Lonza Lifts Midterm Margin Ambition to High 30s and Sets 20% Batch Throughput Target at Vacaville Investor Day
Analyst and Investor Day in Vacaville, California, October 14, 2026
Lonza used its investor day at the former Genentech site in Vacaville to raise its midterm profitability ambition. The company now targets a core EBITDA margin in the high 30s, up from 33% to 34% guided for 2026 and just above 30% reported in 2024. Management also introduced a return framework of 2x to 3x the cost of capital for ROIC, up from a reported 12% in 2025, and kept its organic growth model at low-teens, 10% to 13% on average over time. CEO Wolfgang Wienand framed the upgrade as the payoff from 22 months of operating-model change. CFO Philippe Deecke confirmed 2026 guidance of 11% to 12% constant-currency sales growth after what he described as strong momentum in the third quarter.
A Higher Margin Bar, Built on Manufacturing Discipline
The central message was that Lonza sees itself as a manufacturing company rather than a pharmaceutical company. Wienand said this was the confusion he found when he joined in July 2024. He contrasted the roughly 90% gross margin of large pharma with Lonza's gross margin of around 40%. "We are not a pharmaceutical company. We are a manufacturing company" was, in his words, the most used sentence of his first year. In practice, he said, that meant new productivity KPIs for every asset, tighter capex scoping and a commercial model that prices for value.
Deecke identified three margin drivers. The first is maturing growth projects, which absorb cost better as utilization rises. The second is commercial and operational excellence. The third is operating leverage in administration, where Lonza intends to grow costs well below the rate of sales while continuing to raise R&D spending. Deecke also pointed to portfolio mix optimization, which raises asset margins with little or no incremental capital through repricing, process improvement and moving products to better-suited assets. He cited a more than 10-point margin gain in the former small molecule business as precedent. Management hedged the target, however. Deecke said the high 30s "can also be a couple of points different" depending on how the growth projects deliver, and asked investors for "leeway" to invest ahead of plan when warranted.
The integrated biologics unit offered the most concrete operational number of the day. Gordon Bates, who runs the platform, said Lonza expects a 20% increase in batch throughput from existing assets over the midterm. He said the plan rests on asset-by-asset programs covering cycle times, changeovers and debottlenecking, with modest capex in some cases. One example was automated column packing in Singapore, which replaced manual packing of large-scale columns. Bates did not give a margin drop-through figure when pressed by RBC's Charles Weston, saying only that productivity makes a major contribution to the margin guidance. Bates also said Lonza cut one-third of its R&D programs in the past year because they were "too close to intellectual gratification rather than something that would be translated into benefit in the factories."
Why the Return Target Looks Conservative but Is Framed Carefully
Jefferies' James Vane-Tempest noted that ROIC was described at first-half results as already close to 2x WACC. He asked why the 2x to 3x range has a low end that is essentially today's level. Wienand said the range should be read as progress over time, "surpassing the lower threshold to then further increase." He said he did not want to build in an assumption on a rising WACC. Deecke added that reported ROIC includes capital already committed to future growth, plus infrastructure and goodwill that generate no direct return, so the underlying run-rate return is "much higher." The company said each growth project is designed to reach at least 30% ROIC at peak. Mature large assets already exceed 30%.
Goldman Sachs' James Quigley challenged the capex logic. The model assumes 14% to 19% capex to sales supports 10% to 13% growth, yet Lonza spent 20% to 30% over the past five years. Wienand attributed part of the gap to catch-up spending on infrastructure and systems neglected in the late 2010s. He also credited a new discipline on scope and technical reference costs, which he said has cut capex by 30% to 40% on individual projects. Deecke added that two earlier investments lacked durability: a small biologics site in China that was invested in and later closed, and the Moderna capacity built during COVID. That candor is relevant for investors weighing how clean the historical spending-to-growth link has been.
Growth Algorithm and Capacity Pipeline
Management laid out a build-up for the 10% to 13% corridor. A 6% to 7% underlying pharma market, 1% to 2% from picking high-growth modalities and 1% to 2% from continued outsourcing give an addressable market growing 8% to 10%. Lonza then adds 2% to 3% from what it calls the Lonza engine. On the supply side, Deecke said growth capex of 10% to 13% of sales should generate 8% to 10% growth from new capacity after construction and ramp-up, with base assets supplying the remainder. Lonza runs 25 large growth projects worth CHF 8 billion, 90% of it in commercial and mixed assets, which serve all eight modalities. Deecke said every asset needed for growth through 2030 either exists, is ramping up or is in late-stage construction. The main contributors are Vacaville, the 20,000-liter asset in Visp, Portsmouth, commercial bioconjugation in Visp, and highly potent API and drug product assets in Stein. Three new projects announced this year, in spray drying in the US, a commercial highly potent API linker for ADCs and a second ADC aseptic fill/finish line in Stein, start the next cycle. Wienand said further infrastructure announcements will follow a scenario-planning exercise that tested network setups against nine geopolitical scenarios. He flagged bioconjugation in the US as a gap and a likely investment area.
Wienand pushed back on the view that US onshoring by big pharma will fill the market. He said big pharma capex is growing about 3% a year, below the 6% to 7% market growth, with capex-to-sales unchanged. Only about 5% of the $600 billion of announced US investment has broken ground on manufacturing sites, mostly for GLP-1s. Lonza's proprietary supply model shows continued mammalian undersupply, and Wienand said "nothing that has been announced as of today by any CDMO, including us, or the big pharma companies will fundamentally change" the US shortfall. He also said Lonza is "not in a capacity race" and is competing on manufacturability rather than price. Customer surveys put quality first and price fifth among outsourcing criteria.
