2025 in which the outgrowth engine that anchors the bull case missed entirely (the incentive plan paid 0% on revenue-vs-market).

On price, the market has already extrapolated the happy ending. After a +39% six-session, sector-driven, 2.4-beta spike, Entegris trades at its all-time high — ~9.6x EV/Sales (97.5th percentile of its own ten-year P/S history), ~35x adjusted EBITDA, ~65x non-GAAP EPS, ~69x free cash flow. A reverse-DCF implies the market is underwriting a ~24% free-cash-flow CAGR for a decade, or a mid-teens-to-18% sustained FCF margin Entegris has never earned. The scenario set is negatively skewed: even a bull case treads water. This report takes no position; the body that follows quantifies the moat, the returns, and the embedded expectations, and lets the reader weigh them.

2. Business Overview

What Entegris does. Founded in 1966 and headquartered in Billerica, Massachusetts, Entegris (~8,200 employees) develops, manufactures and supplies the materials and purity products consumed inside semiconductor and other high-technology manufacturing. Its products are the slurries, pads, gases, chemistries, filters, purifiers and handling systems that a fab uses up — or relies on for contamination control — every time it processes a wafer. The company frames itself as uniquely positioned at the intersection of materials science, materials purity and materials handling, and its FY25 10-K is explicit that, unlike the capital-equipment names, its revenue is “predominantly unit-driven or recurring from products consumed during semiconductor manufacturing… more impacted by global semiconductor demand and GDP growth than by semiconductor capital equipment sales, which have historically been more cyclical” (FY25 10-K, Competitive Strengths).

Segment structure (two segments, post-2024 reorg). Effective FY2024, Entegris collapsed its prior three-segment structure (Materials Solutions; Microcontamination Control; Advanced Materials Handling) into two reportable segments, recasting all prior years. Note that third-party data aggregators still display the stale three-segment labels — they should be disregarded.

• Materials Solutions (MS) — CMP slurries and pads, CVD/ALD deposition materials, ion-implant specialty gases, formulated etch and clean chemistries. This is the bulk of what CMC Materials brought. FY25 growth was led by CMP consumables, selective etch and advanced deposition, held back by the absence of $33.9M in divested PIM sales.
• Advanced Purity Solutions (APS) — liquid and gas filtration and purification (the highest-quality, highest-margin product line, which posted a “third consecutive record quarter” in liquid filtration per the Q1’26 call), plus FOUPs, wafer carriers and fluid-handling. The filtration/purification piece is the consumables crown jewel; the FOUP/fluid-handling piece is the cyclical, capex-driven part that fell ~7% in 2025 with fab-construction spending.

The margin story to watch: APS segment margin compressed 28.8% (FY23) → 26.8% (FY24) → 23.7% (FY25) on Taiwan/Colorado facility-ramp costs, lower volumes and capex-driven weakness; MS slipped 17.5% → 20.4% → 19.7%. This compression is cyclical and ramp-driven rather than structural, but it is the reason consolidated economics have not expanded with scale.

Revenue model — unit-driven vs. capex-driven. Management quantifies the split precisely: ~75% unit-driven (correlated to MSI — wafer starts) and ~25% capex-driven, of which roughly two-thirds tracks fab construction and one-third tracks wafer-fab equipment. In FY25 the unit-driven base grew ~2% (“in line with wafer starts”) while the capex-driven piece fell ~7%. In Q1’26 the unit-driven base grew ~7% YoY (liquid filtration, advanced deposition, selective etch). This recurring, consumption-linked mix is the single most important structural feature of the business and the basis for the “less cyclical than the equipment makers” claim.

Customers, end markets, geography. Customer concentration is real and rising: TSMC was 16% of FY25 net sales (16% FY24, 11% FY23), the remaining top-10 another 34%, so top-10 = 50% (48% FY24, 43% FY23). End-market mix is roughly 40% advanced logic / 30% mainstream logic / 30% memory (memory split roughly half NAND, half DRAM), with ~60% of revenue from advanced nodes; advanced-packaging revenue is ~$100M (small but growing on AI/HBM). Geographically, the business is 82% international (US 18%), with China at 21% of revenue (down from 24% in FY24), the focal point of the export-control headwind. Non-semiconductor revenue (life sciences, aerospace, solar, hard-disk-drive, flat panel, biomedical, glass) is a small minority.

Verdict (Business Overview): A focused, high-purity consumables franchise with a genuinely attractive revenue model — recurring, consumption-linked, secularly content-growing — but with concentration risk that is increasing as the leading-edge customer base consolidates around TSMC, Samsung, Intel and the memory majors.

3. Industry Dynamics

Where Entegris sits in the value chain. The semiconductor ecosystem captures its economics in a few favorable layers: the equipment oligopolies (ASML in lithography, AMAT/LRCX in deposition/etch, KLA in process control) and the consumables/materials layer with monopolistic niches (Entegris in filtration and specialty materials; Cabot/CMC legacy in CMP). The squeezed middle — generic subsystems, fragmented suppliers — earns far less. Entegris occupies the favorable consumables seat. A peer comparison makes the contrast explicit, framing Entegris as a holder of “genuine monopolistic niches earning durable high returns” against MKS’s “squeezed middle.”

