Lonza Group

Phase A — Understand the business

Lens 1 · Company Overview

What it is. Lonza is a ~130-year-old Basel-region industrial that has, over the last decade, remade itself into a pure-play CDMO — it does not sell its own drugs; it develops and manufactures other people's medicines under contract, across the full molecule lifecycle from lead optimisation → clinical supply → commercial manufacturing. Think TSMC-for-biologics: the customer owns the IP and the marketing authorisation; Lonza owns the plants, the process know-how, and the regulatory-grade capacity.

The transformation is the story. As of March 2026 Lonza completed its transformation to a pure-play CDMO by agreeing to divest its Capsules & Health Ingredients (CHI) business — the old Capsugel gelatin-capsule + nutraceutical franchise — to Lone Star Funds . What remains is a focused, three-platform CDMO (structure operational since **1 April 2025**) :

Customers. Structural mix: large pharma contributes the majority of CDMO revenue (commercial-stage manufacturing volume), while small/emerging biotech is the larger share of the pipeline — early-stage biotechs outsource by design to conserve capital . Named commercial anchors include Roche/Genentech (the Vacaville transition contract), and historically Moderna (COVID-vaccine mRNA supply, now normalised down). Lonza manufactures the viral vector / cell / gene product for **five commercial cell-and-gene therapies across three continents** .

Contract structure & payment terms. CDMO economics are a spectrum: early-phase is fee-for-service (development, per-batch), commercial is often take-or-pay / minimum-volume reservation on dedicated suites — the Vacaville deal explicitly carries "committed volumes over the medium term" from Roche that phase out as Lonza re-sells the capacity ``. This gives Lonza a contracted, multi-year, high-switching-cost revenue backbone on its large-scale assets — the closest thing biologics has to recurring revenue.

Lens 2 · Supply Chain

Lonza sits in the middle of the pharma value chain — it is itself a chokepoint. The map (names, not generalities):

Upstream inputs → Lonza:

• Bioreactor / single-use hardware & consumables: Cytiva (Danaher), Sartorius, Thermo Fisher — single-use bags, filters, chromatography resins. ``
• Cell-culture media & raw biologics inputs: Thermo Fisher, Merck KGaA/MilliporeSigma. ``
• Payloads / linkers / HPAPI precursors (for the ADC franchise): specialty fine-chem suppliers; Lonza is itself vertically integrating this by building payload-linker capacity at Visp (see Lens 4/11). ``
• Capital equipment & construction: the CHF 1.3bn/yr capex program (Visp, Stein, Portsmouth, Vacaville) ``.

Lonza (the node) → end customers:

• Large pharma / biotech (the buyers): Roche/Genentech, Moderna (declining), plus a long tail of pharma & emerging biotech. These are the OEMs; Lonza is the foundry. The drug then flows Pharma → wholesaler → patient — Lonza's product is the bulk drug substance / finished drug product, not the branded pill.

Chokepoints & single-source dependencies:

1. Lonza itself is a global chokepoint in large-scale mammalian biologics capacity — post-Vacaville it is the largest biologics CDMO in the US and holds an estimated 19–21% global biologics-CDMO share ``. Capacity is scarce, lead times multi-year — that scarcity is the moat (Lens 3).
2. Geographic concentration in Switzerland (Visp) — Visp is the mega-site (mammalian, small-molecule, bioconjugate, HPAPI). Single-campus concentration = execution/utilisation risk, but also a deep skills cluster.
3. Regulatory single-source lock-in: once a customer files a BLA/NDA naming a Lonza facility as the manufacturing site, switching requires an FDA/EMA supplement and re-validation (12–24+ months). This is the switching cost that makes commercial contracts sticky — and the reason a Lonza quality failure (Lens 10) is existential for the customer, not just Lonza.

This lens is not generic: the point is that Lonza is downstream of the single-use-bioprocessing oligopoly (Cytiva/Sartorius/Thermo) and upstream of a fragmented pharma-buyer base — it has more pricing power over its customers than over its suppliers.

