Fanuc

Phase A — Understand the business

Lens 1 · Company Overview

FANUC is the closest thing manufacturing has to a utility. It sells the three things a factory needs to run automatically, and it has quietly owned the highest-margin one for forty years.

What it actually does — four businesses:

1. FA (Factory Automation) — CNC (computer numerical control) systems, servo motors, servo amplifiers, and lasers. This is the brain and muscle inside other people's machine tools. FANUC does not sell you the lathe; it sells the controller and servos that the lathe-maker (DMG Mori, Mazak, Makino, thousands of Chinese builders) embeds. FY-ended-Mar-2025 FA sales ¥194.8B, +8%, 24.4% of consolidated net sales. This is the crown jewel: FANUC commands an estimated 50-60% of the global CNC market — a near-monopoly in high-end controls.
2. Robot — industrial articulated robots (handling, welding, assembly, painting) and CRX/CR collaborative robots (cobots). FANUC has the largest single global industrial-robot share, ~17% (or ~11% on a broader top-10 base that includes Kawasaki/Denso). Robot is now the largest revenue segment — ~38% of total revenue averaged over recent quarters. FY2026 9M Robot revenue ¥269.2B, +11.1% YoY, driven by China EV + general industry.
3. ROBOMACHINE — FANUC's own-branded machines built on its FA components: ROBODRILL (compact machining center), ROBOSHOT (all-electric injection-molding machine), ROBOCUT (wire EDM). FY-ended-Mar-2025 ¥137.6B, +33.1%. This is FANUC eating its own dog food — the segment that proves the components are best-in-class.
4. Service — installed-base maintenance, spare parts, field service, retrofits. The annuity: service revenue grew +74.5% over four years, from ¥77.5B (FY-Mar-2021) to ¥135.2B (FY-Mar-2025). On a ~750,000+ robot installed base and millions of CNCs, this is the recurring cushion under a cyclical hardware book.

Customers & structure. Automotive (esp. EV lines), electronics, general industry, and machine-tool builders. There is no take-or-pay; hardware is transactional and cyclical, but the razor-and-blade economics (sell the CNC/robot once, sell parts + service for 15+ years) create a de-facto recurring layer. Geographic mix is dominated by China (the swing factor in every recent print), Japan, the Americas, and Europe/India.

The one-line business model: FANUC sells the nervous system of the world's factories, keeps ~35-40% gross margins because nobody trusts a cheaper controller on a machine that must not miss a micron, and monetizes the installed base forever through parts and service. ``

Lens 2 · Supply Chain

Map: upstream inputs → FANUC → machine builders / integrators → end factory.

Upstream (into FANUC):

• Precision reduction gears / harmonic drives — the single most important bought-in component for robot joints. Named suppliers: Harmonic Drive Systems (6324.T) and Nabtesco (6268.T) dominate high-torque compact strain-wave and RV gears ``. This is a genuine chokepoint the whole Western robot industry shares.
• Semiconductors / MCUs / power devices — for CNC and servo-amp control boards. FANUC designs its own control ASICs/boards; fab is outsourced.
• Servo motors — FANUC makes these in-house (a core competency and a moat input, not a bought part).
• Steel, aluminium castings, cabling, connectors — commodity, multi-sourced.
• Compute for physical AI — NVIDIA Jetson / Thor edge modules, newly a named input via the March-2026 partnership (see Lens 3).

Midstream (FANUC itself): Vertically integrated at the "yellow factory" complex at the foot of Mt. Fuji, Oshino, Japan — heavily automated, lights-out production. FANUC famously builds robots with robots. This concentrates manufacturing geographically (an earthquake / single-country risk) but delivers the cost structure that funds ~20% operating margins.

