The Trump–Xi Summit: China Secures Strategic Gains While Modest U.S. Wins Required No Presidential Visit

• Donald Trump left Beijing on Friday afternoon.

• The summit produced one announced commercial deal of consequence: a Chinese order for 200 Boeing jets, with the possibility of growing to 750 if “all goes well.”

• (side note: although it is hard to find a public confirmation of these deals at the highest level of Chinese government.)

• By comparison, the 2017 Trump visit produced 300 Boeing planes and a $250 billion package.

• The 2026 visit ended with fewer Boeing planes than 2017, no concrete Chinese commitments on Iran’s blockade of Hormuz, and a Chinese readout that conspicuously omitted the Iran file Trump had spent two months publicly insisting was a summit priority.

• Our post ahead of this summit had argued that Trump was walking into the worse seat and that the summit would likely produce photographs and headline numbers while leaving the structural balance unchanged in Xi’s favor.

• The post-summit data largely confirms the prediction.

The Unrequeinted Flattery

• Trump arrived in flattery mode.

• He told reporters Xi was “a great leader” and “a friend.”

• He told Fox News the visit had been “incredible.”

• He praised Xi as “a man I respect greatly. Become really a fan. We’ve known each other now 11 years, almost 12 years.”

• Xi did not return the favor: he did not call Trump a great leader (something that he has done in relation to Putin before.)

• The Chinese readouts described Trump in formal protocol terms and made no comparable gestures of personal regard.

• Beijing’s response was far more muted.

• Xi delivered substantive policy warnings on Taiwan.

• He did not characterize Trump as a great leader, a friend, or a partner he respected.

• This represents a role reversal.

• The standard sequence over two Trump terms has been that world leaders, including the highest-profile NATO allies (Macron, Starmer, Merz), travel to the Oval Office, deliver praise, and receive transactional engagement in return without reciprocation.

• Trump pockets the flattery.

• He does not return it.

• Beijing inverted this.

• The President of the United States flew to Beijing, repeatedly praised Xi as great, and received no comparable praise back.

• The world watched the leader of the world’s foremost military and economic power perform deference to a host who did not reciprocate.

• Not the best signal to send to the ‘‘Global South’’ or countries that hedge between China and the U.S.

The Genuine Wins

• Three genuine wins:

• 1) No chip deal.

• Trade Representative Jamieson Greer told Bloomberg that Nvidia and semiconductor issues were “not a central focus.”

• However, it seems like Chinese pride and strategic calculations, and not our unwillingness to sell these chips was the defining factor here; Trump told reporters aboard Air Force One that China had “passed on” the H200 chips, saying “they chose not to, they want to try to develop their own.”

• Consequently, no licensing breakthrough was announced for H200.

• Fortunately, the more advanced Blackwell question was deferred.

• This is a strategic win.

• Not selling advanced AI chips to China is the correct outcome.

• (side note: even though we really have Beijing to thank for that.)

• Every chip the U.S. declines to sell maintains the American AI compute advantage and denies Chinese frontier model training the hardware it cannot domestically substitute.

• Per Wall Street Journal reporting from October 2025, when Trump previously tried to put Blackwell on the Busan summit agenda, Rubio, Greer, and Lutnick intervened in the final hours and forced removal.

• Luckily, this posture held at Beijing.

• Any deal that allows Nvidia to sell more chips to China means fewer Nvidia chips for U.S. firms, and a smaller U.S. advantage/edge in AI over China.

• 2) The rare earth truce extension.

• The Busan truce of October 2025 suspended the most extreme wave of Chinese rare earth export controls (a licensing regime that would have extended Chinese authority to any product worldwide containing even trace amounts of Chinese-sourced rare earths) and had been due to expire in November 2026.

• The summit appears to have produced an extension.

• This matters because China’s April 2025 controls already produced a 93% year-on-year drop in rare earth magnet exports to the U.S., and a second wave would have devastated American defense manufacturing, aerospace, semiconductors, and EV production.

• Heavy rare earth export volumes remain roughly 50% below pre-restriction levels even with the truce, which is the structural problem Project Vault is designed to address over years.

• But the extension prevents the more extreme regime from snapping back.

• 3) Agreement that Iran should not have nuclear weapons.

• Per the U.S. readout, both sides agreed Iran should not have nuclear weapons.

• The Chinese readout did not explicitly confirm this language, but it also did not dispute it.

• (side note: the Chinese readout didn’t even have a single reference to Iran at all.)

• The win is real even with the asymmetric documentation.

• Beijing is on record, by tacit acquiescence if not by explicit text, supporting the position that Iran should not develop nuclear weapons.