Lens 3 · Competitive Advantages (moats)

Lonza's moat is real, durable, and industrial — closer to a foundry than a services shop.

1. Scale + regulatory-grade capacity (the core moat). Large-scale biologics manufacturing is a capital + know-how oligopoly. Building a 300k-litre mammalian site takes ~$1–2bn and 4–6 years; qualifying it with regulators takes longer. Lonza's 19–21% share , and its position as **#1 CDMO by revenue (~$7.5bn CDMO scale)** , means it can offer capacity now that a customer cannot build. Vacaville's ~330,000 L bioreactor capacity `` is capacity you cannot replicate on a drug-launch timeline.

2. Switching costs (regulatory lock-in). Once a commercial drug is registered against a Lonza site, moving it out requires regulatory re-filing and re-validation — a multi-year, multi-million-dollar barrier the customer bears. This converts commercial contracts into quasi-annuities.

3. Complexity / modality breadth (the premium-mix moat). The crown jewel is Advanced Synthesis at 41.8% CORE EBITDA margin `` — driven by bioconjugates/ADCs and HPAPIs, the hardest, most-potent, least-commoditised chemistry. Few CDMOs can safely handle cytotoxic payloads at commercial scale. This is where Lonza out-earns cost-competitors like Samsung and WuXi.

4. Bargaining power — asymmetric. Over customers: high, on commercial contracts (lock-in + scarce capacity). Over suppliers: moderate — the single-use-bioprocessing vendors (Cytiva/Sartorius/Thermo) are themselves an oligopoly, so Lonza is a price-taker on consumables. Net: Lonza needs its customers less than they need Lonza, once a molecule is committed.

Where the moat is weaker (bull-check): in early-phase development and in standardised large-scale mammalian runs, Lonza competes on price and slot-availability against Samsung Biologics (scale/cost) and, ex-BIOSECURE, WuXi Biologics (China cost). Cell & Gene is not yet a moat — it's a scale-chasing, currently-loss-of-momentum business (−3% CER, 17% margin). The moat is concentrated in large-scale commercial biologics + high-complexity synthesis, not the whole company.

Lens 4 · Segments

Revenue/EBITDA by platform (FY2025, the first full year of the new structure) — all ``:

Group FY2025: Sales CHF 6.5bn (continuing/CDMO), +21.7% CER; CORE EBITDA CHF 2.1bn, margin 31.6% (+1.4pp) ``.

The critical adjustment (do not skip): the headline +21.7% CER is inflated by Vacaville. Management stated that excluding the Vacaville site, organic CER growth was "low-teens" ``. So the true underlying organic run-rate is ~low-double-digits — which is exactly what the 2026 guide of +11–12% CER codifies once Vacaville anniversaries. The market's January-2026 disappointment (Lens 8) was the gap between the flattering 21.7% headline and the honest low-teens organic reality.

Geography: the strategic thrust is US capacity (Vacaville CA + Portsmouth NH) to serve "the world's largest pharma market," complemented by the Swiss (Visp/Stein) European base ``. Precise geographic revenue split not disclosed in accessible sources → n/a at the sub-segment level.

---

Phase B — Measure performance

Lens 5 · Earnings Result

Two prints frame the current state: the FY2025 result (2026-01-28) — strong on the year, soft on the H2 exit and the 2026 guide — and the Q1-2026 confirmation (2026-05-08).

FY2025 (continuing operations / CDMO) — all ``:

• Sales CHF 6.5bn, +21.7% CER (low-teens ex-Vacaville).
• CORE EBITDA CHF 2.1bn, margin 31.6% (+1.4pp).
• Net income CHF 949m (from CHF 636m FY2024) ``.
• Basic EPS (continuing ops) CHF 13.04 (from CHF 8.51); diluted CHF 13.00 ``.
• FCF CHF 545m — ~doubled YoY, on capex discipline + working-capital ``.
• Net debt CHF 3,258m; ND/CORE EBITDA 1.4x — comfortably levered ``.
• Capex CHF 1.3bn = 19.6% of sales ``.
• Dividend proposed CHF 5.00, +25% ``.

