Hexagon AB

Phase A — Understand the business

Company Overview

Model in plain terms. Hexagon sells the instruments and software that let industrial customers measure physical reality precisely, then act on it — coordinate-measuring machines and laser scanners that check whether a manufactured part matches its CAD model; GNSS/positioning gear that tells a tractor, drill, or drone exactly where it is; and the analytics that close the loop. It is a sensor-to-software platform: hardware pulls the customer in, recurring software and services deepen the relationship. Group net sales €5,426.5M FY2025 (pre-spin, "operating net sales"). Recurring revenue was €567.4M in Q4 2025 with ~3% organic growth — i.e. the model is still majority transactional hardware/perpetual-license, not a SaaS book (a key contrast with the Octave business it just shed).

Post-Octave shape (the entity you buy today). Excluding Octave, Hexagon had ~€3,953M revenue and ~29% adjusted EBIT margin (FY2024 basis), ~17,600 employees. Three reporting divisions remain:

• Manufacturing Intelligence (MI) — metrology hardware + CAE/production software (Q4-25 organic +1%, 28.4% op margin).
• Geosystems — surveying, laser scanning, reality capture (Q4-25 organic −1%, 28.4% op margin).
• Autonomous Solutions (AS) — assured positioning + autonomy + robotics (Q4-25 organic +23%, 34.5% op margin). This is the growth engine and highest-margin leg.

Customers / suppliers / competitors. End-customers span automotive, aerospace & defence, electronics/semis, construction & survey, agriculture, mining, energy — highly diversified, no single-customer concentration disclosed (unlike a hyperscaler-fed name). Key rivals differ by leg: metrology → ZEISS, Mitutoyo, Keyence, KLA, FARO, Renishaw; positioning/geospatial → Trimble, Topcon (Hexagon revenue ≈158% of Trimble's); adjacent industrial software → Dassault, Siemens, PTC, Autodesk, ESRI. Suppliers are electronics/optics/actuator vendors (AEON uses Swiss maxon actuators, NVIDIA Jetson).

Contract structure. Mixed: perpetual + capital-equipment sales (lumpy, cycle-sensitive — visible in the China/automotive softness), plus a growing recurring/subscription tail. No take-or-pay. Cash conversion is strong (121% in Q4-25), which is the model's redeeming financial feature.

Supply Chain

Map: upstream inputs → Hexagon → end customer.

• Upstream: precision optics & photonics (scanners, laser trackers); MEMS/GNSS chipsets and RF for NovAtel positioning; motion/actuator suppliers (maxon for AEON); edge-compute (NVIDIA Jetson Orin in AEON; NVIDIA Isaac/Omniverse for sim-to-real); CMM structures, granite, precision mechanics; contract electronics manufacturing. Software stack is largely in-house (accreted via ~200 acquisitions).
• Hexagon (the integrator): designs the sensor, fuses it with software, sells the system + support. Value-add is calibration, accuracy, and the software layer — not raw component manufacture.
• Downstream / named buyers & partners: automotive OEMs and tier-1s (metrology on the line); Schaeffler and Pilatus as named AEON pilot customers; defence primes and government (NovAtel APNT/anti-jam, Hexagon US Federal); agriculture OEMs (positioning/machine-control kits, NovAtel/Leica ag); survey/construction firms (Leica Geosystems); mining majors (autonomy).
• Chokepoints / single-source risks: GNSS RF and high-end NVIDIA compute are the tightest upstream links for the autonomy/robotics push; precision-optics supply is specialised but multi-sourced. No disclosed single-source dependency that would halt the group. Geographic chokepoint = China on the demand side (see Lens 5) — not a supply chokepoint but a revenue one.

Names present → lens satisfied. Underlying map is -derived (no supply-chain.md entry for Hexagon specifically; the robotics-wiki commercial layer is generic and not company-specific).

Competitive Advantages (moats)

1. Installed base + switching costs (strongest moat). A CMM or survey fleet, once calibrated into a customer's QA/production workflow with Hexagon software, is sticky — requalifying a metrology line on a rival's stack is expensive and risky. Decades of installed hardware create a recurring-software and services annuity.
2. Accuracy/brand in mission-critical measurement. In aerospace/defence and semis, "measured on a Hexagon/Leica" is a trust signal; Leica Geosystems and NovAtel are category brands. Perceived value is high where tolerance failure is catastrophic.
3. Sensor + software vertical integration. Owning both the sensor and the analytics lets Hexagon sell outcomes (autonomy, digital twins) competitors selling only one layer can't fully match — the AS division's 34.5% margins and 23% growth are the proof.
4. Scale in a fragmented field. ~€4bn RemainCo dwarfs FARO/Renishaw; the ~200-acquisition roll-up bought category coverage rivals can't replicate quickly.

