Lonza Wants Higher Margins From Maturing Investments. I’d Like to Meet the Payback Schedule.

A new facility gets a name, a ceremony and photographs of executives holding unusually large scissors.

The return on capital generally gets none of those things. It is expected to arrive later, quietly, after customers qualify their products, production behaves and somebody collects the invoices.

I have nothing against the ceremony. I would simply prefer that the project remain on the investment committee’s agenda after the photographer leaves.

Lonza’s capacity ramp and margin ambitions make that distinction worth examining. Maturing investments can produce attractive economics: more revenue through an existing cost base, better use of equipment and less cash tied up in unfinished projects.

I find that encouraging. What I want to know is which operating milestones turn an investment from an expensive address into an earning asset.

My argument is that Finance should budget the gap between installed capacity and qualified, saleable output. If that gap is missing, the project has acquired a return through scheduling optimism. Operations will eventually send us the correction.

Lonza capacity ramp: the targets and the time frame

At its October 14, 2026 Capital Markets Day, Lonza upgraded its organic growth model for its continuing contract development and manufacturing business. It targets average organic sales growth of 10% to 13% over time at constant exchange rates, core EBITDA growth ahead of sales, and a core EBITDA margin in the high thirties in the medium term.

Its return-on-invested-capital ambition is two to three times today’s weighted average cost of capital, with improving free-cash-flow conversion. It also reiterated its 2026 outlook of 11% to 12% constant-currency sales growth and a 33% to 34% core EBITDA margin.

Lonza’s announcement is the source.

These are disclosed ambitions and outlook figures, not achieved margins for every facility. Lonza does not disclose an 8% WACC in the cited announcement; that rate appears only in my hypothetical examples below.

I would follow the project through several different gates

Construction completion establishes that a facility has been built. It doesn’t establish that every product can be manufactured there, that customers have approved it or that all output can be released for sale.

In a regulated manufacturing setting, qualification and process validation can be crucial. Different products may require different readiness work. I’d avoid assuming one site-opening date converts every planned customer into revenue.

My model would distinguish equipment readiness, qualified capacity, booked demand, scheduled production and released output. Then it would follow billing and collection.

That is how I’d connect Finance to Operations. A facility can be physically ready while the commercial plan is behind. It can produce batches while quality release takes longer than expected. It can recognize revenue while cash arrives later still.

None of that automatically means management has failed. It means the project needs more than one date.

The construction schedule may be green. The cash schedule is allowed to disagree, however inconvenient that is for the update meeting.

First worked example: capacity changes the profit curve

Take a hypothetical facility with $200 million of annual revenue capacity at assumed normal pricing. Variable costs equal 40% of revenue, and fixed annual cash operating costs are $45 million.

At 50% utilization, revenue is $100 million. Variable cost is $40 million, giving $15 million of EBITDA after fixed cash costs.

At 75% utilization, revenue is $150 million. Variable cost is $60 million, leaving $45 million of EBITDA.

At 90% utilization, revenue is $180 million. Variable cost is $72 million, leaving $63 million of EBITDA.

Revenue rises 80% between the first and third cases; EBITDA rises 320%. The facility hasn’t become four times larger. The fixed cost base is being covered by more revenue.

These are fictional plant economics, not Lonza’s facility results. They show why a commercial ramp can be a powerful margin driver and why the reverse movement deserves equal attention.

Operating leverage is very friendly when utilization rises. It becomes less sociable when the customer schedule slips.

EBITDA break-even isn’t an investment return

The hypothetical plant’s contribution margin is 60%. Covering $45 million of fixed cash cost requires $75 million of revenue: $45 million divided by 60%.

That is 37.5% of the stated revenue capacity.

Now assume annual depreciation of $15 million. EBIT break-even requires $100 million of revenue: $60 million of fixed cash cost plus depreciation, divided by 60%. That is 50% utilization.

A plant at 50% utilization can therefore produce positive EBITDA while producing no EBIT in this simplified model. The equipment is being used up, and depreciation has made an unwelcome but entirely legitimate appearance.

Neither break-even measure establishes that the original investment earned an acceptable return. They tell us when particular profit lines turn positive.

I’d label them separately. Otherwise “profitable plant” becomes a remarkably flexible phrase, capable of surviving almost any follow-up question until someone asks about the capital.

Put invested capital beside the margin

Assume the hypothetical facility required $250 million of fixed investment and supports $25 million of operating working capital at maturity. Total invested capital is $275 million.

At 90% utilization, the $63 million EBITDA less $15 million depreciation gives $48 million EBIT. At an assumed 25% operating tax rate, after-tax operating profit is $36 million.

Simplified mature ROIC is approximately 13.1%: $36 million divided by $275 million.

At an illustrative 8% WACC, that is about 1.64 times the cost of capital. The plant’s EBITDA margin is a healthy-looking 35%, but its modeled ROIC does not reach twice that assumed WACC.