Downstream (out of FANUC):

• Machine-tool OEMs — DMG Mori, Yamazaki Mazak, Makino, Okuma (partly), and thousands of Chinese machine-tool builders who embed FANUC CNCs. FANUC's CNC is sold into the tool, so its end-demand is machine-tool capex.
• Robot system integrators — who deploy FANUC arms into automotive/electronics lines.
• End customers — auto OEMs & Tier-1s (EV lines the key swing), electronics/semis, logistics, general industry.

Chokepoints & single-source risk:

Names or it didn't happen — the supply chain here is real and named: Harmonic Drive + Nabtesco upstream, NVIDIA for AI compute, DMG Mori / Mazak / Makino + Chinese builders as the OEM layer, auto/EV as the end. The vulnerability is that FANUC's own leverage over the CNC chokepoint is exactly what Chinese domestics are attacking (Lens 3/13).

Lens 3 · Competitive Advantages (moats)

FANUC has one of the widest moats in industrials — and it is being tested at the low end for the first time.

The moats, ranked by durability:

1. CNC installed-base + switching costs (the deep moat). Machine-tool builders standardize on FANUC because the entire shop-floor workforce, G-code programs, maintenance training, and spare-parts logistics are built around FANUC controls. Ripping out a CNC standard is a multi-year, multi-site retraining cost. 50-60% global CNC share is not a marketing number — it's an ecosystem lock. This is the Keyence-like "we are embedded in your process" moat.
2. Reliability / brand as a purchasing default. In automation, downtime is the enemy; a factory line stopped by a controller failure costs more per hour than the controller itself. FANUC's "yellow" reliability reputation lets it hold price. This is why gross margin sits at 35-40% while Chinese equivalents undercut by 20-40% and still can't fully displace it in mission-critical lines.
3. Vertical integration + scale. In-house servos + lights-out manufacturing = a cost structure competitors buying gears/motors on the open market cannot match at FANUC's volume.
4. Service annuity. A 750k+ robot / multi-million CNC installed base generates parts+service that (a) compounds (+74.5% in 4 yrs) and (b) deepens lock-in.
5. Balance-sheet moat. Zero debt, ¥603B cash means FANUC can out-survive any downturn, out-invest on R&D, and never be forced to sell at the bottom. In a cyclical business, the fortress balance sheet is a competitive weapon.

Bargaining power: Strong over machine-tool OEMs (they need FANUC's CNC to sell a credible high-end tool) and over most customers (mission-critical, hard to switch). Weakening over the price-sensitive Chinese mid-market, where "good enough" domestic controllers/robots now exist.

The moat's soft flank (the thing that matters): The moat is deepest in high-end CNC and mission-critical robots. It is shallow in the commoditizing mid-market — general-industry pick-and-place, price-sensitive Chinese lines — which is precisely the fastest-growing volume segment. `` The bull thinks FANUC's moat is a fortress; the bear thinks it's a fortress guarding the high ground while the valley (volume) floods with Chinese product. Both are right — the question is mix (Lens 12/13).

Lens 4 · Segments

Segment revenue (yen), FY2026 nine-months (Apr-Dec 2025) — the most recent hard breakout:

Full-year FY-ended-Mar-2025 segment marks (for the trend line):

• FA ¥194.8B (+8%, 24.4% of sales); ROBOMACHINE ¥137.6B (+33.1%, 17.3% of sales); Robot down in China/EU/Americas but up in Japan; Service ¥135.2B.
• Note the whipsaw: In an earlier interim window robot sales fell −16.4% to ¥242.4B on China EV weakness, then swung to +11.1% in FY2026 9M as China EV re-accelerated. This is the single most important fact in the whole dossier: FANUC's growth is levered to the Chinese EV/automation capex cycle, and that cycle is both volatile and increasingly served by Chinese domestics.

Geography (directional, not fully sourced to a clean table — n/a on exact splits): China is the swing region across every print; Japan flat; Europe weak (machine tool + auto); Americas steady-to-soft with tariff overhang; India a genuine bright spot for CNC. Exact regional revenue percentages: n/a to a reliable current breakdown.

