Entegris, Inc.（ENTG） · Electronic Materials

Entegris (ENTG) Zen Horizon Research Report

Jun 5, 2026·~31 min read

Quick ReadPlain-language overview · read this first

This is a “picks-and-shovels” supplier to chipmakers, called Entegris. The report’s stance is clear: the business is high quality, but the current price is too expensive. It assigns a “Watch” rating, does not recommend chasing now, and suggests waiting.

What does it mainly do? It does not make chips. It supplies TSMC, Samsung, and other chipmakers with materials that are consumed continuously during manufacturing, such as filtration membranes, polishing materials, and specialty gases. Chipmaking requires an extreme level of cleanliness, where a single dust particle can ruin an entire wafer. Entegris’ job is to purify materials to that extreme.

Why can it earn money steadily? About 75% of revenue comes from these repeatedly consumed materials. Every time a chipmaker processes a wafer, Entegris gets paid, creating a steady stream much like an annual fee. More importantly, its products are embedded in customers’ production lines. Switching to another supplier is costly and may hurt yield, so customers cannot easily move away from it. This is the core strength the report values most.

So why is it expensive? Over the past 1 year, its share price has almost doubled and is close to an all-time high. Yet revenue last year actually slipped slightly, and growth was slower than peers. In other words, the market has already priced several good years ahead into the stock. Buying now leaves almost no bargain, and there is not much cushion if the thesis proves wrong. The report’s calculation is that the stock only has a safety buffer below USD 108.

The biggest thing to watch is the high valuation: if this chip upcycle fades, or if results fail to meet the market’s high expectations, the downside in the share price could be meaningful.

The above is only an explanation of this research report and is not investment advice. Stock markets involve risk; invest with caution.

Lead
Entegris is a global leader in semiconductor purification consumables, with about 75% of revenue tied to annuity-like materials consumed as wafers are produced. Its core thesis rests on certified-in filters, CMP slurries, specialty materials and gases, and wafer carriers, with no single product platform above 3% of sales and very high switching costs once qualified into fab processes. Research rating Watch: a high-quality compounder and one of the cleanest AI pick-and-shovel cash-flow stories, but the current price leaves little margin of safety.

Full report

Prices in the article are as of publication; see the valuation band above for the live price.

Research Summary: A purification leader certified into fab processes and paid through consumable annuities, with price the main issue

Start with the answer. Entegris (NASDAQ: ENTG) is a consumables company selling "purity" and "materials" in the semiconductor manufacturing chain. It does not make chips or lithography tools. It supplies TSMC, Samsung, Intel, Micron and other fabs with items that are continuously consumed during manufacturing: filtration membranes for liquids and gases, slurries and pads for chemical mechanical planarization (CMP), specialty precursor materials for deposition and etch, and wafer carriers such as FOUPs. In one sentence, it is a semiconductor purification pick-and-shovel company: every wafer run through a fab creates another opportunity for Entegris to get paid.

What truly drives its economics comes down to two structural facts. First, about 75% of revenue is unit-driven or recurring consumables. It follows global wafer output rather than the more cyclical capital spending on semiconductor equipment (Entegris FY2025 10-K, 2026-02-11). Management broke down that 75% further on the Q1 2026 call: the remaining roughly 25% is capex-driven, with about two-thirds tied to fab construction and one-third tied to equipment (Entegris Q1 2026 call, 2026-04-30). Second, its products are written into customer processes. The 10-K says its solutions are specified into customer manufacturing flows and customized to unique process conditions, and replacing them is costly, time-consuming, and may threaten yield (FY2025 10-K). This is a switching-cost moat. No single product platform accounts for more than 3% of net sales, so there is no single-point dependency (same source).

What narrative is the market trading now? Two things. First, structural growth in content per wafer under the AI supercycle. AI wafers in advanced logic at 2nm and HBM account for only about 5% of wafer volume but contribute roughly 30% of Entegris revenue. Advanced nodes require more process steps and tighter purity tolerances, so each wafer consumes more of its materials. Second, the market is pricing in deleveraging and margin repair after the CMC Materials acquisition. The stock has nearly doubled over the past year and reached an all-time closing high of $156.33 on 2026-04-27 (Macrotrends historical price), reflecting exactly these two narratives.

What drove the stock historically? In July 2022, Entegris acquired CMC Materials for about $6.5 billion, taking on heavy debt. Net leverage once approached 4.5x. The 2023 semiconductor downturn then hit, while amortization of acquired intangibles and interest expense compressed GAAP earnings, and ROIC fell from 21% to just above 5% (Entegris CMC acquisition closing 8-K, 2022-07-06). The 2024 to 2026 rally is essentially the market believing three things can happen together: cyclical recovery, the final stretch of deleveraging, and AI making the advanced-node content thesis real.

The most important bull-bear debate can be compressed into one sentence: is this a good business? Both bulls and bears agree it is. The disagreement is price. Bulls argue that consumable annuities plus advanced-node content are a decade-long compounding engine, FY2026 adjusted EPS consensus is up 66% year over year in a major recovery year, and UBS has a $205 target price (stockanalysis ratings, 2026-06-04). Bears argue that FY2025 revenue still fell 1%, growth trails the closest peer MKS, margins are low versus peers, EV/EBITDA around 28 sits near the top of the semiconductor materials group and the upper-middle of its history, and Goldman Sachs still rates it Sell with a $115 target (Insider Monkey).

Qualitative label: high-quality compound growth, but a "great business at an expensive price." The business itself is high-quality compound growth, with annuity-like consumables, high switching costs, advanced-node content tailwinds, and improving cash flow. But the current price already embeds much of the long-term thesis and the near-term AI story, leaving a thin margin of safety. The rest of this report builds traceable support for three questions: what is good, how expensive is it, and should investors buy now? The investment stance and rating are left to Section 10, where they follow naturally from the first nine sections.

1. Longitudinal History: From plastic containers to a semiconductor purification leader

1.1 Origins: purity is the hidden necessity of semiconductors

Entegris is built around "materials and purity." The company in its current form was created by the 2005 merger of two lines: Mykrolis, spun out from Millipore's microelectronics business and focused on liquid and gas filtration and purification, and legacy Entegris, focused on handling and storage carriers for wafers and substrates, including FOUPs. Both lines point to the same physical fact: chipmaking has a counterintuitively low tolerance for contamination. A particle that can be ignored at the 28nm node can become a "killer defect" that ruins an entire wafer at 3nm (Entegris Q1 2026 call). Whoever can deliver extreme purity in chemicals, gases, and wafer environments becomes difficult for fabs to replace. The company has been solving that hidden necessity from the beginning, and the essence is unchanged today.

1.2 Development stages: three phases, each driven by category expansion

Dividing Entegris history by business substance explains more than a chronological list.

Phase 1: organic expansion in purity consumables (2005-2018). After the merger, Entegris grew organically along two legs, filtration and materials handling. Revenue rose steadily from about $1.175 billion in 2016 to $1.55 billion in 2018, while gross margin stayed around 43-46% and operating margin around 13-19% (Macrotrends revenue). This phase proved one thing: purity consumables are a slow but steady business with respectable gross margins.

Phase 2: M&A transformation into a global electronic materials leader (2019-2022). This was the turning point. In 2019, Entegris participated in the bidding for Versum, which was ultimately acquired by Merck KGaA for about $6.0 billion (C&EN). After missing that deal, Entegris announced in December 2021 and closed in July 2022 the roughly $6.5 billion acquisition of CMC Materials, adding CMP slurries, polishing pads, and electronic chemicals and positioning itself as a global electronic materials leader with end-to-end CMP solutions (Entegris CMC acquisition closing 8-K, 2022-07-06). The deal pushed revenue from $2.3 billion in 2021 to $3.28 billion in 2022 and $3.52 billion in 2023, reflecting half-year then full-year consolidation, and lifted unit-driven revenue from about 70% to about 80% (CMC acquisition closing 8-K). The cost was substantial debt, including $1.6 billion of 4.75% secured notes plus term loans and bridge financing (Entegris 4.75% notes 8-K, 2022-04).

Phase 3: pruning, deleveraging, and cyclical recovery (2023-present). The company bought CMC just before the 2023 semiconductor downturn, making this the most painful phase. It carried 4.5x net leverage through a downcycle while selling non-core assets to raise cash and repay debt. In 2023, it sold QED, an electronic chemicals business to Quaker Houghton, and another business to Element Solutions, bringing in about $1.1 billion from three divestitures and repaying $1.3 billion of debt that year (Entegris Q1 2024 8-K). In March 2024, it sold the PIM, or Pipeline and Industrial Materials, business acquired with CMC to SCF Partners for up to $285 million (Entegris PIM sale closing release, 2024-03-04). In one sentence: Entegris kept the CMP slurries and pads it truly wanted from CMC, sold lower-strategic-value electronic chemicals and PIM assets, and used the proceeds to repay debt. This was a disciplined digestion of a large acquisition.

The point of this longitudinal story: Entegris did not start with a single hit product. It grew by continuously adding adjacent purity and materials categories, then getting them qualified into fabs. CMC was its largest wager and is also the source of current GAAP earnings distortion and high leverage. Without understanding this, Entegris' P/E can look absurdly high for the wrong reasons.