Those observations are entirely compatible. Margin measures profit relative to revenue. ROIC asks what the profit earns on the capital supporting it.

I’d use consistent definitions and average invested capital for actual reporting. I’d also reconcile company-defined core metrics with statutory and cash measures.

A strong margin can justify interest. It cannot excuse us from looking down the balance sheet.

What would a higher ROIC require?

To earn a 16% ROIC on $275 million, the plant needs $44 million of after-tax operating profit. At the assumed 25% tax rate, required EBIT is about $58.67 million.

Add $15 million of depreciation and $45 million of fixed cash cost, then divide by the 60% contribution margin. Required revenue is approximately $197.78 million, or 98.9% of the stated capacity.

A 24% ROIC under unchanged assumptions would require $66 million after tax, $88 million EBIT and approximately $246.67 million revenue. That exceeds the stated capacity.

The answer cannot be “work harder” pasted into the utilization cell. Reaching the higher return would require different pricing, mix, productivity, costs or capital intensity.

This isn’t an assessment of whether Lonza’s group ambition is feasible. It is a hypothetical demonstration of what an ambitious return requires from a project model.

I’d rather discover the mathematical constraint before asking Operations to achieve 123% of capacity through positive thinking.

Qualified output matters more than a full production schedule

A utilization figure can measure scheduled hours, equipment activity or commercially saleable output. Those are not interchangeable.

I’d define the denominator and the numerator before comparing sites. A plant running many hours with low yield may consume capacity and cash without delivering the planned revenue.

In the model, production throughput should connect to yield, release rates and customer demand. I’d include downtime, maintenance windows and changeovers where they matter to decisions.

The accounting treatment also deserves attention. Producing more inventory can spread fixed manufacturing costs across units. It may improve current-period reported margins while tying up cash in stock that has not sold.

I’d ask whether the margin improvement came from customer shipments or inventory accumulation. Accounting absorption is legitimate within applicable rules; treating extra stock as proof of economic demand is not.

The plant can be busy and the business can remain underutilized commercially. A loud factory is not a cash-flow forecast.

Customer qualification is part of the capital case

I’d map each important customer program to qualification milestones, expected launch timing and realistic volume. A commercial opportunity needs a route through the facility’s constraints.

The finance model should record dependencies. If a customer’s validation is delayed, which production slots become available? Can another program use them? Does the alternative require different equipment or additional setup?

I would also distinguish a contracted minimum from an aspiration. Terms may contain volume flexibility, termination provisions or timing conditions. The relevant revenue assumption depends on those details.

This is where integrated operational and financial planning becomes more useful than another capacity chart.

Sales may be planning a launch. Operations may be planning qualification. Finance may be planning invoices. Until those activities share a calendar, the company can have three sensible forecasts describing three different years.

I’d treat the mismatch as a decision to resolve, not a formula to disguise.

A six-month delay has a bill

Return to the hypothetical plant. Suppose the plan expected 75% utilization for a year but the ramp is delayed, leaving the plant at 50% for the first six months and 75% for the second.

Annual revenue becomes $125 million: half of $100 million plus half of $150 million. EBITDA becomes $30 million: half of $15 million plus half of $45 million.

Compared with a full year at 75%, the delay costs $25 million of revenue and $15 million of EBITDA under the assumed cost behavior.

If additional readiness work costs $4 million during the delay, EBITDA falls to $26 million before any classification adjustments. I wouldn’t quietly exclude the $4 million from the investment’s economic case just because management uses a core earnings measure.

Some fixed costs might be avoidable; some revenue might be shifted rather than lost permanently. I’d show those differences explicitly.

A delay doesn’t have to destroy a project to reduce its value. Time is expensive even when everybody still agrees the project is wonderful.

Second worked example: the lifetime cash schedule

Consider a separate hypothetical investment with $250 million paid at time zero. Assume net project cash flow, after taxes, all further capital expenditure and working-capital movements, is negative $10 million in year one, $15 million in year two and $35 million in year three.

Assume $50 million annually in years four through ten, plus $50 million of net terminal recovery at the end of year ten.

At an 8% discount rate, the present value of the future cash flows and terminal recovery is approximately $261.2 million. Subtract the $250 million investment: NPV is about positive $11.2 million.

The calculation uses each year’s cash flow divided by 1.08 raised to that year. There is no financing interest deduction in those project cash flows because this is an unlevered valuation discounted at the assumed WACC.

The model is useful precisely because it is only modestly positive. A successful mature run rate doesn’t guarantee a large value cushion after a slow start.

Delay and financing assumptions change the answer

If every future cash flow in that separate example arrives one year later, including terminal recovery, while the initial investment remains at time zero, future present value falls to about $241.8 million.

NPV becomes approximately negative $8.2 million. This simplified delay scenario excludes additional idle-period costs, so it is not a worst case.

