01. Overview of Cross-Border BD in Innovative Drugs in the First Half of 2026 and Representative Transactions

According to data from the National Medical Products Administration, as republished on the official website of the National Healthcare Security Administration, as of June 30, 2026, China's innovative drugs had been out-licensed in a total of 81 deals, with an aggregate transaction value of approximately USD 110 billion, reaching 80% of the full-year total for 2025; the transactions covered ten therapeutic areas including oncology, metabolism, immunology and neurology, and the main licensees came from 20 countries and regions.[1] The aforementioned "aggregate transaction value" is the sum of the maximum potential quantifiable amounts disclosed for each transaction, usually including upfront payments, near-term payments, R&D and regulatory milestones, sales milestones and other contingent consideration; it does not include potential sales royalties, nor does it equal the certain value achievable at the time of signing.

Based on publicly disclosed transactions, cross-border BD in China's innovative drugs in the first half of 2026 mainly presented two changes in transaction structure, which further affect the project boundaries, decision-making procedures, governance mechanisms and consideration arrangements in the contracts:

First, the transaction subject matter has extended from a single candidate drug or R&D project determined at signing to multiple projects, technology platforms and follow-on projects nominated or formed during the cooperation period in accordance with the agreement;

Second, the cooperation model has extended from a one-way license in which one party grants and the other takes up, to cooperation divided by capability, such as target nomination, joint discovery, early-stage R&D, co-development and co-commercialization;

Under the above structures, some projects establish exclusive or preferential cooperation relationships at an early R&D stage, and subsequent rights are determined in stages through nodes such as project nomination, candidate confirmation, data review or exercise of options; accordingly, contracts need to allocate governance mechanisms, research funding, near-term payments, nomination fees, option fees and project-level milestones by project and R&D stage, so that the parties' responsibilities, acquisition of rights and payment obligations correspond to the project advancement process.

The following representative transactions reflect the above changes and their different combinations. On this basis, this article further discusses: the changes in transaction subject matter and cooperation models, and the specific legal responses in contract content.

Note: The contents and amounts in the table are presented according to the public disclosure standards of the transaction parties; undisclosed matters do not mean that the transaction documents do not provide for them.

02. Two Changes in Transaction Structure

(1) Transaction Subject Matter Extends from a Single Asset to Multiple Projects and Continuous R&D

Several transactions in this period simultaneously cover candidate drugs or R&D projects already determined at signing, as well as follow-on projects and continuous R&D arrangements nominated by one party, with targets designated by the licensee, or jointly formed by both parties during the cooperation period. For example, the CSPC/AstraZeneca transaction covers both existing preclinical projects and subsequent platform R&D; the Haisco/Lilly transaction centers on drug discovery and early-stage R&D after the licensee designates targets. The transaction subject matter is not limited to patents, compounds and data already formed at signing, but may also include research activities conducted in accordance with the agreement and the candidates and project results subsequently formed therefrom.

For the licensee, multi-project and platform cooperation can provide early access to multiple innovative projects not yet fully formed, in addition to competitive single clinical assets, and allow resources to be invested gradually according to project milestones. What the licensee obtains includes not only specific candidate drugs, but also external discovery capabilities, early-stage R&D efficiency and opportunities to continuously replenish the pipeline. Having the Chinese licensor complete candidate screening and early validation can leverage the Chinese party's R&D speed and technical advantages, and also helps match the licensee's resource investment with the maturity of scientific data.

For the licensor, continuous R&D cooperation can provide funding and external validation for early-stage research, and connect its own advantages in drug discovery, translational research or early-stage clinical work with the licensee's global clinical development, regulatory, manufacturing and commercialization resources. Allocating rights by target or project can also preserve for the licensor non-cooperation fields, other indications and space for subsequent cooperation. Considering that the same platform may apply to different projects, target selection, the scope of exclusivity and ownership of platform improvements will also enter transaction negotiations earlier.