1.3 Key node: the CMC acquisition is the key to understanding Entegris today

The CMC acquisition deserves a closer look. What happened: announced on 2021-12-15 and completed on 2022-07-06, with enterprise value of about $6.5 billion. Consideration was $133 in cash plus 0.4506 Entegris shares for each CMC share, a premium of about 35% to CMC's pre-halt price. After the transaction, legacy Entegris shareholders held about 91% and CMC shareholders about 9% (Nasdaq report). The market reaction at the time was mixed: CMC rose about 34% that day, while Entegris took on heavy debt and then came under pressure through the 2022-2023 downcycle and leverage burden.

In hindsight, was this moment overestimated, underestimated, or genuinely destiny-changing? A bit of all three. It did change the company's destiny by taking Entegris from two legs, filtration and carriers, into a full-stack electronic materials leader across filtration, CMP, specialty materials, gases, and carriers. Unit-driven revenue rose to 80%, and the content-per-wafer opportunity expanded meaningfully. But it paid a rich price near a cyclical peak, leaving the company squeezed by debt and amortization for the next two years. Impact today: GAAP net income is still depressed by roughly $180 million of annual intangible amortization and, at the peak, $300 million of interest expense (FY2025 8-K, 2026-02-10). This is the fundamental reason GAAP P/E is near 80 while forward P/E is roughly half that.

1.4 Management transition: a 13-year veteran retires, a finance-oriented CEO takes over

A recent variable to watch closely: management turnover concentrated in 2025-2026. CEO Bertrand Loy, who led the company for 13 years, retired on 2025-08-18 and transitioned to executive chairman through the end of Q2 2026. Successor David Reeder is a seasoned finance and capital-markets operator. He previously served as GlobalFoundries CFO, where he led the largest IPO in semiconductor history in 2021, as well as Lexmark CEO and Chewy CFO (Nasdaq, citybiz). The CFO role changed twice in 2026: Linda LaGorga left on 2026-02-28, Controller Mike Sauer briefly acted as interim CFO, and new CFO Sukhi Nagesh officially joined on 2026-05-18 with about 30 years of finance and M&A experience (SEC 8-K). A finance and M&A CEO paired with an M&A-oriented CFO tilts the team toward capital allocation and possible renewed acquisitions. At a point when deleveraging is nearing completion and buybacks may restart in 2027, this team's capital allocation bias deserves continued monitoring.

2. Longitudinal Financial Review: Why P/E is 80 yet not necessarily expensive, and why deleveraging is the main story

2.1 Ten-year income statement: CMC lifted revenue to a new tier and created a margin hole

Fiscal year	Revenue ($M)	Gross margin	Operating margin (GAAP)	Net income ($M)	GAAP EPS
2018	1,550	46.5%	18.9%	241	-
2020	1,859	45.7%	21.2%	295	-
2021	2,299	46.1%	24.0%	409	3.00
2022 (CMC consolidated on 7/6)	3,282	42.6%	14.6%	209	1.46
2023	3,524	42.5%	14.2%*	181	1.20
2024	3,241	45.9%	16.5%	293	1.93
2025	3,196.6	44.4%	14.3%	235.6	1.55

Sources: FY2025 8-K (2026-02-10), Macrotrends net income, stockanalysis financials. *2023 operating income included large one-offs, including a $185 million alliance termination gain, a $115 million goodwill impairment, and a $31 million asset impairment. GAAP comparability across years is very noisy.

This table says three things. First, CMC lifted revenue from the $2.3 billion tier to the $3.2 billion tier. That was a one-off acquisition step-up, not organic growth. Second, gross margin was diluted from 46% pre-acquisition to 42-44%, recovered to 45.9% in 2024, then slipped to 44.4% in 2025 as capacity utilization fell. Third, GAAP net income and EPS look poor, falling from 3.00 in 2021 to 1.55 in 2025, but this is an accounting illusion. The next section unpacks why.

2.2 Why GAAP P/E is 80 while forward P/E is half that: amortization and interest are the two accounting scissors

This is the most important section for understanding Entegris valuation. At the current price of $139.88 and GAAP trailing EPS of about $1.55-1.73, GAAP P/E is roughly 80-90x. In isolation, that looks like a bubble. The bridge makes it clearer:

Fiscal year	GAAP net income ($M)	GAAP EPS	Non-GAAP EPS	Intangible amortization ($M)	Net interest ($M)
2024	292.8	1.93	3.01	190.1	207.9
2025	235.6	1.55	2.75	184.4	191.9

Source: FY2025 8-K EPS reconciliation.

The main bridge from GAAP to non-GAAP is about $1.20 per share of annual intangible amortization. FY2025 amortization was $184 million, mainly from customer relationships, technology and other intangibles created by the CMC acquisition. This is non-cash, so after adding it back, FY2025 non-GAAP EPS is $2.75. The second scissor is interest expense, which peaked at $301 million in 2023 after CMC and fell to $192 million in 2025 as the company deleveraged.

The proper valuation framing is therefore:

GAAP P/E ≈ 80-90x, with the denominator depressed by amortization and interest, distorted and not useful;
Non-GAAP trailing P/E ≈ 48x ($139.88 divided by roughly $2.91 non-GAAP trailing EPS);
Forward P/E: about 54x using FY2026 adjusted EPS consensus of $2.58, or about 39x using FY2027 EPS of $3.62 (stockanalysis forecast).

Methodology warning: forward P/E figures in the market range from 36x to 54x, depending entirely on whether the denominator uses NTM, FY2026, or FY2027 EPS. This report uses one consistent frame: about 48x trailing non-GAAP and about 39x on FY2027 recovery earnings. Under any frame, the conclusion is the same: this is priced as a growth-premium company, not a bargain. Amortization explains most of the gap between GAAP P/E of 80 and forward P/E of 39.

2.3 Balance sheet and deleveraging: the real story of 2024-2026

After the CMC acquisition, total debt at face value reached about $5.87 billion at year-end 2022, net debt was about $5.3 billion, and net leverage approached 4.0-4.5x (stockanalysis balance sheet). The story since then is a deleveraging curve:

2023: repaid $1.3 billion of debt, including $1.1 billion from three divestitures;
2024: repaid about $625 million, bringing cumulative debt repayment since the acquisition to about $2.0 billion (FY2024 DEF 14A);
2025: repaid $300 million of term loans, ending the year with about $3.7 billion of total debt, about $3.4 billion of net debt, and net leverage down to 3.8x;
Q1 2026: repaid another $50 million, with net leverage at 3.6x and a target of about 3x by year-end (Q1 2026 call).

The key implication: management is explicitly prioritizing debt repayment with cash flow, pushing buybacks to 2027 (same source). For the next roughly 18 months, shareholder returns consist only of a modest $0.40 per-share dividend, a yield of about 0.29%, with no buyback support. That is a bearish fact.

2.4 Cash flow and capital returns: FCF is improving, ROIC is diluted by goodwill

Free cash flow is a bright spot. FY2022 capex rose to $466 million because of CMC integration and capacity expansion in Colorado and Kulim, Malaysia, and FCF briefly fell to negative $103 million. Since then, capex has fallen to $299 million in 2025, OCF has risen to $695 million, and FCF has recovered to roughly $396-404 million, with FCF margin around 12.4-12.7% (Macrotrends FCF, Q4 2025 call).

Capital returns are the other side of the coin. CMC brought about $3.9 billion of goodwill and intangible assets into invested capital. ROIC fell from about 21% in 2021 to about 1.5% in 2023, including impairments, and recovered to about 5.6-5.9% in 2025 (Macrotrends ROI, stockanalysis ratios). The distinction matters: the underlying business has high capital returns because purity consumables are asset-light and high-gross-margin. The acquisition lifted the denominator, invested capital, and diluted accounting ROIC. As goodwill declines relative to revenue and earnings recover, ROIC should improve slowly. In the near term, this remains a good business whose reported return on capital is acquisition-burdened.

2.5 Last four quarters: revenue has returned to single-digit growth, but the near-term beat includes one-offs

Quarter	Revenue ($M)	YoY	GAAP EPS	Non-GAAP EPS
Q3 2025	807.1	-	0.46	0.70
Q4 2025	823.9	-3.1%	0.32	0.70
Q1 2026	811.9	+5.0%	0.60	0.86 (above guidance high end)

Sources: FY2025 8-K, Q1 2026 8-K.

The near-term picture is turning upward. Q1 2026 revenue rose 5%, non-GAAP EPS of $0.86 exceeded prior guidance of $0.70-0.78, gross margin recovered to 46.9%, and several banks raised target prices after earnings, including Citi at $170, Needham at $165, and UBS at $205. Two discounts are necessary. First, Q1 gross margin included about +50bp of one-time benefit plus an accounting change in asset useful lives, and management said later quarters would not repeat the same magnitude (Investing.com call). Second, the midpoint of Q2 2026 revenue guidance still implies about +5% YoY at $815-845 million. This is a moderate recovery, not a breakout.

3. Business Model and Moat: Annuity-like consumables plus switching costs from being qualified into fabs

3.1 Revenue structure: two segments, with consumables dominant

Note a reporting change: in Q4 2024, the company combined Microcontamination Control (MC) and Advanced Materials Handling (AMH) into Advanced Purity Solutions (APS), so the latest 10-K has only two reportable segments (FY2025 8-K):

Materials Solutions (MS): CVD/ALD deposition precursors, CMP slurries and pads, specialty gases for ion implantation, and formulated etch and clean materials. FY2025 revenue was $1.407 billion, about 44% of total, with segment margin around 19.7%.
Advanced Purity Solutions (APS, formerly MC + AMH): liquid and gas filtration and purification, wafer and substrate cleaning, FOUP wafer carriers, and fluid handling. FY2025 revenue was $1.799 billion, about 56% of total, with segment margin around 23.7%.

