Entegris ENTG

At a glance

One-line thesis. Entegris sells the filters, high-purity chemistries and wafer-handling systems that every advanced fab consumes with every wafer it runs — a genuinely sticky, unit-driven franchise now riding the AI materials-intensity story — but revenue actually shrank in FY24 and FY25, forward growth is only ~8–12%, net-debt/EBITDA is still 3.9× from the CMC Materials deal, and at ~40× FY26E EPS after a +68% year the recovery is already in the price, so this is a Watch: own the business on a pullback toward ~$125, don't chase it at $145.

1. What it is

Entegris (Nasdaq: ENTG) develops, manufactures and supplies critical solutions for microcontamination control, specialty chemicals and advanced material handling for semiconductor fabrication: liquid/gas filtration and purification, ultra-pure process chemistries and gases, CMP slurries and pads (via the 2022 CMC Materials acquisition), and wafer/reticle carriers and fluid-handling systems. Customers include logic and memory chipmakers, semicap equipment builders, chemical/gas producers and wafer growers, with smaller exposure to displays, life sciences and other high-tech end markets. Founded 1966, headquartered in Billerica, MA; CEO David W. Reeder; ~8,200 employees; fiscal year ends December.

Revenue mix (from filings):

• By segment (FY25): the company now reports two segments — Materials Solutions $1.41B (44%) and Advanced Purity Solutions $1.80B (56%) of $3.20B total. (Pre-2024 it reported three segments — SCEM/MC/AMH — so long histories don't splice cleanly.)
• By geography: the FY25 breakdown in our data is partial (China $658M ≈ 21%, South Korea $429M, Japan $318M, SE Asia $256M, Europe $239M — North America and Taiwan rows are missing from the pull). The complete FY24 split: China 20.7% · North America 20.7% · Taiwan 20.4% · South Korea 12.9% · Japan 9.5% · Europe 8.5% · SE Asia 7.2%. Read it as: ~80% Asia/international, with China + Taiwan ~41% — a genuine geopolitical exposure (§11).

The structural story: as nodes shrink and stack (GAA transistors, advanced packaging, more layers), materials intensity per wafer rises — more filtration steps, purer chemistries, more handling — so Entegris' content grows even when unit volumes are flat. That is the thesis; the counterweight is that the reported numbers haven't shown it yet (revenue −8.0% FY24, −1.4% FY25).

2. The expert thesis — why the panel is bullish (traceable)

The Synthos KB holds 6 traceable claims on ENTG — but honest breadth is one net-bullish external voice: Jordi Visser (our highest-skill voice, selection skill 2.0), tracked across three variants, plus half-weighted management guidance. This is thin coverage, so this note is predominantly fundamentals-driven, with Visser as the conviction spark:

• The Visser buy (fresh, basket-level). "Bought chemical/semi-materials names this week; they can double over the next 12 months and have less downside right now than semiconductor stocks" (jordi_visser-Ov2QzUTbhc4:ad846bb44f, conviction 76, 2026-06-04). His tracked variants reiterate it: Entegris named explicitly as part of an "AI-thematic chemicals basket" (jordi_visser_m-Ov2QzUTbhc4:c95fad330a, conviction 70; jordi_visser_ai-Ov2QzUTbhc4:8cc6320cba, conviction 75). Honest weighting: this is one person expressed three ways, the thesis is basket-level ("chemical/semi-materials names") rather than an ENTG-specific dossier, and the "can double" framing is aggressive against a Street high of $205. What's genuinely useful: it was struck ~2026-06-04 near ~$143 — essentially today's price — so unlike many momentum names, the thesis is fresh, not consumed.
• Management's own guidance (half-weighted, ENTG_mgmt, skill 0.5): guides Q2 2026 net sales to $815–845M (midpoint ~$830M), a sequential increase off Q1's $811.9M seasonal trough (ENTG-earnings-2026Q2:00b074cfbe); non-GAAP EPS $0.76–0.84 with the midpoint below Q1's $0.86 actual (ENTG-earnings-2026Q2:71cdc83096, neutral); adjusted EBITDA margin roughly flat at 27.0–28.0% (ENTG-earnings-2026Q2:b40f433117, neutral). Corroborates a grinding recovery, not an inflection.