(2) Cooperation Models Extend from One-Way Licensing to Early-Stage R&D, Co-Development and Co-Commercialization Divided by Capability

The change in cooperation models is reflected not only in the increased number of projects, but also in the parties undertaking work at different stages according to their respective capabilities. In some transactions, the licensee nominates targets, the Chinese company is responsible for drug discovery and early-stage R&D, and subsequent licensing occurs after candidates meet agreed standards; the Hengrui/BMS and Innovent/Pfizer transactions, within the same cooperation framework, allocate exclusive licensing, co-development and co-commercialization rights respectively according to project source, territory, development responsibility and degree of market participation. The parties may assume different roles under different projects, R&D stages and territories.

For the licensee, territorial splitting helps it obtain market rights that can be effectively advanced, and completing Phase I studies or proof of concept in China can leverage the licensor's advantages in drug discovery, early translation, China clinical execution or local regulation. Co-development arrangements, meanwhile, allow the licensee to continuously obtain scientific support from the original R&D team without acquiring the entire asset or platform.

For the licensor, retaining Greater China rights can preserve local product value; co-development and co-commercialization may also provide profit potential higher than mere sales royalties, and build global clinical, regulatory and commercialization experience. Correspondingly, deeper participation also brings risks such as development costs, budget fluctuations, decision-making deadlocks and commercialization losses. Such arrangements reflect the complementarity and integration of the transaction parties around discovery, development, regulatory, manufacturing and commercialization capabilities.

03. Impact of Changes in Transaction Structure on Contract Architecture

(1) Scope of Cooperation and Project Boundaries Need to Be Dynamically Clarified

When cooperation simultaneously covers relatively mature assets, follow-on projects and continuous R&D activities, research cooperation and product licensing usually coexist: the licensor first conducts work within the agreed targets or research scope, and specific candidates or projects are formed at subsequent nodes and brought into the license. The contract therefore cannot merely list the licensed assets at signing, but also needs to separately define the background technology usable during the research stage, the scope of formation of projects and results during the cooperation period, and the boundaries of subsequent product development and commercialization licenses.

First, the agreement should clarify the scope and capacity of cooperation projects. Exclusivity should be delineated along dimensions such as project, target, mechanism of action, indication, technology modality and territory, and should state whether pre-signing existing projects, third-party cooperation and fields retained by the licensor are excluded. For continuous R&D cooperation, it should also clarify the cap on concurrent and cumulative projects, the release of slots after a project stops, and whether projects not advanced continue to occupy the licensor's platform resources.

Second, the research plan and scope of work should not be merely a principled annex. Work packages, candidate standards, deliverables, research records, sample and data formats, timelines, personnel input, budgets and third-party costs should be specific and executable, and sufficient to confirm the parties' responsibility boundaries and cost bearing; plan changes, additional work and acceleration requirements should also be subject to clear approval and cost mechanisms. If the licensee bears research funding, the payment schedule, budget-versus-actual differences and treatment of costs already incurred after a project stops should also be agreed. A commercially reasonable efforts obligation can serve as part of the performance standard, but still needs to be based on a clear research plan and division of responsibilities, so as to distinguish normal scientific research risk from contract performance issues.

Third, platform technology and project results need to be separately defined as to ownership. Platforms, general tools, algorithm models and improvements not specific to cooperation projects existing before signing should in principle remain with the licensor; compounds, sequences, uses, data and inventions formed around specific targets or candidates need to have ownership and licensing determined according to the content of the results, the inventors and the parties' arrangements for project rights. The agreement should also clarify what results constitute platform improvements, what results constitute project results, and the extent to which project results may use platform intellectual property.