FY2025 total revenue was $3.197 billion, down 1.4% year over year. One structural point: MS, which includes CMP consumables, rose 0.5%, while APS fell 2.8%. The APS decline was mainly due to slower fab construction capex weighing on fluid handling and FOUPs, exactly the cyclicality of the 25% capex-driven revenue, partly offset by growth in gas and liquid filtration consumables (FY2025 10-K MD&A).

3.2 Cost structure and operating leverage: consumable annuities provide a floor

Operating leverage is moderate and stable. About 75% of revenue follows wafer output, or unit-driven demand, giving the business annuity-like resilience when demand weakens. In 2025, unit-driven revenue still rose about 2%, while capex-driven revenue fell 7%, neatly confirming the structure of consumables providing a floor while fab construction fluctuates (Q4 2025 call). Gross margin is sensitive to utilization. Lower utilization in 2025 reduced gross margin by about 250bp, but 43-44% has been the recent floor and Q1 2026 recovered to 46.9%. The company must keep spending on capex, about $300 million a year, and R&D to maintain material qualification leadership at advanced nodes. This business requires investment to maintain the moat, but the intensity is manageable.

3.3 Moat: three real moats and one narrative that deserves a discount

Real moat 1: switching costs from being qualified into fab processes, the strongest moat. This mechanism is written into 10-K risk factors: products are specified into customer processes, customized to unique customer conditions, and replacing them is costly, time-consuming, and threatens yield (FY2025 10-K). This moat is also a double-edged sword. Even Entegris changing its own products, including changes intended to improve yield, can trigger customer requalification. This stickiness remains effective through 3-5 years of adverse conditions. It is structural, not just a tailwind-cycle story.

Real moat 2: portfolio breadth plus end-to-end CMP suite. No single product platform exceeds 3% of net sales, so there is no single-point dependency. In CMP, Entegris can provide slurries, pads, post-clean materials, and conditioners as an integrated suite. This "one-stop qualified" value is meaningful for fabs (FY2025 10-K).

Real moat 3: advanced-node design wins through joint development with fabs. Field application engineers participate directly in customer qualification and process optimization. Plan-of-record, or POR, wins at advanced nodes are ahead: CMP slurry POR wins at the N2 node are 2x N5, and molybdenum deposition materials replacing tungsten in 3D NAND have won POR at all major manufacturers (BeyondSPX analysis).

Narrative to discount: "content per wafer rises with process advancement." The direction is sound, but honesty matters. Absolute first-party financial disclosure of "$X content per wafer" is unavailable. The company mostly provides relative quantification: CMP slurry doubling from N3 to N2, double-digit NAND increases, and outgrowing the market by 4-5 percentage points. These come from calls and secondary accounts, so they are management guidance rather than realized financial disclosure (BeyondSPX, Q1 2026 call). Investors should treat this as a highly probable tailwind that must be verified quarter by quarter, not as locked-in numbers.

3.4 Governance and hedge: customer concentration is rising

The offset to the moat is customer concentration, and it is increasing. The only named customer above 10% in the 10-K is TSMC, at 16% of FY2025 revenue. The top ten customers rose from 43% of FY2023 revenue to 48% in FY2024 and 50% in FY2025 (FY2025 10-K). Samsung, Intel, and Micron are major customers, but the 10-K does not disclose their individual shares, so precise figures should not be invented. Rising concentration cuts both ways: being tied to leading advanced-node customers is good because content is highest there, but changes in any large customer's fab plans or lost share are magnified. High switching costs also mean lost share is hard to replace quickly. On governance, the CMC integration period included many one-offs, but there is no evident hard issue such as accounting fraud or audit controversy.

4. Industry and Cycle: Semiconductor materials market, AI supercycle, and localization tailwind

4.1 Industry structure: do not confuse materials with equipment

One scope distinction is essential: the semiconductor "materials" market is not the WFE fab equipment market. According to SEMI, global semiconductor materials revenue reached about $73.2 billion in 2025, up 6.8% and a record high, including $45.8 billion of wafer fabrication materials and $27.4 billion of packaging materials (SEMI press release). WFE equipment is a different, hundred-billion-dollar scale market. Entegris belongs to the former: a moderate-growth, single-digit, highly sticky market that grows steadily with wafer output. Entegris' own target is to outgrow its served market by about 4-5 percentage points through content-per-wafer gains, under its own definitions rather than SEMI's.

The profit pool favors material suppliers' pricing. The top eight CMP slurry suppliers hold about 77% share, DuPont-related players hold more than 50% of polishing pads, and the top five pad conditioner suppliers hold about 90%. Subsegments are highly concentrated (JEES industry comparison). Downstream customers are also concentrated among TSMC, Samsung, Intel, Micron and other giants with strong bargaining power, but long contracts plus qualification create mutual lock-in. In 2023, Entegris signed a multi-year CMP slurry supply agreement with TSMC.

4.2 Cyclicality: semiconductor cycle plus AI capex cycle, currently upswing

Entegris is exposed to both the semiconductor cycle and the capital spending cycle, but consumable annuities make it less volatile than pure equipment stocks. The current cycle position is clear: 2023 was a down year, and 2024-2026 are an AI-capex-driven upcycle. Evidence includes TSMC guiding 2026 capex to $52-56 billion, versus about $30 billion in 2024 and about $41 billion in 2025; 300mm fab equipment investment breaking $100 billion for the first time in 2025 (industry overview); and the five largest hyperscalers guiding about $600-690 billion of combined 2026 capex, with roughly 75% toward AI infrastructure (QuantFlow).

Entegris' leverage to this cycle comes from about 75% of revenue tied to wafer output, including advanced logic 2nm production starting in 2026 and HBM/high-layer NAND output, plus about 25% tied to fab construction capex. Put differently, it participates in the AI cycle through more advanced wafers being manufactured and each wafer consuming more of its materials, not by selling a more expensive tool. That makes its cyclicality milder than equipment stocks, but it also means it will not have the same explosive upside early in an upcycle.

4.3 Policy and geopolitics: China controls are a disturbance, localization is a tailwind

China exposure: Mainland China revenue rose from 15% of FY2022 revenue to about 21% in FY2024-2025, tied with Taiwan as one of the two largest markets (FY2025 10-K). The actual impact of export controls is a single-digit-percentage disturbance, not a collapse: the October 2022 first round of controls reduced Q4 2022 revenue by about $40-50 million, affected 2023 by about $20 million per quarter or about $80 million annualized, and the 2025 outlook included about $30-40 million per year of impact from China restrictions (Q3 2022 8-K, FY2022 10-K). The more important risk is structural. The 10-K says controls "incentivize Chinese domestic semiconductor companies to work more closely with local suppliers to better compete with our products" (FY2025 10-K). The company's hedge is localized supply. By the end of 2025, about 80% of China demand was supplied from local Asian bases, rising to about 90% in 2026.

Localization tailwind: Entegris was the first semiconductor materials supplier to finalize CHIPS Act award terms, receiving up to $77 million in direct funding for a new Colorado Springs facility, with about $600 million of private investment and about 600 jobs. The plant will produce liquid filtration products and FOUPs, which had previously all been manufactured in Asia (NIST press release, 2024-12). Logically, new fabs in the U.S., Europe and Japan bring one-time construction-related materials and facilities demand, then recurring consumable annuities after ramp. The company has not disclosed a precise "one fab equals $X" quantification.

5. Horizontal Peers: Highest multiple, slowest growth, and lower margins within semiconductor materials peers

Entegris' three business lines map to different competitors. The table below uses a unified primary source, stockanalysis.com as of 2026-06-04. Cross-company definitions may differ, with known caveats after the table.

Metric (as of 2026-06-04)	ENTG	MKS Inc (MKSI)	Qnity (Q)	Air Liquide (AI.PA)	Linde (LIN)	Fujimi (5384)	Donaldson (DCI)
Market cap	$21.3B	$22.3B	$32.4B	≈$114B	$234.6B	≈$2.0B	$9.7B
PE-TTM	80*	69.1	49.9	29.9	33.7	35.0	22.6
Forward PE	36-54*	26.0	37.6	25.2	27.8	28.2	20.0
EV/EBITDA	28.0	27.7	24.9	14.9	18.9	18.7	15.1
EV/Sales	7.6	6.4	7.3	4.3	7.4	4.4	2.7
Revenue growth	Nearly flat (FY25 -1%)	+11.5% YoY	n/a (new spin)	Forecast +5.7%	Forecast +5.7%	Forecast +9.2%	Forecast +4.3%
Operating margin	15.7%	14.7%	21.6%	19.4%	28.2%	19.9%	15.1%

Sources: ENTG, MKSI, Q, AI.PA, LIN, Fujimi, DCI. *ENTG's GAAP PE-TTM of 80 is distorted by amortization and interest. The 36-54 forward range depends on which fiscal-year EPS is used, as discussed in §2.2.

