Europe’s China debate rests on the wrong grand strategy. Brussels still describes Beijing as a partner, a competitor, and a systemic rival, and argues over the right balance among these three dimensions of its long-term approach. The formula dates from 2019 and urgently needs revision, because it assumes that China can be handled in separate files: trade and the economy in one, climate and global governance in another, and security in a third. Even in October 2026—when the leaders of France and Germany asked the European Commission for the power to shut Chinese companies out of the single market, and when Brussels and Beijing announced a quota on Chinese hybrid cars days later—the discussion was still framed in terms of growing imbalances and a contested level playing field.
The European Union operates with a vocabulary and a mindset that no longer reflect the situation on the ground. The world has entered a new Cold War 2.0 between the United States and what I have called, since 2015, the DragonBear: the strategic coordination of China and Russia across all systemically relevant domains, without a formal alliance. Within that contest, two proxy wars are being fought simultaneously, in Ukraine since 2022 and in Iran since February 2026. Europe is exposed to both, and in both, China stands on the opposing side as the systemic sustainer of Russia’s war effort and of Iran’s survival strategy.
The future of China-Europe relations is therefore less a matter of market access, tariff levels, or summit calendars than of whether Europe remains the terrain across which the contest is fought or becomes an actor within it. In 2025, I set out three options for Europeans—strategic autonomy, continued dependence on the United States, or a deepening of relations with China—and argued that the decisions of the next few years would fix Europe’s trajectory for 10 to 15 years. 15 months later, the options are unchanged, but the room to choose has narrowed. The first remains the only one that produces unique agency, although it demands a civilizational effort. The second describes where Europe stands today. The third amounts to geopolitical and geoeconomic self-destruction. Europe’s standing in 2035 will follow from which of the three it selects, whether by decision or by default.
DROP CAP: Two Wars, One System
Most Western commentary still frames the present as a great game between Washington and Beijing, with Russia cast as a regional disruption, and this leads to the wrong diagnosis. The contest is better understood as one between two geopolitical ecosystems: the American-led network of finance, technology, and alliances, and the DragonBear’s combination of industrial scale, resources, and military resilience. The DragonBear functions as a division of labor. China contributes manufacturing capacity, critical-minerals processing, dual-use technology, and parallel systems of payment, navigation, and logistics. Russia contributes energy and raw materials, nuclear deterrence, wartime experience, and geography, including the Arctic, which China cannot enter as a serious actor without Russia as its entry ticket.
The partnership is asymmetric, and increasingly so: the Power of Baikal (formerly known as the Power of Siberia 2) pipeline still has no binding gas-sales contract, leaving Beijing in a position to set the price. Some European analysts take this as proof that the axis is hollow, yet the conclusion should be the reverse. A Russia that relies on China for its markets, components, and financial channels is a Russia whose war economy Beijing can either sustain or constrain—which gives China a say over the endgame in Ukraine, the central question for Europe’s future security order.
In Ukraine, China’s role has shifted from diplomatic cover to material enablement. Chinese suppliers provide a large share of the dual-use goods on which Russia’s war effort runs, including microelectronics, optics, and machine tools, while China remains the principal buyer of Russian oil, coal, and gas, and is embedded in Russia’s drone economy through a supply of components that amounts in practice to silent co-production. I have long argued that Moscow would not have launched a full-scale invasion without prior, long-term Chinese backing, and nothing since has weakened that judgment.
In Iran, the same pattern is repeated below the kinetic threshold. Since the American and Israeli strikes of February 28th, 2026, the DragonBear has reportedly provided Tehran with deniable capabilities, among them satellite navigation, radar, and yuan settlement. In April, China and Russia vetoed a Security Council resolution on defensive action to reopen the Strait of Hormuz, and while most traffic through the strait was restricted, tankers under Chinese and Hong Kong flags reportedly continued to move along routes directed by Iran.