Fourth, representations and warranties, indemnities and limitations of liability need to match the formation stage of the project. For projects already determined at signing and for which certain R&D materials have been formed, the licensor may make relatively complete representations regarding the source of rights, upstream licenses, R&D compliance, disclosed data and known adverse matters; for follow-on projects nominated or formed during the cooperation period, the agreement may provide that upon project entry, candidate confirmation or exercise of options, project-level representations and disclosures are updated based on facts verifiable at that time. Scientific failure caused by scientific uncertainty should not in itself automatically constitute a breach of representations and warranties (except concealment of material adverse facts known at signing). The scope of indemnity and limitations of liability should be differentiated according to the nature of the conduct and events, specifically addressing fraud, willful misconduct, confidentiality obligations, intellectual property ownership and data compliance, and should isolate consequences by project as far as possible, avoiding a representation issue in a single project automatically extending to the entire cooperation.

(2) Project Entry, Acquisition of Rights and Project Release Need Staged Arrangements

Once the scope of cooperation extends to future projects, subsequent rights usually cannot be determined all at once at signing. For projects for which no candidate has yet been formed or which have not yet been brought into the formal license at signing, the parties often set project nomination, candidate confirmation, data package delivery and option exercise at different R&D nodes, so that scientific validation, funding investment and acquisition of rights occur in stages. Project entry standards, review periods, exercise procedures and the release of rights when a project does not enter the next stage therefore become important components of the contract.

First, the exercise mechanism should be clarified around each subsequent licensing or expansion right. For projects, candidates, indications or territories that one party decides during R&D whether to include, the agreement needs to clarify the data package for review, the review period, supplementary inquiry arrangements, exercise notices and related payments, and clarify when the rights take effect. In multi-project cooperation, the above arrangements should also be managed separately by project, avoiding the exercise, lapse or termination of one project affecting other projects.

Second, data evaluation rights and research use rights should be defined separately. During the evaluation period, the party obtaining information may need to conduct internal research, validation and regulatory analysis, but its scope of use of samples and data should still match the evaluation purpose. The agreement may further clarify accessible personnel, third-party laboratories, model training, ownership of results, and destruction or retention requirements after a project does not enter the next stage, avoiding evaluation activities extending to external development, commercialization or platform information beyond the necessary scope.

Third, the agreement should provide in advance for a release mechanism for the event that the project does not enter the next stage, including termination of exclusivity, return or licensing of research results, transfer of available data and regulatory communications, the right to disclose to subsequent potential collaborators, and non-challenge or non-obstruction obligations that may apply for a certain period after cessation. In addition, project release involves not only the return of rights but also the need to prevent cross-contamination of R&D data and dilution of platform value. The agreement should expressly provide for data isolation obligations after project termination; for resources that may affect other collaborative projects, a reasonable technical isolation period should be established, during which the terminated project's data may not be used for new project R&D without written consent. At the same time, a third-party collaboration restriction period should be established, restricting the licensor from entering into collaboration with competitors on the same target or indication without fully compensating the licensee for its prior investment, thereby preventing the licensee's prior sunk costs from being improperly diminished by the licensor's separate licensing.

(III) Joint development and commercialization require a governance mechanism that matches the division of capabilities

When the parties separately undertake development and commercialization work across different projects, R&D stages and territories, the contract must not only delineate the scope of work but also determine who has decision-making authority over which matters, who bears the corresponding costs, and how one party's activities interface with the other party's retained rights and regulatory arrangements. Therefore, the governance mechanism for joint development and commercialization should match the parties' division of capabilities, contributions and risk-bearing.

First, governance rights should be allocated according to the nature of the matter, and the authority of the joint committee should be clearly defined. The global development plan, the sequence of indications, key trial designs, annual budgets, CMC changes, regulatory communications, product positioning and pricing do not have the same impact on both parties. The agreement should distinguish item by item between matters requiring mutual consent, matters subject to one party's final decision, and matters of consultation and notification; the final decision right should also take into account factors such as territory, budget, patient safety, regulatory compliance and adverse effects on the other party's retained rights. Matters that the joint committee cannot resolve should be subject to an escalation mechanism through the scientific lead, project lead and senior management in sequence, and the agreement should specify which activities continue or are suspended during a deadlock.