Market size and growth. Demand is the product of wafer starts × materials intensity per wafer. Wafer starts grow at a GDP-plus, AI-accelerated rate; materials intensity grows structurally as process complexity rises. Each node transition (FinFET → gate-all-around at 2nm), each added 3D-NAND layer (250 → ~300+), each shift toward high-NA EUV, and the rise of 2.5D/3D advanced packaging for HBM raises the dollar content of Entegris products consumed per wafer. Management targets ~3–6 points of revenue growth above the underlying market from this content effect — the central secular argument for the business.

Competitive intensity (segment by segment). Moderate-to-favorable in the high-value niches:

• APS (filtration/purification, FOUPs): competitors include Pall (Danaher), Cobetter Filtration, Air Liquide, Linde, Aicello and Shin-Etsu Polymer. Entegris is the leader in liquid microcontamination control and 300mm FOUPs.
• MS (CMP and materials): EMD Performance Materials (Merck KGaA), Qnity Electronics (the 2025 DuPont electronics spin-off), and Fujifilm (which bought Entegris’s divested Electronic Chemicals business). The 10-K’s tell is that “no single competitor competes with us across the full range of our product offerings.”

Regulation — China export controls. China was 21% of FY25 revenue (down from 24%). The structural response is local-for-local manufacturing: ~85% of Q1’26 China revenue is now supplied from Asian facilities, targeted above 90% during 2026. This protects revenue and reframes competition there around yield and performance (Entegris strengths) once supply assurance is solved — but management is conspicuously quiet on the margin cost of regional duplication, which is an open question. China is a managed, persistent headwind, not an existential threat.

Marathon capital-cycle read. Wafer-fab equipment is in a mid-to-late, AI-driven up-cycle attracting heavy capital — classic late-cycle risk for the equipment names. Entegris’s 75% unit-driven base is materially less exposed to the capex cycle and more to production volumes, which dampens the capital-cycle whipsaw. China’s subsidized capacity additions distort the cycle, but Entegris’s qualification moat plus its local-for-local response blunt the share threat better than they would for a fragmented subsystem supplier.

Verdict (Industry): Structurally good. The consumables/purity layer is the high-return, lower-cyclicality seat in an attractive ecosystem, riding a genuine secular intensity tailwind. The caveats — rising customer concentration and Chinese policy risk — are real but do not change the verdict.

4. Competitive Position

Name the moat. In Greenwald’s taxonomy, Entegris’s advantage is primarily demand-side captivity (switching costs), reinforced by intangibles (process IP and materials-science know-how) and niche manufacturing scale. It is not a network-effects business.

• Qualification / switching costs — the core mechanism. Materials are qualified “copy-exact” into a customer’s specific process flow. The 10-K is direct: “Due to the specialized nature of our products, complexity of our customers’ manufacturing processes, customer qualification requirements and costs associated with re-formulation and re-qualification, we believe we have a strong position with our customers,” and switching “may be costly and time consuming for our customers and may introduce risk to their manufacturing yields.” The penalty for switching is yield risk — the most expensive outcome in a fab. With more than 30,000 SKUs and no single product material to total revenue, the captivity is broad and granular.
• Intangibles / co-development. Entegris wins “positions of record” at new nodes years ahead of volume production, co-engineering on customer roadmaps. That is durable, accumulated know-how, not merely patents.
• Niche scale. Class 100–10,000 cleanroom manufacturing and a global local-for-local footprint constitute a genuine cost-and-capability barrier in ultra-pure production.

Pressure-test against the financials. A moat that cannot be tied to a financial outcome is not a moat. Here it can, partially:

• Gross-margin stability through the cycle is the cleanest evidence: ~44.7% (FY19), 46.1% (FY21 peak), held at 42.5% even at the FY23 trough, 44.4% (FY25). A price-taker in a commoditized market does not hold 42%+ gross margin through a downturn.
• APS segment margins of 24–29% and a 12.4% FCF margin (FY25) corroborate real value-add.
• But the moat is masked at the ROIC level. Normalized ROIC is only ~5–7% (FY25) versus 14–18% pre-CMC — not because the underlying franchise deteriorated, but because the debt-funded CMC goodwill and amortization bury the return. The moat is real; the balance sheet obscures it.

One yellow flag. On the Q1’26 call an analyst observed that TSMC revenue was “flat to modestly up” in 2025 — trailing TSMC’s own growth — a possible hint of content or share slippage at the single largest customer. This sits uncomfortably alongside the content-per-wafer claim and warrants validation; it is consistent with the 2025 miss on market-outgrowth.

Verdict (Competitive Position): Durable advantage — a real moat tied to a financial outcome (gross-margin stability through a deep downturn plus structural content-driven outgrowth), currently masked at the return-on-capital level by the CMC debt load. This is a genuine quality business, not a crowded commodity — but the 2025 outgrowth miss is a warning that the moat’s growth dividend is not automatic.

5. Growth History and Forward Opportunities

Historical composition — read past the headline. Reported revenue went $2,299M (FY21) → $3,282M (FY22) → $3,524M (FY23, peak) → $3,241M (FY24) → $3,197M (FY25). Decomposed:

• FY22’s +43% was acquired, not organic — CMC Materials closed July 2022.
• The FY23→FY25 decline was divestiture-led, not a demand collapse. Entegris sold QED (Mar 2023, $134.3M), Electronic Chemicals to Fujifilm (Oct 2023, $675.3M net), terminated the MacDermid Enthone alliance (Jun 2023, +$184.8M pretax gain), and divested PIM to SCF Partners (Mar 2024, $263.2M + up to $25M earn-out). PIM alone was $33.9M of FY24 sales absent in FY25. Underlying organic growth was roughly flat-to-down on the weak 2024–25 semiconductor/capex environment plus the China revenue loss.
• The outgrowth target was missed in 2025. Management aims for 3–6 points above MSI via content per wafer; in 2025 unit revenue grew ~2% against roughly flat wafer starts, and — tellingly — the company’s own incentive plan paid 0% on the “organic revenue growth in excess of markets” metric. The bull engine stalled.