Honest composite note. One high-skill bull with a fresh position is a real signal, but it is not a panel. There is no independent cautionary voice in the KB either — the bear case in §3 is built from the fundamentals (leverage, valuation, decelerating estimates), not from a countervailing expert.

3. Synthos scores & the Bull / Base / Bear cases

The one-glance judgment — three scores, 0–10, each anchored to real metrics (not probabilities we can't honestly calibrate):

The three cases (our own scenario model — assumptions shown; each target is a ~12–18-month fair value). We deliberately do not attach probabilities: the base case is the expected path, so a weighted blend would just restate it with false precision.

Synthos fair value = the base case, ~$165 (+14%) — deliberately anchored on the Street's $165.43 consensus (25 rated analysts) and cross-checked as ~35× FY27E EPS, with the full $110–$210 span as the honest range. Note the range brackets the Street's own $115–$205 target band almost exactly: this is a name where our fundamentals work and the consensus agree, and the disagreement is only about entry price. A +14% base-case gap is not enough margin for a 3.9×-levered cyclical at 40× forward — hence Watch, not Buy. This is a tracked call — the Forecaster Scorecard grades it once it matures.

4. Exponential Potential

Synthos separates compounders (durable returns on captive demand) from exponentials (accelerating, multi-baggers-from-here). ENTG is firmly the former:

• Forward growth: consensus revenue $3.45B (2026E) → $4.56B (2030E), a ~7.2% CAGR; EPS $3.64 (2026E) → $6.06 (2030E), ~13.6% CAGR, with the EPS kicker coming from margin recovery and interest expense rolling off, not from top-line lift.
• Acceleration (the 2nd derivative) is negative: revenue growth +8.0% (2026E) → +12.1% (2027E) → +7.4% (2028E) → +0.9% (2029E) → +8.7% (2030E). The best year is 2027, then it fades. Coverage also thins badly in the outer years (2029/2030 EPS rest on a single analyst; 2028 net-income estimates span $548M–$1.21B) — treat the out-year figures as sketches, not forecasts.
• Room to run: $22.1B cap against a semiconductor-materials market that grows with wafer starts plus materials intensity — real, but a mid-single-digit-growth TAM. A double requires multiple expansion on top of estimate beats, which is exactly what Visser's basket thesis asserts and what the decelerating consensus does not support.
• Reinvestment runway: capex has normalized to $299M (9.4% of revenue, down from ~14% in the 2022–23 build-out), freeing FCF — but the marginal dollar is going to debt paydown ($300M in FY25), not growth. Right for the balance sheet, not exponential fuel.

Exponential Potential: Low (3/10). The honest read: this is a moat story and a cycle-recovery story, not an acceleration story. It can compound; it is very unlikely to multiply from $145 without an estimate regime-change.

