In 2024, the United States imported $439 billion in goods from China. In 2025, that figure dropped to $308 billion. That is a 30 percent fall in a single year. Early 2026 data shows imports running at about $20 billion per month, which puts the full year on track for around $240 billion. Compared to 2024, that is a 45 percent decline in two years.
Numbers this large do not happen by accident. They tell a clear story at first glance. Tariffs went up. Trade broke down. Decoupling is finally happening. The headlines have caught up. Politicians on both sides claim victory. Analysts who predicted a fragmenting world now see vindication. The story is wrong. Or rather, it is incomplete in a way that changes everything.
What the Data Actually Shows
US imports from Vietnam in 2023 totaled $114 billion. In 2025, they reached $191 billion. That is a 67 percent increase in two years. Vietnam’s entire GDP is roughly $400 billion. The country now sells almost half its economic output to the US alone. US imports from Mexico rose from $506 billion in 2024 to $535 billion in 2025. Modest growth, but Mexico already had the largest import relationship with the US. US imports from India climbed from $87 billion in 2024 to $104 billion in 2025. A 19 percent jump.
Add the increases together. Vietnam gained $54 billion. Mexico gained $29 billion. India gained $17 billion. Total: $100 billion in additional imports from these three countries over a single year. The decline in imports from China that same year was $131 billion. Roughly 78 percent of the trade that disappeared from China showed up in the books of three other countries. All three are well-known transit points for Chinese-supplied goods.
The Number That Settles the Question
China’s total exports to the world reached $3.75 trillion in 2024. That is the highest figure in Chinese history. It is higher than the previous peak in 2022. It is higher than any year before or during the Trump or Biden trade wars. China is not selling less to the world. It is selling more.
The bilateral US-China number fell. The global Chinese export number rose. There is only one way both facts can be true at the same time. Chinese goods are still reaching American consumers, just not with “Made in China” stamped on the customs paperwork.
Why This Pattern Exists
The technical name for what is happening is trans-shipment. The mechanics are simple. A Chinese factory produces a finished product, or the components for one. The goods are shipped to Vietnam, Mexico, or India. There, they undergo final assembly, light processing, or in some cases just relabeling. Then they enter the US under the new country of origin.
Tariffs that apply to “Chinese” imports do not apply to “Vietnamese” imports. The economics of the supply chain push aggressively toward this kind of routing whenever tariff differentials are large enough to make it worthwhile. The Trump administration is aware of this. The 2025 trade deal with Vietnam included a 40 percent tariff specifically on goods “shipped through” Vietnam suspected of being Chinese-origin. That policy itself is the strongest possible confirmation that the pattern is real and large enough to require a dedicated countermeasure.
The Three Sectors That Are Actually De-Risking
The trans-shipment pattern explains most of the bilateral trade collapse. But not all of it. Three sectors are genuinely separating from Chinese supply chains. These are the only places where the de-risking story holds up under data.
Semiconductors
The US has spent three years systematically blocking China from advanced chip technology. The October 2022 export controls cut off advanced semiconductor manufacturing equipment. Subsequent rules expanded the scope. On December 31, 2025, the Bureau of Industry and Security revoked Validated End-User authorizations for major Chinese chip fabs. That closed one of the last remaining loopholes.
The flip side is real domestic investment. The CHIPS and Science Act allocated $52 billion in subsidies. TSMC is building fabs in Arizona. Samsung is building in Texas. Intel is expanding in Ohio. Whether these plants produce competitive chips at scale remains unproven. But the capital is flowing. The intent is clear. This is the only sector where decoupling is structurally underway.
Defense and Critical Materials
China controls 90 percent of global rare earth processing. It produces 94 percent of the world’s permanent magnets. These materials power electric vehicles, wind turbines, defense systems, and AI data centers.
In October 2025, China announced its most aggressive export controls to date. Companies tied to foreign militaries faced automatic license rejection. Beijing claimed jurisdiction over products made anywhere in the world that contained Chinese rare earth content. Most of these controls were paused in November 2025 as part of the Trump-Xi Busan deal. The pause runs to November 2026. The US response accelerated. The Defense Production Act now funds domestic rare earth processing. MP Materials, Lynas, and Noveon Magnetics are receiving government support. The Department of Defense has signed multi-year offtake agreements that guarantee demand.
The IEA estimates that rebuilding alternative rare earth supply chains takes 20 to 30 years. The US is starting now. Whether it succeeds is uncertain. The trajectory is set.