What each company really represents:

MKS Inc. (MKSI): the closest comparable, growing faster and cheaper. Formerly MKS Instruments, it scaled through vacuum, gas and optical subsystems plus the Atotech acquisition in plating chemicals. Revenue of $4.07 billion already exceeds Entegris, and TTM revenue growth of +11.5% is clearly faster than Entegris' near-flat growth. The two trade at almost identical EV/EBITDA, 27.7 versus 28.0, but MKSI's forward P/E of 26x is meaningfully below Entegris. MKSI's flaw is also high leverage from the Atotech acquisition. In comparison, MKSI has stronger growth and a cheaper valuation; Entegris has a purer semiconductor consumables mix and steadier margins.

Qnity Electronics (Q): a new DuPont electronics spin-off and higher-quality profit competitor. This is an important correction for this research. The most direct comparable for Entegris in CMP slurries and specialty electronic materials is no longer DuPont (DD), but Qnity, which separated from DuPont on 2025-11-01 and contains the former Semiconductor Technologies and Interconnect Solutions businesses, including CMP slurries and pads and lithography materials. Qnity has $4.95 billion of revenue and a $32.4 billion market cap. Its operating margin of 21.6% and net margin of 13.1% are clearly above Entegris at 15.7% and 8.2%, while its EV/EBITDA of 24.9 is slightly below Entegris. In CMP, Entegris faces a strong competitor with higher margins and no higher valuation (SEC Qnity 10-12B).

Air Liquide and Linde: industrial gas duopoly, superior profitability, lower valuation. Electronic specialty gases are only part of their businesses, and their scale is 5-11 times Entegris. Operating margins of 19-28% are higher across the board, but the market values them as stable cash cows, with EV/EBITDA only 15-19 and forward P/E 25-28. They are both upstream suppliers and competitors to Entegris' specialty gas business. The valuation gap reflects two different narratives: utility-like gas versus semiconductor consumables growth.

Fujimi (5384.TSE) and Donaldson (DCI): narrow-domain specialist and industrial reference point. Fujimi is a Japanese pure-play CMP slurry company, small-cap, with higher margins and a lower valuation. Donaldson is an industrial filtration leader, with semiconductors only a small portion and the lowest valuation at 20x forward P/E and 15x EV/EBITDA. Together, they further show that Entegris sits in the most expensive "semiconductor consumables" bucket.

Ecosystem conclusion: Entegris is a full-stack leader in semiconductor purity and materials, spanning filtration, CMP, specialty materials, gases, and carriers. Its position is solid. But in a peer-group view, one uncomfortable fact stands out: its multiples are high-end among peers, with EV/EBITDA and EV/Sales near the upper bound, while growth (FY25 -1%) and margins (15.7% operating, 8.2% net) are low-end among peers. This is not a judgment; it is the fact pattern the valuation section must answer.

6. Current Fundamentals and Bull-Bear Debate

6.1 What the market is trading

The current stock price mainly reflects two narratives: advanced-node content growth under the AI supercycle, with Q1 2026 revenue up 5% and management emphasizing 2nm, HBM and NAND content, and margin/EPS repair after deleveraging nears completion, with FY2026 adjusted EPS consensus up 66% year over year. It is important to separate real fundamentals from narrative heat. Real positives include consumable annuities, improving FCF, and net leverage down to 3.6x. Possible excesses include a stock that has nearly doubled in a year and is close to historical highs, even though FY2025 revenue declined and the Q1 gross-margin beat included one-offs. The market has pulled forward a lot of future recovery and AI content.

6.2 Bull and bear cases, each with evidence

Bull case:

About 75% consumable annuities, no single-point dependency above 3%, and switching costs from being qualified into fabs equal a high-quality and predictable compounding engine (FY2025 10-K).
Advanced-node content tailwinds are quantifiable: AI wafers are 5% of volume but 30% of revenue; CMP slurry doubles from N3 to N2; N2 POR wins are 2x N5 (BeyondSPX).
2026 is a major EPS recovery year: adjusted EPS consensus is $2.58, up 66.6%, and FY2027 is $3.62, up 40.4% (stockanalysis forecast).
Deleveraging is near the final stretch, with net leverage going from 3.6x toward 3x, and buybacks may restart in 2027, releasing shareholder returns (Q1 2026 call).

Bear case:

Valuation is expensive: EV/EBITDA of 28 is at the upper end of semiconductor materials peers, GF Value judges the stock significantly overvalued with about a 37.6% premium, and Goldman Sachs maintains Sell with a $115 target (gurufocus, Insider Monkey).
Growth trails peers: FY2025 revenue fell 1%, slower than MKS at +11.5%, and margins lag Qnity, Fujimi and the gas giants.
The near-term beat has noise: Q1 gross margin included about +50bp of one-time benefit plus an accounting change, and management said it is not sustainable at the same magnitude (Investing.com call).
There is a shareholder-return vacuum: buybacks are delayed to 2027, dividend yield is only 0.29%, and there is no buyback support for roughly the next 18 months (Q1 2026 call).
Insiders are net sellers, with zero buys. Net sales over the past three months are about $14-26 million depending on source, with the direction consistent (Simply Wall St).

7. Valuation Analysis: A good business, but no margin of safety at the current price

7.1 Historical valuation percentile

Current EV/EBITDA of about 28 is in the upper-middle part of the company's 13-year range of 7.9-34.9, with a median of 16.6 and a five-year average of 19.3 (gurufocus EV/EBITDA). P/E is less reliable as a percentile metric because trough-cycle earnings are distorted by amortization, and the five-year average is as high as 53x due to near-zero EPS quarters. Conclusion: on the cleaner EV/EBITDA basis, current valuation is in the upper historical range and is not cheap. This is consistent with a share price close to historical highs and nearly doubled over the past year.

7.2 Peer valuation: peers being expensive does not make Entegris cheap

Section 5 already showed that among semiconductor materials pure plays, ENTG, MKSI and Qnity, Entegris has EV/EBITDA and EV/Sales at the upper end, while growth and margins are at the lower end. The Zen Horizon Framework discipline applies here: when the entire semiconductor sector is at historically high valuation levels, with IDC forecasting 2026 global semiconductor revenue up 52.8% year over year and AI dominating the narrative, group-wide narrative premiums may be overheated. "Peers are also expensive" cannot prove Entegris is cheap. Absolute valuation must stand on its own.

7.3 Absolute valuation: cash-flow look-through plus three scenarios

Cash-flow look-through: FY2025 operating cash flow was $695 million versus GAAP net income of $236 million, so OCF/net income was about 2.9x. That is far above 1 and shows GAAP profit is heavily understated by non-cash intangible amortization. Owner earnings are much better than accounting earnings. Maintenance capex likely accounts for most of annual capex of about $300 million, with expansion capex falling as new facilities near completion. Using FCF of about $400 million, FCF yield is about $400 million divided by a $21.3 billion market cap, or 1.9%. That absolute level is low and reflects growth-premium pricing.

Using EV/EBITDA as the main anchor, cleaner for a leveraged company, with net debt of about $3.4 billion and about 152.5 million shares, the three scenarios are:

Dimension	Bear	Base	Bull
Revenue/EBITDA assumptions	Recovery stalls, adjusted EBITDA ≈ $880-920 million	2026-27 recovery materializes, EBITDA ≈ $980 million to $1.05 billion, adj EPS $2.6->$3.2	AI supercycle, EBITDA ≈ $1.15-1.20 billion, adj EPS above $3.6
Multiple assumptions	EV/EBITDA 18-20x	EV/EBITDA 21-24x	EV/EBITDA 26-28x
Key catalysts	Cycle rolls down again, China controls tighten, margins fall back	Moderate recovery + deleveraging to 3x + content thesis verified quarter by quarter	Content compounding + buybacks restart in 2027 + valuation stays high
Implied value per share	$85-110	$115-145	$165-205
Implied return (vs $139.88)	Downside -21% to -39%	-18% to +4%	+18% to +47%
Permanent loss risk	Triggered by: FY2026 revenue turns negative + net leverage stuck above 3.5x + AI content thesis falsified	Triggered by: recovery below expectations, gross margin gives back one-time benefit	-

(Upside is shown as return percentage and downside as permanent-loss triggers; the two are not symmetric. This is not investment advice, but scenario analysis under the research framework.)

The current price of $139.88 sits near the upper end of the base-case range of $115-145. That means the market is already pricing in a smooth recovery. The bull-case upside requires AI content to keep exceeding expectations and valuation to stay near historical highs. The bear-case downside, -21% to -39%, is much larger than the remaining base-case upside of +4%.

7.4 Expectation-gap analysis

Market-implied expectations are optimistic: FY2026 adjusted EPS up 66%, paired with a historically upper-middle EV/EBITDA multiple. The most likely sources of expectation gaps are: 1. whether CMP/NAND content actually shows up in reported revenue growth, since this is the core bull assumption and currently rests mainly on management commentary; 2. China revenue trends under controls and local substitution; 3. whether gross margin can hold 44-45%+ after excluding one-offs; 4. capital allocation after deleveraging, meaning buybacks versus renewed M&A. In the next earnings report, Q2 2026, the market will focus most on whether revenue can continue above +5% and on full-year MSI, or materials demand, guidance.

7.5 Margin-of-safety review, independent discipline

The current $139.88 is a premium to the bear-case implied value of $85-110. Margin of safety is zero.
The most fragile assumption in the three scenarios is "valuation multiple remains in the historical upper-middle range." If EV/EBITDA falls 30% from 22x to about 15x, returning to the historical median, base-case value per share would fall to around $90. Downside is large.
If earnings do not grow over the next three years, annualized return at the current price would be far below the risk-free rate. This entry price has no margin of safety.
This is a classic great company but bad price setup. It is worth waiting for a better price rather than buying now.
Margin-of-safety conclusion: insufficient. The current price is above the bear-case range, and stress-testing the most fragile assumption creates downside much larger than upside.

