Novanta

Phase A — Understand the business

Lens 1 · Company Overview

Novanta is a picks-and-shovels component supplier to mission-critical machines — it does not sell finished systems to end users; it sells the laser, the scanner, the encoder, the force-torque sensor, the insufflator, the light engine that an OEM designs into its surgical robot, sequencer, semiconductor tool, or warehouse robot. The "design-in" model is the whole business: once Novanta's part is qualified into an OEM's platform, "there are generally significant barriers to subsequent supplier changes until the end of the product or system life cycle, especially in the medical market". That is the moat and the curse — long, sticky sockets, but multi-quarter qualification cycles and exposure to the OEM's own demand swings.

Two reportable segments (FY2025):

• Automation Enabling Technologies — rev $500.8M, GM 47.8%, segment op income $114.5M (22.9% margin). Product lines: laser beam steering/scanning (scan heads), laser sources (diode/DPSS/CO₂), robotic & precision motion (position/force-torque sensors, servo drives, motion-control SW), robotic end-of-arm tooling (the legacy ATI Industrial Automation franchise — tool changers, collision sensors), and air-bearing spindles (PCB/semiconductor). This is the "robotics/photonics" engine and the AI-datacenter beneficiary.
• Medical Solutions — rev $479.8M, GM 41.6%, segment op income $51.2M (10.7% margin). Product lines: medical insufflators/endoscopic pumps + disposables (W.O.M. franchise), imaging/identification/RFID (now incl. Keonn), advanced motion control, light engines (DNA sequencing), integrated OR technologies.

End markets / customers: medical ~53%, advanced industrial ~47% of revenue. Concentration is real but diversified across a base: in FY2025 two OEM customers each ran ~12% and ~11% of consolidated revenue (both medical), and the top ten customers = ~42% of sales. Contract structure is OEM purchase-order-driven (not take-or-pay, no minimum-purchase commitments) — backlog $481.2M at YE2025 vs $445.5M YE2024. Revenue is "not highly seasonal" but Q1 is the soft quarter (customer capital-budget cycles).

`` framing: management/sell-side describe Novanta as "a global technology partner specializing in mission-critical photonics, vision, and precision motion components and sub-systems" with a rising AI-data-center tilt.

Lens 2 · Supply Chain

Upstream → Novanta → end customer, named where the filings name them:

• Inputs: "a wide variety of raw materials, components and parts… generally available from alternative sources," but with explicit single-source / limited-source dependencies for certain components and subassemblies — Novanta sometimes re-engineers parts to route around shortages. Optics, laser diodes, precision motors, semiconductors and electronics are the sensitive input classes. The 10-K flags that single-source disruption "would not create substantial manufacturing delays and additional cost" cannot be assured.
• Novanta manufacturing (internal-heavy, regionalizing): AET plants — Apex NC, Bedford MA, Marlborough CT, Mukilteo WA, Rocklin CA, Suzhou China, Taunton + Manchester UK. Medical plants — Irvine CA, Syracuse NY, Ludwigsstadt Germany, Manchester UK, Barcelona Spain (Keonn), Přelouč Czech Republic. A 2025 restructuring program is explicitly "regionalizing the manufacturing footprint" and building "centers of excellence" — and a "significant capacity expansion project in China" is inside the $30–40M FY2026 capex plan. Strategic read: dual-purpose — tariff/onshoring hedge and the source of the near-term margin drag (temporary cost duplication during the move).
• Downstream OEM buyers (the actual sockets): surgical-robotics OEMs (the two ~11–12% medical customers), minimally-invasive-surgery OEMs, DNA-sequencing instrument makers, warehouse-automation / humanoid robot OEMs, semiconductor & PCB equipment makers, metrology and additive-manufacturing OEMs, and — increasingly — AI-data-center / silicon-photonics supply chains (~15% of revenue, see Lens 4/8).

Chokepoints: certified medical lines (ISO 13485 / cGMP / FDA) cannot be re-sourced quickly; single-source electronic components; and Novanta's own qualification status inside each OEM platform is itself the chokepoint that protects revenue.

