ECB puts a number on how alike German and Chinese exports have become
Victoria Shinder2 min readA focus box in the ECB's Economic Bulletin, Issue 6/2026, published 22 September 2026 and prepared by Ginevra Aguiari, Francesco Chiacchio, Matteo Falagiarda, Vanessa Gunnella and Emiel Cornelius Marchand, applies an Export Similarity Index — a measure that goes back to Finger and Kreinin (1979) — to the sectoral make-up of Chinese and EU goods exports between 2019 and 2025.
Read country by country, the index rises almost everywhere and rises most where manufacturing is the largest share of the economy; Germany is the case the box names. Countries with narrower, more specialised export baskets moved less. Decomposed by sector, the bulk of the movement sits in one SITC category covering machinery and vehicles — which the box attributes to the speed of China's move into cars and into more technology-intensive goods generally.
The box reads the traffic in the other direction too. China's own import basket has drifted away from what Europe sells, which the authors take as reduced reliance on European industrial goods. Set against export market shares, the pattern the box reports is losses where Chinese competition sharpened, alongside gains for EU exporters in the United States, where China's share fell over the same period — gains concentrated in the higher-value end of the range.
What it means
The phrase the box uses for the whole phenomenon is "China shock 2.0", and the numeral is the argument. The first shock was about cheap consumer goods arriving in Western markets. This one is measured in the categories that European industrial policy has treated as the safe ground: capital goods, vehicles, machinery. A similarity index rising fastest in the most manufacturing-intensive member states is a statement that exposure is not spread evenly across the single currency.
That is why this sits in a central bank bulletin rather than a trade ministry's. The box says so directly: asymmetric exposure to a global shock bears on convergence between member states and on the functioning of a monetary union. One policy rate is set for economies whose external position is being moved by the same force in different directions and by different amounts.
Two limits the box states for itself are worth carrying. The analysis covers goods only and therefore leaves out services, where several European economies hold strong positions. And an overlap in export structures is a measure of where two countries compete, not a measure of who wins — the market-share evidence is reported separately and is mixed, with the US result running the other way. The box also notes that competition can show up in margins and pricing power rather than in volumes, which an index of this kind does not capture at all.