EU Trade Commissioner Maroš Šefčovič and China’s Commerce Minister Wang Wentao smiled for the cameras in Brussels on 29 June as they announced a new Trade and Investment Consultation Mechanism, presented as a platform to address bilateral trade issues.
Beijing’s readout emphasised cooperation to manage differences. Brussels, facing a trade deficit with China worth roughly €1 billion per day, adopted a firmer tone, stating that “the status quo is not an option.”
The diplomatic language, however, masked a clear divergence.
Hours later, China’s Foreign Ministry spokesperson declared that “the root of the EU’s problems does not lie in China” (here). Notably, the Chinese text uses 根源 (gēnyuán), meaning the underlying or fundamental source of a problem, rather than the more general 原因 (yuányīn), meaning “cause.”
Meanwhile, on 1 July, Beijing brought into force its first comprehensive regulation on outbound investment. The rules introduce a national security review covering not only capital, but also data, technology and other intangible transfers abroad, including technical know-how and cross-border engineering support. They also extend the framework to private individuals and provide for countermeasures against foreign restrictions considered discriminatory.
The timing is notable. The regulation follows Beijing’s decision to block Meta’s reported acquisition of Manus, a company founded by Chinese nationals that had closed its mainland offices, laid off most of its China-based staff and relocated its remaining development team to Singapore.
Europe continues to press for a level playing field and has increasingly relied on trade defence instruments.
China, meanwhile, is building a more formal legal framework governing the overseas transfer of technology, data and investment, reinforcing state oversight over strategic assets and capabilities.
Around 2019, much Chinese commentary still treated semiconductor equipment as a catch-up story. Domestic self-sufficiency was low, and lithography remained the hardest bottleneck, but the mood was not one of permanent dependence. Policy funds, China’s large domestic market and the expectation that global technology progress was slowing all fed the belief that domestic suppliers could narrow the gap.
A 2019 Chinese report argued that domestic semiconductor equipment technology was accelerating its catch-up and that domestic substitution had arrived. A 2020 article argued that lithography equipment would be one of the areas benefiting from China’s second national semiconductor fund.
Six years later, the import data tell a more sober story. China is still buying large amounts of chipmaking machinery from the Netherlands. The mix may have changed, and the data do not tell us how advanced each shipment is. But the persistence of the flow shows how difficult the machinery bottleneck has been to close.
When China introduced export licensing for several semiconductor materials in August 2023, the obvious expectation was that Europe would buy less from China.
Instead, the trade barely changed in volume while China’s share of its value almost doubled.
At first sight, that is a surprising result. The licensing regime appears to have affected the value of the trade far more than its physical volume.
On 3 July 2023, China’s Ministry of Commerce announced export licensing requirements for several gallium-related semiconductor materials, citing national security. The measures entered into force on 1 August 2023 (official announcement here). The licensing regime covered gallium nitride (GaN), gallium arsenide (GaAs), gallium phosphide (GaP), gallium oxide and indium gallium arsenide (InGaAs), all classified under CN 38180090.
Rather than analysing all gallium trade, we focused exclusively on EU imports under this customs code between January 2022 and April 2026.
China’s licensing regime did not produce a sustained reduction in its physical presence in the EU market under CN 38180090. Nearly three years later, China’s average share by quantity remains virtually unchanged.
Although volumes changed little, its share of the value of EU imports rose sharply. That suggests the trade became considerably more valuable, even if it did not become much smaller.
The unit-price data reinforce that interpretation. Throughout the period, Chinese imports entered the EU at substantially higher values per kilogram than imports from the rest of the world, with the premium becoming even more pronounced after the licensing regime entered into force. The obvious question is what explains that premium.
Why?
These are not commodity materials. They are advanced semiconductor substrates rather than raw gallium.
GaAs wafers, GaN wafers and InGaAs materials are highly engineered products used in RF communications, power electronics, photonics and defence. Depending on their diameter, crystal quality and epitaxial structure, individual wafers can be worth hundreds or even thousands of euros.
The customs data cannot tell us which mechanism lies behind the shift. It could reflect a richer product mix, greater pricing power under the licensing regime, or continued European dependence on specialised Chinese materials that proved difficult to replace.
Whichever explanation dominates, they all point in a direction that matters.
China’s export licensing appears to have changed the economics of this trade more than its physical volume.
The lesson extends beyond gallium. China’s export controls should not be judged only by whether exports stop. They can also reshape the economics of strategically important trade while allowing China to remain an indispensable supplier.