Lens 3 · Competitive Advantages (moats)

Moat = switching costs + breadth-of-portfolio applications knowledge, not a single patent. The 10-K is explicit that "we do not believe that any individual patent is material to our business as a whole" — protection comes from "market position, technological innovation, know-how, application knowledge and product performance". Translation: the moat is the design-in socket — once qualified into a regulated OEM platform, displacement requires the OEM to re-qualify, re-validate and (for medical) re-certify, which they almost never do mid-lifecycle.

Supporting moat evidence: vitality/new-product index 27% of sales, design wins +~30% YoY, book-to-bill 1.10, bookings +37% YoY in Q1-FY26 — i.e. the company keeps winning new sockets, which is the leading indicator of the moat compounding. Gross margins (AET 47.8%, blended adj GM ~46%) are evidence of pricing power on differentiated parts; the medical segment's stickier disposables (insufflator consumables) add a razor/blade annuity.

Bargaining power is mixed: strong over most suppliers (multi-sourced, re-engineerable inputs) but weaker over its largest OEM customers — two customers at ~11–12% and top-10 at 42% means the biggest buyers have leverage on price at renewal, and "none of our significant customers has entered into an agreement… to purchase any minimum quantity". Durable but not unassailable: the bottleneck is "we are qualified and they don't want to re-qualify," which decays at each platform refresh.

Lens 4 · Segments

All figures `` unless noted.

By segment — FY2025 vs FY2024:

The single most important segment fact: AET margins EXPANDED (21.7%→22.9%) while Medical margins COMPRESSED (12.5%→10.7%) and unallocated corporate cost ballooned +34.4% ($53.4M→$71.7M) — the corporate line, not the operating segments, is what dragged consolidated operating income down 15%. Unallocated grew on "the planning and design phase of our financial and operation system implementation" (an ERP build), 2025 restructuring, and "costs incurred in connection with an insurance recovery claim."

Within AET (FY2025): robotics & automation +$30.3M, precision-manufacturing products −$20.0M — i.e. the robotics/AI side is carrying a declining legacy industrial-precision side. Within Medical (FY2025): advanced surgery +$33.7M (the surgical-robotics/MIS recovery), precision medicine −$12.6M (lower demand) — partly Keonn-aided.

By geography (FY2025, % of revenue): United States $519.0M / 52.9% (rising from 47.4% in 2023), Germany $114.6M / 11.7% (declining), Rest of Europe $137.4M / 14.0%, China $93.2M / 9.5% (rising), Rest of Asia-Pacific $97.0M / 9.9%. US-mix is climbing — favorable on tariffs/onshoring, and consistent with the AI-datacenter tilt.

Latest quarter (Q1-FY26): AET rev $131.2M / seg op inc $31.1M (23.7%); Medical rev $126.5M / seg op inc $16.7M (13.2%); segment op income $47.9M less unallocated −$20.3M (~7.9% of revenue, still elevated) = consolidated op income $27.5M.

Trend: AET decelerating but high-margin and AI-levered; Medical reaccelerating (advanced surgery) but lower-margin; the swing factor for the whole P&L is the corporate/unallocated cost line normalizing once the ERP + regionalization spend rolls off.

---

Phase B — Measure performance

Lens 5 · Earnings Result

FY2025 full year (GAAP):

The headline tension, stated plainly: revenue is at an all-time high and grew every year, yet GAAP net income and EPS have fallen for two straight years ($72.9M→$64.1M→$53.8M; $2.02→$1.77→$1.47). FY2025 operating income fell $16.6M / −15.0%, driven by SG&A +$19.7M, restructuring/acquisition costs +$8.9M, amortization +$1.7M, partly offset by gross profit +$13.7M.

Adjusted (non-GAAP) FY2025: adj EBITDA $221.0M (+5.3%, 23% margin), adj diluted EPS $3.29 (+6.8% vs $3.08), adj GM ~46%. The gap between $1.47 GAAP and $3.29 adjusted EPS is enormous (~$1.82, i.e. adjusted is 2.24× GAAP) — bridged by purchased-intangible amortization ($27.5M), restructuring/acquisition costs ($22.7M), and SBC ($29.5M). That gap is the single biggest analytical fork in the whole name (see Lens 10).

