Entegris

Phase A — Understand the business

Company Overview

Entegris is the purity-and-materials toll-taker of semiconductor manufacturing: it sells the consumables, filters, specialty chemistries, and contamination-control hardware that get used up every time a wafer is run, not the big lithography/etch tools. It frames itself as "a leading supplier of critical advanced materials and process solutions" with "the industry's most comprehensive electronic materials portfolio".

Two operating segments:

• Materials Solutions (MS) — CVD/ALD deposition materials, CMP slurries & pads, ion-implant specialty gases, formulated etch/clean chemistries, advanced materials. FY2025 net sales $1,406.7M.
• Advanced Purity Solutions (APS) — liquid/gas filtration & purification, FOUPs/wafer carriers, EUV reticle pods, high-purity fluid handling (valves/fittings/tubing), drums/packaging. FY2025 net sales $1,799.1M.

The economic quality of the model is in one disclosed line: "In 2025, our revenue was predominantly unit-driven or recurring from products consumed during semiconductor manufacturing… more impacted by global semiconductor demand and GDP growth than by semiconductor capital equipment sales, which have historically been more cyclical". Translation: ENTG is levered to wafers produced, not fabs built — a smoother, more annuity-like exposure than the WFE names (AMAT/LRCX/KLA).

Contract structure — the one weak spot. Supply agreements "typically do not include long-term purchase commitments"; customers work off "non-binding order forecasts" and "may cancel orders, adjust quantities, or delay". So the recurring-consumable revenue is behaviourally sticky (designed-in, qualified, switching-cost-protected) but not contractually take-or-pay. The only take-or-pay is on the buy side (supply purchase obligations of $150.5M).

Customer concentration: TSMC = 16% of net sales (FY2025 and FY2024; up from 11% in 2023); top-ten customers = 50% (48% FY24, 43% FY23). The book is 82% international / 18% U.S..

Supply Chain

Upstream inputs → Entegris → end customer, named at every link:

• Upstream (suppliers / single-source risk): filtration membranes and polymer resins (APS) and engineered abrasive particles, specialty/commodity chemicals, and petroleum coke (MS) are "obtained from a single source, a limited group of suppliers, or from suppliers in a single country," including raw materials from China. ENTG signs multi-year supply agreements for assurance/cost ($150.5M supply purchase obligations).
• Entegris (conversion): global manufacturing in the U.S., Canada, China, Japan, Malaysia, Singapore, South Korea, Taiwan; Class 100–10,000 cleanrooms; engineered-polymer conversion, membrane casting, chemical synthesis/blending/purification, gas transfilling, graphite synthesis. New regional capacity: Kaohsiung Science Park (Taiwan), Colorado Springs (U.S., up to $77.0M CHIPS Act funding), Korea Technology Center — explicitly to sit near customers' new fabs.
• Direct customers: logic & memory fabs (TSMC the named 16% account), plus Samsung/SK Hynix/Intel/Micron-type fabs implied by geography (Taiwan 26%, China 18%, Korea 14%, Japan 11% of Q1-26 sales); also semiconductor equipment makers (ENTG components go into the tools), gas/chemical suppliers, wafer growers, OSATs, flat-panel & HDD makers.
• End demand: AI logic, HBM/DRAM, smartphones, autos, IoT — the wafers ENTG's consumables ran through.

Chokepoints: (1) ENTG's own single-source inputs (membranes, abrasives, petroleum coke, China-sourced raws) — a supply break hits its cost/availability; (2) ENTG is itself a chokepoint for its customers (qualified-in, purity-critical consumables — see Lens 3). The supply chain is the chip industry's, with China dual-exposure (sells into China and sources some raws from China) — the live tariff/export-control fault line (Lens 10/13).

Competitive Advantages (moats)

The moat is real and unusually clean for a materials company: specification lock-in + purity criticality + breadth.