Bargaining power. Over suppliers: moderate-to-high (commodity electronics, multi-sourced) — except NVIDIA compute where Hexagon is a price-taker. Over customers: high in locked-in metrology/positioning niches, weaker in cyclical capital-equipment sales where automotive buyers defer.

Moat caveats (the bear's opening). Much of the "moat" was purchased, not organically compounded — and the Viceroy critique (Lens 10/13) is precisely that acquired growth was dressed as organic. A moat built by continuous M&A needs continuous M&A to look like it's growing. positioning.md/bottlenecks.md on disk are robotics-wiki-generic, not Hexagon-specific, so this lens is +reasoning.

Segments

Group-level, most recent clean quarterly read (Q4 2025, pre-Octave-spin reporting), all:

Q1 2026 (first post-spin quarter, continuing-ops): group revenue €964M, reported ~flat, +8% organic (after −6% FX, −1% structural from the Design & Engineering disposal). By division: MI €433M (+9% organic), Geosystems €349M (+2%), AS €176M (+13%). Group gross margin 62.9% (down from 64.4% Q1-25), EBIT €251M (26.1% margin).

Geography (Q4-25 organic): Americas +11%, EMEA +4%, China −5%, Rest of Asia −14%. Q1-26 China −4%. Read: the Americas are carrying the group; China + Asia + automotive-EMEA are the drag. The trend is a widening Americas/Asia divergence.

Trend + cause. AS is accelerating (positioning/defence tailwind + robotics narrative); MI re-accelerated in Q1-26 (+9% vs +1% in Q4-25) on order book; Geosystems is bottoming as China destocking ends. The overall FY2025 organic figure is where sources diverge — CMD 2026 cites 2.6%; the Year-End Report headline was quoted as "2%". I flag the conflict rather than resolve it; both are low-single-digit and directionally identical (sluggish).

---

Phase B — Measure performance

Earnings Result (latest print = Q1 2026, first post-Octave quarter)

unless noted.

• Revenue €964M, reported growth ~flat YoY; +8% organic — the organic number is genuinely strong and the headline "flat" is entirely FX (−6%) + the Design & Engineering divestiture (−1%). Beat/miss vs consensus not cleanly sourced → n/a on the consensus delta.
• Drivers: MI +9% organic (order-book strength), AS +13% (positioning/autonomy), Geosystems +2% (China destocking ending). Americas strong, China −4%, automotive-EMEA weak.
• Margins: gross 62.9% (−150bps YoY, 64.4%→62.9%) — pressured by a full quarter of tariff impact vs near-zero in the comparison period, partly offset by new-product ramp. EBIT €251M, 26.1% margin.
• Guidance/tone: Management flagged Q2 will still carry FX + tariff headwinds, but expects benefit from the MI order book and the end of Geosystems' China destocking. Attentive to tariffs, currency, Middle East.
• Balance-sheet flags: Cash flow post-investment €250M (+16% YoY); cash conversion 77% (continuing ops). Group net debt €3,395.9M at Dec-2025 (pre-spin; post-spin net debt not yet cleanly disclosed).
• Market reaction: stock −10.2% on the print — the market punished the gross-margin compression and FX despite the strong organic line. Tell: after years of the Viceroy overhang, investors now trade Hexagon on reported margin quality, not the organic-growth headline management leads with.
• Unusual vs own history: the +8% organic is the best in several quarters, yet the stock fell — a sentiment/skepticism discount is in force (see Lens 8).

Earnings Calls (sentiment trend)

No transcripts/ on disk → only, from Q4-2025 and Q1-2026 coverage.