• This confirms the structural limit of China’s relationship with Iran: Bejing backs Iran instrumentally, as an investment in regional leverage and a tool against American power.

• China does not however, back Iran as a strategic ally that Beijing would underwrite to nuclear breakout.

• The Iranian nuclear question is the line China will not cross openly.

• These three wins are real but modest.

• The question is whether they were worth the trip of a U.S. President or if Rubio meeting his counterpart would have been sufficient here..

The Taiwan Issue Did Not Go All Too Well

• Xi delivered his message on Taiwan with unusual directness. Per the Chinese readout, Xi told Trump that “the Taiwan question is the most important issue in China-U.S. relations” and that “if mishandled”, the two nations will experience “clashes and even conflicts.”

• The phrase describes possible armed conflict.

• The framing as “the most important issue” elevates Taiwan above every other agenda item.

• And the construction of this messaging is intentionally asymmetric: clearly, Beijing doesn’t think that it will be the one doing the ‘‘mishandling’’ part.

• As such, only one party can “mishandle”, and that party is the United States.

• Trump’s response in the meeting was to absorb the warning without rebuttal.

• This is, in isolation, defensible.

• No one should be seriously arguing that Trump should have engaged in an open verbal confrontation.

• The bigger issue is however what happened later on Air Force One.

• Trump told reporters that he and Xi had “talked a lot about Taiwan.”

• He confirmed Xi had specifically raised the pending $14 billion arms sale.

• He said he had made “no commitment either way” on it and would “make a determination over the next fairly short period” after speaking to “the person who is running Taiwan.”

• He acknowledged Xi had asked whether the U.S. would defend Taiwan if China attacked, and he had declined to give him an answer (”I don’t talk about that”).

• Then the comment that should be read as the most consequential statement of the summit.

• Asked about the 1982 Six Assurances (the longstanding U.S. commitment that includes a pledge not to consult with China about arms sales to Taiwan), Trump said: “What am I going to do, say I don’t want to talk to you about it because I have an agreement wrote in 1982? No, we discussed arms sales.”

• That is the sentence that is bound to cause some serious worry in Tokyo, Seoul, Canberra, Manila, and Taipei.

• The Six Assurances are not merely symbolic.

• They are the bedrock of the U.S.-Taiwan security relationship that has helped prevent cross-strait conflict for four decades.

• They commit the United States not to set a date for ending arms sales, not to consult with the PRC on arms sales, and not to revise the Taiwan Relations Act.

• Every Republican and Democratic administration since 1982 has reaffirmed them.

• They are the architecture that signals to Beijing that Taiwan’s defense is not a negotiating chip and that signals to Taipei that the United States is reliable.

• Trump just publicly characterized the assurances as an “agreement wrote in 1982”.

• Implying that times have changed and he could be willing to disregard them.

• He confirmed he discussed Taiwan arms sales with Xi, exactly what the assurances pledge not to do.

• He left the $14 billion package, approved by Congress in January 2026 and including missiles and air defense interceptors, in deliberate ambiguity.

• The substantive question of whether the package will be delivered matters less than the precedent of discussing this with Xi.

• Xi now has a seat in conversations about American arms sales to Taiwan.

• That’s a strategic first for China.

• The Six Assurances no longer bind the American president when he finds them inconvenient.

• Taiwan transfers are now effectively subject to negotiation as part of broader U.S.-China management.

• The damage to deterrence is immediate.

• Japan, South Korea, Australia, and the Philippines have integrated their security postures into a U.S.-led Indo-Pacific architecture predicated on reliable American commitments to Taiwan.

• AUKUS, the trilateral U.S.-Japan-South Korea coordination, and the Philippine basing agreement all depend on this assumption.

• The signal that Beijing can influence the U.S. president sufficiently to introduce ambiguity into Taiwan transfers is a signal that the architecture is built on softer ground than allied capitals had assumed.

• Taiwan Foreign Minister Lin Chia-lung publicly thanked Rubio for telling NBC News that “U.S. policy on the issue of Taiwan is unchanged as of today.”

• That public thanks is the diplomatic register of profound concern.

• This is the main development from this summit.

• Not that Trump refused to give a straight answer to whether he would defend Taiwan.

• Because the latter question becomes largely moot if Xi can have a veto over Taiwan arms sales.

• This is because, who is really going to believe that the U.S. is going to fight for Taiwan if it may even refuse arms sales to Taiwan after pressure from Xi?

• Now, one could very well argue that the two questions are actually separate.

• The sale of, or the willingness to sell weapons in this context, is more about the delicate balance of powers and the calculation of whether additional sales would improve deterrence or cause instability and prompt China to expedite its own timeline and attack sooner.

