There is a moment in the life of every blockbuster drug when the science stops being the interesting question and the invoice takes over. GLP-1s reached it this week, in three separate sections of the same newspaper.
Thirteen Million Users, and Who Actually Pays
Roughly 13 million Americans are now taking GLP-1 medications, according to research from JPMorgan published in February. That is the number that makes every other number in this story move. A drug class with 13 million users is no longer a specialty line; it is infrastructure, and infrastructure shows up in somebody’s budget.
This week we learned whose. Medicare drug-coverage premiums are rising because the plans have been hit by growing expenses for GLP-1 medications and other specialty drugs — and because changes implemented under 2022’s Inflation Reduction Act reduced many enrollees’ out-of-pocket costs while requiring insurers to shoulder more of them. Those two pressures compound, and they are landing on the same statement.
The arithmetic, as the administration itself describes next year: around 25% of Part D enrollees will see premiums stay flat or go down; around 30% will see an increase of less than $10 a month; and for the remaining 45%, the increases are largely in the $11 to $20 a month range. Those pressures are expected to continue into 2027, said Juliette Cubanski, a vice president at KFF.
And the cushion is being removed at the same time. The Medicare Payment Advisory Commission, a federal watchdog agency, estimates the stabilization program — the one that was supposed to hold Part D together — reduced the average Part D plan premium by about 40% in 2025, and cut the average rate by an estimated 27% this year. That is the subsidy the administration plans to end. We wrote the policy half of this story separately, from the seniors’ side of the counter; this is the manufacturers’ side.
Put the two together and you get the sentence that matters for a portfolio: the fastest-growing drug class in America is now a named cause of a rising bill for the country’s largest payer. Payers who can name the cause eventually negotiate the price.
Dozens of Companies, and a Founder Willing to Say It
The competitive picture came from an unlikely place: a profile of Regeneron (REGN) co-founder George Yancopoulos, who has spent thirty years building a company on being years ahead of everyone else. On obesity, he was blunt about how crowded the road has become. Nowhere is his strategy clearer than in obesity, the Journal reported, where dozens of companies are racing to launch their own versions of GLP-1 drugs.
He was equally blunt about where the next wave comes from. Chinese biotechs are moving quickly to replicate successful science — described in the piece as a threat to the entire industry, and one that cuts directly at what Regeneron has been able to capitalize on. Yancopoulos, 66, said he hates the copycat model, and put the question plainly: “Do we need 168 GLPs?”
Take that seriously as an investor and one conclusion follows immediately. When an industry founder tells you the science is being replicated by dozens of entrants, the last thing you want to own is the version of the trade whose entire value is the science. Volume growth in a crowded category with a squeezed payer is not a moat. It is a price negotiation with a countdown clock.
The Side Effect That Becomes Its Own Market
The most investable idea in the week was not in the obesity race at all. It was in the thing the obesity race leaves behind. Patients on GLP-1s can lose significant amounts of muscle along with fat — a particular risk for older adults, who need the strength to stay independent.
Regeneron has been developing a pipeline of muscle-preservation drugs that block the body’s natural atrophy systems, and by pairing those therapies with weight-loss medications, Yancopoulos hopes to focus on the quality of weight reduction — so patients get stronger rather than just thinner. The company is also mining its trove of human DNA for the obesity problem directly: Regeneron scientists identified what Yancopoulos calls the “laziness gene,” which dictates whether a person is inherently sedentary, and the company is developing a drug it hopes will switch it off so patients naturally burn more calories.
Read that against the demographics and it stops being a footnote. An aging country, a drug class with 13 million users, and a well-documented consequence that specifically threatens the ability of older adults to stay out of assisted living. That is a durable adjacent problem, and durable adjacent problems are usually cheaper to buy than the headline they attach to.
The Risk That Does Not Appear on the Slide
The Journal also reported the harder story: GLP-1s can sometimes trigger anorexia, inducing disordered eating in some people. The drugs mimic naturally occurring hormones to suppress appetite and cravings — which, for anyone with a history of restrictive eating, is a tantalizing proposition and a dangerous one. Reported cases in the piece include patients who were not overweight to begin with, patients hospitalized after weeks without eating, and off-label use obtained through telehealth.