Latest quarter — Q1-FY26 (ended 2026-04-03): revenue $257.7M (+10.4% YoY, but ~+3% organic; +$9.0M/3.8% acquisition, +$8.2M/3.5% FX); gross profit $113.6M (GM 44.1% vs 44.7%); operating income $27.5M (−15.1% YoY) on SG&A +$8.8M and restructuring swinging from a +$2.5M credit to a −$2.6M cost; net income $21.1M (flat YoY) but diluted EPS $0.51 vs $0.59 — the decline is almost entirely dilution (diluted shares 41.16M vs 36.13M, from the Nov-2025 tangible-equity-unit offering), not operating collapse. Bookings +37% YoY, book-to-bill 1.10, adj EBITDA +14% (margin +70bps).

Balance-sheet flags (YE2025): cash $380.9M (up from $114.0M — the TEU raise), total debt $250.8M (current $38.3M + LT $212.5M, down from $416.6M) → net cash position (~$130M); goodwill $647.3M, net intangibles+goodwill $828.1M; inventories $188.3M (up from $144.6M); AR $184.9M (up from $151.0M). Equity $1,314.3M (APIC jumped $84.2M→$572.1M from TEU prepaid stock-purchase contracts).

Market reaction: Q1-FY26 beat (rev $257.7M vs ~$253.5M est) + raised FY26 revenue guide → stock +~7% pre-market to ~$150, intraday up to ~+13%. The market rewarded the bookings/AI-mix story and looked through the GAAP EPS dip.

Unusual vs own history: the GAAP-vs-adjusted gap widening, the unallocated-cost spike, and the OCF collapse (Lens 10) are the three abnormal items.

Lens 6 · Earnings Calls (sentiment trend)

No transcripts on the research shelf (transcripts=0), so this lens is ``.

Tone arc over the last ~6 quarters: 2022–early-2024 = defensive/destocking ("customers deferring purchases in Life Science and Advanced Industrial," "muted capital spending," recovery "not expected until mid-2025"). Late-2024 into 2025 = inflection ("record revenue," ">50% increase in product launches," "$50M incremental new-product revenue for 2025," book-to-bill recovering). Q1-FY26 = confidently bullish — management leaned hard on "double-digit bookings growth across every segment," "design wins +30%," "vitality index 27%," and explicitly elevated the AI-data-center narrative (~15% of revenue, +20% YoY) while conceding margin recovery is a work-in-progress via "pricing actions and facility closures".

What management now repeats: bookings, design wins, vitality index, AI data center, "high-single-digit organic growth in Q2." What they've stopped emphasizing: destocking / capex deferral (now framed as behind them). Recurring caveat they cannot drop: tariffs and "temporary cost increases from the regionalized manufacturing strategy." Net sentiment: improving and credible, but management's own language concedes the margin story is unfinished.

Lens 7 · Comps

Peer set = precision-photonics / vision / motion / instruments names. Multiples are `` with source/date; where a figure isn't cleanly sourced it is n/a. Pulled mid-June 2026; treat as directional, not tick-accurate.

5-yr average ROE: n/a as a clean series for the set; Novanta's current ROE is only ~5.2% — structurally depressed by the goodwill-heavy balance sheet ($647M goodwill on $1.31B equity) and compressed GAAP earnings, not a clean comparison to asset-light peers.

Read: On EV/EBITDA ~29.6×, Novanta sits below Cognex (~32.6×) but richer than the semicap-levered names (MKSI ~14× trailing), and on forward P/E ~40× it is a premium "quality compounder" multiple. The justification has to be the AI-datacenter mix + margin-recovery optionality, because the organic growth (~3%) does not. Coherent is the cleaner AI-optics pure-play but carries balance-sheet/leverage baggage Novanta doesn't.

Lens 8 · Stock-Price Catalysts (>5% moves, ~5yr)

Mostly ``; pattern matters more than each tick.

• 2021–mid-2023 — the melt-up: serial accretive M&A (ATI Industrial Automation $223.9M and Schneider Electric Motion $118.6M in 2021; Motion Solutions $192.0M announced 2023) + post-COVID capex strength drove NOVT to an all-time high ~$187.61 on 2023-06-30.
• Late-2023 → 2024 — the de-rate: Life-Science + Advanced-Industrial destocking and interest-rate-driven capex deferral pressured organic growth; the stock round-tripped, retesting ~$186 on 2024-07-16 then sliding as the "recovery pushed to mid-2025" message landed.
• 2025 — recovery: bookings/book-to-bill inflected, new-product cycle (>50% more launches) and surgical-robotics OEM re-orders rebuilt the narrative; +25.5% over the trailing 52 weeks.
• 2026-02-23 — FY25 print + FY26 guide: in-line adjusted results, FY26 revenue guide $1.04–1.055B issued.
• 2026-05-12 — Q1-FY26 print: revenue beat + bookings +37% + AI-datacenter mix reveal → +7% to +13% move, the cleanest single up-catalyst in the window.