China’s Commerce Minister Wang Wentao is promoting in London a new Export to China campaign, encouraging foreign companies to view China as a growing destination for their exports.
For the UK, however, the message sounds familiar.
During the Golden Era of UK-China relations, the expectation was that closer economic ties would translate into greater access to China’s market. More than a decade later, the trade data tell a different story.
China’s goods surplus with the UK has increased from US$58 billion in 2023 to an estimated US$77 billion in 2026.
China’s “Export to China” message is not new. This time, as before, the real test will not be the rhetoric but whether the trade balance begins to move in the opposite direction.
The chart should not be read as a simple Taiwan leaves China story. Exports to the Chinese mainland are still growing. The bigger change is that U.S. demand for AI-related hardware has exploded, pulling more of Taiwan’s advanced electronics trade toward the U.S. market. Taiwan sits at the centre of that supply chain, from advanced chips to servers, and the export data now show that shift very clearly.
China’s first shipments of e-cars to Canada
Canada recently shifted its policy to allow the importation of up to 49,000 electric vehicles annually at a reduced 6.1% tariff. This replaces the 100% tariff imposed in 2024 and is part of a broader bilateral trade deal negotiated by Prime Minister Mark Carney.
China shipped 3,500 electric cars to Canada in April, at an average FOB value of about $29,000 per vehicle. That pace would quickly bring Chinese exporters close to the quota ceiling set by the Canadian government.
Despite strong export performance, China’s official PMI employment subindex has remained below the 50 threshold throughout the past four years. The latest reading suggests manufacturing employment is still contracting.
China’s National Bureau of Statistics is scheduled to publish Q2 and first-half GDP data on 15 July. As usual, SOAPBOX is publishing its own forecast ahead of the release. We expect real GDP growth to slow to 4.6% year on year in Q2, from 5.0% in Q1. But nominal growth should remain almost unchanged, as the implied GDP deflator moves from negative or near-zero territory into slightly positive territory. In our forecast, Q1 nominal GDP growth of about 4.94% is followed by 4.86% in Q2. That leaves first-half real growth at 4.8% and nominal growth at 4.9%. Under this assumption, China’s implied first-half GDP deflator turns positive for the first time since H1 2022. However, at +0.1%, the deflator is barely positive and should not yet be read as a strong reflation signal.
On 1 July, the EU introduced a new €3 duty on low-value parcels entering the bloc. At first glance, it looks like another tariff.
In reality, it reflects a much bigger problem.
The EU may know how many parcels are arriving and how much tax should be collected. But it may not know, with the same product-level precision as conventional trade, exactly what is inside those parcels.
That may sound surprising.
When China joined the WTO in 2001, cross-border e-commerce barely existed. Low-value customs exemptions were designed for another world, one in which someone occasionally returned from New York with a new shirt or ordered a fishing lure by post from New Zealand. Small parcels were an exception, not a major trade channel.
That world has gone.
Today, Europe receives billions of low-value parcels every year, worth billions of euros. VAT has applied to all commercial imports since 2021. The new €3 duty is the latest attempt to adapt customs rules to this new reality.
But collecting taxes is only part of the challenge.
Traditional customs systems were designed around containers and commercial consignments supported by detailed customs declarations. Parcel trade works differently. Millions of shipments move through simplified electronic declarations submitted by platforms, postal operators and logistics companies. Customs can count parcels and collect duties. Identifying every individual product with the same statistical precision as conventional imports is much harder.
That matters because discussions about the EU’s trade deficit with China usually assume that all imports pass through the same statistical machinery.
They do not.
The headline trade figures remain robust, but the parcel economy increasingly relies on simplified customs procedures that prioritise efficient clearance over highly granular product classification. As this channel expands, understanding exactly what Europe is importing becomes more challenging.
China recognised this transformation much earlier.
From around 2018, and especially after the pandemic, Beijing progressively built a dedicated customs architecture for cross-border e-commerce. New supervision codes, electronic declarations, simplified clearance procedures, overseas warehouse policies and digital integration between customs, logistics and payment systems were all designed to move millions of parcels efficiently across borders.
The chart illustrates how quickly this new trade channel has expanded. According to China Customs, exports to the EU cleared under simplified procedures increased from US$1.7 billion in 2019 to US$26.8 billion in 2025. An equivalent public series from the EU perspective is, today, much harder to obtain.
The €3 duty is therefore about much more than raising revenue.
Europe is adapting a customs system built for the container age to a trading model that China has spent years redesigning for the parcel age.