The content-per-wafer math (the forward engine). Management frames node transitions as the dominant outperformance driver: 2nm/GAA logic carries higher content with strong Entegris share; the NAND 250 → ~300-layer migration drives a “double-digit increase in NAND per wafer” via more deposition steps, molybdenum and selective etch; and the DRAM→HBM shift adds incremental content. The honest caveat: management discloses no hard dollar-per-wafer figures by node (N5 vs. N3 vs. N2) on its calls — the precise quantification is reserved, likely for the November-2026 Capital Markets Day. The mechanism is credible; the magnitude is asserted, not yet shown.

Capacity for growth is built and paid for. The 2022→2025 capex super-cycle is complete; FY26 capex is guided down to ~$250M (~7–8% of sales). New facilities — Colorado Springs ($600M, up to $77M CHIPS Act support), Kaohsiung, Taiwan ($550M expansion), a Korea Technology Center, and a $700M Illinois US Technology Center announced September 2025 (part of $1.4B total US investment) — give management the confidence to claim the installed base can deliver “>$1B of incremental revenue with limited further investment.” If true, that is significant incremental-margin and free-cash-flow operating leverage as utilization ramps. It is, however, an unproven claim: APS margin actually compressed during the ramp, so the promised leverage has yet to appear.

Verdict (Growth): High-quality growth potential, currently in a trough-to-recovery transition. The growth that exists is recurring, content-intensive and now sits on a built-out, deleveraging base with >$1B of latent revenue capacity. But the 3–6-point outgrowth claim broke in 2025, the supporting dollar-per-wafer disclosure is absent, and the 2026 recovery depends on node-transition timing the company does not control. Quality of the model is high; quality of recent execution is not.

6. Financial Quality

A necessary data correction. Third-party aggregator data for FY25 is corrupted (it shows a fabricated $923.5M operating income / 28.9% margin, a −$17M SG&A line and a phantom $477M of “other non-operating income”). There is no large FY25 divestiture gain. The real GAAP figures, reconciled to the FY25 10-K, are below.

FY25 GAAP P&L (10-K). Net sales $3,196.6M → gross profit $1,419.9M (44.4%) → SG&A $450.6M, R&D $329.0M, intangible amortization $184.4M → GAAP operating income $455.9M (14.3%) → interest $199.8M, other $9.4M → pretax $254.6M → tax $18.0M → net income $235.6M, diluted EPS $1.55.

Company non-GAAP (10-K MD&A reconciliation):

• Adjusted operating income $680.9M = 21.3% margin (FY24 $743.0M / 22.9%). The ~$225M GAAP-to-adjusted bridge is ~82% intangible amortization ($184.4M), plus restructuring $29.7M and a $10.9M loss on the small specialty-chemicals divestiture.
• Adjusted EBITDA $886.2M = 27.7% margin (FY24 $931.1M / 28.7%) — flat-to-down, not expanding.
• Non-GAAP EPS $2.75 (FY24 $3.00) vs. GAAP $1.55 — a 77% add-back uplift, mostly amortization. This is why the headline P/E (>100x) is meaningless and valuation must anchor on EV/EBITDA, EV/Sales and FCF.

Revenue composition and margins. Organic FY25 was roughly flat-to-down; the reported decline is mostly divestitures (above). Gross margin has been range-bound 42–46% (below the FY21 46.1% peak) — no scale-driven margin expansion has materialized from the doubling of revenue. Adjusted operating margin (21.3%) and adjusted EBITDA margin (27.7%) are both below their FY21/FY24 levels.

Free cash flow — the genuine bright spot. Operating cash flow: FY23 $644.5M → FY24 $631.7M → FY25 $695.4M. Capex super-cycle: FY20 $132M → FY21 $211M → FY22 $466M / FY23 $457M (peak, ~13–14% of sales) → FY24 $316M → FY25 $299M (9.4%) → guided ~$250M FY26. True levered FCF (OCF − capex): FY22 −$103M → FY23 $188M → FY24 $316M → FY25 $396M (12.4% margin, 1.68x net-income conversion). This inflection is the core of the investable thesis and should keep rising as capex normalizes and interest falls.

Balance sheet, debt and leverage. Total debt (par) $3,745.0M (FY24 $4,045.0M); cash $360.4M → net debt ~$3.38B. Deleveraging path: FY22 $5.87B → FY23 $4.65B → FY24 $4.05B → FY25 $3.75B. Net debt / adjusted EBITDA ~3.8x (down from ~6–7x at CMC close; still above management’s roughly low-2x aspiration). Maturity profile: nothing in 2026/2027; $400M in 2028; a $2,450M tower in 2029; $895M in 2030 — blended cash rate ~5.0–5.3%. Interest expense has fallen from $312.4M (FY23) to $199.8M (FY25). Coverage is 2.3x on GAAP EBIT and 4.4x on adjusted EBITDA; liquidity is ~$935M ($360M cash + $575M undrawn revolver); current ratio 3.35x; covenant-compliant. Goodwill is $3,946.7M = 47% of total assets, plus ~$1.0B of intangibles, almost all from CMC; impairment risk is real (a $115.2M charge was taken in FY23; the FY25 test passed with thin headroom in the MS unit).