The severity is the part to hold onto. Anorexia kills a higher percentage of patients than any other eating disorder, and it is the second most deadly mental illness after opioid addiction, according to data compiled by the nonprofit National Association of Anorexia Nervosa and Associated Disorders.
The manufacturers responded to the reporting in the language manufacturers use. A Novo Nordisk (NVO) spokesperson said the company promotes only the FDA-approved indications of its medicines for “appropriate patients” who meet the indicated criteria. An Eli Lilly (LLY) spokesperson said patient safety is the company’s top priority and that it is working closely with regulators on potential safety topics, including any data regarding eating disorders.
No label change has been announced, and none may be. But this is exactly the shape of risk that arrives slowly and then all at once: a reported pattern, a nonprofit’s mortality data, a regulator reading the same newspaper. It belongs in the underwriting even though it does not yet belong in the model.
The Franchise War, in Two Numbers
Here is where our own book stops being an observer. Eli Lilly is up 14.3% in the fundamentals books this year. Novo Nordisk is down 2.2%. Same drug class, same calendar, sixteen points apart.
That spread is not a sector call. It is a franchise war with a winner, and it says the market has already decided which company owns the category and which one is defending share. Lilly carries a pipeline — the Alzheimer’s program among it — that Novo does not have, and 13 million users is a base that does not un-prescribe quickly. Novo is the pure-play, and the competition line from the Regeneron profile lands hardest on the pure-play that is already losing ground.
So the answer to the question a reader asked us this week — do the GLP-1 drugs make sense in the books — is yes, but not the way most people own them. Own the franchise leader, own the adjacency, rent the option, and refuse to pay for the tail.
REINFORCE Eli Lilly (LLY). Bring the tactical-tier weights up toward the Income & Quality treatment where the tier’s risk budget allows. It is winning the franchise on our own numbers, the Alzheimer’s pipeline is a second shot Novo does not have, and 13 million users is a base that does not un-prescribe quickly. Funded from cash and the index sleeve, never from a conviction holding.
HOLD, DO NOT ADD — Novo Nordisk (NVO). It stays for the international diversification job it was hired to do — Denmark, the developed-market sleeve — and because selling a 2.2% loser into a franchise war is how you lock in the wrong half of a pair. It gets no new dollars.
REINFORCE as the theme’s second leg — Regeneron (REGN). This is the actual insight of the week. Regeneron is not a GLP-1 stock; it is a bet that the side effect becomes its own market. Muscle loss on GLP-1s, in a population that needs muscle to stay independent, is a durable adjacent problem, and Regeneron is building the combination drug for it. We already own it and reinforced it on July 17.
NEW — a small starter position in Amgen (AMGN). Already on the July 17 reinforce list, so this is a weight decision rather than a new relationship: an obesity program with a differentiated dosing profile, a fraction of the franchise multiple, and a dividend that makes the wait tolerable. Size it small — this is the option, not the position.
AVOID any pure-play, pre-revenue obesity biotech and any name whose entire case is “the next GLP-1.” With dozens of entrants and Chinese replication flagged by an industry founder, the pre-revenue tail is where the losses live.
The risk we are underwriting, named: the payer is the swing factor, not the science. Part D premiums are rising because of these drugs, 45% of enrollees are about to feel $11 to $20 a month, and the pressure runs into 2027. When the payer revolts, the manufacturer’s price is the first thing on the table — which argues for the leader with a pipeline, the adjacency, and the cheap option, and against paying up for volume a formulary committee can reprice. No mandate exclusion applies to any of these names; client screens that prohibit specific pharma are honored automatically, as always.
- LLY · Eli Lilly · reinforce, franchise leader
- NVO · Novo Nordisk · hold, no adds
- REGN · Regeneron · reinforce, muscle-preservation adjacency
- AMGN · Amgen · new starter, cheap option
- HUM · Humana · the payer side, watch
- UNH · UnitedHealth · named in the Part D subsidy math
Related: The same premium story from the other side · Sean’s letter — Your Health Plan Is a Bond You Didn’t Price · See the model books