Second, cost-bearing in joint development should match decision-making rights and participation rights. Joint development involves not only trial costs; unplanned studies, budget overruns, additional trials required by regulatory authorities, supply failures and accelerated development proposed by one party may all change the scope of costs. The agreement needs to specify the annual cost budget, the overrun approval mechanism, the economic arrangements after withdrawal from joint development, and the impact of costs already borne on subsequent profit sharing.

Third, partitioned development should ensure continuous sharing of global data and regulatory materials. The parties need to share clinical, safety, pharmacovigilance and CMC information, and to put in place rights to reference regulatory materials, rights to participate in global trials, and coordination mechanisms for communications with regulatory authorities. The agreement should also address indication priorities, medication safety data, samples and analytical methods, and the spillover effects of one party's regional activities on the other party's pricing, reimbursement or compliance obligations. The agreement should ensure that the licensor can promptly obtain data generated from overseas development that is sufficient to support Chinese registration and commercialization. In addition, given China's regulation of human genetic resources and important data, the agreement must establish a compliant mechanism for cross-border data flow and clearly allocate responsibility.

Fourth, joint commercialization arrangements should center on whether the parties jointly carry out promotion, sales or medical affairs and on the corresponding governance mechanism. The agreement needs to specify who holds the marketing authorization, books sales, leads pricing and market access, which commercialization activities each party undertakes, and how, through the commercialization plan, budget, decision-making rights, information rights and coordination mechanisms, each party's contributions can actually be implemented.

(IV) Consideration needs to correspond item by item to project formation, acquisition of rights and R&D results

Changes in project structure have also further layered the consideration arrangements. Beyond the basic framework of upfront payments, development and sales milestones and sales royalties, platform and early-stage R&D collaborations may also provide for research funding, near-term payments, project nomination fees, option fees and project-level milestones. The different payments are not merely different in name: research funding corresponds to agreed work, nomination fees or option fees correspond to project entry and acquisition of rights, and project-level milestones correspond to subsequent R&D results; different projects within the same collaboration may also separately adopt sales royalties or profit sharing.

First, multi-project transactions should establish relatively independent payment and settlement arrangements on a project-by-project basis. The agreement needs to specify which project, candidate, product, indication and territory the nomination fees, option fees, development milestones and sales milestones respectively apply to, and explain how consideration already paid and not yet triggered is to be handled when a project is stopped, replaced, expanded or terminated. If different projects separately adopt profit sharing or sales royalties, it should also be avoided that adjustment or termination of one project automatically affects other projects.

Second, milestone events should adopt objective and determinable standards. For example, clinical development milestones should specify whether the triggering event is first subject enrollment in a specific clinical trial, first dosing, or completion of enrollment of the agreed number of subjects; early-stage development milestones may consider setting R&D data as the specific determination standard; and regulatory and commercialization milestones should specify whether marketing approval, conditional approval and pricing or reimbursement decisions respectively trigger payment. For new indications, combination regimens or new dosage forms, it should also be explained whether they constitute independent milestones. The agreement should also provide for execution mechanisms such as payment notices, supporting documents, objection periods and exchange rate conventions.

Third, the starting point of sales royalty clauses is to clarify which products constitute royalty-bearing products, and whether the payment obligation is based on patents, know-how or products arising from the collaboration. The agreement should also provide whether circumstances such as expiration of patent protection, generic competition and third-party patent fees lead to a reduction in sales royalties. In platform or target discovery collaborations, the definition of "licensed product" or "collaboration product" should be clearly set out in the agreement, establishing a clear link between the product and the use of the platform, the selected target or the collaboration research results, and corresponding reporting and audit mechanisms should be put in place.

Fourth, sales royalties and profit sharing should adopt different settlement rules according to whether costs and losses are shared. Sales royalties are usually based on net sales, and the licensor does not directly share the licensee's R&D and commercialization costs; profit sharing requires first defining distributable revenue, deductible development, manufacturing and commercialization costs, budget overruns and loss bearing, and then settling according to the agreed ratio. If both parties participate in promotion or sales, the economic arrangements should also match the commercialization division of labor, budget responsibilities, information rights and audit rights.

