Analog Devices

Phase A — Understand the business

Lens 1 · Company Overview

Analog Devices is, in plain terms, the company that builds the bridge between the physical world and the digital one. Incorporated in Massachusetts in 1965, headquartered in Wilmington, MA, listed on Nasdaq as ADI and a member of the S&P 500. It designs, manufactures and sells high-performance analog, mixed-signal, power-management and RF integrated circuits — the chips that sense, measure, condition, convert, power and connect real-world signals (temperature, pressure, motion, sound, current) so that a digital system can act on them.

The product portfolio spans more than 75,000 SKUs across seven families: data converters (ADCs/DACs — its largest and most diverse family, where ADI is a global leader), power management & reference, amplifiers/RF & microwave, sensors & actuators (MEMS), DSPs/system products, interface, and an emerging software/AI layer (CodeFusion Studio 2.0, Power Studio, the "Physical Intelligence" initiative). The defining commercial feature: these are high-mix, low-volume, long-life-cycle parts — "several thousand analog ICs, many of which can have several hundred end customers," with "long product life cycles". A single part can sell for 10–15+ years into industrial equipment that is itself never re-spun.

End markets (FY2025): Industrial 45% ($4.93B), Automotive 30% ($3.28B), Consumer 13% ($1.43B), Communications 13% ($1.38B). Industrial is the heart of the franchise — factory automation, instrumentation & test, aerospace/defense, healthcare, energy. This is a B2B-edge business, not a consumer or AI-logic business.

Customers & channel. No single customer is named as a concentration risk; ADI sells through a direct salesforce plus third-party distributors across ~50 countries. FY2025 channel split: Distributors 56% ($6.14B), Direct/OEM 43% ($4.72B), Other (US government / over-time contracts) 1%.

Contract structure / key terms. Critically, ADI typically does not have long-term sales contracts with customers. It builds to forecast and to non-binding orders, which is the structural source of its cyclicality — orders can be cancelled, leaving inventory. The distributor relationship carries price-protection credits and stock-rotation return rights (variable consideration); the distributor-credit liability was $785M at FY2025 year-end, flagged by E&Y as the single Critical Audit Matter. There is no take-or-pay; revenue is overwhelmingly point-of-shipment recognition of a single performance obligation (the IC sale).

One-liner: ADI is the #2 pure-play high-performance analog franchise on earth — a 60-year-old IP library of sticky, long-life signal-chain and power parts, sold high-mix into industrial and automotive edges, with cyclicality baked in by a no-long-term-contract order model.

Lens 2 · Supply Chain

ADI runs a hybrid (fab-lite-leaning) manufacturing model — it owns front-end fabs but outsources a large share of leading-edge wafers and most assembly/test. Mapping the chain with named stakeholders:

Upstream inputs → ADI:

• Internal wafer fabs — ADI operates front-end fabrication "primarily in the United States, Ireland and Southeast Asia". Net PP&E of $3.32B reflects real but capital-light fab ownership relative to a TSMC or even a TXN. Analog runs on mature/trailing-edge nodes (180nm–28nm-ish), so ADI is not fighting for leading-edge capacity.
• External foundries — for advanced or overflow wafers ADI uses third-party foundries; TSMC is the principal external analog/mixed-signal foundry partner for the industry (filing does not name it explicitly, so ``). ADI's CHIPS-era posture has been "hybrid manufacturing" to add US/EU internal capacity and de-risk geographic concentration.
• A held-for-sale facility at Penang, Malaysia is disclosed — evidence of active back-end footprint rationalization.

ADI → end customer:

• Distributors (56% of revenue) — the dominant route; named industry distributors are Arrow Electronics and Avnet (``; filing names neither). Distributors hold ADI inventory and carry the price-protection/stock-rotation terms.
• Direct OEMs (43%) — industrial-automation, instrumentation, automotive-tier-1, comms-infrastructure and healthcare OEMs. Filing names application areas, not customers (ADI deliberately discloses no customer concentration).
• US government / prime contractors (~1%, "Other") — aerospace/defense over-time contracts.