• Clearly, if the goal is to prevent the invasion and attack, this is an important calculation to make.

• There’s a genuine argument that, depending on the overall context, further sales could lead to perverse outcomes and tempt China to expedite its own timeline while it holds more advantage versus later, when Taiwan becomes even more indigestible.

• But, hat’s a very nuanced take and analysis, and one has to believe that Trump made that very nuanced calculation in discussing this issue.

• A more straightforward analysis, and one that a lot of important decision-makers will surely have to consider, is that there is simply no appetite to move in a direction that increases the likelihood of confrontation with China at any level.

• And if that’s the conclusion, then the question of whether the U.S. would fight for Taiwan largely answers itself…

No Concrete Help On Iran

• The alignment on Iran’s nuclear weapons was a real win.

• But on the two operational questions that mattered (pressuring Iran to lift the Hormuz blockade, and constraining the Chinese supply chain that arms Iran’s drone and missile manufacturing), the summit produced nothing.

• On Hormuz, Trump told reporters Xi had agreed Iran must reopen the strait. He also said, when asked whether Xi had made a firm commitment: “I’m not asking for any favors because when you ask for favors, you have to do favors in return.”

• This is the verbal (implicit) acknowledgment that no commitment exists.

• Beijing’s foreign ministry statement excluded the issues of Hormuz reopening and of Iran blockade entirely.

• China’s ambassador continues to oppose the U.S.-Bahraini draft UN Security Council resolution demanding reopening.

• The Iranian response was definitive.

• On Saturday, the day after Trump left Beijing, Ebrahim Azizi, chairman of the Iranian parliament’s National Security and Foreign Policy Committee, announced Iran would soon unveil a “professional mechanism” to manage Hormuz traffic along a designated route, with fees for “specialized services.”

• Azizi was explicit: “this route will remain closed to the operators of the so-called freedom project,” referencing Trump’s failed Project Freedom operation.

• Iran has already received its first revenue from tolls.

• Iran is not reopening Hormuz.

• Iran is instead institutionalizing the toll regime that maximizes Chinese leverage while imposing costs on American and Israeli vessels.

• Conspicuously absent from the summit’s Iran discussion was any mention of the Chinese supply chain for Iran’s drone and missile manufacturing base.

• This is the substantive issue that should have been raised.

• Wall Street Journal reporting from May 5 documented the scope of the problem.

• Chinese customs data show Chinese companies are shipping hundreds of containers of dual-use goods to Iran (and Russia), with packing lists including engines, computer chips, fiber-optic cables, gyroscopes, and lithium-ion batteries.

• Total shipments are in the hundreds of millions of dollars.

• Between 2023 and 2024 alone, Chinese firms directly supplied at least $63 million in drone-related components to sanctioned Russian companies, with comparable flows reaching Iran.

• The WSJ noted Chinese exporters intentionally mislabeled some shipments to skirt US and European sanctions, but in many instances they no longer bother.

• The transition from concealment to openness is the operational signal that Beijing has stopped pretending to enforce dual-use controls.

• The case of Xiamen Victory Technology illustrates the pattern.

• The small Chinese firm openly markets German-designed Limbach L550 engines for Iran’s Shahed-136 attack drones, the same drones that have struck UAE infrastructure and that Iran’s proxies have used against U.S. forces in Iraq.

• In March, the firm sent an Iranian customer an explicit marketing email during the Iran-Israel war: “We are deeply shocked and outraged by the aggression against Iran, and our hearts are with you,” followed by the L550 sales pitch.

• The Limbach L550 is on the U.S. export ban list specifically because it is used in Iranian and Russian drones. Xiamen Victory continues to sell it.

• Indeed, there is a significant increase in Chinese-manufactured components in Shahed-style drones.

• Iranian-backed militias used fiber-optic-guided drones equipped with Chinese components to destroy a U.S. Black Hawk helicopter and an air defense system in Baghdad in March.

• Every additional Shahed Iran can manufacture is a Shahed that can be fired at American forces, Gulf infrastructure, or Israeli targets.

• There is no indication in any post-summit readout that this was substantively addressed.

• Trump’s claim that Xi “agreed not to supply Iran with military equipment” addresses, at most, direct state-to-state transfers, which were not the primary problem.

• The problem is small Chinese firms operating in dollar-free supply chains through Hong Kong shell companies, which Beijing is not interdicting.

• It is however implausible that Trump could have extracted a real Chinese commitment even if he had pressed.

• This is because arming Iran serves Chinese strategic interest in ways the Hormuz toll regime and the nuclear question do not.