Returns (normalized). ROE 16.0% (FY24 23.1%); pre-CMC FY20–21 ROE was 56–62%. Aggregator ROIC of ~11% is computed off the corrupted NOPAT and should be discarded; a clean recompute puts ROIC at ~5% GAAP / ~7% amortization-adjusted on ~$7.3B of invested capital, versus 14–18% pre-CMC. The core quality problem in one sentence: CMC roughly doubled revenue but permanently diluted returns — ~$4.9B of goodwill and intangibles earns below the pre-deal franchise — and returns improve only at the margin, via deleveraging, off a depressed base.

Quality of earnings. Cash conversion is clean and strong (OCF/NI 2.95x, as expected given $389.7M of non-cash D&A) — not a red flag. SBC is $69.3M (2.2% of sales), stable. Dilution: diluted shares rose 136M (FY20) → 152M (FY25) → ~153M (Q1’26), ~12% cumulative (mostly CMC stock), with no buyback to offset. Working capital is steady (inventory $643M; cash-conversion cycle ~149 days); no accrual or revenue-recognition red flags; accounting is conservative. One caution: restructuring charges have recurred every year post-CMC ($3.9M → $29.7M) and should be treated as semi-recurring, not “special”; FY23 was distorted on both sides (+$184.8M alliance gain, −$115.2M impairment); and a low, volatile tax rate (7.1% in FY25) flatters GAAP EPS versus a normalized ~18–21%.

Verdict (Financial Quality): Economics do not improve with scale — arguably the opposite. Margins did not expand, and normalized ROIC fell from 14–18% to ~5–7% after CMC. The only genuine improvement is financial self-help: capex normalizing, FCF inflecting (−$103M → $396M), and deleveraging (~6–7x → ~3.8x; interest $312M → $200M), which mechanically lifts equity FCF and EPS over time. This is a deleveraging/FCF-recovery story carrying a 47%-of-assets goodwill base and a 2029 refinancing wall — not a scale compounder.

7. Capital Allocation

The defining (and damaging) deal — CMC Materials. Closed 1 July 2022 for ~$6.5B including assumed debt (~$3.8B closing cash; the cash-flow statement shows FY22 “cash for acquisitions” of −$4,474.9M), funded with a senior secured term loan plus secured and unsecured notes, and announced in December 2021 at the cycle peak. The implied multiple was ~14–15x EV/EBITDA on CMC’s ~$430–450M standalone EBITDA — a premium to Entegris’s own ~10–12x at the time, paid into the teeth of the 2022–23 downturn and the October-2022 China export controls. The return dilution is the core problem: pre-CMC ROIC of 14–18% (ROE 56–62%) collapsed to a normalized ~5–7% ROIC / 16% ROE, with ~$3.95B goodwill, ~$1.0B intangibles and ~$3.75B debt layered on, all earning below the pre-deal franchise. In Marathon’s terms this is textbook value-destructive behavior: large, debt-funded M&A at a peak multiple late in the cycle. It is strategically defensible — CMP slurries and pads deepen the demand-side-captivity moat and the cross-sell is real — but the strategic merit does not rescue the return math.

The cleanup — divestitures and deleveraging. Management then sold non-core pieces — much of what it had just bought — within 12–24 months, generally at lower implied multiples than it paid, into a weak tape: QED ($134.3M, Mar 2023), Electronic Chemicals to Fujifilm ($675.3M net, Oct 2023), the MacDermid alliance terminated (+$184.8M pretax, Jun 2023), PIM ($263.2M + up to $25M earn-out, Mar 2024), and a small specialty-chemicals business sold at a $10.9M loss in Q4 2025. The loss on that last disposal confirms part of the CMC bundle was overpaid. That said, the deleveraging execution itself is creditable: net debt/EBITDA fell from ~6–7x to ~3.8x and interest expense from $312M to $200M.

Capex discipline. The super-cycle ($466M FY22 / $457M FY23, ~13–14% of sales → $299M FY25 → ~$250M FY26 guide) was front-loaded into both the levered balance sheet and a demand trough, which drove FY22 levered FCF to −$103M. The cycle is now complete and the FCF inflection is genuine, but the “>$1B incremental revenue with limited further investment” promise is unproven — APS margins compressed during the ramp, so the operating leverage has yet to appear.

Shareholder returns. Dividend of $0.10/quarter ($60.8M FY25, ~26% payout, ~0.2% yield); zero buyback in FY25 and no active authorization — notable because Entegris did repurchase $44.6M (FY20) and $67.1M (FY21) before suspending to fund CMC. With ~12% cumulative dilution and no offset, per-share value has leaked for three years. Routing ~100% of FCF to debt is correct at 3.8x leverage, but a buyback restart at the ~low-2x target is the key forward swing factor — and the obvious lever a new, ex-CFO CEO may pull.