Chokepoints & single-source dependencies:

1. Leading-edge external foundry — to the extent any ADI part needs an advanced node, TSMC is the chokepoint, shared with the entire industry. Low severity for ADI specifically because its mix is trailing-edge.
2. Geographic concentration in Ireland — both a manufacturing AND a tax chokepoint: FY2025 pre-tax income was "primarily generated in Ireland at a tax rate of 12.5%". Irish operations are load-bearing for both COGS and the effective tax rate.
3. Back-end (assembly/test) in Southeast Asia — exposed to the same tariff/trade risk ADI flags repeatedly in the 10-Q forward-looking factors ("recently announced and future tariffs and other trade restrictions").
4. Inventory-related purchase commitments of $269.7M — modest; ADI is not locked into large take-or-pay wafer commitments the way an AI-logic buyer is.

Verdict on the chain: resilient and deliberately un-concentrated. ADI's trailing-edge, fab-lite-hybrid model means it sits outside the CoWoS/HBM/leading-edge bottlenecks that dominate the AI-hardware complex. Its real supply-chain exposure is (a) Irish tax/manufacturing concentration and (b) tariff risk on a globally distributed footprint — not capacity scarcity.

Lens 3 · Competitive Advantages (Moats)

ADI's moat is one of the most durable in semiconductors, and it is not a leading-edge process moat — it is the opposite.

1. The IP library + high-mix catalog (the core moat). 75,000+ SKUs, many with hundreds of end-customers, life cycles of 10–15+ years. Each part is a small, deeply-engineered analog design that took years and elite talent to build. No single part is large enough to attract a focused attacker, but the aggregate catalog is nearly impossible to replicate. This is a scale-of-portfolio moat: breadth, not depth in any one node.

2. Switching costs / design-win lock-in. Analog parts are designed into customer hardware at the schematic level. Once an ADI converter is qualified into an MRI machine, a factory PLC, or an ADAS module, ripping it out means a re-spin, re-qualification, and re-certification — uneconomic for the customer over a 10-year product life. This produces the long life cycles and the pricing power. Switching cost is the mechanism that converts the catalog moat into recurring, high-margin revenue.

3. Engineering talent as a differentiator. ADI explicitly frames "approximately 13,000 engineers" and "the destination for the world's best engineering talent" as a competitive differentiator. High-performance analog design is a craft with a thin global talent pool; this is a genuine, if soft, barrier.

4. Scale & process / manufacturing breadth. Post-Maxim and post-Linear, ADI has the scale to invest ~$1.77B/yr in R&D (16% of revenue) and to run hybrid internal+external manufacturing — a scale only TXN exceeds among Western analog players.

Bargaining power: ADI holds the strong side over both suppliers and customers in normal times. Over customers: design-win lock-in + no second source for many parts = pricing power (61.5% FY2025 gross margin, 67.3% in Q2 FY2026 ). Over suppliers: trailing-edge wafers are not scarce, so ADI is not hostage to a foundry the way an AI-logic firm is. The one place bargaining power flips against ADI is the distributor channel during a downturn — price-protection credits ($785M liability) mean ADI absorbs price risk on inventory sitting in the channel.

The honest moat caveat: the moat protects the installed base and franchise, not the cycle. It does nothing to smooth the boom-bust of industrial/auto demand, because the no-long-term-contract order model deliberately leaves ADI exposed to the inventory cycle. The moat is why ADI survives every downturn with 60%+ gross margins intact; it is not why the stock is volatile.

Lens 4 · Segments

ADI reports as a SINGLE operating segment / single reporting unit — confirmed explicitly in the goodwill policy: "we have determined that the business operates as a single operating segment and has a single reporting unit for the purpose of goodwill impairment testing". There is therefore no product-line revenue or operating-income breakout in the filings — only by end market, channel, and geography. I will not fabricate segment-level operating income; it is n/a — not disclosed (single segment).

By end market:

The trend, and the cause. FY2025 was the bottom-of-cycle recovery year: +17% as customer inventory normalized off a brutal FY2024 destock (FY2024 revenue had fallen to $9.43B from FY2023's $12.31B — a ~23% peak-to-trough collapse ). By Q2 FY2026 the recovery has gone vertical: Industrial +56% and Communications +79% YoY. The drivers are explicit:

• Industrial — broad-based across all sub-markets, highest growth in test equipment and aerospace/defense. This is the destock fully unwinding plus a genuine demand upturn.
• Communications — data-center/wireline expansion for AI infrastructure. ADI's AI exposure is here: not GPUs, but the power and signal-chain content around AI data centers.
• Automotive — the laggard at only +2% YoY in Q2, decelerating from +16% in FY2025. This is the one soft spot: auto is not participating in the upturn, reflecting weak global auto production and ADI's connectivity/A2B content being cycle-exposed.