• A continuing flow of components lengthens the war, drains American munitions stocks, increases the risk of U.S. casualties, raises the political cost of sustained Gulf engagement, and undermines U.S. regional credibility.

• It rewards Chinese investment in Tehran without requiring Beijing to overtly back Iranian aggression.

• It is exactly the kind of indirect strategic harm Chinese strategy is designed to produce: deniable, distributed, structurally damaging.

• The structural limit of the Iran nuclear weapons agreement comes into focus here.

• China is willing to be on the record opposing Iranian nuclear weapons because nuclear breakout would damage Chinese commercial interests and trigger uncontrolled escalation Beijing cannot manage.

• China is not however, willing to cut off the dual-use supply chain, because that supply chain produces exactly the slow-burn American attrition that serves Beijing’s long-term competitive position.

• The nuclear line is the line China will hold.

• Everything below it is fair game.

The Declining Power

• Xi opened the summit by invoking the Thucydides Trap.

• “The world has come to a new crossroads,” Xi told Trump at the Great Hall of the People. “Can China and the United States overcome the so-called Thucydides Trap and create a new paradigm of major-country relations?”

• The Thucydides Trap is Graham Allison’s framework drawing on Thucydides’s account of the Peloponnesian War.

• It describes the structural tension between a rising power and a ruling power.

• In Allison’s tally, 12 of 16 historical cases produced conflict.

• China is the rising power.

• The United States is the ruling power.

• Xi was warning Trump that the historical default is war, and that avoiding it requires Washington to accommodate Beijing’s ascent.

• Trump’s response is where the framing got locked in.

• On Truth Social, he addressed Xi’s reference: “Two years ago, we were, in fact, a Nation in decline. On that, I fully agree with President Xi.”

• The sentence is extraordinary.

• The American president, on the public record, explicitly agreed with the Chinese president’s framing that the United States is (or at least was) the declining power and China is the rising one.

• Trump placed the decline in the Biden years to deflect domestic damage, but the substantive concession was made.

• The framing was not contested.

• The framework will be referenced in every subsequent Chinese diplomatic communication for years.

• Instead of agreeing wit Xi’s framing (just for a cheap and petty opportunity to attack Biden) A good response to Xi’s messaging would have been to rebut the appropriateness of the comparison by highlighting that the U.S. itself is continuously ascending, and that’s its own pinnacle of power is yet to be reached…

Was the Trip Worth It?

• The wins were modest.

• The chip discipline could have been achieved by continuing existing policy without a summit.

• The rare earth truce extension could have been negotiated through Treasury and Commerce channels.

• The Iran nuclear weapons understanding could have been secured by Rubio in a counterpart meeting with Wang Yi.

• None of the wins required Trump’s physical presence.

• None of these required Trump traveling all that way to call Xi a ‘‘great leader’’.

• The costs and strategic downsides all required physical presence.

• The Taiwan precedent could only have been established at a leader-level summit where Trump personally discussed arms sales with Xi.

• The powerful effect of warning a conflict on Taiwan and the Thucydides framing could only have been delivered with Trump in the room.

• The asymmetric flattery could only have been observed at state-visit choreography.

• A Secretary of State counterpart visit would have produced the commercial outcomes without producing the major strategic costs.

• The president would not have been observed performing deference.

• The Six Assurances would not have been characterized as discretionary.

• Xi would not have warned of a “conflict” on Taiwan due to U.S. ‘‘mishandling’’ or invoked the Thucydides Trap at leader level.

• Taiwan arms sales would not have been the subject of bilateral negotiation.

The September Question

• Trump announced Xi will visit Washington for a state visit in late September.

• How that visit turns out depends on what Trump does about the $14 billion Taiwan arms package over the next two to three months.

• If Trump cancels the sales, Xi arrives in Washington having extracted a concrete concession.

• The September trip becomes a victory lap formalizing Chinese authority over American Taiwan policy.

• Beijing benefits enormously.

• If Trump proceeds with the sale, Xi has less obvious incentive to make the September trip to Washington.

• Beyond his already-scheduled UN General Assembly appearance, what would he gain from a separate Washington visit having been seen to lose the most important issue from May?

• The optics would be unfavorable.

• The third possibility is continued delay.

• Trump neither approves nor cancels, and the September visit happens with Taiwan still in ambiguity.

• This is the second most favorable scenario to Beijing because it preserves the precedent without forcing Xi to absorb any visible defeat.

• Trump’s “fairly short period” framing leaves room for this trajectory.

• Three months is a long time.

• The Iran deal could close or collapse.

• The Putin-Xi summit next week could produce framework agreements that change the regional calculus.

• Whether and how the September meeting happens depends on choices Trump has not yet made.