• What management is focused on now: the Octave separation (executed May-2026), the new CMD 2026 financial plan (4-6% organic / 24-26% EBITDA / 90-100% conversion, and the new EBITAC metric), AEON/robotics commercialisation, and defending margins against FX + tariffs.
• Tone shift (Q4-25 → Q1-26): from "record gross margin overshadowed by margin pressure" (Q4-25 framing) to "strong organic growth amid currency headwinds" (Q1-26). The recurring worry across both calls is the narrowing gap between R&D capitalisation and amortisation, which management itself says will "create additional pressure on margins in the near term" — notable because R&D-capitalisation was a core Viceroy allegation, and management is now openly guiding to its margin drag.
• Things they started saying: "EBITAC," "return on R&D," "capital allocation discipline" — a governance-response vocabulary. Things they stopped emphasising: aggressive organic-growth-vs-LVMH-style framing; the tone is more defensive/credibility-rebuilding.
• Leadership instability is itself a sentiment signal: CEO churn (Guglielmini out Nov-2024 → Norbert Hanke interim), and a CFO gap (David Mills stepped down Aug-2025, Hanke covered interim CFO). Calls led by an interim CEO/CFO carry a credibility discount.

Comps

Peer set = precision-measurement / positioning / industrial-software.

Hexagon's own multiple — explicit conflict, do not average: aggregators returned P/E 9.83 (trailing) and 24.39 (forward) from one feed vs 40.33 trailing / 21.58 forward from another; EV/EBITDA 6.27 vs 16.36. These are irreconcilable (the ~6x/9.8x figures almost certainly reflect a spin-off distortion — the Octave distribution inflates trailing EPS / deflates 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. mechanically). The honest read: trust none of the trailing prints across the May-2026 spin; on a clean forward basis Hexagon has historically traded at a premium 20–30x P/E software-adjacent multiple and the ~21–24x forward figures are the only internally-plausible ones. Derived sanity check: market capMarket capitalisationThe share price multiplied by the number of shares. What the market says the equity is worth. ~SEK 222bn + net debt ~SEK ~38bn (€3.4bn) ≈ 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. ~SEK 260bn; against a RemainCo EBIT run-rate of roughly €1.0–1.1bn (~SEK 11–12bn) that is ~22x EV/EBIT — a full, growth-priced multiple, not a value multiple, once you ignore the corrupted aggregator trailing figures.

Stock-Price Catalysts (moves >5%, ~5yr)

Mostly.

• Sep-2021 — all-time high (~16.79 split-adjusted close): peak of the COVID-era industrial-software re-rating.
• 2022 de-rating: rates/multiple compression across European industrial-software; no single event, a secular drawdown.
• Jul–Aug 2023 — Viceroy short reports ("Six Sides to Every Story," "Betting on Short Memories"): sharp drops on fraud/related-party allegations; Hexagon issued a 16-page rebuttal, Greenbridge disputed the front-running claim. This is the defining overhang — the market's skepticism discount dates to here.
• 2024 — CEO change (Guglielmini out, Hanke interim, Nov-2024): leadership uncertainty.
• 2025 — Octave spin-off announcements (Feb→May 2026 execution): the value-unlock catalyst; the split into two listed entities.
• Apr-2026 — Q1 print −10.2%: gross-margin miss + FX.
• Now (Jul-2026): 80.58 SEK, near the 52-wk low (74.58) vs high 102.87; consensus target ~95.78 SEK, "Buy" (5 buy / 2 sell).

Pattern: for Hexagon the market reacts to (1) governance/credibility events (Viceroy, CEO churn) and (2) reported margin quality/FX — far more than to the organic-growth headline. It is a "show me clean numbers" stock. That is the whole investment debate in one sentence.

