Much has been made of the latest U.S. trade figures, which showed imports from China down nearly 24% year-to-date compared to last year. Does this portend an era of nearshoring? Decoupling? Derisking? Or just a shift of tastes and locales?

The depth of the fall of U.S. imports from China was reported as more of a surprising development than perhaps it should have been. As shown in the chart below – which takes the 12-month rolling average of China and HK’s combined share of U.S. nominal imports, as well the shares of Asia including China and Asia excluding China and Hong Kong – the China + HK share actually started to decline from around February 2018 when it was at 22.0%. There was then a Covid era surge that saw it recover to 19.6%, only for the figure to resume sliding when the surge subsided. As of June, it stood at 15.0%, the lowest since October 2005.

From there, it becomes a choose-your-own adventure of explanations. There are up to four we see, some intertwined. If, for example, you want to follow the path to decoupling (or “de-risking,” in the more recent, more diplomatic formulation) taking hold, China’s share reaching a nearly 17-year low seems to be supportive of that. But then you get into the more complex issue of what defines decoupling. Is it U.S. demand just shifting away from production in China? Or from Chinese companies altogether?

If it’s the former, you might be confronted by the curious case of U.S. inventories and the impact they may be having on demand, especially for consumer goods from China. The Covid surge we highlighted above was driven by a number of factors, including a sudden shift in consumer preferences and inventory accumulation as a defense against persistent supply chain disruptions. This led to excess inventories – though just how excessive it’s difficult to say. Once stocks are drawn down, however, we could very well start to see direct imports from China increase again, provided U.S. consumer goods demand remains strong.

If it’s the latter, you might look at the seafoam blue line in the chart and be lured by the China + 1 and diversification claims. Indeed, the share of Asia excluding China and Hong Kong began rising at almost the same time as China + HK started falling. But for those who have tied themselves to the mast of data showing shifts in China’s exports and FDI into the region increasing, the diversification claims – and decoupling from Chinese firms – are somewhat less compelling. One of the restraining knots might also be that China is still the world’s main manufacturer and by a considerable margin.

Relatedly, there is also a long-standing issue with the way trade flows are accounted for. Imports are recorded from the last country of transformation, not by the portion of value added. A short and simple explanation of why capturing value added matters can be found here, using China and early versions of the iPhone. The OECD also estimates global trade in value added, but given the complexity of supply chains, and the time and effort it takes to disentangle where the actual value of the finished product originates, there tends to be a long lag in reporting. Hence the difficulty in determining from the latest trade data the countries and companies that are the actual source of supply.

Finally, one might focus on the top line – Asia’s share including China and Hong Kong – and be led to the nearshore, namely Mexico. While China’s share is falling and the rest of Asia’s is rising, Asia including China is still shrinking, although it remains just slightly under two-fifths of the total as of June. There’s growing evidence that Mexico is filling the gap; we’ve looked at this earlier in the year, using container throughput at Laredo at a time of decreasing seaborne shipments as one measure. But ‘nearshoring’ also remains largely undefined: Does it mean substituting for supply from China and Asia partially or even entirely? When can we conclude that is happening? What does it actually look like in the end?

The key points to take from all of the above are, first, there is a tendency to select the narrative of most interest – to fall prey to confirmation bias. Second, concepts and theories need definitions to be measured. And third is the necessity of accurate data for testing those theories.

In other words, it’s worth exercising some caution when embarking on analytical adventures.

Disclaimer: The contents of this report are made available for informational purposes only and should not be relied upon for any legal, business, or financial decisions. Flexport does not guarantee, represent, or warrant any of the contents of this report because they are based on our current beliefs, expectations, and assumptions, about which there can be no assurance due to various anticipated and unanticipated events that may occur. This report has been prepared to the best of our knowledge and research; however, the information presented herein may not reflect the most current regulatory or industry developments. Neither Flexport nor its advisors or affiliates shall be liable for any losses that arise in any way due to the reliance on the contents contained in this report.