India enters the GLP-1 race | The Daily Brief #449

ELI5/TLDR

On March 20th, Novo Nordisk’s Indian patent on semaglutide — the molecule inside Ozempic and Wegovy — expired, and Indian generic makers rushed in. Prices crashed by up to 90%. But semaglutide is a peptide, not a simple pill, and copying it well enough to pass Western regulators is turning out to be harder than anyone advertised. Meanwhile, Eli Lilly’s next-gen drug Mounjaro had already overtaken Wegovy in India before the patent even expired — the innovators moved upstairs while the generics chased the floor.

The Full Story

Why this patent expiry was such a big deal

Semaglutide is the biggest drug in the world right now — roughly $40 billion a year in global revenue, largely from Americans and Europeans paying $700 to $1,000 a month for a weekly injection that suppresses appetite and manages diabetes. When its Indian patent expired on March 20th, every major generics player had been sprinting toward the date for years. Dr. Reddy’s launched the same day under the brand Obida. Mankind, Zydus, Natco, Glenmark followed. The monthly price dropped from around Rs 8,800–11,175 to as low as Rs 1,290.

“A 90% price collapse in a matter of weeks. But Novo Nordisk didn’t decide to be generous. They were forced to.”

This is the playbook India has run successfully for decades — wait out the patent, copy the molecule, sell it cheap to the world. It’s why India supplies roughly 40% of US generic drug volume.

Why semaglutide isn’t a normal generic target

Here’s the part most coverage glosses over. GLP-1 is a hormone your gut releases after a meal to tell your brain you’re full. Scientists found it in the 1980s but had a problem — natural GLP-1 breaks down in your bloodstream in about two minutes. Not useful as a drug.

Novo Nordisk’s fix was elegant. They swapped one amino acid for a synthetic variant the body’s enzymes can’t easily chew through, and they glued on a fatty acid tail that makes the molecule latch onto albumin, a big stable protein that floats around in blood. Think of it as a passenger hitching a ride on a bus the body doesn’t bother stopping. Two minutes becomes one week.

Those same modifications also made semaglutide a nightmare to copy. The molecule has nearly 600 atoms that all have to be in exactly the right three-dimensional configuration. Compare that with aspirin, which has 21. One wrong amino acid and it won’t bind to the receptor. One change to the fatty acid tail and the albumin camouflage breaks.

The manufacturing problem

Making semaglutide uses a process called solid-phase peptide synthesis — you build the amino acid chain one link at a time. Each coupling step needs specific toxic solvents. The waste per kilogram of active ingredient can weigh up to 14,000 kg. A normal small-molecule generic produces about 300 kg of waste per kg of drug. Roughly 47 times worse.

The math is brutal even when everything works. Thirty-one coupling steps at 99% efficiency each gets you only 73% of chains finished correctly. State-of-the-art synthesis hits around 57% purity in practice, so nearly half of everything you make has to be cleaned out through expensive purification.

Indian companies did build the capability. Dr. Reddy’s spent years setting up peptide synthesis, API manufacturing, sterile fill lines, and a proprietary injector pen, all at its Vizag facility. Zydus became a manufacturing partner for Lupin and Torrent. But then Canada happened.

The Canada wall

Canada is one of the first Western markets where the semaglutide patent had already lapsed. When Dr. Reddy’s and Sandoz filed to sell their generics there, Health Canada asked for additional data. Neither could provide it. Launch postponed.

“Health Canada describes these as complex synthetic products with possible differences that could impact safety and efficacy.”

For a normal generic, you prove bioequivalence — your version delivers the same drug at the same rate into the bloodstream, and you’re done. For peptides, regulators want to scrutinize the impurity profile and the three-dimensional structure itself. Two vials that pass a standard chemical assay can behave differently in a human body. The Indian regulator accepted what Health Canada didn’t.

The number that reframes everything

India’s entire GLP-1 market was worth about Rs 110 million in 2024. Novo Nordisk’s global GLP-1 revenue was over $40 billion. India is roughly 0.3% of the pie. The US patent doesn’t expire until 2032. Six more years of the market that actually funds the company.

The next floor up

The more interesting twist — while Indian firms were building peptide capacity for semaglutide, Eli Lilly was already eating its lunch. Mounjaro is based on tirzepatide, which hits two hormone receptors instead of one. It beat semaglutide head-to-head on weight loss in late 2024. By October 2025, five months before the Indian semaglutide patent even expired, Mounjaro had overtaken Wegovy as India’s top-selling GLP-1 by value. Lilly has roughly a decade of patent protection left on tirzepatide.