---

Phase C — Judge people & books

Management

• Ola Rollén — Chairman (ex-CEO, 1998→2000s architect of the roll-up). The dominant figure. Built Hexagon from a mini-conglomerate into a €5bn measurement group via ~200 acquisitions — a genuinely impressive operator-capital-allocator record by revenue/scale. But Rollén was the subject of an insider-trading case in Norway (acquitted), and Viceroy's central governance allegation is the Greenbridge investment-vehicle overlap with Hexagon top brass (front-running claim, disputed). Chairman with founder-level control → founder-archetype governance in a professional-manager era.
• Norbert Hanke — interim President & CEO (since Nov-2024). Hexagon lifer (joined 2001), ex-COO, ran MI for 10+ years — deep operational credibility, but "interim" for ~1.5 years is a governance yellow flag. Also covered interim CFO after David Mills left (Aug-2025); has since moved to EVP over ventures/ops/strategy per Q1-26. A group this size running with interim C-suite for this long is not a position of strength.
• Ownership / skin in the game: Melker Schörling Tjänste AB (MSAB) ~22% and the controlling voter historically. Melker Schörling died Dec-2023; MSAB (family) remains the anchor holder — so control persists through the estate/vehicle, concentrated and dynastic. No our figures on disk → insider buy/sell trend n/a.
• Capital-allocation history: the defining trait — serial M&A (Brown & Sharpe, Leica Geosystems ~CHF1.3bn 2005, Intergraph ~$2.1bn 2010, Infor EAM ~$0.8bn 2021, Waygate ~$1.45bn Apr-2026). Now pivoting to disciplined bolt-ons + progressive dividend + the Octave spin as the value-unlock. Dividend has actually been declining (~−20% 3yr CAGR; €0.14 FY2025) — capital went to deals, not distribution.
• Red flags: related-party/Greenbridge overlap (disputed), R&D-capitalisation aggressiveness (management now guiding to its unwind), interim C-suite, dynastic control with high-vote structure. Archetype: founder-controlled roll-up transitioning — uneasily — to institutional governance.

Forensic Red Flags

Ground truth from disk (regulatory/regulatory-findings.md): Hexagon has no CIK → no SEC LR/AAER search possible; SEC findings = 0 by construction, not by exoneration. All accounting concern is web/analysis-derived — label accordingly.

Risk map across the statements:

• Revenue recognition / organic-vs-acquired growth (THE issue). Viceroy's core allegation: acquisition growth reported as organic. For a ~200-deal roll-up, the organic/inorganic boundary is genuinely hard to audit from outside, and the company's disclosures were criticised as thin. This is the single most important forensic question and it is structurally unresolved from public data.
• R&D capitalisation (income-statement flattery). Hexagon capitalises development spend; the cap-vs-amortisation gap flatters current EBIT and management itself now says the narrowing gap "will create additional pressure on margins". The new EBITAC metric (adj EBIT excluding capitalised/amortised R&D) is effectively management conceding the point — EBITAC margin was only ~22% in 2025 vs the ~29% EBIT margin. That ~7pt gap is the capitalised-R&D flatter, quantified by the company's own new metric. Enormous tell.
• Goodwill/intangibles (balance-sheet risk). ~200 acquisitions → a goodwill-and-intangible-heavy balance sheet; total assets ~$18.1bn (2023), equity €11,196.1M with a 62.7% equity ratio (2024). Goodwill line not cleanly sourced (n/a on the exact figure) but is unambiguously large. Impairment risk on the "underperforming acquisitions" Viceroy flagged.
• Cash flow vs earnings. The redeeming feature: cash conversion 121% (Q4-25), 77% continuing (Q1-26) — cash generation is real and high, which is the strongest single rebuttal to the "no cash flows" Viceroy line. Where earnings quality is debatable, cash quality is good — that tension is the crux of the bull/bear.
• Receivables/inventory vs revenue, leases, contingencies: not cleanly sourced quarter-by-quarter → n/a; flagged as items to pull from the annual report if the shelf is ever backfilled.

Regulatory findings (required sub-section).

• SEC (EDGAR LR/AAER): none — no CIK, not exonerated, just outside SEC jurisdiction (regulatory-findings.md, 2026-07).
• Non-SEC web search ("Hexagon AB" (FTC OR DOJ OR FDA OR CFPB OR consent decree OR settlement OR fine OR penalty) enforcement): no material government enforcement action found as of 2026-07; the only adversarial event of record is the Viceroy short-seller campaign (2023) — a private research allegation, NOT a regulator action, and Hexagon/Greenbridge publicly disputed it. Rollén's earlier Norwegian insider-trading case ended in acquittal.
• Item 3 (Legal Proceedings): n/a — no 10-K exists (foreign filer, no EDGAR). The Swedish annual-report equivalent is not on the shelf → not sourced.
• Net: No regulator has sanctioned Hexagon. The accounting concerns are (a) a 2023 short-seller thesis the company rebutted and (b) the company's own EBITAC disclosure now quantifying the R&D-capitalisation flatter. Verified via regulatory-findings.md (SEC LR/AAER, nil-by-jurisdiction), web enforcement search, and public 2023–2026 record as of 2026-07-06.

