Global biopharma licensing reached $252 billion in 2025, up from $190.6 billion the year before. Of that, $137.7 billion came from Chinese sellers, a near-tenfold jump from 2021. Korea added another $7.86 billion, a 113% year-on-year rise. Cross-border China outlicensing deal count went from 42 in 2022 to 93 in 2025; average upfront from $52 million to $172 million.
This is not the same business as 2023. The 2026 playbook for moving an APAC-origin asset into an American partnership is no longer about sourcing the molecule. It is about surviving an 8-15-month process where the intellectual-property regime mismatches, where clinical-data transfer rules block what your transaction lawyer assumes is routine, where the valuation gap narrows month by month, and where federal policy may catch the deal in flight. Most American business development teams have not noticed.
Here is what the playbook now looks like.
What is being bought
Oncology dominates. Antibody-drug conjugates and bispecific antibodies together account for half of China-origin licensing volume, and most of the blockbuster deals. Next-generation antibodies alone represented half of cross-border deal volume in 2025. PD-1×VEGF bispecifics in particular drew Pfizer ($1.25 billion upfront for 3SBio's SSGJ-707), AbbVie ($650 million upfront for RemeGen's RC148), and Merck. Metabolic and obesity assets, led by oral GLP-1 candidates, drove the next tier: Merck/Hansoh at $2 billion potential, AstraZeneca/CSPC at $18.5 billion.
Korea is more specialized. Antibody-drug conjugate licensing from Korean biotechs grew 39% from 2021 to 2024, reaching $1.4 billion. Neurodegeneration is a Korean strength, anchored by ABL Bio's $2.78 billion platform deal with GSK.
Nearly three-quarters of China-US cross-border deals in 2025 involved drugs in preclinical or Phase 1 testing. Globally, 61% of all 2025 licensing transactions by count were discovery or preclinical. American buyers are committing real money for assets years away from a Phase 3 readout. The process has to look different from when most deals were Phase 2.
Who is buying
The active list is shorter than the trade press suggests. AstraZeneca is the most prolific, with CSPC twice (June 2025, January 2026) and Harbour BioMed once at $4.68 billion, alongside a $15 billion China investment pledge through 2030. Merck has done at least three China deals. Pfizer paid the largest single-asset upfront of 2025. Bristol Myers Squibb closed what may be the largest bilateral structure on record in May 2026: 13 programs with Hengrui worth up to $15.2 billion, including a bidirectional swap of oncology assets for immunology rights in China. GSK, AbbVie, Gilead, Eli Lilly, Sanofi, Novartis, Roche, and UCB round out the buyer set.
Korea draws a different mix. GSK is the largest Korean dealmaker. Eli Lilly has been active in MASH (OliX, $630 million) and hearing loss (Rznomics, $1.3 billion). AstraZeneca took rights to Alteogen's subcutaneous formulation platform.
What you do not see, almost ever, is a banker-run auction. Most APAC-to-US licensing in this period has been bilateral, conference-initiated, often unfolding from a single JPM or BIO meeting. The volume of transactions, more than 157 China deals in 2025 alone, makes formal sell-side processes the exception. The implication for an American business development head is uncomfortable: by the time your competitor announces a deal, the seller never spoke to you.
The diligence layer that did not exist three years ago
This is where transactions die. There are six diligence questions that did not exist in 2023 and now appear on every term sheet.
First, who owns the licensed intellectual property. Chinese patent law contains statutory pre-emption rights, joint-ownership defaults for certain inventions, and employee-inventor compensation obligations unfamiliar to American counsel.
Second, can you move the clinical data. China's human genetic resource regulations require NMPA review and approval before patient-level data generated in Chinese trials can leave the country. If the seller's Phase 1 was a China-only study, your FDA bridging strategy starts late.
Third, where will the manufacturing be supervised. After the BIOSECURE Act, mapping whether a counterparty's CDMO subcontractor will appear on the next Section 1260H list is standard deal hygiene, even though the Act does not name specific companies.
Fourth, whose law governs and where do disputes go. Chinese sellers increasingly insist on ICC arbitration seated in Hong Kong or Singapore. American buyers prefer New York law. This is the most common letter-of-intent time sink.
Fifth, can the upfront be clawed back. Chinese biotechs treat upfront non-refundability as non-negotiable. American buyers usually want claw-back tied to warranty representations.
Sixth, how is the payment architecture structured. BMS/Hengrui's May 2026 deal pioneered a structure now appearing more often: a $600 million headline upfront, $175 million at the first anniversary, and $175 million contingent at the second. The deferral lets the seller post a larger headline number while the buyer holds payment optionality against early diligence findings. Expect more of this.
Each of these six questions can stall a process by months. None of them appears on any banker's standard checklist.
Process benchmarks, and what the public record will not tell you
Industry advisors put the end-to-end timeline at 8-15 months from first contact to a signed definitive agreement. Vision Lifesciences cites 6-12 months from first contact to definitive. Locust Walk's framework is 1-3 months of preparation plus 6-12 months of execution. Innovent's licensing of IBI363 and IBI343 to Takeda, announced October 21, 2025 and closed December 4, 2025, fits the pattern: a six-week sign-to-close, implying 6-12 months of pre-announcement negotiation. The May 2026 BMS/Hengrui deal, with expected Q3 2026 close, is consistent.
