China Customs will soon publish June trade data. SOAPBOX’s first-half estimate points to a simple result: in China’s own currency, exports to the EU are heading for a new first-half high.
We estimate Chinese shipments to the EU at about CNY 2.12tn in H1 2026, up from CNY 1.51tn in H1 2021. That is an increase of roughly 40%. Put differently, for every CNY 100 China exported to the EU in the first half of 2021, it is now exporting about CNY 140.
The currency choice is deliberate. The chart is shown in yuan, not euros or dollars, because the question here is China’s export machine. In China’s own customs data and China’s own currency, exports to the EU are not shrinking. They are larger than ever.
The pattern is also revealing. The big jump came between 2021 and 2022. Then exports broadly plateaued in 2022–2024. But 2025 and 2026 show a new upward leg. Whatever the language of de-risking, the aggregate numbers still point in the opposite direction:
In yuan terms, Europe has not reduced its exposure to China’s export machine. It has increased it.
Eurostat has not yet released May EU-China trade data, but Germany’s figures are already available and provide a useful early signal.
It is not encouraging. German exports to China fell 13% year on year in Jan-May. May was less severe than the previous three months, but still weak, with exports down 10% year on year. That followed declines of 12% in February, 14% in March and 17% in April.
The direction is clear. Europe’s largest exporter is still losing ground in China.
In April, SOAPBOX published the chart below showing how heavily EU imports of several antibiotic active pharmaceutical ingredients are concentrated in China.
One month later, Sandoz turned that customs picture into an industrial warning. In May, the company filed an anti-dumping complaint over Chinese amoxicillin API, arguing that low-priced imports threaten what remains of Europe’s production capacity.
Amoxicillin does not appear separately in our chart. In the EU customs nomenclature, it is included within the broader category of penicillin-family APIs. In 2025, China supplied 86% of EU imports by quantity in that category.
Eurostat cannot isolate amoxicillin, but China Customs can take us one step further. China uses separate eight-digit codes for amoxicillin and amoxicillin trihydrate. Adding the two, the EU was China’s largest export market by volume in 2025. Yet it accounted for only 16.2% of China’s recorded exports. Thailand followed with 10.2% and India with 7.2%, while the remaining 66.4% was spread across more than 70 countries.
Some Chinese amoxicillin may still reach Europe indirectly after further processing. The 16.2% therefore measures China’s directly recorded exposure to the EU market, not necessarily Europe’s full ultimate consumption of Chinese-origin amoxicillin.
While the comparison is not perfectly like-for-like, the imbalance is clear. China supplies most of Europe’s imported penicillin-family APIs, including amoxicillin.
Europe’s remaining production base is unusually narrow. Sandoz’s Kundl site in Austria is the EU’s only vertically integrated amoxicillin producer, retaining the chain from the basic penicillin intermediate through to the active ingredient and finished medicines.
The EU’s anti-dumping case against Chinese tyres is now visible in the trade data. Under CN 40111000, EU imports of passenger-car tyres from China in Jan-Apr 2026 were only 56% of the level seen in Jan-Apr 2025 by volume. By value, they were down 54%.
The EU opened the investigation in May 2025 and required imports to be registered in July 2025, before imposing definitive duties in July 2026. Registration was not a tariff, but it sent a clear signal to the market that duties could be coming. That is why the Jan-Apr 2026 fall matters. Imports were already dropping sharply before the final duties arrived.
The case is not only about a sudden import surge. It comes after a long period in which Chinese passenger-car tyres became harder for EU producers to match on price. The average CIF price has barely moved for more than a decade, from €29 per unit in 2013 to around €30 in Jan-Apr 2026. In real terms, that means the import price has fallen, not risen.
Some dependencies are hidden because the trade line is small. Magnesium is one of them.
Unwrought magnesium containing at least 99.8% magnesium is an upstream input for aluminium alloys used in cars, aerospace, machinery, packaging and other industrial supply chains. The customs value may be modest, but the industrial exposure is larger than the trade line suggests.
China supplied 93% of the EU’s extra-EU imports of this product by quantity in 2025. By value, the share was 89%.
