Key Takeaways
- • Treasury's proposed COINS outbound-investment rules add biotech to restricted sectors but do not mention pure licensing deals with Chinese biotech companies, a point Treasury confirmed to Endpoints News on a press call.
- • The proposed rules explicitly prohibit joint ventures, defined only by their plain-English meaning (contribution of capital and/or assets by two parties with shared profits and losses), which could capture co-development, co-commercialization and equity-linked structures.
- • The prohibited-technology list names "pharma compounds," enumerated as small molecules, cell and gene therapies, monoclonal antibodies and protein therapeutic recombinants, which Endpoints reporter Max Bayer reads as sweeping in bi-specifics, tri-specifics and in vivo CAR-T deals sourced from China.
- • The NewCo model, in which US venture capital sources Chinese molecules into US startups with equity stakes, is identified as the most exposed structure under the rules.
- • The 45-day comment window is expected to become a lobbying contest over whether the final rule narrows or stays broad, with China expected to escalate rather than negotiate.
Summary
Treasury's proposed COINS (Comprehensive Outbound Investment National Security) rules add biotechnology to the sectors restricted from Chinese investment, yet the text does not address pure licensing, the deal structure that has driven biopharma M&A for roughly two years. According to reporting by Endpoints News reporter Max Bayer, Treasury confirmed the omission on a press call. Bayer argues the omission is thinner protection than it appears, because the proposal explicitly prohibits joint ventures and names "pharma compounds" as a prohibited technology. Together, those clauses could capture much of the dealmaking of recent years.
The joint-venture clause is the main source of legal ambiguity. Treasury declined to define the term and pointed to its plain-English meaning: the contribution of capital and/or assets by two parties, with profits and losses shared. Bayer called this "extremely broad." Endpoints executive editor Drew Armstrong noted that few China deals are pure licensing; co-development pacts, royalty streams and knowledge transfer are routine, and each raises whether equity or asset contribution has effectively created a joint venture. Armstrong cited a Gilead China transaction, announced overnight and carrying a billion-dollar-plus upfront plus milestones, as an example of a narrowly structured deal that appears permissible.
The final text was broader than the sector expected. A week before publication, the prevailing assumption was that biotech would be mentioned but largely unaffected. Bayer said the national-security faction gained the upper hand, and by October 16 he was told pharma was included. Bayer reads the rule's request for public comment on its own breadth as a signal that drafters may have felt constrained by pressure elsewhere in the administration and may narrow the rule after the comment period.
US venture capital running the NewCo model is seen as most exposed. That playbook, which sources a molecule in China, spins it into a US company and sells it to pharma, typically involves equity and often resembles a joint venture. China-focused funds raising from US limited partners face similar issues. Armstrong relayed an argument that clean licensing preserves late-stage access to Chinese innovation while the US loses early-stage participation. Bayer noted the NewCo model may already be losing favor as large pharma builds search-and-evaluation teams in China, citing colleague Kyle's reporting on roughly five or six Roche deals in China since August 2026.
China is expected to respond with escalation rather than negotiation. Per the article, China has discussed restricting exports of certain technologies and there are reports of restrictions on knowledge workers leaving the country. Armstrong, drawing on tariff experience, expects an initial escalatory response, and Bayer said any retaliation would feed into the final rule language and the intensity of lobbying.
On Capitol Hill, Bayer described China restriction as a Republican-led issue with bipartisan support that is shallow. The BINSA bill was introduced by Representative John Moolenaar, Republican of Michigan and chair of the China Select Committee, with Democratic co-sponsor Debbie Dingell; a companion bill came from Representative Elissa Slotkin and Senator Pete Ricketts. The article notes uncertainty over the Select Committee's future funding if Democrats win the House but not the Senate. Armstrong argued that hawkish China policy could be one of few areas of bipartisan agreement after the midterms. With a 45-day comment window and a possible grandfather clause, Bayer expects affected parties to file detailed comments and expects pending China transactions to be accelerated.
The joint-venture clause is the main source of legal ambiguity. Treasury declined to define the term and pointed to its plain-English meaning: the contribution of capital and/or assets by two parties, with profits and losses shared. Bayer called this "extremely broad." Endpoints executive editor Drew Armstrong noted that few China deals are pure licensing; co-development pacts, royalty streams and knowledge transfer are routine, and each raises whether equity or asset contribution has effectively created a joint venture. Armstrong cited a Gilead China transaction, announced overnight and carrying a billion-dollar-plus upfront plus milestones, as an example of a narrowly structured deal that appears permissible.
The final text was broader than the sector expected. A week before publication, the prevailing assumption was that biotech would be mentioned but largely unaffected. Bayer said the national-security faction gained the upper hand, and by October 16 he was told pharma was included. Bayer reads the rule's request for public comment on its own breadth as a signal that drafters may have felt constrained by pressure elsewhere in the administration and may narrow the rule after the comment period.
US venture capital running the NewCo model is seen as most exposed. That playbook, which sources a molecule in China, spins it into a US company and sells it to pharma, typically involves equity and often resembles a joint venture. China-focused funds raising from US limited partners face similar issues. Armstrong relayed an argument that clean licensing preserves late-stage access to Chinese innovation while the US loses early-stage participation. Bayer noted the NewCo model may already be losing favor as large pharma builds search-and-evaluation teams in China, citing colleague Kyle's reporting on roughly five or six Roche deals in China since August 2026.
China is expected to respond with escalation rather than negotiation. Per the article, China has discussed restricting exports of certain technologies and there are reports of restrictions on knowledge workers leaving the country. Armstrong, drawing on tariff experience, expects an initial escalatory response, and Bayer said any retaliation would feed into the final rule language and the intensity of lobbying.
On Capitol Hill, Bayer described China restriction as a Republican-led issue with bipartisan support that is shallow. The BINSA bill was introduced by Representative John Moolenaar, Republican of Michigan and chair of the China Select Committee, with Democratic co-sponsor Debbie Dingell; a companion bill came from Representative Elissa Slotkin and Senator Pete Ricketts. The article notes uncertainty over the Select Committee's future funding if Democrats win the House but not the Senate. Armstrong argued that hawkish China policy could be one of few areas of bipartisan agreement after the midterms. With a 45-day comment window and a possible grandfather clause, Bayer expects affected parties to file detailed comments and expects pending China transactions to be accelerated.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.