• Switching costs / designed-in: "Our solutions are increasingly specified into our customers' manufacturing processes and tailored to their unique process conditions… Switching away from our products may be costly and time-consuming for our customers and may introduce risk to their manufacturing yields". A CMP slurry or implant-gas mix that's been qualified into a node is not re-tendered on price — re-qualification risks yield on a multi-week, multi-hundred-step wafer.
• Breadth as a moat: "no single product platform represented more than 3% of our net sales" in 2025 and "there are no global competitors that compete with us across the full range of our product offerings". ENTG sells across deposition, CMP, etch/clean, transport, and chemical-handling and co-optimizes them — rivals are point players.
• IP estate: ~4,400 active patents (~850 U.S.) + ~2,400 pending — though management says no single patent is material.
• R&D + customer co-development: ~1,600 ER&D employees; collaborations with imec, CEA-LETI, Fraunhofer, MIT, Stanford, Yale. Early design-stage collaboration is the front-end of the lock-in.
• Secular content tailwind: every node transition (FinFET → GAA, 2nm, molybdenum metallization, HBM/3D stacking, High-NA EUV) adds process steps and tightens purity → more ENTG content per wafer. As leading fabs adopt molybdenum, "Entegris is uniquely positioned to support this transition" across precursors, deposition, etch, CMP, contamination control.

Bargaining power: strong vs. most customers on qualified consumables (they need ENTG more than ENTG needs any one of them — except the 16% TSMC account, where power is balanced); moderate vs. single-source suppliers (the genuine vulnerability). Notable competitors: Pall (Danaher), Shin-Etsu Polymer, Cobetter, Gudeng (APS); EMD/Merck KGaA, Qnity Electronics, Air Liquide, Linde, Anji Microelectronics, Mersen (MS). Anji Microelectronics (China CMP-slurry national champion) is the most strategically dangerous — domestic-substitution policy + a Chinese end-market = the clearest moat-erosion vector (Lens 13).

Segments

Revenue and segment profit, three-year, directly from filings:

Geographic mix (FY2025): Taiwan 23%, China 21%, North America 18%, South Korea 13%, Japan 10%, Southeast Asia 8%, Europe 7%.

Trend read:

• MS looks flat 2024→2025 ($1,400→$1,407M) but the 2023→2024 drop ($1,689→$1,400M, −17%) is largely the EC divestiture (sold to Fujifilm Oct-2023) plus the memory downturn; organic MS is now re-accelerating — Q1-26 MS +3% YoY on advanced deposition (moly), selective etch, CMP.
• APS is the profit engine but its segment margin compressed hard: FY2023 28.8% → FY2024 26.8% → FY2025 23.7% ($426.4M / $1,799.1M), on "lower sales, unfavorable plant performance, higher depreciation, and $21.9M restructuring". The Q1-26 turn is dramatic: APS profit +24% YoY to $133.6M on +7% sales — but that swing is partly the depreciation/useful-life change (Lens 5/10), not pure operating leverage.
• Acceleration is genuine and Taiwan-led (Q1-26 Taiwan +18% YoY; Japan +28%; Korea +8% on HBM/DRAM). Deceleration risk is China (Q1-26 China −4% YoY).

---

Phase B — Measure performance

Earnings Result (latest print — Q1 FY2026, quarter ended 2026-03-28)

Beat on the quarter, soft-ish guide, stock fell on a sector headline.

GAAP, from the filing:

• Net sales $811.9M, +5.0% YoY (from $773.2M) — above midpoint of guide.
• Gross profit $380.8M; GM 46.9%, +0.8pt YoY — the highest in years.
• Operating income $141.6M, +15.8% YoY; GAAP net income $92.0M / $0.60 diluted EPS (vs $62.9M / $0.41).
• Non-GAAP: EPS $0.86 (vs $0.67, +28%), Adj operating income $192.0M (margin 23.6%), Adj EBITDA $226.1M (27.8% margin).

vs. consensus: non-GAAP EPS $0.86 beat the ~$0.75 consensus (+~15%); revenue $811.9M edged the ~$808.7M estimate.