That gap points to another problem. China accounted for 93% of EU import volume but only 89% of import value. On that basis, the average customs unit value of China-origin magnesium was roughly 40% lower than the non-China average. Diversification may be possible on paper, but it is likely to come at a higher price.
The pattern is not new. Since 2020, China’s share has stayed above 92% by quantity every year.
The view from China customs is different. The EU is an important market for China, but not the whole market. In 2025, the EU absorbed 34% of China’s exports of this magnesium line, by both quantity and value.
The magnesium trap is a textbook example of the weakness identified by the Draghi report. Europe may possess the mineral resource, but without the processing and metal-making capacity, geology does not translate into industrial security.
Ferro-tungsten is not magnesium. The EU dependency is very serious, but the exposure runs both ways.
China supplied around 69% of EU imports in 2025, while the EU absorbed 43% of China’s ferro-tungsten exports.
China’s export market is also concentrated. The EU and Japan were its two dominant buyers in 2025, together accounting for 85% of Chinese ferro-tungsten exports by volume.
China-origin tungsten ferro-alloys entered the EU at an average customs unit value about 5% lower than supplies from other countries.
Solar panels are the clearest mutual-exposure case in the EU-China trade relationship.
From the EU customs view, the dependencyis almost total. Under CN 85414300, China supplied 99% of extra-EU solar-panel imports by value and quantity in 2025. In Jan-Apr 2026, the share was still 99% by value and effectively 100% by quantity.
But the China customs view changes the story. In 2025, the EU absorbed 37.3% of China’s solar-panel exports by quantity and 37.5% by value. That made the EU China’s largest export market for this line, far ahead of Pakistan, Brazil, India or the UAE.
So solar is not a simple China-leverage story. It is a dependency loop. The EU buys Chinese panels to deploy cheap solar. Chinese manufacturers need large export markets such as Europe because domestic and global overcapacity have pushed the industry into painful losses.
China’s solar manufacturing capacity in 2025 was enough to cover estimated global demand in 2026 almost twice over, while major producers such as Jinko Solar and Trina Solar were still posting losses in early 2026. More than 40 Chinese solar firms have reportedly delisted, gone bankrupt or been acquired since 2024.
The EU depends on China for supply, while China depends on the EU as a critical outlet for an industry trapped in overcapacity.
As Bloomberg Opinion’s Javier Blas has rightly highlighted, China’s oil buying has become one of the key variables shaping the global crude market. But the official GACC data suggest the shock needs to be read carefully.
In Jan-Apr 2026, China’s crude imports averaged 11.18 million barrels per day, the strongest Jan-Apr figure in this series. The weakness came later. In May alone, imports fell to 7.75 million barrels per day, around 3.1 million bpd below both May 2025 and the 2021-2025 May average.
China’s crude-import pullback is real, and it has been large enough to help keep oil prices from rising even more sharply while the Hormuz shock is still unfolding.
But in the official customs data, this is first and foremost a May story, not yet clear evidence of a broad-based collapse in China’s oil demand through 2026.
The June customs data will be important to see whether May was a one-off shock or the start of a more persistent change in Chinese crude buying.
In Jan–May 2026, every major Gulf market shown was down, with the UAE alone accounting for almost 40% of the decline and the UAE, Saudi Arabia and Iraq together explaining about 70%. The total drop is close to $20bn.
The recovery is stronger at the factory gate than at the consumer level. This is partial reflation, led more by production than by households. But there is a catch: factories are charging more for what they sell, while paying even more for what they buy.
China is about to publish its GDP data for the first half of the year (July 15, 10 am, Beijing time). Whatever the headline figure, one caveat is worth recalling.
In its February 2026 Article IV report, the IMF said that China’s statistics continued to have significant gaps. It identified national accounts and government finance statistics as the main areas of concern. National accounts received a C rating, mainly because China does not publish quarterly GDP by expenditure component and provides insufficient detail on the production side.
China’s national-accounts rating looks weak in comparison with other major economies. In recent IMF Article IV consultations, Germany, Japan and the United States were all ratedA. China was ratedC.
That comparison matters. The IMF is not using a special China test. It is applying the same statistical-disclosure framework across countries, and China’s national accounts are rated well below those of other large economies. See IMF Article IV Staff Reports.