Upfront payments represent 6-8% of total disclosed deal value across 2025 and Q1 2026; the rest is deferred into milestones.
What you want to know is not in the public record: what percentage of processes that reach term sheet ever close, or how many bidders the average APAC seller attracts. There is a structural reason the numbers are missing. HKEx-listed Chinese sellers (Innovent, Hengrui, CSPC, 3SBio, RemeGen) do not file SEC-style "background of the merger" disclosures. The negotiation chronology is invisible by design. Survivorship bias is total: BioPharma Dive's tracker shows the closed deals, not the dropped processes behind them.
Institutional knowledge is worth what advisors charge for it. The public data does not exist.
The policy overhang
Three policy lines are converging on these deals.
The BIOSECURE Act was signed into law on December 18, 2025, embedded in the FY2026 National Defense Authorization Act. The enacted version does not name specific companies, relying instead on the Department of Defense's Section 1260H list and an interagency Office of Management and Budget process. Implementation prohibitions are not yet in effect; the initial list must be published by December 18, 2026. The Act targets CMO and CRO supply-chain relationships, not the licensing of drug molecules themselves. Deal pace has been unaffected. Collective China deal value soared in the months after passage.
The Biotech Investment National Security Act, introduced June 2, 2026 by Representatives Moolenaar and Dingell, would catch what BIOSECURE does not. It would require Treasury and Department of Defense review of pharmaceutical licensing deals, joint ventures, and equity investments with Chinese entities, including intellectual property and technology transfer. If enacted, it would cover the $136 billion in 2025 cross-border deals and the roughly $60 billion in Q1 2026 alone. No vote is scheduled.
China responded in April 2026 with State Council Decree No. 834, giving Beijing power to investigate and sanction foreign companies whose commercial decisions harm Chinese industrial-chain security. That is the legal infrastructure for retaliation.
Build a 90-day BINSA-review contingency window into closing mechanics. It may not be needed. If it is, you will be glad it is there.
The other side of the argument
The strongest counter-position is not anti-China. It is that the licensing-out model itself underestimates APAC operator ambition. Zai Lab is the example. The Shanghai-and-Cambridge company is advancing zocilurtatug pelitecan, a DLL3 antibody-drug conjugate, in a global Phase 3 program without licensing it out to a US pharma. Revenues grew 22% year-on-year to $106.5 million in Q1 2025; full-year guidance is $560 million. Samantha Du frames Zai as a "dual-engine company" combining commercial scale in China with global innovation. Daiichi Sankyo's three-ADC deal with Merck is the Japanese template: structurally equal co-development, not pure licensing.
The harder counter-position is American skepticism. STAT News carried a May 2026 op-ed arguing that US pharmas are "trading the future of American biotech for short-term profits." The Trump administration's draft executive order leaked in September 2025 caused a 10% equity hit in Zai Lab and BeOne in a single session. The bargain era is also ending: average China upfronts rose from $102 million in 2024 to $141 million in 2025. The arbitrage that drove the last two years of deals is closing.
An operator's playbook
Seven moves for the business development function that wants to source APAC assets without losing the next deal at letter of intent.
One. Embed bilingual counsel with human-genetic-resource and personal-information-protection law experience at the term-sheet stage, not at the diligence stage. Most LOI failures come from bolting this on later.
Two. Pre-clear governing law in the teaser response, not the term sheet. New York law, ICC arbitration seated where you can both live with. Pick one and surface it before the term sheet conversation. The lawyers' month-long argument is happening either way; have it earlier.
Three. Run a shadow CDMO plan before LOI. Assume the seller's primary CDMO will face supply-chain audit, and have a parallel manufacturer mapped, costed, and timeline-ready before you sign.
Four. Define the bridging-data strategy before pre-IND. If the seller's Phase 1 was generated in a China-only population, FDA will want diverse-population bridging data. House FY2027 FDA appropriations language would bar the agency from accepting China-only IND data at all. Solve this before you commit upfront cash.
Five. Map every counterparty (seller, CDMO, CRO, formulation partner) against the current Section 1260H list and the publicly available OMB draft list. Re-run the map at the second anniversary of any deferred-payment structure.
Six. Build a 90-day BINSA-review contingency window into closing mechanics. Wording matters: a mutual termination right tied to a federal review trigger is different from a buyer walk-away right, and Chinese sellers will negotiate hard on the difference.
Seven. Benchmark the process against data the public record does not contain. Letter-of-intent conversion rates, bidder counts, and competing-offer dynamics for China and Korea processes live with a narrow set of advisors. That asymmetry is the value.
The bargain era is closing. The operator era is starting.
PwC's 2026 Outlook describes China-to-the-West licensing as "a core vector in corporate and business development strategies." Pitchbook expects China's early-stage advantage to "likely persist for at least the next several years." The arbitrage that built the last two years of deals is narrowing, but the underlying market is not going away. It is becoming a normal part of the industry's capital structure.
For American buyers, the question stops being whether to participate. It becomes whether your process can survive the new diligence layer.