What matters for Europe is the simultaneity. The two wars did not merge in 2026; they had never been separate. The war in the East ties down Europe’s security resources and political attention, while the war in the Gulf raises its energy costs and weakens the industrial base that is supposed to pay for rearmament. American stockpiles and attention are divided between the two theaters, which increases the pressure on Kyiv to settle and leaves the European flank more exposed. China gains in each case without firing a shot: it buys discounted energy, sells into markets that sanctions have closed to its competitors, observes Western weapons and their limits, and learns how a war economy is kept running under sanctions.
Two wars at once also have major ramifications for the global system. Their effects are multiplicative rather than additive, because they run through the same channels—energy, fertilizer, food, shipping, insurance, and finance—and each shock arrives while the system is still absorbing the previous one. I describe the result as a global system rupture: the world economy no longer fractures along one line but fragments along several, with parallel architectures for payment, navigation, logistics, and standards taking shape alongside the Western-led ones. The Union was designed as a free and open common market inside a global economy that was integrating. It now has to operate as one inside a global economy that is fracturing, where cascades from two wars cross its borders faster than its institutions can respond. The China question has to be answered in that environment, not in the world of 2019.
SUBHEAD: Dependence as Veto
The current trade figures describe a relationship that is no longer merely commercial. Each side is the other’s second-largest trading partner in goods, with the European Union accounting for about 13 percent of China’s foreign trade and China for nearly 15 percent of the Union’s. In 2025, the EU imported goods worth about €559 billion from China and exported about €200 billion, producing a record deficit of roughly €360 billion as exports fell and imports rose. The Commission president now speaks of a tipping point of about one billion euros a day and calls the situation unsustainable.
European policymakers describe the current wave as a second China shock. The first, in the early 2000s, brought cheap, low-technology goods, whereas the second comes from a China that has moved into advanced manufacturing and now presses on the core of European industry, above all the automotive sector. Countervailing duties on Chinese electric vehicles have partly redirected the flow into hybrids, which the duties do not cover. A portion of the surge is diverted trade, since American tariffs have closed part of the U.S. market, and every arrangement between Washington and Beijing that fixes purchase volumes without addressing Chinese overcapacity sends more of that capacity toward Europe. Europeans absorb the adjustment costs of a bargain they did not negotiate.
The deficit is only the visible half of the problem; the other half lies in what Europe buys. The Union sources nearly all of its rare-earth permanent magnets from China and relies on China for a substantial share of other critical raw materials. Beijing introduced export controls on heavy rare earths in April 2025, followed that October by a wider, extraterritorial package, which is suspended until November 10th, 2026, under a truce agreed between Washington and Beijing. Europe’s access to the inputs for its car plants and arms factories therefore depends on a pause negotiated by others, for reasons unrelated to European interests.
China has already demonstrated its willingness to turn Europe’s industrial dependence into geopolitical leverage. When the European Union adopted its twentieth Russia sanctions package on April 23rd, 2026, including restrictions on Chinese and Hong Kong entities implicated in supporting Russia’s military-industrial complex, Beijing placed 7 European defense-related entities on its export-control list the following day, officially citing their involvement with Taiwan. The connection became explicit three months later. Within 24 hours of the EU’s twenty-first sanctions package on July 23rd, which imposed restrictions on 14 Chinese and Hong Kong entities, Beijing retaliated against 14 European entities across 8 Member States, including Germany’s Rheinmetall, prohibiting supplies of Chinese-origin dual-use goods. These targeted restrictions reinforce a broader coercive mechanism established through China’s rare-earth export controls since April 2025, under which defense-industry end users have reportedly faced categorical licensing refusals.
The strategic contradiction is unmistakable. Europe is attempting to rearm against a Russian war effort sustained in part by Chinese dual-use supplies, while remaining dependent on Chinese-controlled materials and components essential to its own defense production. On October 12th, the EU adopted its largest single set of Russia-related listings, targeting almost 1,600 Russian companies and executives, without announcing new listings of Chinese entities. No corresponding Chinese retaliation had been identified by the following day. Although this does not establish that Brussels deliberately avoided Chinese targets, the preceding episodes demonstrate how Beijing can impose immediate industrial costs when European sanctions affect Chinese interests. What appears to be an asymmetric trade dependency has therefore acquired the characteristics of strategic coercive leverage: China can threaten the supply chains supporting European rearmament, industrial competitiveness, and the energy transition, while raising the political and economic costs of further European action against Chinese support for Russia.