---

Phase B — Measure performance

Lens 5 · Earnings Result

The full-year FY2026 (ended 31-Mar-2026) print — now complete:

• Net sales ¥857.8B, +7.6% YoY.
• Net income ¥166.5B; EPS ¥178.47 (trailing).
• One source cites ordinary income ¥227.5B (+15.6%) for the full year — this looks high vs the ¥172.9B operating-income guidance and I flag it as a possible conflation of ordinary vs operating income, or a data error; treat operating income as ~¥172.9B guided / likely modestly beaten and ordinary income (which includes large FX + interest on the cash pile) as the higher figure. Conflict surfaced, not silently resolved. ``

The nine-month FY2026 result (the clean, reconciled data):

• Revenue ¥623.3B, +6.5%; operating profit ¥127.7B, +15.6%; net income ¥116.9B, +13.7%.
• Beat / raise: FANUC raised full-year sales guidance to ¥840.7B (+5.5%), op income ¥172.9B (+8.8%), NI ¥158.0B (+7.1%) — and the actual full-year came in above even the raised revenue guide (¥857.8B). "Exceeds market expectations" was the headline.
• What drove it: operating leverage — profit grew ~2.4x faster than revenue (op +15.6% on rev +6.5%), on China robot volume + a better mix + the service annuity. This is the bull's core evidence: structural margin strength even with FA demand decelerating.

Margins: Operating margin recovered to ~19.9% (FY-Mar-2025) and is expanding in FY2026. Gross margin 37.0% (FY-Mar-2025), recovered from a 5-yr low of 34.7% (FY-Mar-2024), still below the 40.3% peak (FY-Mar-2022). Data conflict: one aggregator cites "operating margin 24.68% end-2025" — this is inconsistent with the IR-derived ~20% and the ~19.4-21.4% figures from other sources; I prefer the ~20% IR-consistent figure and flag the 24.68% as likely a different metric/period or an error. n/a — the 24.68% is not reconcilable.

Balance-sheet flags: Pristine. Zero debt, ¥603B cash, ¥1,726B equity. No leverage, no covenant, no refinancing risk. Inventory/receivables move with the cycle but there is no distress signal.

Market reaction: The stock has surged — +86.66% one-year total shareholder return, +19.26% in 30 days, and a 52-week range of ¥3,699 → ¥8,880. Current ~¥7,026 (26-Jun-2026). Translation: the market has already priced the cyclical recovery and then some — the beat-and-raise was rewarded, and the stock now sits at a rich multiple (Lens 7).

Lens 6 · Earnings Calls (sentiment trend)

No transcripts on the research shelf (transcripts/ empty; FANUC's IR "prepared remarks" are Japanese-GAAP result PDFs, not Q&A calls, and the IR PDFs are binary-unfetchable). Sentiment reconstructed from result commentaries:

• ~2 years ago (China EV downturn): cautious/defensive — robot sales falling double-digits, EV capex "slowing for a second year," language around "weak demand in automobile-related industries" and destocking.
• ~1 year ago (bottoming): stabilizing — FA firm in India/China, general industry in Japan recovering, ROBOMACHINE surging +33%. Tone shifts from "downturn" to "mixed / early recovery."
• Latest (FY2026, recovery confirmed): constructive and raising — "sales increased significantly due to strong performance in EV-related and general industry applications" (China robot), profit growing faster than revenue, guidance raised, dividend up, buyback resumed.

Recurring phrases: "firm demand in India and China" (CNC), "EV-related," "operating leverage," "shareholder returns." What they stopped saying: the defensive "weak automobile-related demand" refrain of the downturn. What's conspicuously understated: the domestic-Chinese-competition threat — management frames China as demand tailwind, not competitive headwind, which is the exact blind spot the bear presses (Lens 13). Sentiment trend: decisively improved, arguably to the point of complacency on the China-substitution risk.