Lines that drove it: APS +7% (liquid/gas filtration, FOUPs); MS +3% (advanced deposition/moly, selective etch, CMP). Advanced-packaging run-rate now $100M+.

Two flags vs. the company's own history (both in the filing):

1. The margin/EPS beat is partly an accounting-estimate change. In Jan-2026 ENTG extended PP&E useful lives, cutting FY2026 depreciation by ~$73.0M (≈$52M of it lifts gross margin). Q1 total depreciation fell $49.9M → $34.1M YoY (−$15.8M); APS depreciation alone fell $27.5M → $19.9M. A material slice of the +0.8pt GM and the +24% APS-profit swing is this non-cash estimate change, not demand.
2. Effective tax rate collapsed to 1.1% (vs 11.5% PY) on OBBBA + an IRS-guidance UTB release. That flatters GAAP EPS and is not a sustainable run-rate.

Balance sheet / cash: cash $442.7M (up from $360.4M); receivables jumped $458.7M → $529.5M (collections timing, watch DSO); inventory flat at $644.4M; total debt down to $3,651.2M; Q1 OCF $183.0M (+30% YoY), capex just $41.5M (vs $108.0M), → FCF ~$141.5M. CapexCapital expenditureMoney spent on long-lived things — buildings, machines, servers — rather than on running costs. is moderating hard (FY26 guide ~$250M vs ~$304M FY25).

Guidance (Q2 FY2026): revenue $815–845M, GM 46.25–47.25%, non-GAAP EPS $0.76–0.84. Full-year framing: "mid-to-high single-digit semiconductor MSI (market-share-index) growth" with an improved DRAM outlook.

Market reaction: despite the beat, shares fell post-print — guidance was read as only in-line and an SK Hynix HBM-expansion-slowdown report hit the whole AI-chip complex that day. Tells you what the tape cares about (Lens 8): the AI/HBM narrative, not the in-quarter beat.

Earnings Calls (sentiment trend)

No transcripts on the research-layer shelf (transcripts/ empty) — this lens is ``, from call coverage; flagged for backfill.

Tone has shifted from "navigating a trough" (2024) to "constructive and improving" (2026):

• Q4-2024 / early 2025: defensive — guidance baked in a $30–40M China-restriction revenue hit and tariffs threatened "up to $50M" off a quarter; "2+ year downturn and sluggish recovery in most end-markets ex-accelerated-compute".
• Q4-2025 (Feb-2026): revenue −3% YoY but stock rose on the outlook.
• Q1-2026 (new CEO Reeder's first full quarter): "a solid start to the year… constructive and improving semiconductor industry environment"; recurring themes = content-per-wafer, moly deposition in NAND, 2nm/GAA ramp, HBM/DRAM, advanced packaging $100M+ run-rate, mid-to-high-single-digit MSI growth.

What they stopped saying: the 2024 language of "inventory digestion" and "trough." What they now lean on: "market-share-index growth" and "content per wafer" — a cleaner, structural framing. Sentiment trend = clearly improving, but note a new CEO has an incentive to frame the hand-off constructively.

Comps

ENTG own figures + derived. Peer multiples ``, dated; n/a where I could not source a clean current figure (never fabricated).

ENTG `` arithmetic: EVEnterprise valueWhat it would cost to buy the whole company: its market value plus its debt, minus the cash you would get with it. Often very different from market cap. ≈ $26.63B mkt cap + net debt ($3,651.2M debt − $442.7M cash = $3,208.5M) ≈ $29.84B; EV/Sales = 29.84 / 3.24 (TTM sales) = 9.2x; EV/EBITDA = 29.84 / 0.886 (FY25 Adj EBITDA) = 33.7x. GAAP ROE = $235.6M / $3,953.4M equity = 6.0%; non-GAAP ROE ≈ ($2.75 × 152.0M = $418M) / $3,953.4M = 10.6%.

