Novo Nordisk Built a 100-Year Moat. Its Newest Wing Sits on One Molecule.

In a town on the Danish coast called Kalundborg, one industrial site quietly makes about half the world's insulin.⁶ You cannot stand it up in a year. You cannot stand it up in ten. It took a hundred - a chain that runs from a technique August Krogh carried back from Canada, through the first Danish patients treated in March 1923, to a plant so embedded in global supply that other companies' patients depend on it.⁶ That is what a real moat looks like: not a clever feature, but a thing the world has organized itself around. And it is the part of Novo Nordisk's story that almost nobody is talking about.

The official story is that Novo Nordisk is a 100-year-old diabetes fortress, protected on every side. The truer story is that it owns two very different moats stacked on top of each other - and the one generating roughly half the revenue today is the shallow one.

The deep moat is the one built before anyone was watching

Start with ownership, because it is the most underrated structural advantage in the company. Novo Holdings, wholly owned by the Novo Nordisk Foundation, held 28.1% of the share capital but 77.3% of the votes at end-2025 - and the Foundation is obligated by its own Articles of Association to keep a controlling interest.⁵ Read that twice. The owner of last resort is legally forbidden from selling control. No activist can force a breakup; no acquirer can buy the company over management's objection. That is a moat against the capital markets themselves, and it is the kind of advantage you cannot replicate with money, only with a charter written decades ago.

Then there is the manufacturing and the heritage, which compound the same way. The modern company is not even as old as the marketing implies - it was formed in 1989 from the merger of Nordisk Insulinlaboratorium (1923) and Novo Terapeutisk Laboratorium (1925), two rivals that spent decades racing each other on insulin chemistry.⁴ That race produced NPH insulin in 1946, the NovoPen in 1985, and the Kalundborg base that still anchors world supply.⁶ In FY2024, insulin sales rose 15% to DKK 55.4 billion¹ - a business commoditizing on price yet still defended by sheer scale and decades of process know-how that a newcomer would have to rebuild from scratch.

The shallow moat is the one paying the bills

Now the part the century-old narrative quietly borrows credibility for. The growth - the reason the stock soared - is GLP-1: Ozempic and Wegovy, both built on a single molecule, semaglutide. Ozempic alone did DKK 120.3 billion in FY2024 sales, more than twice the entire insulin franchise, and obesity care surged 56–57% to DKK 65.1 billion.¹³ That is not a hundred-year asset. Ozempic only launched for type 2 diabetes in the U.S. and Canada in February 2018.² The thing carrying the bulk of the company — Ozempic's DKK 120.3 billion alone exceeds twice the insulin franchise, and obesity care added DKK 65.1 billion more — is six years old, and it is protected the way a young drug is protected: by patents, which expire on a clock.

Here is the part the 'protected through the early 2030s' headlines skip. Semaglutide's core composition-of-matter patent - the foundational claim on the molecule itself - expired around March–April 2026 in the U.S. and in several international markets.⁷ What remains is a thicket of secondary formulation and delivery-device patents running into the early 2030s, which can block branded biosimilars in the U.S. until at least 2031.⁷ But a thicket is not a wall. Secondary patents are contestable, jurisdiction-specific, and far weaker than a composition claim. The proof is already on the factory floor: after Novo Nordisk's semaglutide patent expired in India in March 2026, at least five domestic manufacturers launched generic versions, with Sun Pharma's Noveltreat available for as little as 750 rupees per weekly injection.¹⁰¹¹ The molecule moat is finite, and the clock has already started.

And the competitive pressure isn't waiting for the patents to lapse. Novo's GLP-1 value market share didn't just slip - it compressed from 85.6% in Q1 2024 to 81.3% by Q4 2024,⁸ and multiple reports citing Q2 2025 investor presentation data show Eli Lilly overtaking Novo to hold roughly 57% of the GLP-1 market.⁸⁹ So the moat is being excavated from two directions at once: a branded rival is already inside it, and the generics are massing at the gate. The depth Novo's century-long story implies is being actively dug out beneath the franchise that matters most.

Isn't a thicket of patents and a first-mover head-start still a moat?

The fair objection: a 55.1% value share of the GLP-1 diabetes sub-segment, plus a manufacturing base making half the world's insulin, plus a controlling foundation owner - that is plainly a defended position, not a vulnerable one.¹ True. The mistake is treating all three as one moat. The honest read is that they decay at completely different speeds. The ownership structure is near-permanent. The insulin infrastructure erodes slowly, on the gentle slope of biosimilar pricing. But the GLP-1 patent thicket is the wrong kind of moat for the franchise leaning on it hardest - it is a legal lease, not an owned asset, and the lease is running out. Note, too, the precision trap in the company's own numbers: the headline '55.1%' is the diabetes sub-segment, not the total GLP-1 market, where Lilly has already pulled ahead.⁸ When the impressive figure quietly narrows the category, the moat is being measured at its widest point, not its weakest.

Novo Nordisk spent a century building the rarest thing in pharma: a moat the market literally cannot force open, fed by a plant the world cannot route around. Then it built a second fortune on top of it - faster, larger, and resting on a single molecule whose foundational patent has already lapsed and whose lead a rival has already taken. Both are real. But only one of them is a hundred years deep. The genius and the exposure are the same fact: the company that proved a moat can outlast the capital markets is now discovering how quickly a molecule can outrun one.