Lens 7 · Comps

FANUC vs global automation/robotics peers.

*FANUC 5-yr avg ROE: one source cites current ROE ~9.35%, ROIC ~12.19% — historically FANUC ran mid-teens ROE in up-cycles and high-single-digits in downturns; the large cash pile structurally depresses ROE (equity is bloated by ¥603B of idle cash). n/a — clean 5-yr ROE series not sourced to one place.

Read: FANUC trades at a premium P/E (~45x) to most Western automation peers (~20-30x) and near Keyence-like multiples without Keyence-like margins or growth. The market is paying for (a) the CNC monopoly, (b) the balance sheet, and (c) a bet that the cyclical recovery + physical-AI story extends. It is not cheap on any absolute or relative basis — the entire bear case (Lens 12/13) is that ~45x for a mid-single-digit structural grower with a commoditizing volume engine is a demanding price. Do not fabricate the peers' exact multiples — many are n/a above precisely because I could not source them cleanly today.

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

Pattern over recent years:

• China EV / automation capex inflections — the dominant driver. Robot sales −16% → stock derated in the 2023-24 downturn; robot +11% → stock nearly doubled off the ¥3,699 low to ¥8,880 into 2026. China is the beta.
• Earnings beats/raises — the FY2026 beat-and-raise + dividend hike + buyback drove the +19% 30-day move.
• Global machine-tool order data (JMTBA) — a leading indicator FANUC's FA book tracks; machine-tool order turns move the stock ahead of the print.
• Yen — a huge exporter; yen weakness flatters translated earnings and the stock, yen strength pressures both.
• Sector/thematic flows — the 2025-26 "physical AI / humanoid / reindustrialization" theme (NVIDIA GTC 2026, SoftBank-ABB) lifted the whole robotics complex, FANUC included.
• Capital-return announcements — buyback resumption + share retirements (13M+ shares retired May 2025) are repeatable catalysts.

What the market reacts to, distilled: China automation demand + yen + the AI/robotics narrative. It reacts less to the slow-burn competitive-erosion story — which is why that risk is under-priced (the bear's edge).

---

Phase C — Judge people & books

Lens 9 · Management

Structure & tenure:

• Kenji Yamaguchi — President & CEO since 2019 (director; joined FANUC 1993, ~30-yr company lifer).
• Founding family: FANUC spun out of Fujitsu in 1956 under Dr. Seiuemon Inaba, the father of Japanese numerical control. His son Yoshiharu Inaba chaired the board — and retired as Chairman/Director on 27-June-2025, marking the end of direct Inaba-family operating control. The family remains "culturally influential but not controlling"; ownership is now dominated by Japanese trust banks + global institutions. This is a genuine inflection: the founder-family era is over; FANUC is now a professionally-managed, institution-owned company.

Track record: Yamaguchi steered FANUC through the 2023-24 China downturn without balance-sheet stress and into a record-ish FY2026 with margins expanding — a competent cyclical stewardship. The multi-decade record (Inaba era) is one of the great industrial franchises: turning a Fujitsu lab into the global CNC monopoly.

Capital allocation: Conservative-to-a-fault, now loosening.

• Zero debt, ¥603B cash — historically criticized as a lazy balance sheet (idle cash depresses ROE; this is the classic Japanese-corporate cash-hoard critique).
• Response to activists/governance pressure: targets a 60% consolidated dividend payout ratio, raised DPS to ¥107.09 (from ¥94.39), retired 13M+ shares (May 2025), and authorized a buyback of up to 12.5M shares / ¥50B (Apr-2025). Capital return is improving but the base rate is still a company that sits on a mountain of cash.
• ROE ~9-13% is structurally capped by the cash — a leaner balance sheet could lift it meaningfully; the market gives partial credit for the improving returns.