5. Financials (real numbers — FMP annual/quarterly)

• Revenue: FY25 $3.20B, −1.4% (FY24 $3.24B, −8.0%; FY23 $3.52B; FY22 $3.28B, +42.8% on the CMC Materials acquisition; FY21 $2.30B). Two consecutive down years — the downturn is real and recent.
• Quarterly trajectory (the turn): Q1'25 $773.2M → Q2 $792.4M → Q3 $807.1M → Q4 $823.9M → Q1'26 $811.9M (+5.0% YoY) — the first positive YoY comp in the window, with Q2'26 guided to $815–845M. A grind higher, sequential and shallow.
• Margins: gross 43.2% TTM; EBIT margin 14.6% TTM; adjusted EBITDA margin guided 27–28%. FY25 GAAP figures need care: the reported FY25 "operating income" of $923.5M includes a ~$469M non-operating item booked in Q4'25 (Q4 also shows a −$334M SG&A line — offsetting artifacts in the feed); the cleaner FY25 read is EBIT $454.4M (~14.2% margin) and EBITDA $844.1M.
• Earnings: FY25 GAAP net income $235.6M / diluted EPS $1.55 (FY24 $1.93; FY23 $1.20). Trailing non-GAAP EPS — the basis the Street prices — is ~$2.94 (Q2'25 $0.66 + Q3'25 $0.72 + Q4'25 $0.70 + Q1'26 $0.86). The GAAP/non-GAAP gap is mostly acquisition-intangible amortization and interest.
• Cash flow (the strong suit): FY25 operating CF $695.4M, capex −$299.2M, FCF $396.2M — income quality 1.64 (cash comfortably above accrual income). FCF yield ~3.1% at this price.
• Balance sheet (the weak suit): total debt $3.89B, cash $360M, net debt $3.53B → net-debt/EBITDA 3.9×, interest coverage 4.9×, interest paid $185M FY25. Deleveraging is steady — total debt $4.66B (2023) → $4.07B (2024) → $3.89B (2025) — but goodwill + intangibles are $4.85B of $8.35B total assets (58%) and tangible book per share is negative (−$4.98). This is a post-LBO-shaped balance sheet on a cyclical business.
• Earnings-beat cadence: four of the last five prints met or beat non-GAAP estimates (Q1'26 $0.86 vs $0.745 est; Q4'25 $0.70 vs $0.667; Q3'25 in line; Q2'25 $0.66 vs $0.645; Q1'25 slight miss).

6. Valuation — priced in or room?

The multiple does the arguing: 84× trailing GAAP EPS (~49× trailing non-GAAP), 30.1× EV/EBITDA, 7.9× EV/sales, 5.5× book — against ~8–12% forward revenue growth. FMP's letter rating is C+ (overall 2/5; P/E score 1/5, debt/equity score 1/5). On forward consensus the P/E compresses to ~40× (2026E $3.64) → ~31× (2027E $4.66) → ~27× (2028E $5.35) → ~24× (2030E $6.06) — but unlike a true exponential, it never gets cheap: even four years out on thin single-analyst estimates you're paying a market-plus multiple. A reverse read: $145 today requires the 2027 EPS recovery (+28%) to land and the market to keep paying ~31× for high-single-digit growth thereafter, all on 3.9× leverage. Street targets (context and, here, our anchor): consensus $165.43, median $160, high $205 / low $115 — a moderate-dispersion band whose midpoint sits +14% above spot. That is a fine business at a full price: the valuation is the reason for the Watch verdict, not the franchise.

7. Technicals (from the tech block)

• Trend: intact but cooling. $145.17 sits just below the 50-DMA ($147.86) and far above the 200-DMA ($115.53); 50 well above 200 (golden-cross posture). MACD +4.7 (positive but fading versus the spring run).
• Location: −21.1% off the 52-week high ($184) — that is also the max drawdown from peak — and +111% off the 52-week low ($68.8). A big recovery leg that has given back a fifth.
• Momentum: RSI(14) 48 — dead neutral; the overbought condition of the spring has fully reset.
• Relative strength: +67.9% 12-mo vs SPY +21.1% / QQQ +31.2%; +24.7% 3-mo vs SPY +14.6% / QQQ +23.6% — long-lens leadership, but the 3-month edge over QQQ has narrowed to roughly nothing.
• Read: a leadership name consolidating: below the 50-DMA on a neutral RSI is neither breakout nor breakdown. For a Watch name this is actually useful — the setup that would trigger interest is either a base near $125–130 (prior consolidation, ~27–28× FY27E) or a reclaim of the 50-DMA on an estimate-raising print.