Conclusion

In the first half of 2026, technology and product licensing still constituted the foundation of cross-border BD for China's innovative drugs. Compared with single-asset licensing, some representative transactions in this period, on the one hand, brought relatively mature assets, subsequent projects and continued R&D into the same scope of collaboration, and on the other hand, allocated work and rights according to the parties' capabilities in drug discovery, early-stage R&D, global clinical development, registration, manufacturing and commercialization. Changes in transaction structure have made the roles, contributions and revenue arrangements of the parties across different projects, R&D stages and territories more diverse.

When the collaboration content includes future projects that have not yet been formed, the contract cannot determine the licensing scope and total consideration only once at signing; it also needs to continuously answer which projects enter, which party advances them, how decisions are made, when rights are acquired and consideration is paid, and how to handle matters when a project does not continue to advance. For the licensor, the boundaries of the platform and projects, the project entry and release mechanisms, the availability of data and IP, single-project economics, participation in decision-making and whether regulatory materials can be clearly implemented remain the foundation for whether transaction value can be continuously realized.

Note: The first draft of this article was completed by Zhou Yingchun; intern Ye Xinmeng (law graduate student at Sichuan University) provided legal research support.

Article Appendix

Questions and answers

What were the aggregate value and number of China's innovative drug out-licensing deals in the first half of 2026?

As of June 30, 2026, China's innovative drugs had been out-licensed in 81 deals with an aggregate transaction value of approximately USD 110 billion, reaching 80% of the full-year total for 2025. The deals covered ten therapeutic areas and licensees from 20 countries and regions.

How did the transaction subject matter in cross-border BD deals change in the first half of 2026?

The subject matter extended from a single candidate drug or R&D project determined at signing to multiple projects, technology platforms and follow-on projects nominated or formed during the cooperation period. Transactions may cover existing preclinical projects and subsequent platform R&D, or drug discovery and early-stage R&D after target designation.

What contract mechanisms are needed when cooperation covers future projects not yet formed at signing?

Contracts need staged arrangements for project nomination, candidate confirmation, data package delivery and option exercise, with separate management by project. They should also provide for data evaluation rights, release mechanisms if a project does not advance, data isolation obligations, and third-party collaboration restrictions to protect prior investment.

How should consideration be structured in multi-project innovative drug collaborations?

Consideration should correspond item by item to project formation, acquisition of rights and R&D results. Research funding, nomination fees, option fees and project-level milestones should be allocated by project, with objective milestone standards and independent payment arrangements to avoid one project's adjustment affecting others.

What governance issues arise in joint development and commercialization of innovative drugs?

Governance rights should be allocated by the nature of the matter, with joint committee authority clearly defined and escalation mechanisms for deadlocks. Cost-bearing should match decision-making rights, global data and regulatory materials must be shared, and joint commercialization requires clarity on marketing authorization, sales booking, pricing and coordination mechanisms.

Laws and cases cited

Other

  • National Medical Products Administration
  • National Healthcare Security Administration

References

  1. National Healthcare Security Administration, citing the National Medical Products Administration: "China's Innovative Drug Outbound Licensing Transaction Value Exceeded US$100 Billion in the First Half of the Year".
  2. CSPC Pharmaceutical Group Hong Kong Stock Exchange announcement (English); CSPC Pharmaceutical Group Hong Kong Stock Exchange announcement (Chinese).
  3. Hengrui Pharmaceuticals Hong Kong Stock Exchange announcement; BMS official announcement.
  4. Innovent Biologics Hong Kong Stock Exchange announcement: Global strategic collaboration with Pfizer; Pfizer official announcement.
  5. RemeGen Hong Kong Stock Exchange announcement; AbbVie official announcement.
  6. Haisco announcement.