Founder vs professional manager: As of mid-2025, decisively professional-manager (Inaba exit). Implication: more governance-driven capital return, less founder-idiosyncrasy, but also less of the founder's paranoid technical edge that built the moat — a subtle long-term watch item.

Red flags: Low. No related-party scandal surfaced; comp is not egregious by global standards; the main "flag" is the historic under-utilization of capital (a value-additive, not a fraud, concern). Governance is improving under TSE reform pressure (majority-ish outside directors).

Lens 10 · Forensic Red Flags

Accounting-risk scan (Japanese GAAP; web-only, no filings on shelf — figures ``):

• Revenue recognition: Hardware sold at delivery/acceptance; low aggressive-rev-rec risk (no long-dated software subscriptions inflating the top line). Service revenue is genuinely recurring. Low risk.
• Cash vs earnings: FANUC is a strong cash generator (EV/FCF ~30 implies real FCF; zero debt). No sign of earnings running ahead of cash — the opposite of the classic red flag.
• Receivables/inventory vs revenue: Move with the cycle (inventory built in the downturn, works down in recovery); no evidence of receivables outrunning sales into a channel-stuffing pattern. n/a — exact working-capital ratios not sourced today; flag as a "verify from the annual securities report (有価証券報告書)" item.
• SBC / non-GAAP flattery: Not a US-style SBC-heavy model; Japanese GAAP, modest stock comp. Low risk — margins are "real," not adjusted-real.
• Goodwill/intangibles: FANUC grows organically, not by acquisition — minimal goodwill, so no impairment landmine. (Contrast: many Western peers carry large deal goodwill.)
• The cash pile itself — ¥603B idle cash is a capital-efficiency issue, not an accounting one, but worth naming: it makes headline ROE look worse than the operating business is.
• FX exposure — a large exporter; ordinary income swings on yen (the ¥227.5B ordinary vs ¥172.9B operating gap is partly FX/interest). Not a red flag, but a reason headline "profit" figures need the operating line pulled out.

Regulatory findings (required sub-section):

• SEC (EDGAR EFTS — LR + AAER): Zero findings. ``. FANUC is not an SEC registrant.
• Non-SEC enforcement: Web scan ("Fanuc" (FTC OR DOJ OR export control OR settlement OR fine) enforcement) surfaced no material current enforcement action. The relevant category risk for a Japanese robot/CNC exporter is export-control / dual-use (CNC and robots have military-adjacent uses; Japan's METI and US/allied export regimes govern high-end machine-tool controllers to sanctioned end-users). No specific FANUC penalty found, but this is the enforcement axis to monitor. ``.
• 10-K Item 3 equivalent: N/a — FANUC files a Japanese 有価証券報告書, not a 10-K; its legal-proceedings disclosure was not fetched (IR PDFs binary-unreadable). Verify litigation from the annual securities report.
• Verdict: No material regulatory or legal findings — verified via SEC EDGAR EFTS (LR/AAER, zero), web search (no material hits), as of 2026-07-06. The only structural exposure is export-control on dual-use CNC/robotics, which is industry-wide and currently shows no FANUC-specific action.

---

Phase D — Project & stress-test

Lens 11 · Forward Projection

Base actuals: FY2026 (ended Mar-2026) — net sales ¥857.8B, net income ¥166.5B, EPS ¥178.47. Analyst consensus: FY2027 (ending Mar-2027) revenue ¥927.7B (+8.1%), EPS ¥206 (+15%), 20 analysts.

Three-year EPS path (``, arithmetic shown; do not treat as sourced consensus beyond the FY2027 line):

• FY2027 (ending Mar-2027): EPS ~¥206 — anchor on the sell-side. Revenue ¥928B, op margin ~20-21%, buyback shrinking share count ~0.5-1%/yr.
• FY2028 (ending Mar-2028) — base ~¥228 ``. Assumes China automation stays a net tailwind and mix holds.
• Bull ~¥245 ``.
• Bear ~¥185 ``.
• FY2029 (ending Mar-2029) — base ~¥248 ``; bull ~¥280; bear ~¥185 (structural-share-loss scenario flatlines EPS).