By geography (FY2025): US $3.24B (+14%), China $2.86B (+34%), Europe $2.29B (+8%), Rest of Asia $1.49B (+24%), Japan $0.99B (−9%), Rest of Americas $0.16B. China at 26% of revenue and +34% is the standout — and the standout risk (tariff/export-control exposure).

Read: the segment story is a textbook semiconductor cyclical recovery, led by Industrial and AI-data-center Comms, with Automotive as the conspicuous non-participant. The acceleration is real but is being compared against a deeply depressed FY2025 base — the YoY rates flatter the underlying run-rate.

---

Phase B — Measure performance

Lens 5 · Earnings Result (latest print: Q2 FY2026, quarter ended 2026-05-02)

The Q2 FY2026 print was a blowout on the numbers and a sell-off on the tape — the single most important fact in this dossier.

Revenue & profit vs. consensus:

• Revenue $3,623M, +37% YoY (vs. $2,640M Q2 FY2025), and above the high end of guidance.
• Gross margin 67.3% (vs. 61.0% PY) — +630 bps, on manufacturing-fixed-cost absorption + favorable mix.
• Operating income $1,380M, +104% YoY; operating margin 38.1%.
• Net income $1,176M, +106%; net margin 32.5%.
• Diluted EPS $2.40 vs. $1.14 PY, +111%.
• Q3 FY2026 guidance: revenue ~$3.63B, adjusted EPS ~$3.00 (GAAP-vs-adjusted gap is amortization of acquired intangibles). Next print expected 2026-08-19, sell-side ~$3.31 adj EPS.

Which lines drove it: Industrial (+56%) and Communications/AI-data-center (+79%); Automotive flat (+2%) was the drag. Management/CFO Puccio flagged "record bookings across B2B markets" — the order book, not just shipments, inflected.

Margin moves: the operating leverage is the story. R&D fell to 14% of revenue (from 17%) and SMG&A to 10% (from 11%) purely on revenue denominator — ADI is dropping incremental revenue to the bottom line at a very high rate as it climbs out of the trough. Gross margin at 67.3% is above ADI's own through-cycle model.

Guidance & tone: management struck a confident, "record demand and operational discipline" tone. The forward-looking risk language in the 10-Q newly and prominently elevates tariffs and trade restrictions to the top of the factor list — a tonal addition vs. the 10-K.

Balance-sheet flags:

• Cash + ST investments $3.4B (down from $3.7B at FY2025-end, as buybacks/dividends outran FCF).
• Receivables jumped +43% to $2.05B vs. $1.44B at FY2025-end — but DSO actually fell to 43 days (from 44); the dollar jump is revenue-driven, not a collections problem. Clean.
• Inventory +12% to $1.85B; days-cost-of-sales-in-inventory rose to 139 from 130 — ADI is deliberately building inventory to support demand. Worth watching: if demand rolls over, this is where the next write-down risk lives, but 139 days is normal-to-healthy for this franchise.
• Total debt $8.77B; $0.9B of 2026 notes reclassified to current. Net debt modest against $3.4B cash and $4.8B+ annual operating cash flow.

Market reaction: despite the beat-and-raise, the stock fell >5%. The reason was entirely valuation/expectations: ADI had rallied ~95% off its late-2025 low of ~$217, was trading near its 52-week high at a trailing P/E in the high-60s/low-70s, and the formal sell-side targets had been outrun by the price. The print was priced in; there was nothing left for the beat to do. This is the central tension of the name: the business is firing and the stock has already paid for it.

Unusual vs. ADI's own history: 67.3% gross margin and 38% operating margin are peak-class — ADI is earning above its long-run model because utilization is recovering off a low base. The +37% revenue growth is not a steady-state rate; it is a base-effect snapback from the FY2024 destock. Normalizing, ADI's "real" revenue run-rate is ~$14.5B annualized (Q2 × 4), well above FY2025's $11.0B but the YoY optics overstate the underlying trend-line.

Lens 6 · Earnings Calls (sentiment trend)

No transcripts on disk (transcripts/ empty), so this is ``-grounded across the recent calls.