Executive comp and incentive alignment (2026 proxy, FY25 comp). The annual cash plan (0–200%) weights Adjusted EBITDA Margin 50% / Organic Revenue Growth in Excess of Markets 25% / Free Cash Flow Margin 15% / Customer Experience 10%. FY25 actual payouts by metric were Adj-EBITDA margin 35.2% / revenue-vs-market 0.0% (missed entirely) / FCF margin 15.6% / customer experience 18.2% → 69.0% total (below target) — the proxy itself confirming Entegris did not outgrow its markets in 2025. LTI for 2025 was 40% RSUs / 30% options / 30% PSUs, the PSUs on 3-year relative TSR vs. the Philadelphia Semiconductor Index (the 2023–2025 grant vested at 90%, the 45th percentile). After say-on-pay fell to 77.2% (from 91.2%), management responded with a pro-shareholder 2026 redesign: PSU weight raised to 60% for the CEO, a 3-year FCF-margin PSU metric added, and options eliminated. Alignment grade: mixed. Metrics tie to real value drivers and the formula actually paid below target in a weak year; but there is no ROIC/return-on-capital metric anywhere — management is conspicuously not paid on the returns the CMC deal destroyed — and the heavy reliance on adjusted EBITDA adds back the very CMC amortization, flattering the metric for the dilutive deal. Insiders and directors as a group own under 1.0% of shares — thin skin in the game — and insiders only ever sell (zero open-market purchases; the former CEO disposed of 110K+ shares into strength in February 2026).

Verdict (Capital Allocation): Below-average on the one decision that mattered, on an improving trajectory. A poor capital-deployment decision (CMC at peak, debt-funded) followed by three years of competent damage control. The franchise assets are good; the price, timing and leverage were not. The forward question is whether new management resists the temptation to re-lever for another peak-multiple deal and instead returns capital as leverage normalizes.

8. Changes and Headwinds — Last Two Years

Segment reorganization (3 → 2). Effective FY2024, Microcontamination Control and Advanced Materials Handling were combined into Advanced Purity Solutions, with Materials Solutions retained — a customer-facing realignment that is also a source of the recurring restructuring charges ($29.7M in FY25).

Divestiture program. The non-core CMC pruning described in, executed to deleverage.

Leadership turnover — the biggest soft risk. CEO Bertrand Loy (in the role since 2012) handed off to David Reeder (ex-GlobalFoundries CFO, ex-Chewy CFO, an Entegris director since March 2024) effective 18 August 2025, with Loy moving to Executive Chair through July 2026. The CFO seat then churned: Linda LaGorga departed (January 2026, stated as not the result of any disagreement), Michael Sauer served as interim (March 2026), and Sukhi Nagesh became permanent CFO on 18 May 2026 (the May-2026 Form 3 is Nagesh’s initial-ownership filing, not an open-market purchase). A new CEO plus three CFOs in roughly fourteen months, at the precise moment a turnaround needs steady hands, is a meaningful execution risk.

China / export controls. From 24% (FY24) to 21% (FY25) of revenue, mitigated by the local-for-local strategy (~85% locally supplied, target >90%). A persistent geopolitical drag, concentrated in APS.

Capex build-out complete; AI/HBM inflection. FY26 capex guided to ~$250M against the unproven “>$1B incremental revenue” claim. The 10-K cites AI driving advanced-logic and HBM demand and 2.5D/3D packaging; advanced-packaging revenue (~$100M FY25) is a tailwind but not yet a needle-mover.

Deleveraging milestones and governance. Debt cut from $5.87B to $3.75B par; leverage ~6–7x → ~3.8x; a $300M term-loan repayment in FY25; no 2026/2027 maturities but a $2.45B 2029 wall. Say-on-pay of 77.2% prompted the constructive 2026 comp reset.

Verdict (Changes/Headwinds): Net mixed, leaning modestly positive on financial trajectory but with elevated execution risk. Strengthening the thesis: deleveraging on track, capex done, FCF inflecting, AI/HBM tailwind, better comp. Weakening it: management instability at the worst moment, the 0% revenue-growth payout confirming the missed outgrowth, the China drag, and APS margin compression from the very ramp the capex created.

9. Risk Analysis

Catastrophic-loss assessment. The probability of a permanent total loss is low: the franchise is cash-generative, gross margins are durable, and there are no near-term maturities. The realistic downside is not bankruptcy but multiple compression from a record level into a sector turn (risks 9 + 4 + 8 acting together), where the 2.4 beta turns a sector drawdown into a large equity loss. The highest-conviction risks are valuation/multiple compression (High/High), China/export controls (High/Med-High), and the combination of technology-transition timing and leadership execution (both Med-High) — these are what would break the “deleveraging + AI-content self-help” thesis.

10. Valuation Discussion (Embedded Expectations)

No price target and no recommendation. This section quantifies what the current price embeds.

Live capitalization. At $178.77 × ~153.2M diluted shares, market cap is ~$27.4B; with ~$3.38B net debt, EV ≈ $30.8B. (Aggregator EVs that key off the December-2025 fiscal-year-end close of ~$84.25 are stale — the stock has since roughly doubled.)

Current trailing multiples (px $178.77):

Own-history context. A valuation-percentile screen (2026-06-18) puts the composite at the 89th percentile and P/S at the 97.5th percentile of its own ten-year history (P/B 78th; the P/E 93rd is distorted by depressed GAAP EPS and should be disregarded). The fiscal-year-end EV/Sales series ran 2.0x (FY15) → 8.4x (FY21 super-cycle peak) → 5.1x at FY25 year-end; since the stock has doubled, the live ~9.6x EV/Sales and ~8.6x P/S now sit at or above the FY21 super-cycle peak — Entegris is, on sales, at or near the most expensive it has ever been.

Peer comp set.

The read: Entegris’s ~35x EV/EBITDA is in-line-to-rich versus the cohort, but Entegris is structurally worse on returns than the net-cash, 37–67%-ROIC equipment oligopolists. Its true analog is MKS Instruments — both levered (~3.5–3.8x), both with ROIC diluted by a ~$6.5B peak-cycle debt-funded materials acquisition, both melted into near-record multiples. Entegris is not a net-cash, 40%-ROIC compounder, and should not be priced as one.

