Novo Nordisk A/S

Two years ago this was the most valuable company in Europe, worth more than LVMH, and its product had become a verb, a punchline, and a red-carpet whisper all at once. Today it trades ~63% below that peak, and the question this report must answer is the one Mr. Market asks every time a great business collapses: is this a wonderful company on sale, or a falling knife with a patent cliff at the bottom?

Novo Nordisk makes semaglutide — sold as Ozempic (a weekly injection for type-2 diabetes), Wegovy (the same molecule at a higher dose, for obesity), and Rybelsus/oral Wegovy (the pill). It is a Danish company with a hundred-year history in insulin that, almost by accident, discovered its diabetes drug also melted away body weight — and stumbled into the largest new drug market of the century. The economics are the kind Buffett dreams about: 82% gross margins, a 45% operating margin, and a 66% return on equity — a pharmaceutical royalty on human metabolism. In the year ended December 2025 it earned 102 billion Danish kroner (~$15.7 billion) on 309 billion kroner (~$47 billion) of revenue. And it is controlled by a charitable foundation that legally cannot sell it — the ultimate long-term owner.

"I don't believe that obesity prevention and treatment is a medical problem, in general. It's a social and cultural and societal problem…" — Lars Rebien Sørensen, Chairman (and former CEO)

It is a startling thing for the chairman of the company selling the medical fix to say — and it frames the whole investment. This is a franchise whose demand is close to certain (the world is not getting thinner), whose science is proven, and whose economics are extraordinary. What is not certain is how much of that enormous market Novo keeps — because a fierce competitor, a patent clock, and the politics of American drug pricing all now sit between this company and its former glory. Note the timing, too: weeks ago we named GLP-1 drugs the central long-term risk to McDonald's. Today the maker of that very molecule trades at a lower earnings multiple than the burger chain it threatens. The disruptor is cheaper than the disrupted.

The history matters because it explains both the durability of the franchise and the paradox at its heart: a company chartered to defeat the very diseases it profits from.

Two lessons for an owner. First, the franchise has survived a century — two world wars, the insulin-patent era, decades of competition — because the underlying need never went away; that is the definition of durability, and it is why the base business deserves respect even now. Second, the fall was not one shock but a cascade: a pipeline disappointment, a competitor's efficacy win, a pricing squeeze, and a management purge, stacked over eighteen months. When you buy today, you are betting that cascade is mostly in the price — and that the oral-pill launch is the first chapter of the next act, not a last hurrah.

The machine, in five steps:

This is the honest boundary. Buffett famously avoids pharmaceutical single-product bets because a molecule's ten-year economics turn on trial data, patents and regulators he cannot handicap — and semaglutide is a single-molecule concentration: it is the majority of revenue and essentially all of the growth. The counter-argument, which a fair analyst must weigh, is that the diabetes franchise beneath it is genuinely durable (a hundred years and counting), and that the obesity market is so vast and so under-penetrated that even a losing No. 2 in a rational duopoly compounds for years. You are not buying certainty here. You are buying a wonderful business at a price that assumes the worst — which is a different, and often better, bet.

Novo reports in Danish kroner; its ADR trades in dollars (≈6.5 kroner to the dollar). The revenue splits two ways — by disease and by geography — and both tell you where the risk lives:

The concentration is the whole story, both ways. Semaglutide is the majority of revenue and nearly all of the growth — a single molecule carrying a $226 billion company. That is the strength (extraordinary margins, focus, scale) and the risk (one patent, one competitor's data point, one pricing decision can move the whole thesis). And the profit geography is lopsided: North America is ~63% of revenue but a far larger share of profit, because a Wegovy pen that nets Novo a few hundred dollars in America nets a fraction of that in Europe. So when Washington forces US GLP-1 prices down — as it now is (Part XI) — it hits the most profitable dollars first. The bull's answer is volume: even at lower prices, the number of treated patients is a small fraction of the eligible population, and the oral pill is pulling in people who would never have taken an injection. Price down, volume up — the question is which wins the race, and 2026's guided decline says price is winning for now.

Novo's moat is real but — like Alphabet's — contested for the first time in a generation. Three walls, honestly graded:

1 · Manufacturing scale in peptides. This is the strongest and least appreciated wall. Injectable GLP-1s are complex peptides that are genuinely hard to make at billions-of-doses scale; the binding constraint on the whole industry has been capacity, not demand. Novo has a century of fill-finish and sterile-manufacturing expertise and just spent enormously — including its sister foundation's ~$16.5B purchase of Catalent — to lock in supply. Trend: widening, and it is why the free-cash-flow number looks so ugly this year (Part IX).

2 · Brand and the diabetes franchise. 'Ozempic' is the category verb; a hundred years of diabetes relationships give Novo prescriber trust no start-up can buy. Trend: stable, though the obesity brand is now a two-horse race.

3 · The patent on semaglutide. The weakest wall, because it has an expiry date. US composition-of-matter protection runs to roughly 2032, but generics arrive far sooner elsewhere — Canada, Brazil and China face semaglutide generics in 2026. A moat with a countdown clock is a moat you must value differently.