8. Risk Analysis

Business risks: 1. another semiconductor cycle downturn, medium probability and high impact, observable through fab utilization, TSMC capex guidance, and memory prices; 2. rising customer concentration, with top ten at 50% and TSMC at 16%, magnifying the impact of lost share or changes in major-customer fab plans, low probability and high impact; 3. Chinese local substitution eroding share, medium probability and medium impact, with controls encouraging local supply chains over the long term.

Financial risks: 1. net leverage is still 3.6x and interest expense is about $190 million a year. If cash flow worsens, deleveraging stalls and buybacks are delayed further, medium probability and medium impact; 2. goodwill is about $3.9 billion, and impairment of acquired assets would hit book value, low probability and medium impact.

Valuation risks, currently the largest risk: 1. EV/EBITDA of 28 sits in the historical upper-middle range, so rising rates or factor rotation could compress the multiple, medium probability and high impact; 2. if the AI theme cools, semiconductor valuations could reset as a group, and Entegris would have higher sensitivity as a high-multiple stock, medium probability and high impact; 3. high growth expectations may be missed, including FY2026 EPS growth of +66%, medium probability and high impact.

Governance and external risks: 1. concentrated management transition, with new CEO/CFO integration and capital allocation direction still to observe, medium probability and medium impact; 2. continued insider net selling, a negative signal; 3. further tightening of export controls and tariff disruption, medium probability and medium-to-low impact; 4. key raw materials dependent on single sources or single countries, a supply-chain risk listed in the 10-K.

9. Catalysts and Tracking Dashboard

Positive catalysts: revenue growth returns to and accelerates within the mid-to-high single digits; gross margin holds above 45% after excluding one-offs; net leverage falls below 3x; buybacks restart in 2027; 2nm/HBM content shows up in revenue; China controls ease at the margin.

Negative catalysts: guidance cuts; gross margin reversal; China revenue declines due to controls or local substitution; AI theme cools and compresses valuation; major-customer order volatility; insider selling intensifies.

Tracking dashboard:

Metric	Why it matters	Normal/healthy range	Risk signal
Quarterly revenue YoY	Litmus test for recovery and content thesis	Mid-to-high single digits and accelerating	Turns negative for two consecutive quarters
Adjusted gross margin	Real profitability after excluding one-offs	Holds 44-46%	Falls below 43%
Net leverage (Net Debt/EBITDA)	Deleveraging main line, buyback prerequisite	Declines quarter by quarter to ≤3x	Stuck above 3.5x
Unit-driven revenue growth	Consumable-annuity resilience	Positive growth with wafer output	Turns negative
China revenue share and growth	Exposure to controls and local substitution	Share around 21%, growth stabilizes	Share and growth both decline
Top ten customer share	Concentration risk	≤50% and no longer rising	Keeps rising
EV/EBITDA	Valuation percentile	Safer if back to historical median (~16-20x)	No margin of safety if maintained at 28x+
Capital allocation moves	New management direction	Buybacks after deleveraging	Renewed high-price M&A with leverage

10. Horizontal-Vertical Synthesis: A proven good business at a price that fully tells the good story

10.1 Horizontal-vertical synthesis

Longitudinally, Entegris has truly proven one capability: turning "purity and materials," a hidden necessity in semiconductors, into a high-quality business qualified into fabs and paid through consumable annuities, while expanding categories through repeated acquisitions, with CMC the largest, to lift content per wafer and the unit-driven revenue share. Its historical success came mainly from structural advantages in the business itself, including switching costs, category breadth and advanced-node design-win leadership, plus the secular tailwind of semiconductor penetration. It was not merely cyclical luck or financial leverage. The CMC wager was even a burden for a time. These success factors still exist today: consumable annuities, switching costs and content tailwinds remain intact, while FCF and leverage are improving.

Horizontally, its real advantage versus peers is full-stack purity and materials leadership plus the purest semiconductor consumables mix. But its relative weaknesses are also real: growth is slower than the closest competitor MKS, and margins are lower than Qnity and the gas giants, while the market gives it a multiple at the top of this peer set. Current valuation rewards past success, meaning a proven good business, while also pulling forward the future, meaning full credit for the 2026-27 recovery and AI content story.

The market's most likely error is equating "good business" with "good price today." Business quality and entry point are separate questions. That is the discipline emphasized repeatedly in this report and in §7.5. Over the next year, the key variables are whether revenue growth can stay above +5% and whether gross margin can hold after stripping out one-offs. Over three years, the key questions are capital allocation after deleveraging and whether the content thesis materializes. Over five years, the question is whether Entegris can defend advanced-node POR leadership and pricing power against Qnity and MKS.

10.2 Pre-mortem: if the stock is down 50% in three years, what is the script?

Script 1: valuation normalization plus cyclical giveback, the most likely. In the second half of 2026, AI capex growth downshifts. In 2027, semiconductors enter a mild downturn. Entegris revenue growth slips from +5-8% back toward zero, and adjusted EPS stalls at $2.6-2.8 instead of moving toward $3.6. At the same time, the AI theme cools and the market cuts EV/EBITDA from 28x back to the historical median of about 16x. EBITDA of about $900 million x 16 = $14.4 billion EV, minus $3.4 billion net debt = $11.0 billion equity value, divided by 152.5 million shares = about $72 per share, down about 49% from the current $139.88. This path does not require the company to "break." It only requires growth disappointment plus valuation normalization.

Script 2: competition plus China double hit. Qnity, with higher margins, and Chinese local suppliers together take share in CMP and specialty materials. Export controls tighten further, structurally eroding part of the 21% China exposure. Gross margin slides from 44% toward 40% due to price competition. The market discounts the pricing-power narrative and compresses the multiple at the same time, again pointing to a roughly halving outcome.

10.3 Final research conclusion

【Company Profile Scorecard】

Fundamental quality: High (consumable annuities + switching costs + no single-point dependency above 3%)
Growth: Medium (long-term tailwinds, but near-term growth is slower than peers and FY25 was -1%)
Moat: Strong (switching costs from being qualified into fabs are structural)
Financial resilience: Medium (FCF improving, but net leverage still 3.6x and buybacks delayed)
Management credibility: Medium (good deleveraging discipline; new CEO/CFO integration and capital allocation still to observe)
Valuation attractiveness: Low (EV/EBITDA of 28 is at the peer upper end and historical upper-middle)
Risk level: Medium (valuation risk dominates; business risk is manageable)
Suitable investor type: long-term growth investors, but only at a better price; unsuitable for short-term chase buying

【Investment Rating】

Rating: Watch
One-sentence investment thesis: the semiconductor purification pick-and-shovel business is first-tier in quality, but the current price is close to historical highs, EV/EBITDA is at the peer upper end, and margin of safety is zero, so wait for a better price rather than chasing now.
Three price signals (endpoints from §7.3):

Ideal Buy Price: below $108 (below the upper end of bear-case implied value and providing ≥20% margin of safety)

Holdable price: $115-145 (base-case implied value range)
Clearly overvalued price: above $205 (above the upper end of bull-case implied value)
Current price classification: $139.88 is near the upper end of the "holdable" band and close to the clearly overvalued end. For new capital, the margin of safety is insufficient.
Is it worth waiting for a better price: yes. Specific buy triggers: share price back below $108, or EV/EBITDA returning to the historical median of about 16-20x while revenue growth is confirmed back in the mid-to-high single digits. The opportunity cost of waiting is possibly missing part of the bull-case upside, but downside risk of -21% to -49% is much larger than remaining upside. Discipline argues for waiting.
Target holding period: 3-5 years. The business is a compounder, but entry price matters.
Expected annualized return: negative in the bear case, low single digits in the base case, high single digits to double digits in the bull case, heavily dependent on valuation remaining high.
Maximum loss risk: as in §10.2, the worst case, growth failing to materialize plus valuation normalization, is roughly a halving at about -49%.
Signals that trigger reassessment: 1. share price back below $108, turning more constructive; 2. revenue turns negative for two consecutive quarters or gross margin falls below 43%, turning more cautious; 3. net leverage stays above 3.5x and does not decline; 4. China revenue share and growth both decline; 5. after deleveraging, management pursues high-price M&A and adds leverage.

Again, this report is research analysis based on public information and is not investment advice.