---

Phase D — Project & stress-test

Forward Projection

No our model logged (unattended watchlist rule). All, built off the Q1-2026 continuing-ops run-rate + CMD 2026 plan. Note: EPS in EUR; Hexagon reports EPS ~€0.11–0.12/quarter pre-spin — post-spin RemainCo EPS base is lower and not yet cleanly disclosed, so I project at the group-margin/growth level and give an indicative EPS band, not a false-precision figure.

Anchor facts: Q1-26 revenue €964M, +8% organic, 26.1% EBIT margin; CMD plan 4-6% organic / 24-26% EBITDA (2026-2030).

Key input lines (base): industry growth mid-single-digit; AS the share-gainer (+13% now); tariffs a ~150bp gross-margin headwind through FY26 then easing; R&D cap-vs-amortisation unwind ≈ −100 to −200bps EBIT drag (the EBITAC/EBIT ~7pt gap closing gradually); FX the swing factor (−6% revenue in Q1-26). DilutionDilutionIssuing new shares, so each existing share owns a smaller slice of the same company. negligible (buybacks/dividends modest). Base call: Hexagon RemainCo compounds revenue mid-single-digit organic with EBIT defended near 26%, EPS high-single-digit — a steady-but-unspectacular industrial compounder, NOT a re-rating growth story unless AS/robotics inflects.

If committing (not done here): <HEXA-B.ST> FY26 organic revenue growth >= 5%, p≈0.55, resolves 2027-01-31.

Bull vs Bear

Bull case. A cleaner, more focused business post-Octave: measurement + the highest-margin, fastest-growing Autonomous Solutions leg (23% organic, 34.5% margins) now a bigger share of the mix. Deep moats in mission-critical metrology/positioning; ~121% cash conversion says the earnings turn into cash. The AEON humanoid + NVIDIA partnership is a credible optionality call on industrial embodied-AI — free upside not in the price. New CMD 2026 plan (4-6% organic, EBITAC discipline) signals a governance/capital-allocation reset. Stock near a 52-wk low with a "Buy" consensus (~95.78 vs 80.58 SEK, ~19% upside) — the FX/tariff/China drags are cyclical and self-liquidating (China destocking ending, tariffs annualising). Contrarian upside: the market is still applying a Viceroy-era skepticism discount to a company that just introduced the exact transparency metric (EBITAC) the shorts demanded.

Bear case. Three ways it permanently impairs: (1) The growth model is the M&A treadmill — strip acquisitions and organic growth is ~2-3%; the RemainCo must now prove it can grow without the deal machine, and the CMD 4-6% floor is aspirational vs a 2.6% 2025 actual. (2) Earnings quality unwinds in slow motion — the EBITAC/EBIT ~7pt gap says ~a quarter of "profit" is capitalised R&D; as that cap-vs-amort gap narrows (management-guided), EBIT margins grind down 100-200bps regardless of operations. (3) Governance/control overhang persists — dynastic 22% high-vote control, interim CEO/CFO for ~1.5 years, unresolved related-party questions; a "show me" stock stays discounted until leadership is permanent and disclosures deepen. Expectations baked into ~22x forward EV/EBIT are full, so any organic disappointment de-rates hard (the −10% Q1 reaction is the template).

Pre-mortem (18 months out, thesis broke): China/auto never recovered, AS decelerated as the defence/positioning cycle cooled, the R&D-cap unwind dragged EBIT below 24%, FX stayed adverse, AEON stayed a demo — and the stock re-rated toward the low-70s (52-wk low) or below as the "compounder" label gave way to "ex-growth industrial with accounting questions."

Are multiples too high? For a mid-single-digit organic compounder, ~22x forward EV/EBIT is priced for re-acceleration that hasn't shown up yet. Fair-to-full, not cheap — the 52-wk-low price is a sentiment discount, not a valuation discount.

Contrarian view (what the market refuses to see): Both bulls and bears are fighting the 2023 war. The real question isn't "was Viceroy right" — it's "can the RemainCo, shorn of its software crutch (Octave) and its M&A treadmill (new discipline), actually grow organically at 4-6%?" Nobody has priced the answer because there's only one post-spin quarter of data. The +8% Q1-26 organic is either the start of the proof or a pre-spin flatter — and that ambiguity is the entire edge.