Read: ENTG's ~45x forward P/E sits above AMAT's ~33x 5-yr-avg and far above MKSI's ~17x NTM EV/EBITDA. The richest-multiple framing is partly justified (ENTG is consumable/recurring, less cyclical than WFE) — but it is not growing faster than AMAT in 2026 (AMAT guiding >20% WFE growth; ENTG guiding mid-to-high-single-digit MSI). ENTG is being valued like a secular-compounding consumables franchise while it is still posting flat-to-down revenue and carrying 3.6x net leverage. MKSI is the closest structural comp (semi materials/instruments + post-deal leverage) and trades at roughly half ENTG's EV/EBITDA — a deserved-quality premium, but a wide one.

Stock-Price Catalysts (what moves the tape)

Over the last ~2 years the >5% moves cluster on three things:

• China / export-control / tariff headlines — e.g. ENTG −9.7% on a single day in a sector-wide rout on US/China export fears; guidance repeatedly carried explicit China-hit call-outs.
• The AI/HBM narrative (often via peers, not ENTG's own print) — shares fell after a Q1-26 beat because SK Hynix was reported slowing HBM expansion. ENTG trades as an AI-memory derivative.
• The cycle turn — the 52-week range is $67.97 → $186.94, a ~2.5x trough-to-peak swing as the market priced the 2024 trough → 2026 recovery; all-time-high close ~$184 on 2026-06-22.

Pattern: the market reacts to (a) China/geopolitics and (b) the AI-memory demand signal far more than to ENTG's in-quarter beat/miss. This is a high-beta proxy on the leading-edge + HBM capex narrative, with a China-policy tail.

---

Phase C — Judge people & books

Management

A fresh CEO transition is the single most important people-fact, and it's recent.

• Bertrand Loy retired as President & CEO effective Aug 18, 2025 after 13 years as CEO. He oversaw the ATMI (2014) and CMC Materials (2022, ~$6.5B EV) acquisitions and the portfolio-pruning era (QED, EC, PIM divestitures). He stays Executive Chair through end of Q2-2026 (now ending) — so the founder-era operator is just now fully stepping back.

• David Reeder is the new CEO (board member, transitioned in). Background: ex-CFO of GlobalFoundries, ex-CFO of Chewy, prior senior roles at Texas Instruments and Broadcom, 20+ yrs semis. A CFO-archetype, capital-markets-fluent operator — fitting for a company whose central task is deleveraging a post-CMC balance sheet while reinvesting in node-transition R&D.

• Track record: the prior team's record is strong on strategy (built the breadth moat, the CMC deal made ENTG the electronic-materials leader, then de-risked the portfolio) but the CMC integration left 4.0x leverage that is still ~3.6x net four years later — deleveraging has been slower than the original "<3.0x" target (Lens 5/13). Reeder's own ENTG track record is <1 year — unproven here.

• Skin in the game / capital allocation: dividend paid every quarter since Q4-2017, now $0.10/qtr (yield 0.23% — token). No buybacks ("did not repurchase any common stock" in Q1-26). Capital allocation = debt paydown first, dividend, R&D/capacity — disciplined and appropriate for the leverage, but it means no shareholder-return lever is being pulled at a ~45x multiple. Insider ownership: n/a (no our figures on shelf; proxy not ingested).

• Red flags (governance/comp): none material surfaced; KPMG auditor, clean controls attestation. The CEO transition itself adds execution/strategy-continuity risk (a CFO-CEO may run the playbook differently than a 13-yr operator).

• Archetype: transitioning from founder-era long-tenure operator (Loy) → professional capital-allocator CEO (Reeder). Implication: expect continued deleveraging discipline and possibly more financial optimization; watch for any strategy pivot (M&A appetite? portfolio moves?) in Reeder's first analyst day.