What the market actually reacts to for NOVT: (1) bookings / book-to-bill (the leading-indicator the bulls trade), (2) the AI-data-center narrative (new, high-beta), and (3) the destocking/recovery cycle in medical + industrial capex. It reacts less to GAAP EPS prints (it looks through them to adjusted). Beta 1.68 means it amplifies the broader risk-on/risk-off tape.

---

Phase C — Judge people & books

Lens 9 · Management

CEO — Matthijs Glastra (57), Chair + CEO since September 2016, joined Novanta 2012; described as a 25+ year technology-operator. CFO — Robert J. Buckley (51). Operating leadership is split into two Co-COOs: John Lesica (Medical Solutions) and Chuck Ravetto (Automation Enabling Technologies).

1. Track record (quantified): Glastra has run the modern Novanta playbook — disciplined, programmatic M&A of precision-technology niches + lean operating system ("Novanta Growth System"). Revenue roughly doubled over his tenure to ~$981M; he navigated the 2022–24 destocking without a guidance blow-up and engineered the 2025 recovery (record revenue, vitality index 27%). But GAAP EPS is lower in 2025 ($1.47) than several years prior — the operating record is "grew the franchise and the adjusted earnings; let GAAP earnings and ROE erode under amortization + corporate spend."
2. Tenure & skin in the game: long-tenured (CEO ~10 yrs). Insider ownership: insider-transactions.csv is absent from the shelf, so precise % is n/a; public reporting shows routine 10b5-1 sales (e.g. "CEO Glastra sells $1.35M in shares") rather than open-market buying. No evidence of large unusual selling, but no conviction-buying signal either.
3. Capital-allocation history: acquisitions are the stated #1 priority — ~$1.1B deployed across ~20 deals in a decade. The 10-K acquisition table confirms the cadence: Keonn 2025 ($75.1M), Motion Solutions 2024 ($192.0M), ATI 2021 ($223.9M), SEM 2021 ($118.6M), W.O.M. 2017 ($134.9M), etc.. No dividend (reinvest-everything model). Newer wrinkle: a $200M buyback authorization (Sept-2025), $191.6M remaining, explicitly to offset equity-comp dilution — plus $39.3M repurchased in FY2025. ROE/ROIC trend is the weak spot: current ROE ~5.2% reflects goodwill-heavy capital and depressed GAAP earnings — the M&A has bought growth and adjusted EBITDA, not high returns on the book capital.
4. Red flags: none egregious. The 2025 SG&A spike tied to an ERP/"financial and operation system implementation" and an insurance-recovery claim is worth watching (corporate-cost discipline). Comp is performance/market-condition PSU-heavy (Monte-Carlo / TSR vs Russell 2000) — reasonable design. Board is unusually strong for a $5B name: Matthew Farrell (ex-CEO Church & Dwight), R. Matthew Johnson (CEO Silicon Labs), Darlene Solomon (ex-CTO Agilent), Frank Wilson (ex-CFO PerkinElmer, Audit chair) — instruments/industrial pedigree that fits the strategy.
5. Archetype: professional-manager / serial-compounder operator (a Danaher/AMETEK-school capital allocator), not a founder. Implication for this stage: execution-and-integration competence is high; the risk is the classic roll-up risk — that reported returns on capital stay mediocre while the multiple prices serial accretion.

Lens 10 · Forensic Red Flags

Forensic lens — every figure labeled.

1. Cash-flow vs earnings divergence — the biggest flag. Operating cash flow collapsed to $64.1M in FY2025 from $158.5M in FY2024 while net income was $53.8M — so OCF/NI fell from ~2.47× to ~1.19×. The entire delta is working capital: AR −$27.3M, inventories −$36.1M, taxes −$12.4M. Management attributes it to "higher inventory levels resulting from our regionalized manufacturing strategy" + receivables timing. FCF = OCF $64.1M − capex $15.6M = ~$48.4M (vs ~$141.4M in 2024). Interpretation: mostly explainable (deliberate dual-running inventory during the footprint move), but it bears watching — if inventory doesn't convert back to cash in 2026, the "temporary" story becomes a quality-of-earnings problem.