The miss that mattered — H2/Q4 2025: on the second-half slice the company missed consensus hard: H2 EPS CHF 8.62 vs CHF 9.45 est (−8.8%) and H2 revenue CHF 3.48bn vs CHF 3.98bn est (−12.6%) . The stock fell **~9% intraday** (closed ~−4.6% at CHF 557.8) . The negative surprise was not the year — it was the 2026 guide (+11–12% vs the 21.7% just delivered) landing as a growth deceleration, compounded by H1-weighted 2026 phasing (planned shutdowns, campaign timing) ``.

Q1-2026 (2026-05-08) — all ``:

• In-line, 2026 outlook confirmed: +11–12% CER sales, CORE EBITDA margin >32%.
• Momentum sustained in Integrated Biologics, strong in Advanced Synthesis, Specialized Modalities returned to healthy growth (the C&G recovery continuing).
• First-half-weighted growth reiterated (easier prior-year base, campaign/release timing, planned shutdowns).
• Visp large-scale mammalian facility → full commercial operations mid-2026 — a key ramp milestone.

Balance-sheet flags: none acute. ND/EBITDA 1.4x is conservative for an industrial; capex intensity (~20% of sales, guided 25–30% through 2026 ) is the cash-flow swing factor — FCF is thin (CHF 545m on CHF 6.5bn sales ≈ 8.4% FCF margin ) precisely because the plants are still being built. Watch receivables/inventory vs. the Vacaville ramp as suites re-fill.

Lens 6 · Earnings Calls (sentiment trend)

No transcripts on the research-layer shelf (web-only). From accessible call summaries across the last ~4 prints ``:

• Tone has shifted from "digesting overextension" (2023–24) to "executing a de-risked plan" (2025–26). The recurring management theme in 2025–26 is operational execution, growth-project ramp, and margin discipline — "maturing growth projects," "operating leverage," "disciplined capex."
• Phrases they now lead with: pure-play CDMO, the Lonza Engine (their operating-model branding), growth projects ramping, first-half-weighted, mid-term ambition intact.
• Phrases/topics they've de-emphasised vs. 2023: COVID / Moderna (normalised away), and the defensive tone around capacity utilisation and CEO stability that dominated the Ruffieux-exit period. Cell & Gene has gone from "growth engine" language to a more measured "return to healthy growth" after the −3% CER year.
• Credibility caveat: management guided down for H2/Q4 into the January print and still missed the Street's H2 number — so the "execution" narrative is only partly earned. The 2026 guide is deliberately conservative-looking and H1-loaded, which sets up a second-half credibility test.

Net sentiment: improving and more disciplined than the 2023 trough, but not yet "trust-us" — the tape (Lens 8) shows the market still punishes any guide-down instantly.

Lens 7 · Comps

CDMO / biologics-manufacturing peer table. Multiples are `` with source/date, or n/a. No multiple is fabricated.

5-yr average ROE: n/a for the peer set at a reliable, consistent basis (Lonza's own ROIC ~6.5% vs WACC ~5.9% `` is the more decision-relevant capital-efficiency read — see Lens 9/13).

Read of the table: Lonza trades rich in absolute terms (≈27x forward earnings, ≈18x EV/EBITDA) but at roughly half Samsung Biologics' multiple and a premium to WuXi (which carries a China/BIOSECURE discount). The market is paying up for Lonza's complexity-mix quality + #1 scale + Western-supply-chain "safe-harbour" status, while Samsung is priced for even faster capacity-led growth. The bull says Lonza is the cheap high-quality CDMO; the bear says ~18x EV/EBITDA is a full price for a business earning barely above WACC (Lens 13).