Forensic Red Flags

Forensic lens — every figure labeled. The accounting is broadly clean (Big-4 auditor, unqualified, no restatement, no SEC findings), but there are three things a forensic analyst flags:

1. Useful-life extension flattering 2026 earnings (the big one). Jan-2026 ENTG lengthened PP&E useful lives (buildings to 40 yrs, mfg equipment to 14 yrs, canisters to 19 yrs, etc.), cutting FY2026 depreciation ~$73.0M (~$52M to gross margin, ~$11M to R&D). This is a prospective accounting-estimate change that mechanically lifts GM, EPS, and segment profit — Q1-26 depreciation fell $49.9M → $34.1M YoY. It is disclosed and defensible, but a chunk of the "margin expansion" story is non-economic. Adjust for it before extrapolating the Q1 margin.
2. 1.1% effective tax rate (Q1-26) on OBBBA + UTB release — flatters GAAP net income; not a run-rate.
3. Goodwill + intangibles = $4.81B of an $8.48B balance sheet (~57%) — Goodwill $3,947.6M + intangibles-net $860.7M, almost all from CMC/ATMI. Annual goodwill test is qualitative for APS, quantitative for MS; MS passed with no impairment and a stated 50bp-sensitivity cushion. A demand or rate shock that pressures the MS reporting unit is the impairment tail-risk. Amortization is a real, recurring $184.4M/yr drag that makes GAAP EPS ($1.55 FY25) look far worse than non-GAAP ($2.75) — the gap between GAAP and non-GAAP is structurally wide and investors must decide which they believe.

Other checks: cash flow exceeds GAAP earnings (FY25 OCF $695.4M >> NI $235.6M — healthy, driven by D&A addbacks), so no earnings-quality alarm there; receivables outran sales in Q1 (+$70.8M on +5% sales) — watch DSO but flagged as collections timing. SBC is embedded in the non-GAAP adjustments (standard). Leases modest. No related-party concerns surfaced.

Regulatory findings (required):

• SEC Litigation Releases: None for Entegris since 2021-06-30.
• SEC AAERs: None.
• Non-SEC enforcement (web): No material FTC/DOJ/FDA/export-control enforcement actions or penalties against Entegris surfaced. The only adjacent items are (i) the terminated PIM/Infineum sale that never cleared HSR antitrust (deal abandoned 2023-02 — a clearance failure, not an enforcement action) and (ii) ENTG being subject to US/China export-control and tariff regimes that constrain China revenue (a market headwind, not an action against the company) ``.
• 10-K Item 3 / 10-Q Item 1 (Legal Proceedings): "from time to time involved in various claims, proceedings and lawsuits… arising in the ordinary course… the final outcome of these matters will not have a material adverse effect" — i.e. nothing material disclosed. (Note: FY24 had a one-off $20.0M gain from settling patent-infringement litigation in ENTG's favor.)
• Conclusion: No material regulatory or legal findings — verified via SEC EDGAR EFTS (LR, AAER), web search, and 10-K Item 3 / 10-Q Item 1 as of 2026-06-30.

---

Phase D — Project & stress-test

Forward Projection (EPS, next 3 fiscal years; FY = calendar Dec)

Built bottom-up from the latest actuals + guidance; every line labeled; outputs ``. Base on non-GAAP EPS (the metric the Street and ENTG's own comp plan use); FY2025 actual non-GAAP EPS = $2.75.

Inputs:

• FY26 already-printed Q1 non-GAAP EPS $0.86; Q2 guide midpoint $0.80. H1 ≈ $1.66; with the useful-life depreciation tailwind ($73M FY26) and a 2H MSI ramp, full-year FY26 base ≈ $3.40–3.55.
• FY27/28 drivers: industry MSI growth mid-to-high-single-digit + ENTG share-index growth (content/wafer, moly, 2nm/GAA, HBM) → revenue growth ~+8–10%/yr; operating leverage on a ~47% GM and capex moderation; continued interest-expense decline as debt is repaid (FY25 $199.8M, falling); tax normalizing back toward ~10–13% from the 1.1% Q1 anomaly; ~flat share count (no buyback, modest SBC dilutionDilutionIssuing new shares, so each existing share owns a smaller slice of the same company.).

