Strip away the cultural noise around obesity drugs — the celebrity anecdotes, the think pieces, the shortages at your local pharmacy — and you are left with a manufacturing problem. GLP-1 medicines are biologics filled into glass cartridges, assembled into pen devices, and kept between two and eight degrees from the filling line to the patient's fridge. Every step of that journey runs through capacity that is finite, specialised and, at present, almost entirely spoken for.
This is the fact the demand forecasts glide over. The commonly cited models — 15 million, 25 million, 40 million patients by 2030 — all describe willingness to take the drug. None of them describe the industry's ability to fill it. The second number is smaller, and it is the one that matters.
Counting filling lines, not patients
Sterile fill-finish capacity for cartridge-based biologics is concentrated in fewer than thirty facilities worldwide that can operate at commercial scale and regulatory grade. Lumera's capacity tracker, updated in December 2025, counts 41 expansion projects announced since 2022 across Novo Nordisk, Eli Lilly and eleven contract manufacturers. The committed capital is roughly $23 billion. Even so, the lead times are brutal: a greenfield fill-finish line takes three to four years from sanction to validated commercial runs, and validation itself cannot be rushed because regulators will not let it be.
Our base case puts effective industry output for the GLP-1 class at about 1.9 billion cartridge-equivalents in 2028 — enough, depending on dosing mix, for somewhere between 28 and 34 million patients on therapy. That is an enormous number by any pharmaceutical standard. It is also below the mid-point of the demand models, which means allocation, not marketing, decides who gets treated for at least the next three years.
Demand is the least interesting variable in the obesity market. Capacity is the whole story, and the story runs through a filling line in Clayton or Kalundborg, not a prescriber's office.
Allocation is visible everywhere once you look for it. Novo's Wegovy launches have been sequenced country by country with starting-dose rationing. Lily's Zepbound supply updates read like logistics memos. And the compounding pharmacies that filled the gap in the US are being squeezed out ingredient by ingredient as the FDA shortage database clears.
What breaks first: the pen
If one component deserves its own paragraph it is the autoinjector. Each weekly pen requires a spring, a needle sub-assembly and a drive mechanism manufactured to tolerances the pharma industry used to associate with aerospace. Two firms, Ypsomed and SHL, dominate the supply of the underlying platforms, and both are sold out on relevant lines through 2027. When a company tells investors its obesity supply is constrained by 'device components', this is what it means.
The pricing question follows from the same arithmetic. Conventional wisdom expects obesity prices to halve once competition arrives. But orals using small-molecule GLP-1s — the event that would crash injectable pricing — require their own manufacturing scale-up, with the added complication of vastly larger API quantities per dose. Our pricing model has injectable list prices falling just 9 to 14 percent by 2028, with the real deflation arriving 2030 and beyond.
The companies that will look cleverest in hindsight are not the ones with the best molecules. They are the ones that booked filling lines in 2022, while everyone else was writing demand forecasts.