Tone trajectory (last ~4 calls):

• FY2024 calls (trough): defensive — inventory destocking, "customer inventory rebalancing," cautious bookings, "navigating the bottom." Management's recurring framing was resilience of the model and capital return as the through-cycle anchor.
• Q1 FY2026 (Feb 2026): inflection — alongside the print ADI raised the dividend 11% and added $10B to the buyback. The act is the sentiment: you do not raise the payout 11% at a cycle bottom unless you have conviction the recovery is durable.
• Q2 FY2026 (May 2026): confident/peak-tone — "record" revenue, "record bookings across B2B markets," "record demand and operational discipline". CEO Roche and CFO Puccio leaned into breadth of recovery and AI-data-center pull in Communications.

What they started saying: "record bookings," "AI-driven infrastructure" (Communications/data-center), "Physical Intelligence" (the AI-at-the-edge product vision), "operational discipline / operating leverage."

What they stopped saying: "inventory rebalancing / destocking," "cautious," "macro uncertainty weighing on bookings" — the trough vocabulary is gone.

The one tonal hedge: tariffs/trade moved up the risk stack — the one place management is adding caution even as everything else turns up.

Sentiment read: decisively positive and confirmed by capital-allocation action (the 11% dividend raise is the tell). The risk in a "record everything" tone at a cyclical high is that it marks the optimism peak — exactly the setup the stock's post-print drop is pricing.

Lens 7 · Comps

Peer set: the high-performance / broad-line analog & mixed-signal cohort. Multiples are `` as of mid-June 2026; where a figure isn't cleanly sourced it is marked n/a rather than fabricated.

5-year average ROE: n/a for the peer set; for ADI specifically, FY2025 ROE ≈ 6.7% — depressed because (a) earnings are still recovering and (b) the equity base is bloated by ~$27B goodwill + $8B intangibles from the Maxim/Linear deals. On tangible equity ROE is far higher; on a normalized-earnings basis (Q2 annualized ~$4.7B NI ÷ $33.8B equity) ROE would be ~14%.

The comps read. Every multiple in this cohort is distorted right now because the entire group is climbing off trough earnings — trailing P/E and EV/EBITDA are mechanically inflated (small denominators). ADI at ~35x forward sits at a clear premium to TXN (~33x trailing) and a large premium to NXP (~13x EV/EBITDA), and roughly in line with MCHP/MPWR's growth-premium multiples. The market is paying up for ADI's franchise quality and Industrial/AI-comms leverage. The provenance-honest conclusion: ADI is not cheap on any sourced metric; it is priced as the quality compounder of the group, with the cyclical recovery already in the multiple.

Lens 8 · Stock-Price Catalysts (what moves ADI >5%)

``-grounded pattern over the recent cycle:

• Earnings prints + guidance are the dominant >5% movers — but asymmetrically tied to expectations, not absolutes. The clearest data point: Q2 FY2026 beat and raised and the stock fell >5%. The market trades ADI on the delta vs. an already-high bar, not on the headline beat.
• The cyclical turn itself — ADI rallied ~95% off the ~$217 late-2025 low as the industrial destock ended. The single biggest move of the cycle was the recognition that the bottom was in.
• Capital-return announcements — the Feb 2026 11% dividend raise + $10B buyback authorization was a positive catalyst and a sentiment marker.
• Macro / industrial-PMI / auto-production data — as a 45%-industrial / 30%-auto name, ADI moves with the broad industrial cycle and global PMIs.
• Tariff / China-trade headlines — given 26% China revenue and a global footprint, trade-policy shocks are a live macro catalyst (newly elevated in the 10-Q risk language).

What the pattern reveals: ADI is an expectations-and-cycle stock, not a single-customer or single-product stock. There is no NVIDIA-style "one hyperscaler order" swing factor. The market reacts to (1) where we are in the industrial/auto inventory cycle and (2) the print relative to a bar that rises with the price. At a cyclical high with a rich multiple, the bar is the enemy — exactly what the post-Q2 drop showed.

---

Phase C — Judge people & books

Lens 9 · Management

Vincent Roche — CEO & Chair. A 35-year+ ADI lifer (joined 1988), CEO since 2013, also Chair. The archetype: long-tenured insider-operator, not a founder, not a parachuted-in turnaround manager. His record is the defining one: he executed the two transformational deals — Linear Technology ($14.8B, 2017) and Maxim Integrated ($21B all-stock, closed Aug 2021) — that turned ADI from a converter specialist into the #2 broad-line analog franchise. Under Roche, ADI roughly tripled revenue and built the high-mix catalog that is the moat. Capital-allocation track record is strong: 22 consecutive years of dividend increases, ~$32B returned over the program, a stated ~100% FCF-return commitment, and 29 consecutive years of positive FCF. He is the right CEO for a mature, cash-generative compounder.