Key Data Table

Item	Value (as of 2026-06-05)
Share price / market cap	$139.88 / about $21.3-21.4 billion
52-week range / all-time closing high	$66.72-$159.15 / $156.33 (2026-04-27)
Revenue TTM / FY2025	about $3.24 billion / $3.197 billion (YoY -1.4%)
Gross margin / operating margin (GAAP)	44.4% / 14.3%
GAAP EPS / Non-GAAP EPS (FY2025)	$1.55 / $2.75
P/E: GAAP-TTM / Non-GAAP-TTM / forward (FY27)	about 80 / about 48 / about 39
EV/EBITDA / EV/Sales	about 28 / about 7.6
FCF / FCF margin (FY2025)	about $400 million / about 12.5%
Net debt / net leverage (2026Q1)	about $3.4 billion / 3.6x
Dividend (annual) / dividend yield	$0.40 / about 0.29%
Unit-driven revenue share	about 75%
China revenue share (FY2025)	about 21%
Largest customer (TSMC) share / top ten share	16% / 50%
FY2026E / FY2027E adjusted EPS consensus	$2.58 (+66.6%) / $3.62 (+40.4%)
Analyst target price (average/range)	about $160 / $115-$205

Research Uncertainties (Known Blind Spots)

No first-party absolute "content per wafer" figure: CMP N3->N2 doubling, NAND double-digit growth, and 4-5pp market outgrowth are management or secondary-source figures. No first-party filing discloses an absolute "$X per wafer" value. The core bull assumption is mainly guidance-supported and must be validated quarter by quarter through revenue growth.
Forward P/E methodology divergence: 36x-54x depends on whether NTM, FY2026, or FY2027 EPS is used. This report consistently uses about 48x trailing non-GAAP and about 39x FY2027. Pay attention to methodology when citing other sources.
No first-party disclosure for Samsung/Intel/Micron shares: the 10-K only names TSMC at 16%. Other major-customer shares are industry knowledge rather than disclosed figures, so no precise assumption is made.
Historical EV/EBITDA differs across sources (18x vs 28x), due to TTM versus fiscal-year definitions and treatment of one-offs. This report uses current enterprise value divided by TTM adjusted EBITDA.
Synergy delivery and new-fab ramp contribution are unquantified: the actual delivery of CMC's original $75M cost synergy target and incremental revenue from the Colorado new facility have not been explicitly quantified by the company.
Some call details come from secondary transcripts because SEC/IR pages repeatedly blocked scraping. Key quarterly figures have been cross-checked against primary 8-Ks, but certain export-control revenue-impact figures, such as about $30-40M per year in 2025, come from secondary call transcripts and should be rechecked against official transcripts.

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.

Mentioned Tickers

MKSI Q AI LIN 5384 DCI DD

Baillie Framework · Ten Questions for Growth Investing

Hunting ten-year five-baggers among great growth stocks — pressing the upside question: "Can it get much bigger?"

How high is its market ceiling? Is it expanding an existing pie, or creating an entirely new market?5/10

Bottom line: it is expanding a large, moderately growing existing pie, not creating a new market. The ceiling is broad and gradual, not the kind of explosive new category Baillie Gifford prefers. Entegris sells a consumables annuity around “purity and materials” to fabs: filtration membranes, CMP slurries and pads, deposition and etch precursors, specialty gases, and FOUP carriers. Its arena is the semiconductor materials market that has existed for decades. Its excess growth comes from “higher content per wafer at advanced nodes + outgrowing the market by 4–5 percentage points,” not from opening a market that previously did not exist.

How big is the pie, and how fast is it growing? On SEMI’s definition, global semiconductor materials revenue reached a record $73.2 billion in 2025, up +6.8% year over year ($45.8 billion in wafer-fabrication materials + $27.4 billion in packaging materials). The key is growth rate: this is a single-digit growth market with very high stickiness, not a new blue ocean capable of doubling quickly. The report also clarifies a common misread: the semiconductor “materials” market is not the WFE equipment market, which is a hundred-billion-dollar-scale market. Do not conflate the two.

Ceiling and penetration: Entegris FY2025 revenue was about $3.2 billion. Against a $73.2 billion materials market, its share is only around the single digits, roughly ~4–5%. That means the long runway of “expanding an existing pie” is real and far from capped. Its growth engine is structural rather than market-share grabs: advanced nodes require more process steps and tighter purity tolerances. In Q1 2026, Entegris said wins in molybdenum deposition, doubled CMP content, IPA purification, and related technologies drove “roughly 4–5 percentage points” of market outgrowth and increased content per wafer. AI wafers are only about 5% of volume but contribute about 30% of revenue, and spend per wafer from 14nm to 2nm has doubled to tripled, while advanced logic accounts for roughly 40% of its revenue.

The honest framing: the TAM is larger than a pure niche and the long runway is clear, tied to AI and advanced processes. But the market itself grows only at a single-digit rate, and Entegris has only about $3.2 billion of revenue while FY2025 still showed −1%. That makes its ceiling a “broad and gradual” one: it slowly expands a moderately growing stock of demand through higher content, rather than creating or defining a brand-new market. Measured against Baillie Gifford’s yardstick of “5× in 10 years, driven by an explosive new market,” this dimension is a clear weakness.

Jun 5, 2026
Can its revenue at least double over the next five years? Will growth be driven mainly by volume, price, or new businesses?4/10

Bottom line: a five-year revenue double is very difficult and probably will not happen. A more realistic range is five-year +40%~60% (about 7%~10% CAGR), not a doubling. And “revenue” must be kept strictly separate from “EPS”: the large EPS recovery the market sees in 2026 does not mean revenue is doubling.

Start with how low the base is. The report shows FY2025 revenue of about $3.197 billion, down −1.4% year over year. Doubling in five years requires roughly a 15% CAGR. That is a hard move from negative growth straight to 15% sustained for five years. Recent conditions point to a mild recovery rather than a breakout: 2026Q1 revenue was +5%, and Q2 2026 guidance was $815–845 million, with the midpoint only about +5% year over year; see Entegris Q1 2026 8-K. Sell-side consensus is similarly modest: FY2026 about $3.45 billion, FY2027 about $3.84 billion, and FY2028 about $4.13 billion. Simply Wall St explicitly gives future three-year revenue growth of about 7.8% per year. Management’s own long-term target is also only a 7–9% CAGR. At an 8% CAGR, the five-year cumulative increase is about +47%, far short of a double. The report’s “outgrow the materials market by 4–5pp” framing points to high-single-digit to low-double-digit growth in an upcycle, which is consistent with this.

One easily misread point must be cleared up: the report discloses consensus adjusted EPS growth of +66.6% for FY2026 and +40.4% for FY2027, but that is one-off earnings leverage from deleveraging, with interest falling from a $301 million peak to about $192 million, plus margin recovery. It is not a revenue double. Simply Wall St directly identifies the gap between “earnings growth of 20.4% per year vs revenue growth of 7.8% per year,” driven by margin expansion rather than revenue.

By driver: growth is mainly volume. About 75% of unit-driven consumables follows wafer output, with added AI content per wafer: AI wafers account for about 5% of volume and about 30% of revenue, while CMP slurry grew 14% in 2024 and pads grew 24%. Price is secondary, with higher ASPs but not the main driver. New businesses are potential upside: advanced packaging and 3D NAND molybdenum deposition are upside options, but the report notes that many content figures come from management commentary and must be validated quarter by quarter through revenue growth. They should not be treated as locked in. In one sentence: a good business with moderate growth; a five-year double is the ceiling of an optimistic scenario, not the base case.

Jun 5, 2026
After five years, what will take over as the next growth engine? Does that “second curve” exist today?4/10

Bottom line: the second curve “exists, but it is fragmented and shares the same roots.” Its existence is moderate. Do not dress up “content extension” as a disruptive new curve. As a multi-product consumables platform, with Materials Solutions covering deposition precursors/CMP/specialty gases/etch and cleaning at $1.407 billion in FY2025, and Advanced Purity Solutions covering filtration and purification/wafer cleaning/FOUP/fluid handling at $1.799 billion, while no single product platform accounts for more than ≤3% of net sales, Entegris has more succession options than a single-product company. But most of those options are existing categories extending content along advanced processes, not a standalone second curve built from scratch.

Look at the candidate engines one by one: ① Molybdenum replacing tungsten in deposition, the most concrete one. In 3D NAND, molybdenum is replacing tungsten. Entegris has won PORs at all major manufacturers, creating cascading content across molybdenum etch chemicals + molybdenum CMP slurry + supporting filtration. Each added layer adds content per wafer. This is a new material transition, but it still lands inside the MS segment’s existing categories. ② Advanced packaging materials. The packaging materials market is about $27.4 billion (SEMI 2025). Entegris started this business from about $100 million in 2024, targeting wafer-level and panel-level packaging with filtration, gas/chemical delivery, specialty coatings, ultra-clean carriers, and hybrid-bonding purification. This is the closest thing to a “new use case,” but the base is still small and multiple product lines remain in sampling and qualification. ③ HBM-related demand. It uses IPA purification to reduce trace metals from 3ppt to <1ppt and improve Korean HBM yield. This is essentially deeper content in the core filtration and purification business at a new node. ④ CHIPS-driven localization of new plants. Colorado Springs received up to $75 million in grants, with about $600 million of investment and about 600 jobs, localizing FOUP and liquid filtration that had previously all been manufactured in Asia. This is capacity/geographic expansion, not a new growth curve.

The honest judgment: the strongest growth engine is still the deepening of the report’s main logic, namely higher content per wafer: CMP slurry doubles from N3 to N2, and AI wafers are about 5% of volume but about 30% of revenue. Advanced packaging is the only candidate with the outline of an independent “second curve,” but FY2025 revenue was still −1%, and much of the above remains management guidance rather than financial results already delivered. The second curve is fragmented but same-rooted, with moderate existence. It must be validated quarter by quarter through revenue growth and should not be treated as a locked-in disruptive new engine.

Jun 5, 2026
What is its core competitive advantage? Will this moat widen or narrow over the next three to five years?5/10

Bottom line: the core advantage is a three-part package of switching costs once certified into fab processes + a full-stack materials portfolio with no single-point dependence + leading POR wins at advanced nodes. Over the next three to five years, 2nm/molybdenum replacing tungsten/advanced packaging should make the moat slightly wider at advanced nodes, but the widening is limited. This is not a Lasertec-style single-point monopoly; it is broad rather than deep, stable but not overpowering.