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

What actually moves LONN ``:

• 2021–22 — COVID/Moderna super-cycle up, then normalisation down. The Moderna mRNA supply relationship inflated growth and expectations; the post-COVID demand normalisation began the multiple de-rate. ``
• 2023-07 — guidance cut (−). Cut 2023 outlook citing weak nutraceutical-capsule demand + biotech customers pursuing fewer early-stage and cell-&-gene projects. ``
• 2023-09-18 — CEO Ruffieux abrupt exit (−13%). Shares fell from ~$55.49 to $48.10 (−13%) on the surprise departure "by mutual agreement," chairman citing "challenging times." Third surprise CEO exit in ~5 years. ``
• 2024-03-20 → 10-01 — Vacaville acquisition (+, strategic). The $1.2bn Roche/Genentech site purchase reframed the story as offensive capacity-grab; broadly well-received. ``
• 2026-01-28 — FY2025 print / 2026 guide (−~9% intraday). Strong FY headline but H2 miss + decelerating 2026 guide → sharp drop. ``
• 2026-03-06 — CHI divestment to Lone Star + CHF 500m buyback (+, clarity). Completed the pure-play pivot; capital-return signal. ``

Pattern (what the market reacts to for this name): (1) guidance/forward growth-rate revisions (the single biggest driver — the stock trades on the second derivative of growth, not the level); (2) CEO/management stability (a genuinely idiosyncratic risk given the turnover history); (3) strategic capacity M&A (Vacaville, CHI exit). It reacts less to any single quarter's beat than to what the guide implies about the trajectory. This is a "prove the ramp" stock — it re-rates on evidence that growth-project capex is converting to margin, and de-rates instantly on any hint the ramp is slipping.

---

Phase C — Judge people & books

Lens 9 · Management

CEO — Dr Wolfgang Wienand (from summer 2024) ``.

1. Track record: Career CDMO/specialty-chem operator — CEO of Siegfried Holding (a rival Swiss CDMO) 2019–2024, previously Siegfried's CSO then CSO/Chief Strategy Officer; before that senior roles at Evonik. He is a domain-native operator, not a generalist parachuted in — a deliberate contrast to the prior instability. ``
2. Tenure & skin in the game: ~2 years in seat. Insider-ownership figures n/a (no research-layer insider file; Swiss disclosure less granular than US). Flag to verify.
3. Capital-allocation history (on his watch): the defining moves are (a) completing the Vacaville integration, (b) divesting CHI to Lone Star for CHF 2.3bn EV / CHF 1.7bn upfront + a retained 40% stake with preferential exit participation , **(c) a CHF 500m buyback**, and **(d) reinvesting proceeds into organic CDMO growth + bolt-ons**. Coherent and focused — turning a sprawling conglomerate into a pure-play. **But the capital-efficiency verdict is unflattering: ROIC ~6.5% vs WACC ~5.9%** — the company is barely creating value per franc invested, and it is investing enormous francs (capex ~20% of sales). The whole thesis rests on that spread widening as growth projects mature.
4. Red flags: the serial CEO turnover — Ridinger (retired abruptly 2019) → Funk (lasted ~9 months) → Ruffieux (abrupt exit 2023) → Baehny (interim) → Wienand ``. That is a governance/culture red flag independent of any one person; it signals a demanding board (Chairman Baehny was long the power center) and a hard-to-run business. No related-party or accounting red flags surfaced.
5. Archetype: professional manager / industry veteran (not founder). Correct archetype for this stage — the job is disciplined execution of a de-risked, capacity-ramp plan, not visionary reinvention.

Lens 10 · Forensic Red Flags

Web-only; no filings to run divergence analytics against, so this is qualitative + disclosure-scan, all ``.

Accounting / quality-of-earnings watch-items:

• CORE vs. IFRS gap. Lonza reports on "CORE" (adjusted) EBITDA/EPS as its headline. FY2025 CORE EBITDA CHF 2.1bn / 31.6% is the flattered metric; IFRS net income was CHF 949m `` — the two are reconcilable but the narrative runs on CORE. Standard for the sector, but the adjustment bridge (restructuring, amortisation of acquired intangibles from Vacaville/Capsugel, M&A costs) deserves scrutiny each period. n/a — bridge not itemised in accessible sources.
• Discontinued-operations optics. CHI has been reclassified to discontinued operations, which is why continuing sales are CHF 6.5bn vs the ~CHF 8.7bn group total ``. Legitimate under IFRS 5, but it re-bases the growth-rate optics upward (you're comparing a cleaner, faster CDMO). Cross-check any "growth" claim against whether it's continuing-ops.
• Vacaville "committed volumes" tail. Roche's committed volumes phase out over the medium term ``. Revenue quality on that CHF ~0.6bn declines as the contract runs off and Lonza must re-sell the capacity to new customers — a real forward risk that the +32% Integrated Biologics optic masks. If re-fill lags, that segment decelerates sharply.
• Capex-to-FCF gap. FCF (CHF 545m) is far below CORE EBITDA (CHF 2.1bn) because capex (CHF 1.3bn) is high ``. Not a red flag per se (growth investment), but it means reported EBITDA overstates near-term cash generation — the bull case requires capex intensity to fall (guided down from ~30% toward high-teens) for FCF to inflect.
• SBC / goodwill: n/a in detail. Goodwill/intangibles from Capsugel (2017, ~$5.5bn) and Vacaville warrant an impairment watch given the C&G softness, but no impairment was flagged.

Regulatory findings (required sub-section). Read regulatory/regulatory-findings.md (Step-0 output).

• SEC (EDGAR EFTS — LR + AAER): none possible. Lonza has no CIK and files nothing with the SEC; the fetch returned 0 findings and the file explicitly notes no EDGAR search is possible ``. This is a jurisdiction fact, not a clean bill of health.
• FDA (web search): Lonza's manufacturing footprint has a normal-for-a-CDMO inspection history. Recent (2023–24) inspections at Tampa FL, Portsmouth NH, and Fuzhou CN resolved as VAI / NAI / No-Action . Historically, the **Lonza Houston cell-therapy plant drew an FDA Form 483 / warning letter** that contributed to delaying Bristol Myers' liso-cel timeline (~2019–20) , and Lonza Walkersville received a 2017 warning letter ``. For a CDMO, facility quality is the existential risk (a warning letter can halt a customer's commercial product), so this is a permanent operational risk category, not a one-off. No open, material FDA action surfaced as of 2026-07.
• Other agencies (legacy, immaterial): Lonza America $150k EEOC disability settlement ; **Lonza/EPA disinfectant-label settlement** and a chemical-environmental settlement — these attach to the old consumer/chemicals/Capsugel legacy, are small, and are largely being divested with CHI.
• Verdict: No material financial-reporting enforcement found — verified via SEC EDGAR EFTS (LR/AAER: not applicable, no CIK), FDA inspection databases, and web search as of 2026-07-06. The live regulatory risk is operational (FDA facility quality), not accounting.

---

Phase D — Project & stress-test

Lens 11 · Forward Projection

Base actuals anchor : FY2025 basic EPS (continuing) **CHF 13.04**; consensus 2026 EPS **~CHF 18.21** . Note the 2026 consensus jump reflects continuing-operations re-basing + the CHI-stake accounting + margin expansion, not a ~40% organic earnings surge — treat with care. Management framework: 2026 CER sales +11–12%, CORE EBITDA margin >32%; mid-term 2028: 11–13% sales CAGR, 32–34% margin, double-digit ROIC ``.

Three-fiscal-year EPS path (FY2026 / FY2027 / FY2028), all `` — arithmetic shown, off a continuing-ops CORE-EPS proxy:

Caveat (provenance-critical): these are `` projections anchored to a sourced consensus base (CHF 18.21) and management's own sales/margin framework; the exact CORE-EPS bridge (Vacaville amortisation, CHI-stake equity accounting, FX) is not fully sourced, so treat the levels as directional and the scenario spread as the real output. No forecast.ts create is run — this is the unattended --watchlist loop, and the base case is not yet a committed, human-signed forecast.

The number that actually matters is not EPS — it's ROIC. The stock re-rates if ROIC (~6.5%) climbs toward the "double-digit" 2028 ambition as the CHF-billions of growth-project capex convert from cash-drain to margin. Track ROIC and FCF conversion, not just the EPS line.