The projection's load-bearing assumption: that China robot demand remains a net positive for FANUC. If Chinese domestics (growing 25-40%/yr) take FANUC's China robot volume, the base case is too high and the bear case is the truth. This is a mix/share bet, not a demand bet — Chinese factory automation demand is rising; the question is whose robots fill it.

Forecast (Brier) — not logged per --watchlist unattended rules (skip forecast.ts create). If promoted to a thesis, the base call to log would be: "FANUC FY2028 non-GAAP EPS ≥ ¥225, p≈0.55, resolves 2028-05-15."

Lens 12 · Bull vs Bear

Bull case. FANUC is a compounding monopoly on the picks-and-shovels of automation. The CNC franchise (50-60% share) is close to un-attackable at the high end; the service annuity compounds double-digits regardless of cycle; the balance sheet (zero debt, ¥603B cash) makes it the last man standing in every downturn; and the physical-AI wave is a tailwind, not a threat — the March-2026 NVIDIA partnership (Jetson/Thor edge compute, Isaac/Omniverse simulation, OpenUSD digital twins) lets FANUC's 750k+ install base become smarter without FANUC having to win the foundation-model race. Add improving capital return (60% payout, buybacks, share retirements) and a cyclical automation up-turn, and you have a quality industrial re-rating higher. Operating leverage is proven (FY2026 profit +15.6% on rev +6.5%).

Bear case (2-3 things that could permanently impair the franchise):

1. China commoditization of the volume segment. Chinese domestics passed >51% of their home market in 2024; Estun (~9.5%) and Inovance (~8.8%) are now within ~1.5pts of FANUC's own ~11% China share, growing 25-40%/yr, pricing 20-40% below, and beginning to export (Chinese robot exports +60% H1-2025). FANUC's FY2026 growth came from China robot demand — the exact market being taken. If FANUC's China robot share halves over 5 years, its largest, fastest-growing segment stalls.
2. Multiple compression. At ~45x trailing, FANUC is priced for durable growth. A cyclical rollover or visible China share loss could de-rate it toward Yaskawa/ABB-like ~20x — a ~50% valuation air-pocket even with flat earnings.
3. Cyclicality mistaken for growth. The 52-week range (¥3,699→¥8,880) shows this is a highly cyclical name that the market periodically re-prices as a secular grower at the top of the cycle. Buying at ~¥7,000 after an 86% run risks buying the cycle peak.

Pre-mortem (18 months out, thesis broke): China robot orders roll over as domestics take share and the physical-AI narrative cools; FA stays soft in Europe/Japan; yen strengthens, gutting translated earnings; the stock de-rates from ~45x to ~28x on flat EPS — a 35-40% drawdown from ¥7,000. The tell will be China robot segment YoY decelerating while Chinese-domestic share rises in the same quarter.

Are multiples too high? Yes, on the numbers — ~45x for a company consensus has growing mid-single-digit to low-teens is demanding; the premium is a quality/monopoly premium that leaves no margin of safety.

Contrarian view (what the market refuses to see): The market is trading FANUC as a physical-AI winner when the physical-AI wave may actually accelerate its core-market commoditization — cheap Chinese robots + NVIDIA/open foundation models lower the barrier to "good enough" automation, eroding the reliability premium that is FANUC's whole moat. FANUC could win the narrative (NVIDIA partner) and lose the volume war simultaneously.