Richard Puccio — EVP & CFO. The notable recent hire: Puccio joined from AWS, where he was CFO. An outside-the-industry, hyperscale-finance background — an interesting signal that ADI wants data-center/AI fluency and capital-markets discipline at the top of finance. Reports to Roche.

Board refresh: Dr. Yoky Matsuoka (ex-Google/Nest/Panasonic robotics & AI) joined the board as independent director, effective Jan 2026 — reinforcing the AI/Physical-Intelligence strategic tilt.

(1) Track record: strong and quantified — the Linear + Maxim integrations delivered the scale and margin structure ADI runs on today; $275M Maxim synergy target, and gross margins have structurally risen post-deals (57–62%+ range vs. pre-Maxim mid-60s on a smaller base — and 67% at peak utilization now). (2) Tenure & skin in the game: very high tenure (Roche 35+ yrs). Insider ownership % is n/a (would be in the DEF 14A, which is not on disk); for a non-founder it is presumably modest in absolute terms but accumulated over decades. (3) Capital allocation: A-grade for a mature compounder — disciplined buybacks (216.5M shares / $17.0B repurchased program-to-date ), a 22-year dividend-raise streak, sensible M&A that built the moat. The one critique: ROE/ROIC optics are depressed by the $35B of goodwill+intangibles the deals put on the books — the M&A created a strategic franchise but a heavy balance sheet. (4) Red flags: none material. No related-party issues, no promotional behavior, no strategy whipsaw, clean E&Y audit (auditor since 1967), no executive 10b5-1 churn flagged in Q4. (5) Archetype: seasoned professional-manager/insider running a cash-return compounder — exactly fit-for-stage.

Lens 10 · Forensic Red Flags

Acting as a forensic analyst across the income statement, balance sheet and cash-flow statement. ADI is, bluntly, one of the cleaner large-cap accounting profiles in the sector. Findings:

Income statement:

• GAAP vs. adjusted EPS gap is large and structural — driven by amortization of acquired intangibles of ~$1.59B in FY2025 ($1.74B FY2024) running through the cash-flow reconciliation, plus $750M on the income statement. Adjusted/non-GAAP EPS (what sell-side quotes, e.g. Q3 ~$3.00) excludes this. This is legitimate (real M&A amortization, not a gimmick) but it means the headline "adjusted" numbers flatter GAAP materially — a Q2 GAAP diluted EPS of $2.40 vs. an adjusted figure that is higher. Always anchor on which metric is being quoted.
• SBC is modest and well-controlled — $321.6M FY2025 (~2.9% of revenue). Not used to flatter non-GAAP aggressively; small relative to a typical tech-hardware name. Green flag.

Balance sheet:

• Goodwill $26.9B + intangibles $8.0B = $34.9B, i.e. ~73% of total assets ($48.0B) is acquisition-related intangible value. This is the single largest quality-of-balance-sheet flag: ADI's book is dominated by Maxim/Linear purchase accounting. Goodwill is tested as a single reporting unit, and ADI itself notes "there would need to be substantial negative developments… for there to be potential impairment". No impairment to date, and the market cap (~$184B) vastly exceeds book — so impairment risk is low — but a severe, prolonged downturn is the scenario that could threaten it.
• Receivables/inventory vs. revenue: clean. Q2 receivables +43% but DSO down to 43 days; inventory +12% with days at 139 (deliberate build). Neither is outrunning revenue in a way that signals channel-stuffing or demand-pull-forward.
• Distributor price-protection liability ($785M) — the E&Y Critical Audit Matter. Legitimately judgment-heavy (estimating future price credits to distributors), and the place where a demand reversal would force a true-up. ADI states "actual distributor claims activity has been materially consistent" with estimates — but this is the line item to watch if the cycle turns.

Cash-flow statement:

• Cash flow exceeds earnings (the good direction): FY2025 operating cash flow $4.81B vs. net income $2.27B — the gap is the $1.59B intangible amortization + $0.41B depreciation, all non-cash. There is no divergence where earnings outrun cash — the opposite, which is the hallmark of a clean, cash-generative franchise. FCF (OCF − $533.6M capex) ≈ $4.28B FY2025.

