A different way of looking at trade -- trade in kilos Europe now imports (by weight) more machinery, electronics and cars from China than from the rest of the world combined 1/3

There is a clear asymmetry at play. The long term correlation is very low. What do I mean? Look at right side of the graph: the currency declined and exports grew above trend. That makes sense and it's observable. Now look at the left side: the currency grew significantly more expensive but exports kept growing on trend. This tells is two key things. First, there is a clear asymmetry in how Chinese trade responds to changes in the value of the RMB. Just because exports increase when the currency declined, do not expect exports to decline (or though not pictured here imports to increase). Second, there are clearly missing variables that would help explain why we are seeing the results we are seeing. As a simple example, the mass of Chinese over capacity is playing a significant role in boosting exports. This concept would be especially true considering imports. In short, do not expect a symmetric bounce in imports

Worth noting the Trump admin has succeeded in reducing the trade deficit vs GDP - US trade deficit for 2026 is tracking at 2.5% the lowest it’s been for a while- Huge success to be able to reduce this to 2% by the end of the presidency? Exports / particularly AI led will increase materially

For those wondering why Setser strips gold imports out of China’s trade data, the reason is that gold can be purchased for two very different purposes. You can buy it for jewelry or manufacturing, or you can buy it as a store of value, much like buying a gold ETF, government bonds or property. The former is clearly a current-account transaction; the latter, when the asset is foreign, is a financial-account transaction. The official balance-of-payments rules, however, do not make this distinction for physical gold. Nonmonetary gold is recorded as a good whether it is used for jewelry or held as a store of value. A foreign gold ETF is unambiguously an acquisition of a foreign financial asset, but an imported gold bar can be an almost perfect economic substitute and is nonetheless recorded as an import of goods. This matters for China, especially in recent years. Chinese jewelry demand has collapsed relative to investment demand: in the first half of 2026, bar-and-coin demand reached a record 314 tons, more than twice the 136 tons of jewelry demand. So Setser is making an analytical adjustment. Much of the recent surge in Chinese gold imports appears to reflect a shift in household and investor portfolios toward gold rather than a comparable increase in demand for goods and services. That is why stripping gold out of the trade balance can give us a better picture of the underlying goods imbalance.