DROP CAP: A Deal That Defers
Brussels has recognized most of this. In June, the European Council mandated the Commission to design new instruments against overcapacity and for diversification, and on October 9th—after a second round of consultations in Beijing that Brussels had billed as make-or-break—Trade Commissioner Maroš Šefčovič and Commerce Minister Wang Wentao announced a list of deliverables. Its centerpiece is an understanding on hybrid vehicles, the segment into which Chinese exporters moved after the duties on battery-electric cars: a cap of 350,000 cars in the first year, rising by 6 percent annually, for four years, with both sides pledging to abide by World Trade Organization rules. Around it sit a dozen softer items. China is “willing to continue facilitating” rare-earth and magnet licenses through its fast-track mechanism; both sides will give advance warning of entity listings; several market-access dialogues continue; and a third ministerial meeting is set for March 2027.
The arrangement is better than nothing, and much less than it appears. Almost every item is a commitment to keep talking, and the hard questions, from rare-earth licensing to the subsidies regime, have been moved to next year or beyond rather than resolved. The quota has a loophole wide enough to sail a car carrier through, as journalist Bill Bishop has highlighted: under WTO rules of origin, and unless the final measure provides otherwise, a Chinese brand’s hybrid assembled in a third country counts against that country’s quota, so a BYD built in Turkey does not count toward China’s cap—and Chinese firms are building exactly such plants.
More fundamentally, as the economist Michael Pettis has argued, the problem is systemic, not sectoral. China runs its surpluses not because it produces too many hybrids, but because of a deep consumption imbalance and a very low household income share of GDP. A product-by-product deal does nothing about that imbalance or about Europe’s role in absorbing it; it merely shifts the locus of net imports, so that European hybrid producers gain at the expense of other European manufacturers. In a world of deep and persistent imbalances, according to Pettis, the Union’s only alternatives to accepting deindustrialization are to lower its own household income share of GDP, which would be politically disruptive and bad for the world economy, or to intervene in its external accounts to offset the competitive effect of China’s model. Neither is on the table in Brussels.
The Member States had gone further, and that, in my view, is why Beijing conceded at all. On October 5th, three days before the Beijing meeting, French President Emmanuel Macron and German Chancellor Friedrich Merz wrote jointly to the Commission president calling for urgent action against the “weaponization of trade” and, according to press reports, for authority to expel Chinese companies from the single market through expedited procedures—a proposal quickly dubbed the “kill switch.” The shift in Berlin was the significant part, since for two decades German industry had acted as the brake on every European attempt to confront Chinese practices. As a standing deterrent, however, the instrument is far-fetched. A halt in magnet deliveries would stop European assembly lines within weeks, whereas excluding companies from a market of 27 states would take months of procedure and years to bite, contested in court at every step. The Union has never used the Anti-Coercion Instrument it adopted for exactly this situation. And the kill switch itself, which the letter proposed to activate by reversed qualified majority, does not yet exist—a gap Beijing can afford to test.
The deeper weakness is conceptual, and it has two parts. The first is that Europe does have leverage. The Union is the largest open market still available to Chinese exporters, the principal destination for the capacity that American tariffs have displaced, and the holder of compound power as a market, a regulator, a standard-setter, and a source of capital all at once. The October deal shows that Beijing concedes on terms when Brussels sets a deadline and Franco-German pressure stands behind it. But leverage becomes power only when it is exercised, and that requires the political will and institutional competence the Union has so far withheld: an Anti-Coercion Instrument never used, trade-defense procedures measured in years, and 27 governments among which Beijing can always find a dissenter.