2. Receivables/inventory outrunning revenue. Revenue +3.3% but AR +22.4% ($151.0M→$184.9M) and inventory +30.2% ($144.6M→$188.3M). DSO and DIO both rising materially faster than sales — classic working-capital build. One customer = ~17% of the AR balance at YE2025 (concentration in collections).

3. GAAP-vs-non-GAAP gap (SBC + amortization flattering adjusted). Adjusted diluted EPS $3.29 is 2.24× GAAP $1.47. The add-backs: SBC $29.5M (a real, recurring cash-comp substitute), purchased-intangible amortization $27.5M (non-cash but reflects real acquisition cost), restructuring/acquisition $22.7M (recurring because the M&A engine is recurring). A skeptic discounts adjusted EBITDA toward GAAP; a bull treats amortization as non-economic. This single choice swings the multiple from "~30× EV/EBITDA, fine for a compounder" to "~107× GAAP P/E, absurd."

4. Goodwill / intangibles load. $828.1M net intangibles+goodwill (incl. $647.3M goodwill) on a $1.31B equity base — impairment risk if a reporting unit underperforms its DCF (the 10-K runs annual goodwill tests at the reporting-unit level). No impairment taken to date.

5. Segment reporting / unallocated. The unallocated corporate bucket (−$71.7M, +34.4% YoY) absorbs ERP, most restructuring, and acquisition costs — convenient for showing clean segment margins while the consolidated margin erodes. Not improper (it's disclosed), but it's where the cost pressure is parked.

Controls / auditor: ICFR concluded effective at YE2025 and Q1-FY26; auditor Deloitte & Touche LLP; no restatements, no material weaknesses, no clawback triggers. SBC accounting uses standard Monte-Carlo/Black-Scholes models.

Regulatory findings (required sub-section):

• SEC Litigation Releases: none naming Novanta (EDGAR EFTS LR, 2021–2026).
• SEC AAERs: none naming Novanta (EDGAR EFTS AAER, 2021–2026).
• Item 3 Legal Proceedings (FY2025 10-K): only "various other legal proceedings and claims that arise in the ordinary course of business"; management "does not believe the outcome… will have a material adverse effect." A 2013 pension Guarantee for Novanta Technologies U.K. Ltd. in favor of the U.K. Plan trustees is disclosed. Q1-FY26 10-Q reports no material change.
• Non-SEC (FTC/DOJ/FDA/etc.): Novanta is FDA-regulated as a medical-device maker (QMSR, EU MDR/CE) — a compliance burden, not an enforcement finding. No material FTC/DOJ/FDA enforcement action surfaced in web search as of 2026-06-24 ``.
• Verdict: No material regulatory or legal findings — verified via SEC EDGAR EFTS (LR, AAER), web search, and 10-K Item 3 as of 2026-06-24.

---

Phase D — Project & stress-test

Lens 11 · Forward Projection

Built bottom-up from FY2025 actuals + FY2026 guidance. Every input labeled; outputs ``. Novanta reports on adjusted diluted EPS, so the base case is framed in those terms (and the GAAP gap is flagged).

Anchors:

• FY2025 actual: revenue $980.6M, adj EBITDA $221.0M (23%), adj diluted EPS $3.29.
• FY2026 company guidance: revenue $1.04–1.055B (midpoint ~$1.048B, +6.9%), adj diluted EPS $3.50–3.65 (midpoint $3.58). Consensus adj EPS ~$3.59–3.73.
• Share count rising to ~41M diluted post-TEU (vs 36.7M FY2025) — the dilution headwind is structural until TEU settlement (Nov-2028).

Three-year adjusted-diluted-EPS path (FY2026 → FY2028) ``:

GAAP EPS runs roughly $1.50–$1.80 below adjusted on this path (amortization + SBC + restructuring), so a GAAP investor sees ~$2.00–$2.80 by FY2028 ``.