Lens 12 · Bull vs Bear

Bull case. Lonza is the #1 pure-play CDMO in the world, sitting in front of a multi-year structural demand wave: (1) BIOSECURE Act (signed Dec 2025) is forcing a Western-capacity re-shoring away from WuXi, and Lonza is repeatedly named the best-positioned beneficiary given service overlap in biologics + C&G ``; (2) the ADC/bioconjugate super-cycle flows straight into Advanced Synthesis (41.8% EBITDA, the highest-margin, least-contestable franchise); (3) the pure-play transformation is done — CHI divested, ~CHF 3bn of capital to redeploy, a CHF 500m buyback, and a de-cluttered story; (4) capacity that cannot be replicated on a launch timeline (Vacaville 330k L; Visp mega-site) plus regulatory switching-cost lock-in; (5) a credible, domain-native CEO executing a de-risked ramp with a 2028 double-digit-ROIC target that, if hit, justifies the premium. Potential positive surprise: a WuXi customer stampede that fills Vacaville's re-sell gap faster than modelled.

Bear case (2–3 permanent-impairment risks). (1) Capital efficiency never inflects — ROIC ~6.5% vs WACC ~5.9% means the business is a near-value-neutral capital hog; if the growth-project ramp disappoints or capex stays elevated, you own a low-teens-grower earning ~cost-of-capital, trading at ~18x EV/EBITDA / ~27x forward earnings — a de-rating waiting to happen. (2) Customer-side fragility — a large share of the pipeline is small/emerging biotech, and biotech funding is under pressure (XBI at pre-COVID lows, ~$9.5bn NIH cuts by mid-2026) ``; fewer funded programs = a thinner future book. (3) Cell & Gene structural disappointment — the modality that was supposed to be the growth engine printed −3% CER; if C&G stays stalled, a whole platform underdelivers. Pre-mortem (18 months out, thesis broken): the 2026 H1-loaded guide met its easy first half, then H2 slipped (Vacaville refill lagged as Roche volumes rolled off, C&G stayed soft), management guided 2027 below the 11–13% CAGR, ROIC failed to climb, and a ~27x-forward multiple compressed to ~20x — a 25–30% de-rating on unchanged "quality." Are multiples too high? For the quality, defensible; for the current returns on capital, yes — the multiple is borrowing from a 2028 ROIC that hasn't shown up yet. Contrarian view (what the market refuses to see): everyone owns the BIOSECURE/reshoring tailwind as a given — but BIOSECURE enforcement isn't expected before ~2028, Vacaville's committed Roche volumes run off in the same window, and there is a real air-pocket risk in 2027 between the two. The tailwind is real but later and lumpier than the price implies.

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

Dismantling the bull case.