Scenario analysis (FY25 → FY28 exit, 3-year; ~0.7%/yr SBC dilution, no buyback):

The set is negatively asymmetric — the same bind as the rest of the WFE cohort. Even the bull case treads water (~−1%/yr), because ~35x EV/EBITDA leaves no room: to make money you need an above-bull operating outcome and further multiple expansion from a near-record level. A perfectly respectable base case still loses ~14%/yr on normalization. The one genuine distinction from the equipment majors is that Entegris’s earnings are not at a cycle peak — but here the multiple, not the earnings, is doing all the work.

Embedded-expectations / reverse-DCF (10% WACC) — the killer datapoint. A single-stage perpetuity off trailing $396M FCF requires ~8.6% FCF growth in perpetuity to justify a $30.8B EV (at 4% perpetual growth, fair EV is only ~$6.9B; at 6%, ~$10.5B). A two-stage model (10 years + 4% terminal) off $396M requires a ~24% FCF CAGR for a decade — FCF growing ~8.5x to ~$3.4B. Even crediting a bull-case forward FCF base of ~$766M (an 18% margin Entegris has never earned) still needs ~19% FCF CAGR over seven years. Interpretation: the market is underwriting a full, durable FCF inflection — capex to $250M, falling interest, content-per-wafer plus AI/HBM growth, FCF margin to mid-teens-to-18% — sustained for 7–10 years. That is aggressive: Entegris has never earned a mid-teens FCF margin (FY25 was 12.4%), its 3–6-point outgrowth target was missed in 2025, and its ROIC is structurally diluted by CMC. The price is reasonable only if you fully believe the self-help and secular-content story compounds cleanly and uninterrupted.

11. Variant Perception

Consensus. The market has re-coded Entegris from over-levered, post-CMC laggard (the $68 narrative of late 2025) into a “deleveraging + FCF-inflection + AI/HBM content-per-wafer compounder,” and the tape agrees — an all-time high — paying ~35x adjusted EBITDA / 97.5th-percentile P/S for the story.

Strongest bull case. The deleveraging self-help is real and largely mechanical (FCF −$103M → $396M; capex $466M → ~$250M; interest $312M → $200M and still falling). Content-per-wafer is a genuine secular grower at node transitions, partly independent of wafer-start growth. Durable 42–46% gross margins through a deep downturn prove a real demand-side-captivity moat. And ~75% unit-driven revenue makes the business less cyclical than the equipment majors — so the cycle risk that caps the AMAT/LRCX/KLAC multiples should bite Entegris less.

Strongest bear case. A near-record multiple (P/S 97.5th, ~35x EV/EBITDA, ~69x P/FCF; reverse-DCF needs ~24% FCF CAGR for a decade); normalized ROIC of only ~5–7% (CMC permanently diluted returns; goodwill + intangibles are 47% of assets); the outgrowth thesis missed in 2025 with no hard dollar-per-wafer disclosure and a hint of content slippage at TSMC; a negatively-skewed scenario set; a 2.4 beta that amplifies any sector drawdown; leadership turnover (new CEO plus a CFO carousel) at the turnaround moment; China at ~21% with export-control drag; insiders net-selling with zero buys; no buyback; and a $2.45B 2029 maturity wall.

The 3–5 assumptions that matter most: (1) does content-per-wafer outgrowth actually resume after missing in 2025? (2) does FCF margin reach and hold the mid-teens-to-18% the price implies? (3) does the ~35x multiple hold, or de-rate toward the cohort/own-history mean? (4) does the AI/MSI demand cycle run clean, given the 2.4 beta? (5) does new management avoid another peak-multiple, re-levering acquisition?

Falsification tests. The bull case is falsified if 2026 again tracks MSI with no outperformance, if FCF margin stalls around 12–13%, if a semiconductor air-pocket arrives, or if management announces another over-priced re-levering deal. The bear case is falsified by two-to-three quarters of clear MSI outperformance with margin recovery to 30%+ and FCF to mid-teens, deleveraging below 3x, and a buyback restart — which would re-base earnings power and begin to justify the premium.

Factor-positioning read. Entegris is a crowded, high-beta, sector/factor-driven (R² 0.75) AI-semiconductor momentum trade at its five-year high and a near-record multiple, with negative alpha (it has lagged what its factor exposures alone would predict — the rally is the sector, not Entegris outperforming). This is the same crowded-cohort setup as MKSI, LRCX, KLAC, AMAT and TER, and it is where consensus is most likely offsides on near-term downside risk: a 2.4-beta name at an all-time high prices in the continuation of the rally, not its reversal.

12. Fact vs. Interpretation

13. Open Questions

1. Content-per-wafer in dollars. Management discloses no hard $/wafer content figures by node (N5 / N3 / N2) — the quantification that would validate the 3–6-point outgrowth claim. Likely reserved for the November-2026 Capital Markets Day.
2. TSMC content/share. TSMC revenue was “flat to modestly up” in 2025 while TSMC itself grew — is Entegris losing content or share at its largest customer, or is this timing?
3. China local-for-local margin cost. What is the margin penalty of duplicating manufacturing regionally to defend the ~21% China revenue?
4. The “>$1B incremental revenue” claim. What utilization and timeline convert the built-out capacity into revenue, and at what incremental margin — given APS margins compressed during the ramp?
5. Capital-return policy under new management. At what leverage does a buyback restart, and will the new CEO resist another peak-multiple acquisition?
6. 2029 refinancing. At what rate does the $2.45B 2029 tower refinance, and how much does that re-set interest expense?