Where it cracks — the Lilly ladder. Eli Lilly's tirzepatide beat semaglutide head-to-head (~20% vs ~14% weight loss in the SURMOUNT-5 trial), its triple-agonist retatrutide reaches ~24%, and — most dangerously — its orforglipron is an oral small molecule, far cheaper to mass-manufacture than Novo's peptide pill and free of the food-and-water timing rules. If the future of this market is cheap oral small molecules, Novo's peptide-manufacturing moat — its best wall — is partly bypassed. Lilly is now worth roughly three times Novo. That reversal, from a standing start in 2023, is the single most important fact in this section. Score the moat a 7: three real walls, one with a clock, and a rival climbing the strongest one.

Everything turns on this. The GLP-1 market is variously sized at $100–150 billion by 2030, and penetration today is a small single-digit fraction of the eligible population — the volume runway is enormous. The question is whether that runway is shared rationally by two profitable kings, or fought over in a price war that ruins the economics for both.

Our read: this is a genuine two-sided question, which is precisely why the stock is cheap — but the weight of the 2026 evidence tilts constructive. The single most important data point of the year is that the oral pill's scripts are >80% new-to-GLP-1 patients: that is a market expanding, not a zero-sum share war. Novo can lose the efficacy crown to Lilly and still compound handsomely if the pie triples — the way Pepsi compounded for decades as a permanent, profitable No. 2 to Coca-Cola. What it cannot survive is a true price collapse combined with a share collapse; the buyer at $51 is betting that both happening together is already priced in. It is the honest crux, and no one — not Novo, not Lilly, not I — can prove the volume-versus-price race before it is run.

Strip away the noise and there is really one competitor: Eli Lilly, and the story of the last two years is Lilly's ascent from peer to victor. Novo's answer is not to out-efficacy Lilly — it may not be able to — but to out-manufacture and out-distribute: win on capacity, on the pill's convenience, and on the sheer size of an expanding market. One telling defeat sits in this section, though: in late 2025 Novo tried to break up Pfizer's purchase of the obesity biotech Metsera, escalated its bid to ~$7.6 billion, and lost — Metsera's board took Pfizer's ~$10B offer, citing antitrust risk in Novo's unusual deal structure. The former king, outbid and blocked. It is a small event with a large signal: Novo is no longer the automatic winner of anything in this market.

The ownership structure is the most unusual on our board, and understanding it is half the investment case.

Why this matters to a value investor. The Foundation cannot be raided, cannot be pressured by activists, and cannot sell — its charter binds it to control Novo Nordisk 'for as long as the Foundation exists,' and to spend the proceeds fighting the very diseases the company treats. That gives management the rarest luxury in modern business: the freedom to invest through a crisis for a decade-long payoff, exactly what a wounded franchise needs. The flip side, on honest display in 2025, is that the Foundation rules absolutely — it purged more than half the board in a single meeting when it wanted faster change, and it installed its own chairman over the operating company. You are a genuine minority partner here. Capital allocation is otherwise exemplary: a 66% return on equity, dividends raised through the crash (up ~2.6% in 2025 even as the stock halved), and buybacks — though buybacks were sensibly cut from ~DKK 20B to a smaller DKK 15B programme to fund the manufacturing build-out. One blemish: the failed, faintly desperate Metsera chase. Score management a 7 — a superb, patient, mission-locked owner, freshly purged, with one clumsy M&A miss.

Two things to hold together. First, the profitability is genuinely elite — 82% gross margins, a 66% return on equity, 37% of every krone of revenue dropping to net income. Businesses this profitable are vanishingly rare, and they do not usually trade at twelve times earnings. Second, the two flagged lines explain why the stock looks cheaper than it is on cash flow: capex nearly doubled (to DKK 90B) as Novo raced to build peptide capacity, and free cash flow was halved as a result — from DKK 70B to DKK 29B. That is not a business in decline; it is a business spending its cash flow to build the factories that defend its moat, exactly the pattern we untangled for the AI capex giants. The dividend, tellingly, was raised through the worst of the crash — a foundation-owned company signalling that it sees a trough, not a cliff.

You have seen this shape four times now. On free cash flow, Novo looks expensive (48×) — but that is the same illusion we corrected for Microsoft, Alphabet and Amazon: a company pouring its cash into growth capacity shows a depressed FCF that understates its true earning power. Strip the growth capex out and the picture inverts. The automated DCF, in its natural habitat — a durable, high-margin cash compounder — reads $89.24, a striking +73% above the $51 price; unlike Tesla's DCF, this one is not confused, it is looking through the capex bulge to the earning power beneath. Against that sits the market's fear, and one genuinely awkward fact: 2026 earnings will fall — the guided revenue decline drops FY2026 EPS to roughly $3.30 per ADR before the recovery to ~$4.30 by 2030. So you are buying a declining year at ~16× forward, cheap trailing earnings at ~12×, and a DCF that says the whole thing is worth 70% more. The margin of safety is real but not free: it is compensation for the patent clock and the Lilly race. The zone deepens below ~$45, near the lows, where the yield crosses 4% and you are paid to wait.