• What structurally breaks the money-machine: a utilisation air-pocket. This is a fixed-cost, capacity-reservation business — margins are a function of how full the plants are. Lonza is bringing on enormous new capacity (Visp mammalian mid-2026, Stein 2027, Visp bioconjugation 2028) at the same time Vacaville's committed Roche volumes roll off. If new-customer demand doesn't land on schedule (biotech funding is soft, BIOSECURE re-shoring is slow to actually move molecules given re-validation timelines), you get empty new suites + a runoff tail = margin compression on a stock priced for margin expansion.
• Revenue concentration: the Vacaville/Roche transition contract (~CHF 0.6bn) and any large-pharma commercial anchors are concentration points. The −3% Specialized Modalities year shows how fast a platform de-rates when its anchor demand softens.
• Why the moat is weaker than bulls think: in standardised large-scale mammalian, Samsung Biologics is out-building Lonza on cost/scale and trades as if it will take share; ex-BIOSECURE, WuXi is a persistent price threat; and Fujifilm Diosynth is aggressively adding US/EU biologics capacity. The premium moat is real only in ADC/HPAPI + registered-commercial lock-in — perhaps half the company. The other half is contestable, cyclical, capacity-priced commodity.
• Worst capital-allocation reality: paying $1.2bn for Vacaville with a runoff committed-volume tail, then another ~CHF 500m to upgrade it ``, into a business already earning ~6.5% ROIC. The whole model assumes the spread over WACC widens — a short's core bet is that capital-intensive CDMO economics structurally cap ROIC in the high-single/low-double digits, and the market is paying a ~27x-forward, TSMC-like multiple for a business that is not TSMC on returns.
• Assumptions that must hold for today's price: (a) 2028 margin reaches 32–34% and ROIC goes "double-digit"; (b) BIOSECURE actually re-routes molecules to Lonza before the Vacaville tail runs off; (c) C&G recovers into a real growth platform; (d) no facility quality event (a warning letter on a commercial line is a customer-halting, reputation-denting tail risk given the Houston/Walkersville history).
• If growth disappoints 20–30%: a slide to ~7% sales growth with flat margins pushes FY2028 EPS toward ~CHF 18.5 (bear) vs ~CHF 25 (bull) — and the multiple would compress with the growth, a double-hit. Plausible downside: 30–40% from a ~27x-forward multiple on a re-rate to ~18–20x on ~CHF 17–18 EPS.
• Single scenario that permanently impairs: a major FDA quality failure at a large-scale commercial biologics site (import alert / consent decree) — it would halt customer products, trigger contract flight (regulatory switching cost cuts both ways once trust breaks), and permanently dent the "safe Western supplier" premium that is the entire BIOSECURE bull case. Low probability, catastrophic payoff — the definition of the risk you must underwrite.

Lens 14 · Management Questions (ordered by information value)

1. Vacaville re-fill vs. Roche runoff: as the committed Genentech volumes phase out, what is the contracted new-customer backlog replacing them, and what utilisation do you underwrite for that site in 2027–2028?
2. ROIC bridge: you target double-digit ROIC by 2028 from ~6.5% today — decompose exactly how (margin mix, utilisation, capex-intensity decline) and name the single biggest lever.
3. The 2027 air-pocket: BIOSECURE enforcement isn't expected before ~2028 while your Vacaville tail runs off sooner — how do you bridge a potential 2027 demand gap between the two?
4. BIOSECURE conversion: how much of the "WuXi displacement" opportunity has actually converted into signed contracts vs. RFPs, and what's the realistic lead time from win to revenue given re-validation?
5. Cell & Gene decision: after a −3% CER year, is C&G a keep-and-scale, fix, or eventual-exit? What data by when decides it?
6. Capex intensity glide-path: you guide ~25–30% of sales through 2026 — what is the steady-state, and in which year does FCF conversion (vs CORE EBITDA) inflect?
7. CHI 40% retained stake: why retain 40% rather than a clean exit, and what governance/exit mechanics protect the "preferential participation"?
8. Pricing power reality: in standardised large-scale mammalian, what pricing trend are you seeing vs. Samsung and (ex-BIOSECURE) WuXi — are you a price-maker or price-taker there?
9. CORE-to-IFRS bridge: walk through the FY2025 adjustments (acquired-intangible amortisation, restructuring, M&A) and how that bridge trends as Vacaville/Capsugel amortisation runs.
10. Customer-base fragility: what share of your pipeline revenue depends on small/emerging biotech, and how are you hedging the funding downturn (XBI lows, NIH cuts)?
11. CEO stability / culture: given four CEOs in ~six years, what has structurally changed in board–management operating cadence to break that pattern?
12. Bolt-on M&A discipline: with ~CHF 3bn of CHI proceeds to redeploy, what ROIC hurdle must a bolt-on clear, and in which modality/geography are you hunting?
13. Quality/regulatory: given the Houston/Walkersville warning-letter history, what has changed in the global quality system to make a commercial-line FDA action a tail rather than a live risk?
14. Geographic/currency: with the US buildout (Vacaville/Portsmouth) and a CHF cost base, how do you manage the FX and cost-competitiveness gap vs. Korean/Chinese capacity?
15. Mid-term guide confidence: what would have to go right for you to hit the top of 11–13% CAGR + 34% margin, and what is the single most likely reason you'd land at the bottom?

---