The second is that the relationship has become geoeconomically unsustainable, regardless of the next quota. The global economy is bifurcating by function into parallel architectures, and the Union has set itself the goals of reindustrialization, manufacturing rejuvenation, and rearmament, which cannot be met while Europe absorbs China’s surplus capacity. No sectoral settlement reconciles these goals, and none addresses China’s role as the sustainer of Russia’s war economy and the destination of the Iranian oil that still moves through Hormuz. A trade policy conducted in isolation from these facts will be fought on Beijing’s terms and at a time of Beijing’s choosing. The quota and the kill switch are instruments, and Europe has yet to decide which strategy they serve.
DROP CAP: Europe’s Three Roads
The first option is strategic autonomy, understood as a capacity rather than a slogan. It means catching up with America and China in the domains that decide systemic competition—energy, defense, industry, and technology—and it is a generational effort. Europe needs energy independence built on a pragmatic, state-specific mix of domestic and allied gas, expanded liquefied natural gas infrastructure, and nuclear power as a core pillar, because energy is the precondition of industrial power, of defense production, and of the green transition itself. Military credibility requires production depth in ammunition, drones, and air defense, much of it in Central and Eastern Europe. Industrial sovereignty has to extend to mining, processing, magnets, and mature semiconductors; the 2025 dispute over the Dutch chipmaker Nexperia showed that the vulnerability lies in ordinary chips more than in advanced ones.
The order of magnitude is the €750 to €800 billion of additional annual investment that the Draghi report identified across strategic sectors, alongside the up to €800 billion that ReArm Europe aims to mobilize for defense over four years. And money is not the only constraint: the effort also requires longer working lives, less bureaucracy, and an education system oriented toward science and engineering. Autonomy in this sense would change the relationship with China at its root. Critical dependencies would be reduced by building or buying elsewhere, and partnerships with India and Southeast Asia would give European firms alternatives to China-centered supply chains. Engagement with Beijing would continue where interests overlap, but from a position in which a refused export license is an inconvenience rather than a production stoppage. The difficulty is that the costs fall within electoral cycles while the returns arrive after them, and nobody in European politics currently pays a price for deferring the decision.
The second option is the status quo, which in the long run means growing dependence on the United States. For three decades, Europe spread its dependencies across three suppliers: Russian energy, Chinese manufacturing, and American security. That arrangement has collapsed. Russian energy has been significantly reduced since 2022, and Chinese manufacturing has turned from an input into a competitor and a lever, which leaves America as the provider of energy, technology, and market access as well as defense. In a cold war, dependence of this kind is alignment by default, not by design. Europe today accepts a tariff ceiling of 15 percent on most of its exports to the United States and duties of up to 50 percent on steel and aluminum, while it is expected to follow Washington on export controls and on sanctions against Russia and Iran. This is strategic submission more than sovereignty.
The consequences for China policy are direct. Washington and Beijing have built a bilateral machinery for managing their rivalry: a truce that now runs to January 10th, 2027, standing boards on trade and investment, and leaders’ meetings in Beijing in May and in Washington in September. Europe takes part in none of these formats, although the conditions under which its factories receive rare earths are settled there. The second option assumes that American and European interests regarding China coincide; in fact, they only overlap, and the American approach is transactional. Washington may trade stability in the Pacific for outcomes that Europeans must live with in Ukraine, in the Gulf, or in their own markets. None of this is an argument against the transatlantic alliance, which remains a condition of European security for as long as Europe cannot defend itself. The argument concerns the terms. Alignment chosen by a capable Europe is a strategy; alignment that results from the absence of alternatives is a condition whose terms are set by others.
The third option is to deepen relations with China, and it presents itself as realism: if the United States is unreliable and coercive, Europe should hedge by restoring its economic relationship with Beijing. The option has advocates among Member States such as Spain, and it has material backing. Chinese direct investment in the EU and the United Kingdom rose by about two-thirds in 2025, to its highest level since 2018, with record greenfield spending on battery and vehicle plants. Each plant creates a local constituency of jobs, suppliers, and tax revenue, which makes its host government more reluctant to support measures Beijing dislikes. Investment of this kind serves both as a route around European tariffs and as a source of leverage inside European decision-making.