Novo and Lilly are both advancing oral GLP-1 pills through late-stage trials. Novo is also developing a next-gen molecule that combines semaglutide with a second hormone. Each new drug comes with its own decade of protection.

“While generics approach the bottom, the innovators seem to have moved to the next floor.”

The second story — did India’s 2002 patent reform actually help exports?

A 2025 paper from IIFT’s WTO Studies centre (Kachru, Ashok, Banerjee) uses the 2002 Patents Amendment Act as a natural experiment across ~2,500 Indian manufacturing firms. The finding: high-tech firms (those who spent above the industry median on R&D and licensing in 1996–2001) exported about 17% more than low-tech firms after the reform, and tripled their technology-adoption spend. Low-tech firms, which are 62% of the sample, saw modest gains.

The headline mechanism isn’t just that patents protect innovations abroad — it’s that stronger IP enforcement makes foreign partners willing to share sophisticated inputs and license real technology. India got pulled deeper into global value chains, but only where capability already existed. Stronger patents gave low-tech firms “more secure ownership of very little.” A useful lesson for the current PLI and ANRF innovation push — reform amplifies capacity, it doesn’t create it.

Key Takeaways

Novo Nordisk’s Indian semaglutide patent expired March 20th; monthly prices fell from Rs 8,800–11,175 to as low as Rs 1,290 (up to 90% drop)
India is only ~0.3% of global GLP-1 revenue; the US patent doesn’t expire until 2032
Semaglutide is a peptide (~600 atoms) made by solid-phase peptide synthesis — 31 coupling steps, ~57% state-of-the-art purity, 14,000 kg of waste per kg of API
The chemistry trick that works: swap one amino acid for an enzyme-resistant variant, add a fatty acid tail that binds to albumin so the molecule hitches a ride and survives a week instead of two minutes
Health Canada rejected Dr. Reddy’s and Sandoz generic filings on insufficient data — peptide bioequivalence has a higher regulatory bar than small-molecule generics
Eli Lilly’s tirzepatide (Mounjaro) already overtook Wegovy in India by value in October 2025, five months before the semaglutide patent even expired
Tirzepatide has ~10 years of patent left; oral GLP-1 pills and next-gen Novo molecules are in late-stage trials
Theoretical floor for semaglutide manufacturing: ~$28 per patient per year (vs $10,000+ annual US cost)
2002 Indian patent reform: high-tech firms (above-median pre-reform R&D) exported 17% more post-reform and tripled tech-adoption spend; 62% of firms (low-tech) saw modest gains — reform amplified capacity, didn’t create it
Capital goods imports actually fell across the full sample after 2002 — stronger patents also gave foreign suppliers bargaining power

Claude’s Take

The best thing about this episode is that it refuses the easy narrative. “India breaks GLP-1 monopoly” would get clicks. What actually happened is that India got a tiny slice of a market the innovators are already migrating out of, and even that slice comes with a Canada-sized question mark about whether these generics can be exported to regulated Western markets when they finally open. The 0.3% figure is the one to hold onto — India’s whole domestic GLP-1 market is a rounding error for Novo Nordisk, and the US, which pays actual money, is protected until 2032.

The peptide chemistry section is genuinely well done. Zerodha’s team resisted the temptation to hand-wave over the manufacturing difficulty, and the comparison to aspirin (21 atoms vs 600) is exactly the kind of concrete anchor that makes the point stick. The 14,000 kg of waste per kg of drug is the number I’d quote at a dinner party.

The second story gets less airtime but is arguably the more durable insight — policy reform amplifies existing capability rather than creating it. That’s a useful prior for thinking about any “India will catch up via X policy” story, whether the X is semiconductors, batteries, or defence. The firms that were already spending on R&D before 2002 captured almost all the gains. Everyone else got a nicer legal environment and not much else.

Docked two points from a 10 because the narration quality is rough — repeated words, garbled names (Ozempic becomes “Ompic,” albumin becomes “albamin” and “ibupin,” tirzepatide becomes “tzapatide”), and the Razorpay tidbit has a typo that turns 1,029 crore into 1,29 cr. The substance is very good; the editing isn’t.