The hardest part of the moat is switching cost. The report cites FY2025 10-K risk factors: products are specified into customers’ manufacturing processes and customized for customers’ unique conditions. Replacing them is costly, time-consuming, and may endanger yield. This moat is such a double-edged sword that even Entegris changing its own products could trigger customer requalification. The second part is portfolio breadth: no single product platform accounts for more than ≤3% of net sales, there is no single-point dependence, and in CMP it can offer a full certified solution across slurry + pad + post-clean + conditioner. The third is leading advanced-node plan-of-record wins: N2 CMP slurry POR wins are 2 times N5, and in 3D NAND molybdenum replacing tungsten, its high-purity MoO2Cl2 precursor + selective etchant has won PORs at major manufacturers. 2nm ramp, HBM, and advanced packaging should make this part slightly wider at the most advanced nodes.

But it must be discounted honestly: this is not a single-point monopoly. Stronger rivals sit on every product line. In the main CMP battlefield, DuPont spin-off Qnity has an operating margin of 21.6% and net margin of 13.1%, above Entegris at 15.7% and 8.2%, with EV/EBITDA of 24.9 and not more expensive, and is actively attacking CMP slurry/pads. The most direct peer MKS reported FY2026 first-quarter revenue of $1.078B, with +11.5% growth far faster than Entegris (FY25 −1%) and a cheaper forward PE of 26×. In specialty gases, Linde/Air Products/Air Liquide together have >35% share, are 5–11 times Entegris’s size, and have higher margins across the board. Among peers, Entegris sits toward the high end on multiples, the low end on growth, and the low end on margins. Add rising customer concentration, with TSMC at 16% and the top ten at 50%, plus export controls that indirectly incentivize Chinese substitution. The moat is stable, but the “widening” occurs only in a small area around advanced nodes, and the overall magnitude is limited.

Jun 5, 2026
If its core business is disrupted, does it have the DNA to reinvent itself? How does it handle mistakes and bad news?4/10

Bottom line: neutral to positive. There is evidence of reinvention DNA, but it is acquisition-driven rather than internally disruptive; its handling of bad news is relatively candid, which is a positive.

Reinvention DNA: evidenced, but it is “bought reinvention.” Entegris has proven that it can reshape its business mix, but the route has been almost entirely capital allocation rather than internal disruption. Historically, it did not begin as a single hit product. It has kept acquiring adjacent purity/materials categories and getting them certified into fabs. The 2005 merger of Mykrolis and the old Entegris formed the current company. In July 2022, it acquired CMC Materials for about $6.5 billion, bringing CMP slurry/pads and specialty materials into the company in one stroke and turning itself into a global electronic materials leader, with unit-driven revenue rising from about 70% to about 80%. The real evidence of “reinvention discipline” came after the purchase: during the 2023 downturn, carrying about 4.5× net leverage, it decisively divested non-core assets to raise cash and repay debt. In 2023, three divestitures including QED and electronic chemicals, the latter sold to Fujifilm for $700 million on 2023-10-02, contributed about $1.1 billion, and the company repaid $1.3 billion of debt that year. In March 2024, it also sold the PIM business acquired with CMC to SCF for up to $285 million. It kept the CMP assets it wanted, sold lower-strategic-value assets, and continued to deliver deleveraging: Q1 2026 net leverage fell to 3.6×, the year-end target is about 3×, and $400 million of term loans remains to be cleared. This “category expansion + disciplined digestion” proves it can reshape the portfolio through capital allocation. But it is external, not internal reinvention after disruption. And buying CMC near the cycle peak at an expensive price, which left the company constrained by debt and amortization for two years and cut ROIC from 21% to about 5%, is itself the negative case study for its reinvention ability.

Handling bad news: relatively candid, a positive. The report shows that its GAAP-to-non-GAAP bridge is clear, and the company actively explains that profits are distorted by the “two accounting scissors” of intangible amortization + interest. More importantly, it does not hide the low-quality part of good news: when 2026Q1 gross margin of 46.9% exceeded the high end of guidance, management proactively stated that about +50bp came from one-off items + a useful-life accounting change that reduced depreciation, and would not recur to the same extent. Breaking out one-off components during a beat, instead of packaging an accounting benefit as sustainable earnings power, shows decent transparency. Overall: moderate reinvention, driven by acquisitions rather than internally; fairly good candor.

Jun 5, 2026
Does management, especially the founder, have a long-term horizon and deep alignment with the company? Is it willing to sacrifice current profits for five to ten years out?4/10

Bottom line: this is not founder-led and not an owner-operator company. This is Entegris’s weakest area in the Baillie Gifford framework. Alignment and long-termism are weak to moderate: the long-term horizon and capital-allocation discipline are acceptable, but the new team’s dual finance + M&A background leans toward capital operations and further acquisitions. Combined with the prior CMC purchase near the cycle peak, this is “worth watching,” not “safe to entrust blindly.”

1. Founder/owner dimension: absent. Entegris is a large-cap company created through the 2005 merger of two companies. There is no incumbent founder. Bertrand Loy, the former CEO who led the company for 13 years, retired on 2025-08-18 and will transition as executive chairman through the end of 2026Q2; see report §1.4. Successor David Reeder is a professional manager with a finance/capital-markets background: former GlobalFoundries CFO, where he led the largest IPO in semiconductor history, Lexmark CEO, and Chewy CFO. New CFO Sukhi Nagesh took office on 2026-05-18 and has about 30 years of finance and M&A experience. A CEO from finance + M&A paired with a CFO from M&A is a combination that clearly leans toward “capital allocation/further acquisitions,” not product-led long-termism.

2. Alignment: weak. Insider ownership is only about 0.61%, far below the founder-heavy long-term ownership Baillie Gifford tends to prefer. Report §6.2 also states that over the past three months insiders were net sellers with zero buying, about $14 million–$26 million. The signal is negative and far from “deeply aligned interests.”

3. Long-term horizon and “sacrificing current profits for five to ten years out”: acceptable, but not a model case. The positive is capital-allocation discipline: after CMC, the company disciplinedly deleveraged, with net leverage moving from 4.5× to 3.6× and a target of about 3× by the end of 2026; it divested non-core assets to repay debt, deprioritized buybacks in favor of debt reduction, and still maintained R&D and capacity expansion to protect advanced-node certification leadership (report §2.3 and §3.2). This is indeed a long-term posture of “building the foundation before rushing to distribute cash.” But the negatives are also real: CMC was bought at an expensive price near the cycle peak (report §1.3), and management has explicitly delayed buybacks to 2027 before it will “consider other options”. Overall: the long-termism is capital-discipline long-termism, not founder-vision long-termism. After deleveraging is complete, the direction of capital allocation, whether more M&A or buybacks, is the key item to watch.

VERDICT: Not founder-led, not owner-operated. The 13-year former CEO has just retired; the new CEO + CFO team has a dual finance/M&A background; insider ownership is only about 0.61% and insiders have been net sellers. That is far from the deep founder alignment Baillie Gifford prefers. The long-term horizon and capital discipline are acceptable, including deleveraging and sustained R&D, but the company leans toward capital operations/further acquisitions. Together with the prior overpayment for CMC near the cycle peak, alignment and long-termism are weak to moderate: worth watching, not safe to entrust blindly.

Jun 5, 2026
If it disappeared tomorrow, how much would customers miss it? Is its growth model sustainable, and does it avoid relying on social or regulatory harm?5/10

Bottom line: indispensability is high but not absolute. The growth model does not harm society; its sustainability is mainly constrained by three factors: cyclicality, export controls, and customer concentration. Do not exaggerate it into “irreplaceable.”

① If Entegris disappeared tomorrow, how much would customers miss it? Quite a lot, but this is not a shutdown-level single-point monopoly. The reasons customers would miss it are solid. The report states that its products are specified into customer processes and customized to customers’ unique conditions; replacing them is costly, time-consuming, and may endanger yield. That is structural switching cost. Third-party estimates support the magnitude: materials qualification usually takes 6–18 months, and at the 1.4nm node, a 1% yield defect can cost a fab up to $800M per year. So TSMC/Samsung/Intel/Micron rely heavily on it in advanced fabs, and a sudden removal would disrupt yield ramp and production rhythm. But honesty matters: it is not a Lasertec-style sole source. The report states that no single product platform accounts for more than ≤3% of net sales and there is no single-point dependence. Each product line has alternative suppliers: in CMP, there are Fujimi, DuPont spin-off Qnity, Cabot, Versum/Merck, and Chinese local suppliers such as Anji; advanced fabs commonly dual-qualify multiple vendors (Entegris CMP share is about 22%). In specialty gases, alternatives include Air Liquide/Linde/Air Products. Its indispensability is therefore stronger than an ordinary supplier’s but weaker than a single-point monopoly: high switching costs create stickiness and short-term pain, but customers have migration paths that are expensive and slow, not impossible.

