China's GLP-1 Export Machine: How the Obesity Pipeline's Biggest Deals Now Start in China

When Kailera Therapeutics reported positive Phase 3 data on July 7, 2026 for an oral obesity drug, the molecule was not American. HRS-7535 was discovered by Jiangsu Hengrui Pharmaceuticals in China and licensed to Kailera, a Bain Capital-backed US company, for every market outside Greater China. That arrangement is now the template for the obesity race, not the exception.

Over the past two years, the largest names in Western pharma have gone shopping in China for GLP-1 and next-generation obesity drugs. AstraZeneca signed a deal with CSPC worth up to $18.5 billion in January 2026, its largest China-origin obesity pact. Merck licensed an oral GLP-1 from Hansoh, Novo Nordisk bought rights to a triple agonist from United Laboratories, and a startup called Verdiva launched on a portfolio out of Sciwind. Behind the individual deals is a structural change: by one Jefferies count, China-originated assets made up 32% of global out-licensing deal value in the first quarter of 2025, up from 8% in 2021. Chinese biotechs did not invent the GLP-1. They learned to develop fast-follower versions faster and cheaper than anyone else, and the obesity gold rush turned that speed into the industry's busiest deal pipeline.

The deal that set the pattern closed in May 2024. Jiangsu Hengrui, one of China's largest drugmakers, licensed its entire GLP-1 portfolio for all markets outside Greater China to a Delaware startup later named Kailera Therapeutics. The terms show the shape of these arrangements: about $110 million paid up front, up to $200 million in development milestones, as much as $5.725 billion in sales milestones, low-single to low-double-digit royalties, and a 19.9% equity stake kept by Hengrui. Kailera launched with a $400 million Series A led by Bain Capital Life Sciences.

The portfolio has three named assets: ribupatide (an injectable GLP-1/GIP dual agonist), HRS-7535 (a once-daily oral small-molecule GLP-1), and a GLP-1/GIP/glucagon triple agonist. Kailera went public on Nasdaq in April 2026 at $16 a share, raising about $625 million, or $718.8 million once underwriters exercised their over-allotment. The July 7, 2026 readout covered the oral drug, HRS-7535: in the 556-patient HARBOR-1 obesity trial run in China, the 180 mg dose produced 10.9% weight loss at week 44 against 2.5% for placebo, and a companion diabetes trial, OUTSTAND-2, beat dapagliflozin on blood-sugar control. The injectable ribupatide had already posted up to 17.7% weight loss in a Chinese Phase 3 the previous July, and Kailera is now running that asset through a global Phase 3 program, KaiNETIC, with data expected in 2028.

Hengrui was only the opening move. The roster of China-origin obesity assets now sitting in Western pipelines has filled out fast, and the dollar figures have climbed.

AstraZeneca has moved twice. In November 2023 it licensed the oral GLP-1 ECC5004 (now AZD5004) from Shanghai's Eccogene for $185 million up front and up to about $2 billion in total; the drug is now in Phase 2b. Then in January 2026 it signed a far larger pact with CSPC for a portfolio of long-acting injectable weight-management programs, worth up to $18.5 billion with $1.2 billion up front, the biggest China-origin obesity deal to date. Merck licensed the preclinical oral GLP-1 HS-10535 from Hansoh in December 2024 ($112 million up front, up to about $2 billion). Novo Nordisk paid $200 million up front, up to about $2 billion in all, for the triple agonist UBT251 from United Laboratories in March 2025, after the drug showed nearly 20% weight loss in a Chinese Phase 2. And the UK-US startup Verdiva Bio launched in January 2025 on a portfolio of oral GLP-1 and amylin drugs licensed from Sciwind, backed by a $411 million Series A. Sciwind's own injectable, ecnoglutide, won Chinese approval for weight management in March 2026.

The pull is speed and cost. A Phase 1 trial runs in roughly seven months in China against seventeen or more in the United States, at a cost 32 to 52% lower, with recruitment several times faster. China's annual count of investigational-drug applications for innovative medicines rose from 688 in 2019 to 2,298 in 2023. For a drug class like GLP-1, where the biology is validated and the race is about producing a competitive molecule quickly, that clock advantage is decisive.