8. Moat & competitive position

The moat is real and of the best kind: consumables + qualification lock-in. Filters, purifiers, high-purity chemistries and CMP slurries are consumed per wafer, cost a trivial fraction of a fab's output value, and are qualified into process flows that chipmakers will not touch mid-node — switching risk is asymmetric (tiny savings, catastrophic contamination downside). Materials intensity per wafer rises with each node transition, giving Entegris content growth on top of unit growth. Limits of the moat: ROIC of 11.4% is good, not elite — the CMC deal's goodwill dilutes returns; Japanese incumbents (high-purity chemicals, filtration) and in-region Chinese substitution efforts are persistent competitive and policy threats; and none of the moat shields it from the wafer-start cycle, as FY24–25 proved.

Peer set (FMP-supplied, market cap): the supplied list is largely unusable as comps — Tower Semi $24.5B, F5 $23.6B, Amkor $17.3B, Kaspi.kz $17.1B, Nova $14.8B, Zebra $12.9B, Rambus $12.2B, Bentley $9.5B, Skyworks $9.3B, DocuSign $9.0B — a grab-bag of mixed tech. The relevant comparators (semiconductor materials & consumables: Shin-Etsu, FUJIFILM electronic materials, JSR, MKS, Element Solutions) are not in the supplied set — a data caveat: judge ENTG against the materials cohort, where 30× EV/EBITDA is the rich end.

9. Management, capital allocation & guidance

• Capital allocation: the playbook is deleverage first — $300M of debt retired in FY25 (after $624M in FY24 and $1.26B in FY23), a small dividend ($60.8M, $0.40/sh, ~0.28% yield, 23% payout) and no buybacks. Given 3.9× net leverage that is the right priority, and capex discipline (9.4% of revenue, down from the 2022–23 build-out) is visible in the FCF.
• Insider activity: no open-market buys in the recent file. Recent Form 4s: division president Clinton Haris exercised options at $98.11 and sold 6,848 shares at $149.23 (2026-05-27); director James Lederer sold 3,569 shares at $143.59 (2026-06-03); new CFO Sukhi Nagesh filed a Form 3 and received an 8,254-share RSU award (2026-05-18) — note the CFO seat changed hands this spring, worth watching through the next two prints. The sales look like routine option-cycle diversification, but the pattern is sell-only.
• Management guidance (half-weighted, they talk their book): Q2 2026 sales $815–845M, non-GAAP EPS $0.76–0.84 (midpoint below Q1's $0.86), adjusted EBITDA margin 27–28% flat (ENTG-earnings-2026Q2:00b074cfbe, :71cdc83096, :b40f433117). Ingested from the 8-K release; the full analyst Q&A is not on our FMP plan (same gap flagged in prior notes).

10. Catalysts & what to watch

• Next earnings: 2026-07-29 (Q2 2026; Street EPS $0.81, revenue ~$833M vs guide $815–845M). The key lines: whether YoY growth accelerates past Q1's +5%, and whether adjusted EBITDA margin holds the 27–28% band.
• Leading-edge ramp signals: GAA-node and advanced-packaging materials intensity showing up as content-per-wafer commentary — the mechanism that would break the consensus deceleration and re-rate the stock.
• Deleveraging milestones: net-debt/EBITDA sustainably under ~3.5× (and a path to 3×) de-risks the equity and is the cheapest re-rating catalyst available.
• China / export controls: ~21% of revenue ships to China; any tightening of materials-adjacent export rules (or accelerated domestic substitution) hits both the numbers and the multiple.
• Estimate revisions: 2027E EPS (~$4.66) is the fulcrum year — upgrades there move our base case mechanically.

Thesis tripwires (what would flip Watch → Buy): price into ~$125–130 with the story intact (~27× FY27E); or an estimate-raising print that shows YoY growth compounding past +8–10% with margin expansion. (What would flip Watch → Avoid): a guide-down below ~$800M/quarter, adjusted EBITDA margin below ~26%, or leverage stalling above 4× into a downcycle.