Regulatory findings (required sub-section).

• SEC Litigation Releases / AAERs: None. Per regulatory/regulatory-findings.md (generated 2026-06-22 via EDGAR EFTS, LR + AAER, 2021-06-22→2026-06-22): "total_sec_findings: 0" — no LR and no AAER naming Analog Devices in the search window.
• Item 3 Legal Proceedings (FY2025 10-K): routine only — "From time to time in the ordinary course… various claims, charges and litigation… We do not believe that any current legal matters will have a material adverse effect". No material litigation disclosed.
• Non-SEC enforcement (web search): no material FTC/DOJ/FDA/CFPB consent decree, settlement, fine or penalty surfaced for Analog Devices in the recovery-period searches. (The recurring non-SEC item for analog peers is export-control/EAR compliance given China exposure — ADI flags it as a risk, not a finding.)
• Conclusion: No material regulatory or legal findings — verified via SEC EDGAR EFTS (LR, AAER), web search, and 10-K Item 3 as of 2026-06-22.

Forensic verdict: clean. The only two things a skeptic legitimately flags are (1) the GAAP-vs-adjusted amortization gap (cosmetic, not fraudulent) and (2) the goodwill-heavy balance sheet (a structural consequence of M&A, low near-term impairment risk). Neither is an integrity red flag. Audit clean, controls clean, cash > earnings.

---

Phase D — Project & stress-test

Lens 11 · Forward Projection (FY2026 / FY2027 / FY2028 EPS)

Built bottom-up from the latest actuals + guidance. ADI fiscal year ends late-Oct/early-Nov; "FY2026" ends ~2026-11-01. All outputs `` with arithmetic shown. No forecast.ts create (watchlist loop). Note: figures below are adjusted/non-GAAP EPS where benchmarked to sell-side, since that is how ADI guides and is quoted; GAAP would be ~$1.10–1.30/yr lower on intangible amortization.

Anchors:

• H1 FY2026 GAAP diluted EPS already booked: $4.09 (Q1+Q2).
• Q3 FY2026 guide: revenue ~$3.63B, adjusted EPS ~$3.00; sell-side ~$3.31.
• Share count ~490M diluted and falling on buybacks.

FY2026 (base): H1 GAAP $4.09 + Q3 + Q4. If Q3 GAAP ≈ $2.45 and Q4 GAAP ≈ $2.55 (flat-to-modestly-up sequential, in line with the ~$3.63B revenue plateau the company guides), FY2026 GAAP EPS ≈ $9.10; FY2026 adjusted EPS ≈ $12.0–12.2. This frames the ~35x forward P/E: $371 ÷ ~$10.5 blended-forward ≈ 35x, consistent with the `` 35.46x.

• Bull: demand keeps accelerating, GM holds 67%+ → FY2026 adj EPS ~$12.6.
• Bear: auto stays weak, a comms-AI air-pocket, GM mean-reverts toward 64% → FY2026 adj EPS ~$11.3.

FY2027 (base): the key judgment is normalization, not continued +37% growth. A reasonable base assumes revenue grows ~8–10% off the ~$14.5B FY2026 exit run-rate (industrial mid-cycle + AI-comms secular − auto drag), GM ~66%, buyback shrinking share count ~2%/yr. FY2027 adj EPS ≈ $13.7–14.0.

• Bull (~$15.5): AI-data-center comms content compounds, industrial runs hot, GM 67%+.
• Bear (~$11.5): a fresh inventory correction (the cycle is real); revenue declines mid-single-digits as the snapback overshot true demand — the single most important downside scenario for this name.

FY2028 (base): mid-cycle steady-state. Revenue +7–8%, GM ~66%, share count −2%. FY2028 adj EPS ≈ $15.0–15.5.

• Bull ~$17.5; Bear ~$11.0 (a 2028 downturn).

Base-case forecast (for tracking, NOT logged in watchlist loop): ADI FY2026 adjusted EPS ≥ $12.00, p≈0.62, resolves 2026-11-01. Rationale: H1 GAAP $4.09 already in hand + a guided ~$3.63B Q3 makes ~$12 adjusted highly reachable barring a demand shock.

The valuation implication: at $371 and ~35x forward, the stock is discounting the mid-cycle, not the bottom. For ADI to work from here on multiple, you need FY2027–28 EPS to compound and the multiple to hold ~33–35x — i.e. the recovery to extend into a durable up-cycle without a destock. The edge is not at this price; it is in buying the next inventory correction, when ADI's 60%+ trough margins and 22-year dividend streak let you own the franchise at 20–25x trough earnings.