Vacaville: Roche Backfill Done, Peak Sales Still Years Away
Vacaville closed on October 1, with roughly 95% staff retention. Wienand said contracts signed by January 2026 fully replace the outgoing Roche business. That secures revenue of roughly CHF 0.5 billion until 2028, after which sales should grow toward a peak of up to CHF 2 billion in the first half of the 2030s. The site has completed its first technology transfer and passed a US FDA inspection.
Investors hoping for further de-risking got a measured answer. Bates said the site is "ahead of where we thought we needed to be," but described considerable adaptation work. This includes gas delivery to bioreactors for certain cell lines, more buffer and media hold capacity, larger purification columns, and expanded warehouse, quality control and control-system capacity. Many products will transfer over the next three years, and engineering and validation batches mean volumes will not arrive on day one. Asked by Bank of America's Charlie Haywood about earlier sales, Bates said he did not expect significant acceleration of the 2028 inflection, unless demand for early products is unusually high. Deecke said the construction shutdown of one Vacaville asset is on schedule, and the Visp large-scale mammalian facility is ramping as planned.
Large-Scale Capacity Still Matters, but Titers Will Slowly Shift the Mix
Bates offered the most analytically useful argument for biologics investors on scale mix. The average titer across 465 marketed CHO-derived products is 3.1 grams per liter. He estimated it could take 15 to 25 years for that average to reach 6 grams per liter, because the very high titers seen in newer programs sit in early development, and the median time to approval is 11 years. Legacy products make up a large share of blockbuster volume, and he doubts innovators will spend resources reformulating old molecules. He said "rumors of the death of large-scale mammalian capacity are overexaggerated," though mid-scale and small-scale capacity will grow in relevance. About 22% of commercial products need more than 250 kilograms a year and account for about 80% of volume. Lonza has 75% of its biologics capacity at large scale, 20% at mid-scale and 5% at small scale. Bates forecast total biologics volume rising about 40%, from more than 60 tonnes to more than 100 tonnes. Biosimilars account for only around 15% of the volume.
JPMorgan's Zain Ebrahim questioned the 3.1 grams per liter figure because one pharma company has talked of moving from below 2 to 4.8 grams per liter. Bates said the number is industry data, and it matches Lonza's own commercial portfolio.
Integrated Offering and Early Evidence on AI
Lonza says 70% of new molecules signed are integrated drug substance and drug product offerings, and 100% of new programs where it does early development are integrated. Phase II and III integrated programs are up 35% and 70%. Revenue in integrated biologics is 70% large pharma and 30% biotech, though biotechs make up 75% of the customer count. Commercial products dominate revenue, and a large share of molecules arrive at the commercial stage through added demand, dual sourcing or what Bates called a "rescue mission" after another supplier's failure. Wienand said 99% of molecules that enter Lonza's system stay through development stages once clinical failures are excluded.
The AI evidence was early but specific. Wienand cited a project with a large pharma customer where historical batch data and a human-in-the-loop expert lifted output by 15%. Bates said resin screening in downstream development has fallen from 12 weeks to 4 weeks using data. A predictive model due next year aims to bring it to under a week with a confirmatory experiment. He added that gene-to-INB timelines for complex molecules have fallen to 6 months, from 13 to 15 months about 18 months ago.
China Innovation, M&A and Other Questions
Morgan Stanley's Thibault Boutherin asked how rising Chinese biotech innovation affects the model. Wienand said Lonza already manufactures molecules originated in China and gave an example of a client that bought a Phase II asset and called to say "the supply chain is horrible." Lonza provides GS and Synaffix technology to Chinese innovators and runs a small molecule site in Nansha. It will not develop Chinese innovation for the Chinese market, and any further China exposure will be asset-light.
On M&A, Wienand said Lonza evaluates deals against two tests: whether the business model creates value for customers and Lonza, and whether the Lonza engine can add something distinctive. He said the strategy team is actively approaching owners of assets that may not yet be for sale. Deals for underutilized pharma assets, he said, may carry a transformation period before they grow at the 10% to 13% rate. Vacaville is the archetype. He did not announce any transaction.
On capital allocation, Deecke said about CHF 1.7 billion of proceeds from the partial exit of Capsules and Health Ingredients to Lone Star Funds are expected this year, with closing before year-end. The company has committed to a CHF 500 million share buyback once the deal closes and retains a 40% stake. Leverage is expected to fall significantly as margins and cash conversion improve, though free cash flow conversion was described only as roughly double the 2024 rate.
Third-Quarter Color and Near-Term Watch Items
Deecke said Advanced Synthesis, which had a very strong first half, should be roughly flat in the second half against a strong prior-year comparison. He said this was expected and does not signal the direction for 2027. Bioscience is normalizing after large first-half orders that were not expected to repeat. BNP Paribas' Odysseas Manesiotis noted that bioprocessing peers have not flagged similar weakness. Deecke called the business niche, with concentrated customers. Wienand cautioned investors against extrapolating the first-half and second-half split and advised using 2026 as the base year. Deecke guided to a 2% to 3% currency headwind to sales from the US dollar against the Swiss franc.
The day left two questions open. Management did not quantify the margin drop-through from the 20% throughput gain. It also declined to give a market share figure in ADC conjugation, saying only that Lonza is the leader in commercial volumes and toolbox breadth. Bates also declined to give a full capex update on Vacaville beyond saying the work is on track. The margin ambition and return framework are now explicit, and investors will want to see how much of the step-up comes from throughput gains, portfolio repricing and administrative leverage as the growth projects mature.