② Is the growth model sustainable and free from reliance on social or regulatory harm? It does not harm society, but sustainability has three constraints. The growth model does not harm society: about 75% of revenue is a recurring consumables annuity tied to wafer output, playing the role of purity/yield “gatekeeper” for advanced chips. CHIPS Act localization, as the first semiconductor materials company to finalize a grant agreement, with up to $77M for a Colorado plant producing FOUPs/liquid filtration, is also a tailwind. There is no negative social externality. The constraints sit in three places: cyclicality, with about 25% of revenue tied to fab construction capex and FY25 revenue down −1%; export controls/geopolitics, where restrictions on advanced processes in China affect its advanced materials, and the 10-K cited in the report says controls conversely “incentivize Chinese local companies and local suppliers to cooperate to compete with us.” This is a structural risk vector, partly hedged by localized supply, with China demand localization rising from about 80% to 90%; and customer concentration, with TSMC at about 16% and the top ten at about 50%, so changes in major customers’ fab plans or lost awards are magnified, while high switching costs also mean lost business is hard to replace quickly. In sum: a good business with clean growth, but “indispensable” needs a discount, and sustainability mainly depends on the cycle and geopolitics.

Jun 5, 2026
What are the unit economics of this business, including gross margin and incremental returns? Do they get better or worse with scale? Where does the money it earns go?5/10

Bottom line: the unit economics are good, but not top-tier. A ~47% gross margin + ~75% recurring consumables annuity + very strong cash conversion, with GAAP profits systematically understated by acquisition amortization, are real positives. But operating/net margins are toward the low end of peers, ROIC is still recovering from a trough, 3.6× net leverage has not been cleared, and FCF Yield is only ~1.9%. As scale rises, unit economics should improve through deleveraging + margin recovery, with incremental returns recovering. But this is not Lasertec-style extreme economics with ~49% operating margin and near-zero leverage.

Gross margin and incremental returns: respectable, not extreme. Gross margin returned to 46.9% in 2026Q1, including about +50bp from one-off items + a useful-life accounting change that management explicitly said was not sustainable; see report §2.5. Structurally, ~75% of revenue is a unit-driven consumables annuity that follows wafer output, and no single product platform is more than ≤3% of revenue. That gives annuity-like resilience: in 2025, unit-driven revenue was still +about 2%, while capex-driven revenue was −7%. The issue is depressed margins: GAAP operating margin was 15.7% and net margin 8.2%, clearly below Qnity’s 21.6%/13.1% and also below gas majors (report §5). What is depressed is not the underlying business, but about $184 million of annual intangible amortization + interest from the CMC acquisition. FY2025 adjusted operating margin was actually 21.3%. ROIC was 21% before the acquisition and was cut to just above 5% after CMC; it is slowly recovering with deleveraging and margin repair. This is a good business whose accounting returns were dragged down by the acquisition.

Cash conversion is the real highlight. FY2025 OCF was $695M vs GAAP net income of $236M, about 2.9×. In the latest quarter, Q3 2025 OCF was $249.5M vs net income of $70.5M, also ~3.5×. This confirms that accounting earnings are heavily understated by non-cash amortization. FCF has recovered to about $400 million, with a margin of ~12.4%. 2026Q1 FCF was $144M, or 18% of revenue, and capex is falling back to ~300 million per year as new plants near completion.

Where the money goes: debt reduction is the current priority. Cash flow is clearly being directed first to deleveraging: net leverage is 3.6×, the year-end target is ~3×, term loans have fallen to $400 million, and another $50 million was repaid this quarter. The rest goes to R&D and capex to maintain advanced-process certification leadership. Buybacks are delayed to 2027. For now, shareholders only receive a $0.40/share dividend, with a dividend yield of ~0.29%. FCF Yield ≈ $400 million ÷ $21.3 billion market cap ≈ 1.9%, low and reflective of a growth-premium valuation (report §7.3).

Jun 5, 2026
What conditions must all hold for the stock to rise fivefold over ten years? Are those conditions realistic? What expectations are embedded in today’s share price?3/10

Bottom line: from the current price (~$140), a fivefold rise over ten years is very unrealistic. It requires two conflicting conditions to hold at the same time: earnings must rise about 10 times, far beyond what a ~7–9% revenue CAGR can support, and the high valuation multiple must avoid major compression over ten years. The report itself already frames it as “good business, expensive price, zero margin of safety.” GuruFocus’s GF Value judges it significantly overvalued, and Goldman maintains a Sell rating with a $115 price target below the current price. This is the question where cold water is most necessary.

Write the math explicitly. Fivefold means $140 → ~$700, annualized = 5^(1/10) − 1 ≈ 17.5% per year, uninterrupted for 10 years. Price = earnings × multiple. The current forward PE is ~39–54× (report definitions: non-GAAP rolling ~48×, or ~39× using FY2027 EPS of $3.62), and EV/EBITDA of ~28 sits in the upper-middle part of its 13-year range of 7.9–34.9, with a median of 16.6. Over a 10-year horizon, the multiple is likely to compress. If the multiple falls from ~40× to ~20×, a halving that merely returns it near the historical median, then supporting a 5× share price requires earnings to grow about 10 times (5 ÷ 0.5 = 10). Spread over 10 years, that is 10^(1/10) − 1 ≈ 26% earnings CAGR.

Why this is unrealistic: revenue is only compounding at ~7–9% (report: FY2025 −1%, Q1/Q2 2026 ~+5%, long-term target 7–9%). Over 10 years, 1.08^10 is only about 2.2 times. To squeeze 10 times earnings out of 2.2 times revenue, net margin would need to expand by ≈ 10 ÷ 2.2 ≈ 4.6 times. Entegris margins are already not low enough for that to be plausible: operating margin is 15.7%, and the report says it is below Qnity/gas majors but nowhere near having 4 times improvement room. A structural doubling would already be optimistic; a 4 times expansion is close to impossible. Even if revenue reaches the 9% upper bound and the multiple compresses only mildly, the two routes conflict with each other: either depend on no multiple compression, which contradicts the high starting point relative to history, or depend on a margin miracle, which contradicts peer comparisons.

What today’s price embeds: not a “cheap compounding machine,” but “AI content growth + deleveraging-driven EPS recovery, including FY2026 adjusted EPS +66%, priced quite fully in advance.” In other words, the market is already pricing a favorable combination of high-single-digit revenue growth + one-off EPS repair + sustained growth-premium multiples. The report’s neutral scenario implies value of $115–145, and the current price is already near the upper end; conservative-scenario downside of −21% to −49% is much larger than remaining upside. A 10-year fivefold rise is the optimistic tail of an optimistic scenario, requiring many consecutive years of “growth does not step down + multiple does not compress.” Probability is low.

Jun 5, 2026
Why has the market not recognized all of this yet? Does it fail to understand, look down on it, or fail to look far enough ahead? What will be the “narrative inflection point”?3/10

Bottom line, stated plainly: this Baillie Gifford question assumes “the market has not yet recognized an undervalued great growth stock,” but Entegris is the counterexample. “The market does not understand / looks down on it / does not look far enough ahead” is mostly unconvincing. The market has already priced AI content growth + deleveraging recovery quite fully, and arguably a little too hot. Its difference from a pure monopoly growth stock is large bull-bear dispersion, but the dispute is about “price,” not about “whether this is a good business.”

Why “the market has not recognized it” does not hold: this is not a neglected stock. The share price nearly doubled over the past year and reached an all-time closing high of $156.33 on 2026-04-27 (report §Research Summary/Macrotrends). The current price is about $140, close to historical highs. On valuation, EV/EBITDA≈28 is in the upper-middle of the company’s 13-year range, whose median is 16.6; forward PE is 39–54× depending on definition; and trailing PE is as high as 75–80×. Report §6.1 directly says “the market has priced future recovery + AI content quite fully in advance.” This is not “failure to look far enough ahead”: the logic that AI wafers are 5% of volume but 30% of revenue, plus the 2026 adjusted EPS recovery of +66%, is already in the price.

Disprove the three Baillie Gifford prompts one by one: “Does the market fail to understand?” A consumables annuity + switching costs from being certified into fabs is a textbook simple business, and sell-side coverage is sufficient (4 major investment banks + consensus target around $162, range $115–205). “Does the market look down on it?” The market is giving it peer-high multiples (report §5: EV/EBITDA and EV/Sales are both at the high end of the semiconductor materials group). That is reward, not neglect. “Does the market fail to look far enough ahead?” The AI supercycle + content-per-wafer compounding story is the mainstream bull case; nobody has missed it.

What is truly distinctive: dispersion, not neglect. The sell side is split: Goldman maintains Sell with a $115 target vs UBS at $205, with Citi/Needham in the middle. But report §Research Summary makes the point: bulls and bears agree “it is a good business”; the dispute is only price. That is different from “the market has not understood a good business.”

Narrative inflection points, both directions, with downside more realistic for today’s buyer: upside = deleveraging completion followed by delivered margin/EPS recovery (2026 adjusted EPS +66%), gross margin holding above 45%+ after stripping out one-offs, buybacks restarting in 2027, 2nm/HBM content translating into revenue quarter by quarter, and possible further M&A. Downside = AI capex growth shifting down, revenue recovery missing expectations (FY2025 was actually −1%, and Q1 gross margin included one-offs), valuation multiples normalizing from upper-middle levels toward historical medians (~16–20×), and new management doing another high-priced acquisition that releveres the balance sheet. Report §10.2 estimates that “growth disappointment + valuation normalization” together could produce about a halving (−49%), with downside (−21%~−49%) materially larger than remaining upside. For a buyer today, the Baillie Gifford idea of “waiting for the market to wake up” does not apply; the more realistic inflection point is downward valuation reversion.

Jun 5, 2026

Ask about this report

Members can ask about this report; once answered it appears under "Reader Q&A" on this page. You can also highlight a passage in the text to ask about it directly.