Lens 12 · Bull vs Bear

Bull case. ADI is a fortress-quality compounder bought correctly only on weakness, and right now it is firing on every cylinder. The moat — 75,000 long-life SKUs, design-win lock-in, 13,000 elite analog engineers — is among the most durable in semis and is widening as the catalog scales post-Maxim. The cyclical recovery is broad and real: Industrial +56%, Comms +79%, record bookings, gross margin at a peak-class 67.3%, and operating leverage dropping incremental revenue to the bottom line at >50% flow-through. ADI now has a genuine AI growth leg — not GPUs, but the power and signal-chain content around AI data centers (the +79% Communications driver), plus the "Physical Intelligence" edge-AI vision and an ex-AWS CFO + AI-roboticist board member to execute it. Capital return is elite and reliable: 22 straight years of dividend hikes, ~100% FCF returned, 29 years of positive FCF, an $11.5B buyback in hand. Secular tailwinds — electrification, factory automation, AI-edge sensing, data-center power — all expand ADI's content-per-system. Earnings surprise potential is real if the up-cycle extends.

Bear case (the 2–3 things that could permanently or durably impair the thesis from this price):

1. Valuation at a cyclical high. ~35x forward / ~65x trailing, near the 52-week high, after a ~95% rally — the stock fell on a beat-and-raise because the bar now rises with the price. You are paying mid-cycle multiples for peak-class margins; the asymmetry is poor.
2. The cycle is structurally real and will turn again. ADI's no-long-term-contract order model guarantees future inventory corrections. FY2024 saw a ~23% peak-to-trough revenue collapse. The current +37% is a base-effect snapback off that trough; the "real" run-rate is lower than the optics. A 2027 destock — entirely normal for this franchise — would take EPS down 15–25% and the multiple with it.
3. Automotive deceleration + China concentration. Auto (30% of revenue) grew just +2% YoY in Q2 — the franchise's growth is leaning hard on Industrial and AI-comms while a third of the business stalls. Layer on 26% China revenue and the newly-elevated tariff/export-control risk, and a single trade-policy shock or auto air-pocket hits a large slice of the book.

Pre-mortem (18 months out, thesis broke — what happened?): It is late 2027. The industrial recovery that drove FY2026 turned out to have over-shipped true demand — distributors and OEMs double-ordered into the upturn, and a fresh destock arrived. ADI revenue is down high-single-digits YoY, gross margin slipped back toward 63%, the $785M price-protection liability needed a true-up, and adjusted EPS missed the ~$14 the Street had penciled. The stock, having been priced at 35x peak earnings, de-rated to 25x trough earnings — a >40% drawdown. The franchise is fine; the entry price was the mistake.

Are multiples too high? From a cycle-timing standpoint, yes — ~35x forward at the top of an up-cycle prices in continuation that the franchise's own order model makes unreliable.

Contrarian view (what the market refuses to see): the consensus is anchored on "best-in-class analog, AI-data-center leg, buy quality" — and it's right about the business. What it's under-weighting is that ADI's defining feature is cyclicality the moat does not fix. The smart way to own ADI is not to chase the recovery at 35x; it is to pre-commit to buying the next destock — the franchise quality is precisely what lets you do that safely. The market is treating a deeply cyclical name as a secular-growth name at the wrong point in the cycle.

Lens 13 · Devil's Advocate (short-seller)

Dismantling the bull case. I am not short the company — the business is excellent — I am short the setup. Here is the case:

• What structurally breaks the money-machine? Nothing breaks the franchise; what breaks the stock is the inventory cycle the bulls are wishing away. ADI explicitly builds to non-binding orders with no long-term contracts — the +37% you're celebrating is the same mechanism that produced −23% in FY2024. You are buying the top of the saw-tooth.
• Where is revenue concentrated, and what shifts? Industrial (45%, now 50%) is carrying the recovery while Automotive (30%) grows +2%. If Industrial double-ordered (it always does into an upturn), the concentration that's helping you reverses violently. Plus 26% China into a live tariff/export-control regime.
• Why is the moat weaker than bulls think? It isn't weaker on the franchise — it's that the moat is irrelevant to the thing that drives the stock. A 60%+ trough gross margin is great, but the moat does nothing to stop a 40% drawdown when a 35x multiple meets a destock. Bulls conflate "great business" with "great stock at this price."
• Most dangerous competitor bulls underestimate? Texas Instruments — the only player with greater scale, running 300mm captive analog fabs for structural cost leadership. In a downturn TXN's cost position lets it compete on price in the broad/catalog overlap, pressuring ADI's mix exactly when ADI needs pricing.
• Worst capital-allocation moves / accounting aggression? No accounting aggression (clean audit). The fair critique is that $35B of goodwill+intangibles (73% of assets) from Maxim/Linear sits on the book, depressing ROE to ~7% GAAP and creating a (low-probability but real) single-reporting-unit impairment tail in a severe downturn.
• What must hold for today's price? That the up-cycle extends through FY2027–28 without a destock, GM holds ~66%+, and the multiple stays ~33–35x. That is three things going right simultaneously, at a cyclical high, in a business whose own order model makes the first one unreliable.
• Valuation if growth disappoints 20–30%? If FY2027 revenue comes in ~25% below the bull path (a normal destock), adj EPS lands ~$11 not ~$14, and a de-rate to ~25x puts the stock near $275 — roughly −26% from $371. The downside is a normal cyclical event, not a tail.
• Single scenario that permanently impairs? True permanent impairment is hard for this franchise — even a goodwill write-down wouldn't break the cash machine. The realistic worst case is durable not permanent: a 2027 destock + auto weakness + a China-trade shock landing together, EPS −25%, multiple −10 turns, a multi-quarter ~40% drawdown. Plausibility: moderate-to-high over a 2–3 year window — because the cycle always turns.

Short-seller's bottom line: great company, wrong price, wrong point in the cycle. The trade is not to short it outright (the franchise + buyback put a floor under it); it is to wait for the destock and buy it 30–40% lower.

Lens 14 · Management Questions (15, ordered by information value)

1. The +37% Q2 growth is a snapback off the FY2024 destock — what is your read on true end-demand vs. channel re-stocking, and how much double-ordering do you believe is in the current bookings? (The single question whose answer most changes the view.)
2. Automotive grew just +2% YoY while Industrial grew +56% — is auto a timing lag or a structural content/share problem, and when do you expect it to re-accelerate?
3. Gross margin is 67.3%, above your through-cycle model — what is the sustainable mid-cycle gross-margin range, and how much of today's 67% is utilization that reverses in a downturn?
4. With 26% of revenue in China and tariffs newly elevated in your risk language, what is the realistic revenue-at-risk under a severe export-control or tariff scenario, and what is the mitigation?
5. You raised the dividend 11% and added $10B to the buyback at what looks like a cycle inflection — what conviction signals drove that timing, and would you keep buying back aggressively at today's ~35x forward multiple?
6. The Communications +79% is AI-data-center driven — quantify your AI-data-center content per rack/MW today and where it goes over three years; how durable vs. cyclical is this leg?
7. $35B of goodwill+intangibles sits on the balance sheet as a single reporting unit — what downturn severity would put goodwill at impairment risk, and how do you think about the tested headroom?
8. The no-long-term-contract order model is the source of your cyclicality — are you pursuing any structural change (LTAs, capacity reservations) to dampen the inventory cycle, or is volatility a permanent feature?
9. "Physical Intelligence" and the AI/software push (CodeFusion, Power Studio) — what is the revenue model and timeline; is this a real P&L line or a design-enablement/stickiness play?
10. Richard Puccio came from AWS — what does hyperscale-finance discipline change about how ADI sets capital-return, M&A, and capacity decisions?
11. The distributor price-protection liability is $785M and your Critical Audit Matter — how would that line behave in a demand reversal, and what's the worst-case true-up?
12. Capex is guided to 4–6% of revenue (fab-lite-hybrid) — how much of your wafer supply is external/TSMC, and is your internal-capacity strategy changing under CHIPS/onshoring incentives?
13. M&A built today's ADI (Linear, Maxim) — at this point in the cycle and this valuation, is there another transformational deal in the pipeline, or is it organic + buyback from here?
14. Inventory days rose to 139 as you build for demand — what is your trigger discipline for cutting production if bookings roll over, and how fast can you flex?
15. The stock fell on a beat-and-raise — how do you think about communicating with a market that has priced in continuation; what would you most want investors to under-appreciate today?

---