Brier forecast (logged conceptually; not written in --watchlist): base case → NOVT FY2026 adjusted diluted EPS ≥ $3.58, p≈0.55, resolves 2027-02 on the FY26 print. (Per skill rules, forecast.ts create is skipped in the watchlist loop.)

The honest read: even the base case is ~40× forward FY26 adj EPS and ~30× EV/EBITDA on ~7% revenue growth — the multiple already capitalizes the bull's margin recovery. The bull EPS path ($5.30 by FY28) only gets you to ~$151 / $5.30 ≈ 28× FY28 — still not cheap.

Lens 12 · Bull vs Bear

Bull case. A best-in-class precision-technology compounder with a genuine, decaying-slowly moat (regulated design-in sockets, 27% vitality index, +37% bookings, 1.10 book-to-bill). Three growth legs lining up at once: (1) AI-data-center photonics — ~15% of revenue growing ~20% YoY, leveraging laser/optics/precision-motion IP into the optical-interconnect buildout (the same secular wave lifting Coherent/Lumentum); (2) surgical-robotics / advanced-surgery OEM re-acceleration (advanced surgery +$33.7M in FY2025); (3) margin self-help — once the ERP and manufacturing-regionalization spend rolls off, the unallocated cost line normalizes and 200–300bps of operating margin comes back "for free." Net-cash balance sheet ($380.9M cash, net cash) + reloaded M&A firepower + $191.6M buyback = optionality. Capital allocator (Glastra) with a decade of accretive deals and a blue-chip board. The earnings-expansion narrative: adjusted EPS compounds mid-teens off a depressed base as growth + margin recovery + buyback stack.

Bear case (2–3 permanent-impairment / de-rating risks). (1) The multiple is the risk — ~40× forward / ~30× EV/EBITDA / ~107× trailing GAAP on ~3% organic growth and 5-year GAAP profit growth of ~2.3%; the DCF fair value cited is ~$104.77 vs ~$151 price — i.e. ~30% downside to a sober DCF if the margin-recovery story slips. (2) Margin recovery may not come — tariffs, the China-capacity build, and sticky corporate cost (ERP) could keep the unallocated line elevated; management has guided this for multiple quarters and GAAP operating income has fallen two years running. (3) Customer + end-market concentration — top-10 = 42%, two medical OEMs ~11–12%; a single surgical-robotics OEM's platform pause or in-sourcing decision dents a whole growth leg, and medical/industrial capex is rate-sensitive (the 2022–24 destocking can recur).

Pre-mortem (18 months out, thesis broke): It's late 2027. NOVT is ~$110. What happened? The AI-data-center revenue turned out to be lower-margin commoditized optics/motion content that grew the top line but not the mix; the "temporary" regionalization costs proved structural (tariffs + dual-running never fully unwound); one large medical OEM pushed out a platform; organic growth printed ~2–4% three quarters running; and the market re-rated a ~3%-organic-growth industrial from ~40× to ~22× forward. None of it required an accounting scandal — just the premium multiple meeting mediocre organic growth.

Are multiples too high? On any GAAP or DCF basis, yes — the stock is priced for the bull margin-recovery + AI-mix case to execute. On an adjusted-EBITDA-compounder basis it's "expensive but in-line with quality peers (below Cognex)."

Contrarian view (what the market refuses to see): The bulls are trading bookings (+37%) and the AI-datacenter label as if they're high-incremental-margin secular growth, while the financials say Novanta's real problem for two years has been its own cost base, not demand — and that's a self-help story management has repeatedly promised and not yet delivered. The contrarian bet isn't "demand is bad"; it's "the margin inflection arrives a year later and smaller than the multiple assumes."

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

Dismantling the bull case.