If the schedule stays unchanged but the discount rate rises from 8% to 10%, NPV falls to approximately negative $18.2 million.

Neither scenario proves that a real project should be stopped. It shows how a small initial cushion depends on timing and required return.

I’d also reconcile liquidity separately. A company can fund an economically attractive project poorly, or fund an unattractive project comfortably. Debt availability doesn’t make the project’s cash flows more valuable.

The cost of capital is allowed to change while construction is underway. The original business case does not get to preserve a cheaper rate because it preferred that version.

EBITDA needs a bridge to the bank balance

At maturity, suppose our first hypothetical plant earns $63 million of EBITDA. With $15 million depreciation, operating taxes at 25% of $48 million EBIT are $12 million.

Subtract $8 million of maintenance capital expenditure and $10 million of incremental working capital. Illustrative unlevered cash flow is $33 million: $63 million less $12 million, $8 million and $10 million.

That is substantially below EBITDA. It also excludes any growth capital expenditure and other specified project costs.

I’d distinguish maintenance from expansion spending without allowing the classification to become a beauty contest. If the plant needs a recurring investment to preserve current output, the economic model needs it regardless of the label.

I’d monitor capital expenditure and operating benefits together. The investment may still be attractive. Its cash return should be described honestly.

EBITDA is useful. It is not the amount available to distribute simply because depreciation has been politely excused from the room.

I would compare several ways to improve the ramp

The first option is accelerating commercial qualification and customer onboarding. That may require technical staff, validation work or contract changes. I’d compare the spending with incremental contribution and the probability of moving the timeline.

The second is improving yield, changeovers or reliability within existing capacity. That can create saleable output without another building. The cost and effect depend on the actual bottleneck.

The third is deferring the next expansion phase while using current facilities more fully. That preserves capital but may surrender customer opportunities or increase later construction costs.

A fourth option is changing the customer or product mix. It can improve returns, but it may also require new qualification or displace valuable long-term relationships.

I’d build the alternatives around the limiting factor. More sales effort will not resolve an equipment constraint. More equipment will not resolve the absence of qualified demand.

The company needs a diagnosis before it needs another capital request.

A focused intervention can deserve more funding

Suppose a hypothetical $5 million technical-readiness project brings forward $20 million of otherwise delayed revenue. At a 60% contribution margin, gross incremental contribution is $12 million.

Allow $2 million of additional operating support and $3 million of one-time intervention expense. Incremental pre-depreciation contribution after those expenses is $7 million before tax and the $5 million capital outlay.

I’d then model tax, depreciation, working capital and the duration of the timing benefit. Revenue brought forward may be offset by less revenue later; it shouldn’t be counted as permanent additional volume unless the evidence supports it.

Even with that qualification, accelerating an existing investment can be a better use of capital than adding another one.

This is why I’d resist a blanket spending freeze when a project falls behind. Some interventions improve the remaining economics materially.

Equally, I won’t endorse “one more investment” every time the original plan misses. A rescue request needs an incremental case. The project’s age does not qualify it for unlimited extensions.

Review the remaining choice without erasing the original mistake

If a project has already consumed $200 million, that history matters for accountability. It should not determine whether spending the next $30 million is worthwhile.

I’d evaluate the remaining alternatives using avoidable future costs and future benefits. Stopping can incur contractual penalties, remediation or lost customer value. Continuing can preserve returns or deepen the loss.

Separately, I’d compare the total project result with the original approval. Management needs to learn whether timing, capital requirements, demand or productivity assumptions were wrong.

These are two reviews with different purposes. The forward decision asks what to do now. The retrospective review asks what should change in the next approval process.

My argument about updating a historical planning baseline fits here: accurate historical spending does not establish a representative future return.

We can own a mistake without spending more money to make it look less lonely.

My conclusion: keep the return on the agenda

Lonza’s ambition to connect margin, ROIC and cash generation is the kind of capital discipline I want to see. The useful test is how the operating assumptions connect those measures over time.

I’d support additional investment when qualified demand, execution milestones and remaining cash economics justify it. I’d prioritize removing a demonstrated bottleneck over funding capacity merely because the strategic plan contains a growth arrow.

I’d also keep the full capital base visible. A mature facility’s strong margin must earn a return on everything required to create and operate it, not just on the last expenditure approved.

The distinction between a good operation and a good investment remains essential.

Before the next capital review, ask for the date of profitable output, the date of positive cash contribution and the date the original investment earns its required return.

The ribbon can be cut on one day. The return may take years. I’d like management to stay interested for both.

Sources and calculation notes

Disclosed targets are from Lonza’s October 14, 2026 Capital Markets Day announcement. All facility economics, tax rates, discount rates, cash schedules and interventions are hypothetical. They do not estimate Lonza’s specific project costs, WACC, qualification timelines or realized facility returns.