Devil's Advocate (short-seller)

Dismantling the bull case, channeling and extending Viceroy:

• What structurally breaks the model: Hexagon "grows" by buying growth. Post-Octave and post-M&A-discipline, the organic engine is exposed — and it's 2-3% ex-deals. A roll-up that stops rolling reveals its true growth rate, and here that rate barely beats GDP.
• Revenue concentration risk: not customer-concentrated, but geography- and cycle-concentrated — automotive + China are structural soft spots, and Americas strength is doing all the lifting. If the US industrial cycle turns, there's no China to offset.
• Why the moat is weaker than bulls think: switching costs are real in installed metrology, but new-system sales are competitive capexCapital expenditureMoney spent on long-lived things — buildings, machines, servers — rather than on running costs. decisions where ZEISS/Keyence/Trimble compete hard on price and where customers defer in downturns. The "moat" doesn't protect growth, only base.
• Most dangerous underestimated competitor: Trimble (positioning/geospatial overlap) and Keyence (metrology, famously higher-margin and more efficient) — Keyence earns Hexagon-beating margins without the goodwill mountain, exposing how much of Hexagon's "quality" is financial engineering.
• Worst capital-allocation / accounting: R&D capitalisation (now self-admitted via EBITAC), ~200 acquisitions with "a vast number underperforming" (Viceroy), the Greenbridge related-party overlap, and a declining dividend despite "progressive dividend" language. The Octave spin can also be read cynically — offloading the slower-growth software into a separate entity to make RemainCo's headline growth look better.
• Assumptions that must hold for the price: 4-6% organic (vs 2.6% actual), EBIT defended near 26% (vs an admitted R&D-cap unwind), FX normalising, AS staying at 20%+ growth.
• If growth disappoints 20-30% (i.e. organic +3-4% instead of 5-6% with margins to 24%): the ~22x forward multiple is unjustifiable; fair value compresses toward SEK 70 (the 52-wk-low zone) — ~15% downside from 80.58.
• Single scenario that permanently impairs: a goodwill impairment cycle on the "underperforming acquisitions" combined with the R&D-cap unwind, in a China/auto downturn — turning the "compounder" into an ex-growth industrial re-rated to mid-teens EV/EBIT. Plausibility: moderate — no catalyst forces it, but the ingredients are all on the balance sheet.

Management Questions (ordered by information value)

1. Post-Octave and under the new M&A-discipline policy, what is the RemainCo's organic revenue growth algorithm — and how do you bridge from 2025's 2.6% actual to the 4-6% CMD floor without acquisitions? (Answer most changes the thesis.)
2. You introduced EBITAC (excl. capitalised/amortised R&D). Over the next 3 years, how far does the cap-vs-amortisation gap narrow, and what is the EBIT-margin drag path as it does?
3. Of the ~200 acquisitions, what fraction are underperforming their acquisition-case IRR, and what is the resulting goodwill/intangible impairment risk under a China/auto-downturn scenario?
4. When does the interim C-suite become permanent — is Norbert Hanke the CEO, and who is the permanent CFO? What does the timeline signal about board conviction?
5. Precisely how do you delineate organic vs. acquired growth in disclosure, and will you publish the acquisition-contribution reconciliation the 2023 critics asked for?
6. Autonomous Solutions is 23% organic / 34.5% margin — how much is the durable defence/positioning cycle vs. a cyclical peak, and what's the normalised growth/margin?
7. What is the realistic revenue timeline for AEON/robotics — pilot-to-production with Schaeffler/Pilatus — and what capex/opex does scaling it require before it's P&L-positive?
8. On Greenbridge and related-party arrangements involving current/former leadership: what governance changes have you made since 2023 to eliminate the conflict-of-interest perception?
9. China is −4/−5% organic — is this cyclical destocking (your framing) or structural share loss to domestic competitors, and what's the evidence?
10. Capital allocation: you call the dividend "progressive," yet it's declined ~20%/yr over three years — reconcile that, and give the buyback-vs-dividend-vs-M&A framework post-spin.
11. Tariffs cost ~150bps of gross margin in Q1-26 — what's the mitigation (pricing, footprint, sourcing) and the residual structural drag?
12. What is RemainCo's post-spin net debt and leverage, and how does the separation change your M&A firepower?
13. How exposed is the autonomy/robotics roadmap to NVIDIA compute supply and pricing, and what's the second-source plan?
14. Which single division would you divest or de-emphasise next, and what's the portfolio logic now that Octave is gone?
15. What KPI should investors hold you to that would prove the "measurement-and-autonomy compounder" thesis over the software-roll-up skepticism — and what's your line in the sand on it?

---