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

Dismantling the bull case:

• Where revenue is concentrated & the shift: Robot (~38% of revenue) is the growth engine, and it is disproportionately China-and-EV. That is the single most contested market on earth — Chinese domestics are taking it structurally, not cyclically. A short's thesis writes itself: the segment carrying the growth is the segment being commoditized.
• Why the moat is weaker than bulls think: The moat is high-end CNC + mission-critical reliability. But the marginal robot sold in China in 2026 is a general-industry arm where "20-40% cheaper and good enough" wins — and Estun/Inovance are in-house-integrated (their own servos + controllers), so they're not dependent on FANUC-style components. The moat protects the past (installed base) better than the future (net-new volume).
• The most dangerous competitor bulls underestimate: Inovance (300124.SZ) — not just a robot maker but a vertically-integrated servo/drive/PLC champion attacking FANUC's FA crown jewel (servos + controls) from below in China, with Estun taking the robot arm. Two-front war on both core segments, in FANUC's most important market.
• Worst capital-allocation history: The decade-long cash hoard (¥603B idle) — value destroyed via opportunity cost and depressed ROE. Only now, under TSE governance pressure, is it loosening. A short frames this as "management that under-earned on capital for years, re-rated to 45x by a momentum crowd."
• Assumptions that must hold for ¥7,000: (i) China robot demand stays a net FANUC tailwind; (ii) op margin holds ~20%+ despite price competition; (iii) the yen doesn't strengthen materially; (iv) the physical-AI narrative persists. Break any one and 45x is untenable.
• If growth disappoints 20-30%: consensus FY2027 EPS ¥206 → ~¥145-165; at a de-rated ~28x that's ¥4,000-4,600 — a 35-45% downside from ¥7,026.
• The single scenario that permanently impairs: FANUC becomes the "high-end niche" incumbent — retains prestige CNC + auto-OEM robots but cedes the entire high-volume general-industry/China mid-market to domestics, so its TAM share shrinks even as the robot market grows. Not bankruptcy (the balance sheet forbids it) but a structural de-rating from growth-monopoly to ex-growth quality cyclical. Plausibility: medium-high on a 5-year view given the China share data already printing.

Lens 14 · Management Questions (ordered by information value)

1. China robot share: What is FANUC's unit market share in China robots today vs 3 years ago, and what is your internal assumption for it in 2028 given Estun/Inovance growth? (The answer that most changes the thesis.)
2. When Chinese domestics price 20-40% below you in general-industry robots, do you defend share on price (margin hit) or cede volume and hold price (share loss)? Which, explicitly?
3. Is the ~¥600B net cash a permanent buffer or a transitional balance you intend to return down? What is the target net-cash level, and by when?
4. FA/servo is being attacked in China by vertically-integrated players like Inovance. How defensible is your CNC + servo position in the Chinese mid-market specifically (not the high end)?
5. What share of FY2026 robot growth was price/mix vs volume, and how much of the volume was China EV specifically?
6. On the NVIDIA physical-AI partnership: does open/foundation-model robotics lower the barrier for cheaper competitors and erode your reliability premium? How do you monetize AI without commoditizing yourself?
7. What is your through-cycle operating-margin floor, and how low did contribution margins go at the 2023-24 trough?
8. Post-Inaba-family exit, how does capital allocation and long-term R&D strategy change under fully professional management + institutional ownership?
9. What is your realistic ROE/ROIC target on a normalized balance sheet, and what gets you there?
10. Export-control/dual-use: how much revenue is exposed to tightening restrictions on high-end CNC/robots to sanctioned end-users, and how do you manage it?
11. Service is +74.5% over 4 years — is that attach-rate gains (structural) or installed-base catch-up (one-time)? What's the steady-state growth?
12. Manufacturing is concentrated at Oshino — what is the single-site / single-country (Japan earthquake, yen) risk plan?
13. Where do cobots (CRX) and non-factory automation fit — a growth vector or a defensive hedge against arm commoditization?
14. Given the 52-week range doubled the stock, do you think the current multiple reflects FANUC's normalized earnings power, or the top of a China capex cycle?
15. What is the one competitive development (Chinese domestic, a new entrant, or an AI platform shift) that keeps you up at night?

---