• Base: FY26 $3.45 (H1 actual+guide $1.66 + H2 ~$1.79); FY27 = $3.45 × ~1.17 (≈+9% rev, ~+3pt op-leverage, ~$25M lower interest) ≈ $4.05; FY28 = $4.05 × ~1.16 ≈ $4.70.
• Bull: outsized HBM/GAA/moly share capture, China stabilizes, full operating leverage on ~48% GM, faster deleveraging → ~+13–15% EPS CAGR → FY28 ~$5.80.
• Bear: AI-capex digestion + China export-controls tighten + the depreciation tailwind laps out in FY27 (no further estimate benefit) → revenue flattish, EPS stalls ~$3.30–3.40.

Forecast NOT logged (per --watchlist rule: skip our model create in the breadth loop). If promoted to a thesis, the base call to track would be: "ENTG FY2027 non-GAAP EPS ≥ $4.00, p≈0.55."

Bull vs Bear

Bull case. Entegris is the cleanest "toll on every wafer" in the supply chain: recurring, designed-in consumables whose content grows with every node transition (GAA, 2nm, molybdenum, HBM/3D, High-NA EUV), insulated from the lumpiness of WFE because it's tied to wafers run, not fabs built. The moat — spec lock-in + purity criticality + portfolio breadth (no platform >3% of sales; no full-range global competitor) — is durable and hard to attack point-by-point. The cycle has turned (Q1-26 +5%, APS +7%, Taiwan +18%, advanced-packaging $100M+ run-rate), margins are expanding (GM 46.9%), FCF is strong ($141M in Q1, capex moderating to ~$250M), and a deleveraging balance sheet hands a growing interest-expense tailwind straight to EPS. New CEO Reeder is a balance-sheet operator for exactly this chapter. Earnings could surprise up as DRAM/HBM re-accelerates and China normalizes.

Bear case (2–3 permanent-impairment / de-rating risks).

1. Valuation is the bear case. ~45x forward / 9x EV/sales / 34x EV/EBITDA for a company that shrank revenue 1% in 2025 and guides mid-to-high-single-digit growth — and the stock ($174) trades above the consensus price target ($160) with downside to it. Much of the 2026 margin lift is a depreciation accounting-estimate change, not demand. The multiple prices a secular-compounder; the financials show a cyclical recovering off a trough.
2. China structural de-risking. China is ~18–21% of sales and a single-source for some raws; export controls + tariffs + a domestic-substitution champion (Anji Microelectronics in CMP slurries) can permanently shrink the China TAM and pressure cost.
3. Leverage + amortization overhang. ~3.6x net leverage and $184M/yr intangible amortization keep GAAP earnings (and GAAP ROE ~6%) thin; a demand/rate shock raises MS-goodwill impairment risk on $4.8B of intangibles+goodwill.

Pre-mortem (18 months out, thesis broke): AI-memory capex digested through 2026–27 (the SK-Hynix-HBM-slowdown signal proved real), China revenue stepped down on tightened controls, the one-time depreciation tailwind lapped, EPS stalled near $3.40 — and a ~45x multiple compressed to ~25x as the market re-rated ENTG from "secular compounder" to "good cyclical," taking the stock back toward the analyst targets in the $130s.

Are multiples too high? Yes, on the current fundamentals — ENTG is priced for a growth/quality outcome it must still deliver, and the tape already sits above sell-side targets.

Contrarian view (what the market is refusing to see): the market is treating the Q1 margin expansion as proof of operating-leverage inflection while under-weighting that ~$73M of it is a non-cash useful-life estimate change — so the "earnings power" being capitalized at 45x is partly accounting, and the real organic margin lift is smaller than the headline.

Devil's Advocate (short-seller)

Dismantling the bull case.

