Novartis Spent a Decade Buying a $38 Billion Business — Then Gave It Away on Purpose

In April 2008 Novartis began assembling one of the largest eye-care companies on earth, signing an agreement with Nestlé to acquire a 25% stake in Alcon with options on the remaining majority.⁹ It bought into Alcon in stages, paid Nestlé USD 28.3 billion for the controlling majority alone, and ended up with a blended cost of roughly USD 38.5 billion — about USD 168 a share — for full ownership.⁶ Then, eleven years later, it gave the whole thing back to its own shareholders for nothing. On April 9, 2019, Novartis handed every investor one Alcon share for every five Novartis shares they held and let it float away onto the NYSE.⁴⁵ Four years after that, it did the same to Sandoz, a business doing USD 9.2 billion a year.¹⁷ This is not the story of a company unwinding a mistake. It is the story of a company deciding, very deliberately, what kind of revenue it refused to keep.

The easy reading is that Alcon and Sandoz were laggards — units that disappointed and got cut loose. That reading is wrong on the facts. Alcon's unit sales had grown for two straight years before the 2019 separation, and Novartis never claimed a turnaround had failed.⁸ The real story is colder and more interesting: these businesses were healthy, and Novartis let them go anyway — because being healthy was not the bar. Earning like a patented medicine was the bar.

The number that decided everything: 32 versus 18

Strip the strategy decks away and the entire decade-long pivot rests on a margin gap. Novartis' pharmaceuticals division ran a core profit margin around 32%. Alcon, the surgical and vision-care business, ran around 18%. Sandoz, the generics and biosimilars arm, sat near 20%.⁸ Those are not rounding errors — they are different physics. A patented drug is a temporary monopoly: the molecule is protected, the price is defended, and the margin reflects scarcity. A generic is the opposite by design — it exists to be cheap, interchangeable, and competed down to the bone. Medtech sits in between, dragged by hardware costs and slower replacement cycles. Bolting an 18% business onto a 32% business doesn't average to something fine. It dilutes the whole, and it forces a 32%-return company to ration its capital across opportunities that can never clear its own hurdle.

Notice what that sentence is really admitting. Capital and attention are finite, and inside a conglomerate they are spent by committee. Every dollar a 32%-margin pharma business deploys into an 18%-margin eye-care unit is a dollar not deployed into the next gene therapy. The dilution is not just on the income statement — it is in the boardroom, in the time senior leaders give to surgical lasers instead of oncology pipelines. The spinoff doesn't fix Alcon. It frees Novartis.

The cannibalization nobody forced

What makes this an archetype rather than housekeeping is the willingness to give up scale. Sandoz alone was 18.3% of Novartis' total net sales of USD 50.5 billion in 2022.⁷ Most companies guard revenue like territory; shrinking by nearly a fifth on purpose is the kind of move that gets a CEO questioned. Novartis ran a formal strategic review and concluded a full 100% spinoff — not a sale, not a carve-out keeping a stake — was in shareholders' best interests, announcing the intention on August 25, 2022.² Shareholders approved it at an extraordinary meeting on September 15, 2023, and Sandoz began trading independently on the SIX Swiss Exchange under SDZ on October 4, 2023, again at one new share for every five Novartis shares.¹³ The deal was structured to be tax-neutral, which tells you the goal was never to harvest cash — it was to subtract a business cleanly from the body.³

The honest objection: diversification was the safety net it just cut

The fair counter is that Novartis just dismantled its own hedge. A pharma company lives and dies by the patent cliff — every blockbuster has an expiry date, after which generics gut its price overnight. Sandoz was, quite literally, the business that profited from those cliffs; Alcon's slower, hardware-anchored cash flows smoothed the lumpy economics of drug discovery. By going pure-play, Novartis has removed the ballast and left itself entirely exposed to the one thing it can least control: whether the next molecules in its pipeline actually work. There is no generics arm to soften a failed Phase III now. The bet is total.

But that objection assumes diversification was actually buying insurance. It wasn't — it was buying a discount. Markets have widely been observed to price diversified conglomerates below the sum of their parts — a pattern finance scholars call the diversification discount — which in Novartis's case meant a 32%-margin pharma engine was being valued partly through the lens of a 20%-margin generics business and an 18%-margin medtech unit. The argument was that the 'hedge' was costing Novartis the premium multiple a focused drugmaker earns. The pivot only makes sense if you believe the pipeline can stand on its own — and Novartis is betting it can. That is the whole wager: not that focus is safer, but that its remaining medicines can sustain the premium margins generics and medtech structurally cannot, and that a pure-play deserves to be priced like one.

Novartis spent eleven years and nearly forty billion dollars learning that owning a great eye-care company doesn't make you a better drug company — it just makes you a more diluted one. So it ran the trade in reverse: shed Alcon, shed Sandoz, shed almost a fifth of its revenue, and kept only the part that earns like a monopoly because it briefly is one. The genius of the pivot isn't the businesses it built. It's the conviction to give them away, and to want nothing back but a sharper definition of what Novartis is for. Whether that conviction was wisdom or hubris will be settled by the pipeline — which is exactly where Novartis has now staked everything.