14. What Must Be True

For the bull case to work (and its falsification test):

• Content-per-wafer outgrowth must resume — Entegris must grow 3–6 points above MSI for several consecutive quarters, driven by the 2nm/GAA and 3D-NAND transitions. Falsified if: 2026 revenue again merely tracks the market (a repeat of the 2025 0%-outgrowth result).
• Free-cash-flow margin must reach and hold the mid-teens-to-18% the price implies, lifting FCF toward $600–800M+. Falsified if: FCF margin stalls around 12–13% as capex/working-capital or APS-margin drag persists.
• The ~35x adjusted-EBITDA multiple must broadly hold while earnings grow into it. Falsified if: a sector de-rating or air-pocket compresses the multiple from its near-record level (the 2.4 beta makes this acute).

For the bear case to work (and its falsification test):

• The multiple must compress from a record level toward the cohort/own-history mean as growth disappoints or the sector cycle turns, with the 2.4 beta amplifying the equity loss. Falsified if: Entegris delivers two-to-three quarters of clear MSI outperformance with adjusted-EBITDA margin recovering to 30%+ and FCF to mid-teens, deleveraging below 3x and restarting buybacks — re-basing earnings power and justifying the premium.

15. Source Appendix

See the separate ENTG_source_appendix.md (Appendix B in the combined report) for the full source list with URLs and access dates. Primary sources: Entegris FY2021–FY2025 Form 10-Ks (CIK 0001101302); Q1 2026 Form 10-Q; FY2025 8-K earnings release and material-event 8-Ks; 2026 DEF 14A proxy; Form 3/4/5 insider filings; Q4 2025 and Q1 2026 earnings-call transcripts (a third-party fundamentals aggregator). Quantitative cross-checks: a third-party fundamentals aggregator (statements, ratios, enterprise value — with the noted FY25 income-statement data-integrity issue corrected against the 10-K); the valuation index and price history; a quantitative factor model. Peer comparables: public filings and market data for MKSI, AMAT, LRCX, KLAC and TER.

APPENDIX A — Standard Diligence Questionnaire — Entegris, Inc. (NASDAQ: ENTG)

Supplemental to the research memo. Answers grounded in the underlying analysis, with Fact / Interpretation / Assumption labels where it matters. As-of 2026-06-19; price $178.77.

General

What thoughtful questions have other investors asked about this company? The recurring institutional questions cluster on five points: (1) whether the content-per-wafer “outgrowth” of 3–6 points above the market is real and durable, after it missed in 2025 (the incentive plan paid 0% on revenue-vs-market) — Fact; (2) whether the CMC Materials acquisition will ever earn its cost of capital, given normalized ROIC fell from 14–18% to ~5–7% — Fact/Interpretation; (3) the deleveraging timeline and when capital return (a buyback) restarts; (4) China exposure (~21% of revenue) and the margin cost of local-for-local manufacturing; and (5) on the Q1’26 call specifically, whether Entegris is under-growing TSMC (its 16% customer), which was “flat to modestly up” while TSMC itself grew — a possible content/share-slippage tell.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Interpretation: Below mid-cycle, not at a peak — this is the one genuine distinction from the equipment majors. FY25 adjusted EBITDA ($886.2M, 27.7% margin) and adjusted EPS ($2.75) are down from FY24 ($931.1M / 28.7%; $3.00) and below the FY21 super-cycle peak. The ~25% capex-driven revenue fell ~7% in 2025; the ~75% unit-driven base grew only ~2%. So earnings have room to recover — but the multiple is at a record, doing the opposite.

Driven by the external environment or internal actions? Both. External: wafer starts, the semi/AI demand cycle, China policy. Internal: the deleveraging/FCF self-help (capex normalization, interest reduction) is management-controlled and mechanical; the outgrowth depends on node-transition timing it does not control.

How stable are revenues? Fact: More stable than the equipment makers because ~75% is consumed per wafer (unit-driven), and gross margin held 42.5% even at the FY23 trough. But reported revenue fell FY23→FY25 — mostly divestitures (QED, EC/Fujifilm, PIM), not demand collapse.

Outlook for products/services; how big is the market? The served market grows with wafer starts × rising materials intensity. Secular drivers: 2nm/GAA logic, 3D-NAND layer scaling (250→300+), high-NA EUV, and AI/HBM advanced packaging — all raise content per wafer. Growing, global (82% international), with China a managed ~21% headwind.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? Stable-to-favorable in the high-value niches (filtration/purification, CMP); the 2025 DuPont electronics spin-off (Qnity) is a new pure-play competitor in MS. “No single competitor competes with us across the full range” (10-K).

How profitable is the business (ROIC, ROE)? Fact: GAAP ROE 16.0% (FY25); normalized ROIC ~5–7% (vs 14–18% pre-CMC). Gross margin 44.4%; adjusted EBITDA margin 27.7%. The business is profitable; the returns on the post-CMC capital base are mediocre and only improve at the margin via deleveraging.

Barriers to entry / nature of competition? Interpretation: Real demand-side captivity — materials qualified “copy-exact” into customer process flows; switching risks yield. Plus process-IP intangibles and niche ultra-pure manufacturing scale. Greenwald type: demand/captivity (primary) + intangibles + scale; not network effects.