• Where revenue is concentrated: TSMC 16%, top-10 50%, 82% international, ~18–21% China. A TSMC capex/qualification shift, or a China step-down, moves the whole P&L. Contracts are non-binding — customers can cancel/delay. The "recurring" revenue is behaviourally sticky but not contractually guaranteed.
• Why the moat may be weaker than bulls think: in China, the moat is being actively dismantled by policy — domestic-substitution mandates + funded national champions (Anji in CMP slurries, Cobetter in filtration) don't need to beat ENTG on merit, only be "good enough + local." That's a structural haircut to ~1/5 of the book.
• Most dangerous competitor bulls underestimate: Anji Microelectronics — China CMP-slurry incumbent with state backing aimed squarely at ENTG's highest-margin MS franchise in its largest-by-units region.
• Worst capital-allocation / accounting items: (a) the CMC deal levered the company to 4.0x and it's still ~3.6x net four years later — slower deleveraging than promised; (b) the Jan-2026 useful-life extension is textbook earnings-management optics (legal, disclosed, but it lifts FY26 EPS by a non-cash ~$73M right as a new CEO takes over and needs a clean print); (c) GAAP ROE of ~6% says the accounting returns on the capital deployed (incl. $4.8B goodwill/intangibles) are mediocre.
• Assumptions that must hold for ~$174: mid-to-high-single-digit MSI growth sustains into FY27–28, margins hold the depreciation-aided level after the benefit laps, China doesn't step down further, and the multiple stays ~45x. Several are fragile.
• If growth disappoints 20–30%: revenue grows ~0–3% instead of ~9%; EPS stalls ~$3.30–3.40; a ~45x → ~25–28x de-rate ≈ stock toward $90–110 (the analyst low is $95).
• Single scenario that permanently impairs: a hard US–China decoupling that bifurcates the leading-edge supply chain — China fully substitutes ENTG consumables domestically and retaliates on ENTG's China-sourced raws — permanently removing ~1/5 of revenue and triggering an MS-goodwill impairment. Plausibility: moderate and rising given the live tariff/export-control trajectory.

Management Questions (15, ordered by information value)

1. Of the ~+0.8pt gross-margin and +24% APS-profit improvement in Q1-26, how much is the useful-life depreciation change versus underlying operating leverage — and what is the organic margin trajectory once the ~$73M FY26 benefit laps in FY27?
2. China is ~1/5 of revenue and a domestic-substitution target. What is your realistic 3-year China revenue path under continued export controls, and which product lines are most/least defensible against Anji and other local champions?
3. CMC closed at ~4.0x leverage in 2022 with a "<3.0x" target — why is net leverage still ~3.6x, and what is the explicit deleveraging schedule and the leverage level at which you'd restart buybacks?
4. At ~45x forward earnings and a token 0.23% yield with no buyback, how do you think about per-share value creation — is debt paydown genuinely the highest-return use of FCF at this multiple?
5. David — as a CFO-archetype CEO succeeding a 13-year operator, what will you do differently in strategy, M&A appetite, and portfolio over the next three years?
6. Walk through content-per-wafer at 2nm/GAA vs. the prior node, and quantify the molybdenum and HBM incremental dollar content you actually capture.
7. The advanced-packaging run-rate just crossed $100M — what's the realistic 3-year size, and what's your share/right-to-win there versus the materials incumbents?
8. Which single-source raw-material dependencies (membranes, abrasives, petroleum coke, China-sourced inputs) are the highest-risk, and what's the mitigation timeline and cost?
9. TSMC is 16% and rising; how do you protect pricing/qualification power as that account concentrates further?
10. What MS-reporting-unit assumptions underpin no goodwill impairment, and what demand/discount-rate shock would breach the test?
11. With contracts explicitly non-binding, what share of revenue is genuinely consumable/recurring versus exposed to cancellation in a downturn — and how did that mix behave in the 2024 trough?
12. The IEEPA-tariff refund and Section 301 investigations — quantify the net tariff cost in your cost base today and the range under plausible 2026–27 scenarios.
13. How sustainable is the tax rate — where does the effective rate normalize post-OBBBA, and what was one-time in the Q1-26 1.1%?
14. CHIPS Act funding (Colorado Springs $77M) and new Taiwan/Korea capacity — what's the expected ROIC and utilization ramp, and the downside if regional fab build-outs slip?
15. Where could you most disappoint the Street's mid-to-high-single-digit MSI assumption, and what's your contingency if AI-memory capex digests through 2026–27?

---