Obesity is where the trend concentrates. China accounted for about 27% of the world's GLP-1 clinical trials in 2025, second only to the United States, and held dozens of late-stage obesity assets. The money followed. Beyond the Jefferies figure of 32% of out-licensing deal value in early 2025, the analyst Tim Opler of Stifel estimated that large drugmakers now source roughly a third of their in-licensed molecules from China, up from 10 to 12% at the start of the decade. One industry tally counted $137.7 billion across 186 cross-border out-licensing deals originating in China in 2025, close to ten times the 2021 total. Takeda's research head put the appeal plainly: in China, he said, you can move fast and pay a fraction of US costs.

The direction matters, and it is easy to get wrong. The dominant flow is now Chinese discovery moving out to Western developers: Hengrui to Kailera, Eccogene and CSPC to AstraZeneca, Hansoh to Merck, United Laboratories to Novo. But two prominent obesity drugs run the other way and are worth separating out, because they are often lumped in.

Mazdutide, the GLP-1/glucagon dual agonist that Innovent won Chinese approval for in 2025, was licensed into China from Eli Lilly, a US molecule developed for the Chinese market rather than a Chinese one heading west. And in February 2026 Pfizer took the China-only commercial rights to Sciwind's ecnoglutide for up to $495 million, a Western company licensing a Chinese drug for use inside China. Both cut against the export thesis and are better read as evidence that the traffic now runs in both directions. The net balance, though, has tipped hard toward China as a source rather than only a market.

The bottleneck comes after the deal is signed: whether a drug tested largely in China can win approval in the United States. The FDA has grown wary of applications built on China-only efficacy data, pushing sponsors toward multi-regional trials, and this is precisely why Kailera is running its injectable ribupatide through the global KaiNETIC program rather than leaning on Hengrui's Chinese results. A drug that only ran in one country now carries regulatory risk that a global trial removes.

The politics add pressure. A House Appropriations Committee report in 2025 directed the FDA to stop accepting clinical data generated at sites in China and a handful of other countries in new drug applications, citing the agency's limited inspection reach; the language is not binding, but it signals intent. The BIOSECURE Act, signed into law in December 2025, is often invoked here, but it aims at Chinese contract manufacturers and genomics firms and federal contracting, not at drug licensing, and its prohibitions phase in only after later rulemaking. Licensing a Chinese molecule is legal and common. Getting its trial data across the FDA's bar is the real work, and it is why the smartest deals now come with a global development plan attached.

More shots on goal is good for anyone waiting on obesity treatment. A pipeline that runs through Lilly and Novo alone is a near-duopoly with the pricing power that implies. A dozen additional GLP-1 and multi-agonist programs, many of them oral and several aimed at cheaper manufacturing, widen the field that will eventually compete on price and access. Over time, that competition is the main lever likely to bring monthly costs down.

The timeline tempers the optimism. Most China-origin assets are still in early or mid-stage testing, and the FDA's caution on foreign-only data means the ones that reach US patients will have to run the same multi-year global trials as everyone else. Realistically, the bulk of this wave lands after 2028. There is also concentration risk worth naming: if a large share of the next obesity generation depends on molecules and, in some cases, manufacturing tied to China, trade friction and shifting data-acceptance rules turn into supply questions. The export machine is real, and its payoff for patients lands later this decade rather than next year.

The July 7 obesity-pill data from Kailera and Hengrui is a clean illustration of how the obesity race now works: a molecule discovered in China, developed by a Western company, tested on two continents, and headed toward a US filing. It is one of many. AstraZeneca, Merck, and Novo Nordisk have all placed similar bets, and the money behind them reflects a genuine shift, with China supplying roughly a third of global drug-licensing value in early 2025 and more than a quarter of the world's GLP-1 trials.

The reasons are structural and not going away: faster, cheaper trials and a deep bench of fast-follower obesity drugs. The constraints are just as real: most assets are years from approval, the FDA is tightening its stance on China-only data, and geopolitics could turn a licensing story into a supply story. For patients, the payoff is a more crowded field that should eventually pressure prices. For the industry, the obesity pipeline's center of gravity has already moved, and the deals signed in 2025 and 2026 will decide who owns the back half of the decade.