Can it be undermined by foreign low-cost labor? Low risk — the product is ultra-high-purity materials science qualified to the customer’s process, not labor-cost-sensitive commodity manufacturing. China competition (Cobetter, etc.) is the more relevant threat, contained so far by qualification + local-for-local.

Do brands matter? Switching costs? Brand per se is secondary; qualification is the switching cost — high and the core of the moat.

Financial Condition & Balance Sheet

Assets not fully recognized / off-balance-sheet liabilities? No material off-balance-sheet liabilities flagged. The dominant balance-sheet feature is the opposite: $3.95B goodwill (47% of assets) + ~$1.0B intangibles from CMC, an impairment risk (a $115.2M charge was taken in FY23; FY25 test passed with thin MS-unit headroom).

How conservative is the accounting? Fact: Conservative — clean cash conversion (OCF/NI 2.95x), no accrual/revenue-recognition red flags, steady working capital. One caution: restructuring has recurred every year post-CMC ($29.7M FY25) and should be treated as semi-recurring, not “special”; a low/volatile tax rate (7.1% FY25) flatters GAAP EPS.

How CapEx-hungry is the business? A super-cycle just ended: capex peaked at $466M (FY22, ~13–14% of sales) and is normalizing to ~$250M (FY26 guide, ~7–8%). Going forward, moderately capital-intensive but with the build-out paid for.

Capital Allocation & Management

How much FCF, and how is it used? Fact: FY25 levered FCF $396M (12.4% margin), inflecting from −$103M (FY22). Essentially 100% goes to debt paydown plus a small dividend ($60.8M, ~26% payout); no buyback.

Significant acquisitions? CMC Materials (~$6.5B, Jul 2022) — the defining, return-dilutive deal. Followed by a divestiture program (QED, EC, PIM, MacDermid alliance) to deleverage.

Buying back shares / issuing to insiders? Fact: No buyback since FY21 (suspended for CMC). Diluted shares rose ~12% (136M→152M, mostly CMC stock); ongoing SBC $69.3M (2.2% of sales). Net dilution, no offset.

Compensation policy / motivations of management? Fact: Annual cash plan on Adj-EBITDA margin (50%) / revenue-vs-market (25%, paid 0% in FY25) / FCF margin (15%) / customer experience (10%); 2026 redesign raised PSU weight to 60% and added an FCF-margin PSU after say-on-pay fell to 77.2%. Interpretation: Mixed alignment — no ROIC metric (management is not paid on the returns CMC destroyed), heavy reliance on adjusted EBITDA (which adds back CMC amortization), and <1% insider ownership with only-ever selling (zero open-market buys).

Valuation & Market Data

ADR / MLP / K-1? No — ordinary US common stock, NASDAQ, files standard 10-K/10-Q; no K-1.

Dividend policy? $0.10/quarter ($0.40/yr), ~0.22% yield, ~26% payout — token; capital priority is debt paydown.

Is net income diverging from cash from operations? Fact: OCF ($695M) far exceeds GAAP NI ($236M) — a healthy gap driven by ~$390M non-cash D&A (incl. CMC amortization), not a quality concern. Note the GAAP P/E (>100x) is meaningless; anchor on EV/EBITDA, EV/Sales, P/FCF, non-GAAP EPS.

Risks & Downside

What would cause the stock to decline? Multiple compression from a record level (P/S 97.5th percentile) into a sector turn — amplified by the 2.4 beta; another year of no market-outgrowth; a semiconductor air-pocket; an escalation of China export controls; a goodwill impairment; or a new re-levering acquisition.

Risk of catastrophic / total loss? Interpretation: Low. The franchise is cash-generative, gross margins durable, no near-term maturities (2029 wall manageable). The realistic downside is a large equity drawdown (the 2.4 beta cuts both ways), not insolvency.

Recent News & Events

Has the business environment changed recently? Fact: (1) The June 2026 +39% six-session move was a sector-wide AI-semiconductor rally (Trump pro-semi posts, Intel/Nvidia/Apple, US-Iran de-escalation), not Entegris-specific — no pending M&A. (2) New CEO David Reeder (Aug 2025) and permanent CFO Sukhi Nagesh (May 2026) after a CFO carousel. (3) Capex super-cycle complete; FY26 capex guided ~$250M. (4) Deleveraging continues (~3.8x). (5) Illinois US Technology Center ($700M) announced Sep 2025.

Significant acquisitions / accounting changes / new markets? No new acquisitions (the era is divestiture/deleveraging). Segment reporting changed from three segments to two (MS + APS) effective FY2024. New facilities ramping in Colorado Springs, Taiwan (Kaohsiung), Korea and Illinois.

APPENDIX B — Source Appendix — Entegris, Inc. (NASDAQ: ENTG)

All sources accessed 2026-06-19 unless noted. Primary sources prioritized. Facts reconciled to filings; third-party aggregated data flagged.

Primary — SEC filings (CIK 0001101302), via SEC EDGAR

Primary — Earnings-call transcripts (a third-party fundamentals aggregator)

Quantitative cross-checks (third-party aggregated — reconciled to filings)

Peer / industry comparables

Frameworks

• Greenwald & Kahn, Competition Demystified — moat taxonomy (demand-side captivity / switching costs applied to materials qualification).
• Chancellor (Marathon), Capital Returns — capital-cycle lens on the peak-cycle CMC acquisition and the WFE up-cycle.