This option fails on its own terms, because it repeats the logic of Wandel durch Handel, the belief that deep commercial ties moderate an authoritarian power. Europe tested that belief on Russia for two decades and received its answer in February 2022. Applying it now to Russia’s principal enabler would mean moving from de-risking to re-risking. Closer ties would reinforce the DragonBear, since European capital, technology, and market access would strengthen the industrial base that supplies Russia’s war. They would erode Europe’s own security order, because the Union cannot credibly sanction the suppliers of that war while courting their government. And they would end what remains of the American guarantee before Europe has anything to put in its place.
SUBHEAD: The Closing Window
The most probable course over the next one to three years is managed escalation, with neither rupture nor reset. The October arrangement is this path in its purest form: a quota that buys a year, licensing promises that Beijing can slow at will, and the structural questions deferred to a meeting in March. Europe will add trade-defense instruments in increments and continue to list Chinese entities that supply Russia’s war. Beijing will answer each step with calibrated measures against European firms, chosen for political effect in particular capitals, while stopping short of a full embargo that would unite the continent against it.
Two other paths are possible. The first is the trade war that the Franco-German letter anticipates, in which Europe adopts a strong instrument against overcapacity, China reinstates its extraterritorial rare-earth controls for European users while extending the pause for Americans, and Europe reaches for the kill switch. Such a sequence would disrupt the car and defense industries within weeks and would test European unity more severely than any tariff has done. The pause on those controls expires in November, and the U.S.-China truce comes up for renewal in January; neither date is set by Europeans.
The second is a transactional settlement. A ceasefire in Ukraine and a stable truce between Washington and Beijing would allow China to offer Europe export restraint, investment, and purchase commitments in return for delistings and a role in reconstruction. This path is the more dangerous precisely because it is the more comfortable: a ceasefire would be the moment at which the third option returns in respectable form. Member States with Chinese plants on their soil would press for acceptance, and Europe would be invited to reward the power that sustained the war for its contribution to ending it.
Behind all three paths stands a latent third theater. A crisis over Taiwan would directly affect European manufacturing, shipping, and chip supply; shape Russian planning on the European flank; and confront the Union with a demand for sanctions against a partner with which it trades more than €750 billion in goods a year. China’s leverage in rare earths, mature chips, and pharmaceutical ingredients is positioned for that day, and Europe has no playbook for it.
Several decisions nonetheless lie in European hands and can be taken now. Frozen Russian assets can be transferred to Ukraine, which would alter the balance of the war more than any other single step. Every listing of a Chinese entity can be paired with protection for the European firms likely to be targeted in response, through stockpiles, joint purchasing, and accelerated permits. A market-exclusion instrument can be tied explicitly to China’s support for Russia’s war, not only to trade distortions. As a general deterrent against overcapacity, such a tool would be too slow to matter; as a targeted sanction against the specific firms that sustain Russia’s war economy, it would rest on firmer legal and political ground than any trade measure. And any rare-earth arrangement that Washington negotiates can be made to cover allies as a condition of European alignment. The change of government in Budapest has cost Beijing its most dependable advocate inside the Union, though no vote has yet tested the change, and that opening will not last indefinitely, as Spain’s approach to China already indicates.
The window of opportunity is finite. Strategic autonomy is expensive, slow, and politically difficult, yet it is the only course that leaves Europeans in charge of their own relationship with China. Dependence on the United States will hold only for as long as Washington finds it useful, and deeper ties with China would strengthen the axis that is waging war on Europe’s border. Industrial bases, energy systems, and armed forces are built over 10 to 15 years, and the decisions that determine them are taken in the first two or three—the years Europe is now spending. A Europe that builds capacity will deal with Beijing as a party in its own right, with instruments it is prepared to use and partners that give it alternatives. A Europe that defers will find its China policy written in Washington when the two capitals agree, and in Beijing when they do not. The bottom line is that Europe is not a pole in the new Cold War 2.0 between America and the DragonBear. For now, it remains the battlefield across which that contest is being fought.