• What structurally breaks the model: Novanta sells components, not systems — its content-per-platform and pricing reset every time an OEM redesigns. If AI-datacenter optics commoditize (multiple suppliers, hyperscaler price pressure), the ~15% "growth leg" becomes low-margin volume that dilutes the 46% adj-GM, not a mix-up story.
• Revenue concentration: top-10 = 42%, two medical OEMs ~11–12%, one customer = 17% of AR. A surgical-robotics OEM deciding to dual-source or in-source motion/vision content — exactly the kind of vertical integration Intuitive-class players pursue — removes a double-digit chunk that took years to qualify in and can't be replaced quickly ("lengthy qualification periods").
• Moat weaker than bulls think: "no individual patent is material" — the moat is qualification inertia, which is strong mid-lifecycle and zero at the next platform refresh. In commoditizing industrial-precision (already −$20M in FY2025), the moat is visibly eroding.
• Most dangerous competitor bulls underrate: in AI-datacenter optics, Coherent (NVIDIA-anchored, $2B, dedicated optics scale) and Lumentum can out-invest Novanta's optics franchise; in precision motion, vertically-integrated robotics OEMs and the large diversified industrials (AMETEK, Parker) can bundle. Novanta is a high-quality niche player, not a scale leader, in the very market it's now valued on.
• Worst capital-allocation / accounting tells: the ~$1.82 GAAP-to-adjusted EPS gap; restructuring add-backs that recur because the M&A engine recurs; OCF collapsing to 1.19× net income on a working-capital build; ROE of ~5% on a goodwill-heavy book. None are fraud — but together they argue the quality of the compounding is lower than the multiple implies.
• Assumptions that must hold for ~$151: mid-teens adjusted-EPS CAGR, 200–300bps margin recovery, AI-mix staying high-margin, no medical-OEM concentration shock, and the multiple staying ~40× forward.
• Valuation if growth disappoints 20–30%: if FY27 organic prints ~2% and margin recovery stalls, adj EPS lands ~$3.55 (the bear path) and a re-rate to 22–25× forward (industrial-grower multiple) implies ~$80–90, i.e. ~40–45% downside ``. The cited DCF fair value ($105) is the midpoint of that gravity.
• Single scenario that permanently impairs: a major surgical-robotics OEM in-sources Novanta's motion/vision content and AI-datacenter optics commoditize simultaneously — removing the two legs the multiple is built on. Plausibility: moderate-low individually, low jointly, but either alone would compress the multiple hard.

Lens 14 · Management Questions (ordered by information value)

1. The unallocated corporate cost line rose 34% to $71.7M on ERP and restructuring — what is the exact quarter you expect it to normalize, and to what run-rate, and what is the resulting consolidated operating-margin bridge?
2. Operating cash flow fell to $64M from $159M on a working-capital build — how much of the $188M inventory converts back to cash in 2026, and what's your normalized FCF-conversion target?
3. Of the ~15% of revenue tied to AI-data-center applications, what is its gross margin versus the corporate average, and is it accretive or dilutive to mix as it scales?
4. Top-10 customers are 42% of revenue and two medical OEMs are ~11–12% each — what is your concentration trajectory, and what protects you from a large OEM in-sourcing precision-motion/vision content?
5. You've guided margin recovery for several quarters while GAAP operating income fell two years running — what specifically is different now that makes the 2026 inflection credible?
6. Adjusted EPS is ~2.2× GAAP EPS — at what point do restructuring/acquisition add-backs stop, given M&A is a permanent part of the model?
7. With ~$1.1B deployed across ~20 deals and ROE at ~5%, what return-on-invested-capital hurdle do you underwrite new acquisitions to, and how have the last three deals performed against it?
8. The China capacity-expansion project inside the $30–40M 2026 capex — how do you weigh that against tariff/onshoring risk, and is it for local-for-local or export?
9. Book-to-bill is 1.10 and bookings +37% — how much is durable demand versus OEM order-pattern normalization / pull-forward after the destocking?
10. What is the organic (ex-FX, ex-M&A) growth algorithm you're underwriting for 2026–2028, and what end-market mix gets you there?
11. The $200M buyback is framed as dilution-offset — given the stock trades near ~40× forward, why buy back stock versus deploying into M&A or debt reduction?
12. The TEU dilutes diluted share count to ~41M until Nov-2028 — walk through the settlement mechanics and the EPS path as those contracts settle.
13. Precision-manufacturing products fell ~$20M in AET and precision-medicine fell ~$12.6M — are these cyclical troughs or structural share loss, and to whom?
14. The 10-K notes "we do not have personnel fully dedicated to IP management" — how do you defend a moat you describe as know-how rather than patents against well-capitalized photonics scale players?
15. What is your succession plan beyond the CEO and the two Co-COOs, and how is the board thinking about it given the CEO's ~10-year tenure?

---