11. Key risks

• Valuation / de-rating (the dominant risk): ~40× forward EPS and 30× EV/EBITDA for high-single-digit growth leaves no cushion; if 2027's +28% EPS recovery slips, both numerator and multiple compress.
• Leverage: net-debt/EBITDA 3.9× with 4.9× interest coverage on a cyclical business — the balance sheet turns an earnings wobble into an equity event. Tangible book is negative.
• Cyclicality: FY24–25 revenue declines are the proof; consensus itself pencils 2029E growth at just +0.9%.
• China / geopolitics: ~21% of revenue from China (plus Taiwan ~20% in FY24) → export-control and domestic-substitution risk on both demand and policy fronts.
• Estimate fragility: out-year coverage is thin (single-analyst 2029/2030 EPS; 2028 net-income spread of $548M–$1.2B) — the "cheap on 2030" argument rests on numbers barely anyone has modeled.
• Data caveats: FY25 Q4 GAAP lines carry offsetting artifacts (a ~$469M non-operating item and a negative SG&A line); FY25 geographic segmentation is incomplete in our feed (no North America/Taiwan rows); the FMP peer list is not a usable comp set.
• Single-voice conviction: the only external bull is one (high-skill) voice with a basket-level thesis — no independent corroboration, no cautionary short thesis to test it against.

12. Verdict, position sizing & monitoring

Watch. Entegris is exactly the kind of business the flagship likes to own — per-wafer consumables, qualification lock-in, rising materials intensity, cash conversion of 1.64, disciplined deleveraging — and the one expert signal we have (Jordi Visser, skill 2.0, bought ~2026-06-04 near $143) is fresh rather than consumed. But the entry math fails: ~$165 base-case fair value gives only +14% against a name at ~40× FY26E EPS with 3.9× net leverage, decelerating consensus growth, and a +68% trailing-year run already in the price action. We want the business, not this price.

• Sizing: none today. If the trigger hits — ~$125–130 (roughly 27–28× FY27E, near prior consolidation) or a genuine estimate-raising print — enter as a satellite, ~1–2% of the flagship, sized for a levered cyclical, not a core anchor.
• Monitoring: Q2 print 2026-07-29 (guide vs $833M/$0.81); §10 tripwires each quarter; formal re-score on any ±15% move or estimate revision cycle. This verdict is logged as a tracked Synthos call as of 2026-07-06 at $145.17.
• Single biggest risk: a full multiple on a levered cyclical — if fab utilization or China policy breaks the 2027 recovery, the downside case (~$110) is a −24% air-pocket with no valuation floor above it.

Provenance & disclosures

• Traceability: 6 KB claims, breadth 1 net-bullish external voice (Jordi Visser across 3 tracked variants, top skill 2.0) plus half-weighted ENTG_mgmt guidance, last claim 2026-06-04 — all reconciled to real claim_ids (cited inline). Fabricated conviction is structurally impossible (claim-ID reconciliation). The KB reports per-voice convictions but not a signed net aggregate, so kb_net_conviction is left null rather than invented.
• Data as-of: fundamentals 2026-03-28 (Q1 2026) · estimates & prices 2026-07-06 · expert claims through 2026-06-04. Forward figures are analyst consensus (FMP), labeled as estimates; 2029–2030 EPS rest on a single analyst.
• Valuation anchor: the base case is anchored on the Street's $165.43 consensus target and cross-checked as ~35× FY27E EPS; bull/bear multiples (32× / 30×) and EPS paths are our own labeled assumptions, not consensus.
• Earnings-quality caveat: FY25 GAAP operating income includes a ~$469M Q4 non-operating item with an offsetting negative SG&A line (feed artifacts); underwrite EBIT ($454M) and the non-GAAP/cash figures.
• Coverage caveats: the FY25 geographic split in our feed is incomplete (North America and Taiwan rows missing — FY24 used for the full picture); the FMP-supplied peer list omits the relevant semiconductor-materials comps; full earnings-call Q&A is not on our FMP plan.
• Management caveat: ENTG_mgmt guidance is management's own book, half-weighted by design.
• Not investment advice. Independent research, educational and informational only, never personalized. Hypothetical/forward figures are labeled; the only performance numbers Synthos will headline are the live, real-money Flagship's.
• Version: 2026-07-06. Prior versions available via the deep-dive version dropdown ("based on the info at the time").