Politically Symbolic Consumer Electronics
Apple is the company most aggressively diversifying away from China. The numbers are striking. Apple assembled $22 billion of iPhones in India in the twelve months ending March 2025. That was a 60 percent increase over the prior year. India became the leading exporter of smartphones to the US for the first time in the second quarter of 2025. But Apple’s stated long-term goal is roughly a 50/50 split between India and China. Not exit from China. Production in India costs 5 to 10 percent more. Yield rates at Indian facilities sit around 50 percent, compared to near-perfect yields in China.
Most consumer electronics companies have not made similar shifts. The economics do not work for products without Apple’s brand value or political exposure. The “Made in India” label often describes final assembly of substantially Chinese-made components. For investors, this is a narrow story. Branded leaders with high US political exposure are genuinely shifting. The rest of the industry is not.
The Much Larger Universe Where Nothing Is Changing
Outside these three sectors, the supply chain story is closer to “business as usual.” China still produces about 80 percent of the world’s solar panels. That share has grown, not shrunk, over the past five years. Lithium-ion batteries follow the same pattern. China produces around 75 percent of global capacity. Western EV manufacturers continue to source extensively from Chinese suppliers. Planned battery facilities in the US and Europe rely heavily on Chinese-licensed technology and Chinese-supplied processing equipment. In furniture, textiles, basic chemicals, generic pharmaceuticals, and machine tools, China’s role as the world’s manufacturing base remains essentially unchanged.
The Trump-Xi Busan deal in November 2025 acknowledged this reality. The deal extended lowered tariffs. China announced tariff reductions on 935 separate items. Both sides agreed to suspend ship fees and various retaliatory measures. This was not the behavior of governments dismantling their trade relationship. It was the behavior of governments stabilizing one.
Why Both Governments Want the Story to Continue
The trans-shipment pattern is not hidden. Trade economists have been writing about it for two years. The Trump administration’s 40 percent Vietnam tariff is policy confirmation. Chinese export totals are public. Yet the “decoupling” narrative persists in headlines and political speeches. There is a reason for this.
For Washington, the rhetoric supports a coherent set of goals. Tariffs generate revenue. Industrial policy produces jobs in politically important states. Security framing justifies defense spending. Tough-on-China posturing polls well across the political spectrum. None of these benefits require decoupling to actually happen. They only require it to be plausibly in progress.
For Beijing, the narrative is equally useful. Being targeted justifies “self-reliance” investment. It supports the Communist Party’s domestic narrative of foreign hostility. It gives China leverage in negotiations. The October 2025 rare earth export controls were the most aggressive Beijing has ever deployed. They were paused six weeks later in exchange for tariff reductions. The threat had more value than its execution.
The result is a stable equilibrium. Tariffs go up and come down. Export controls escalate and pause. Bilateral trade flows redirect through third countries. The actual supply chain dependence continues. Both governments claim victory at home. Neither wants to break the relationship. The economic damage would be too large.
Distinguish the Three Real Sectors From Everything Else
Genuine de-risking is happening in semiconductors, defense materials, and politically symbolic consumer electronics. Investment theses built around these sectors have structural support. They should persist regardless of which party controls Washington.
Theses built around “China alternative” plays in furniture, low-end electronics, or basic manufacturing have weaker foundations. The underlying economics still favor China. The only thing that changed is the country stamped on the customs paperwork.
Be Skeptical of “China Alternative” Pure Plays
Vietnam, India, and Mexico have benefited from the rebalancing. The benefit is concentrated in specific industries with real capability rather than spread across their economies.
Vietnamese final assembly is real. Vietnamese deep-tier component manufacturing is largely Chinese-supplied. Indian iPhone assembly is real. Indian semiconductor manufacturing is largely aspirational. Investors who assume the “China+1” strategy lifts every developing Asia economy equally are likely to be disappointed.
Watch for Trans-Shipment Enforcement
The single biggest risk to the current equilibrium is whether the US starts enforcing rules-of-origin requirements seriously. The 2025 Vietnam tariff of 40 percent on suspected trans-shipped goods exists on paper. Enforcement has been limited. If enforcement tightens through stricter customs inspections, expanded rules-of-origin tests, or new legislation, the cost of routing Chinese goods through Vietnam could rise sharply.
Companies whose business model depends on tariff arbitrage through third countries are exposed. Companies with genuine local supply chains are not.
Position for Reversibility in Both Directions
The November 2026 expiry of the rare earth export control suspension is a real date. So is the US midterm election that month, which historically reshapes trade policy. Investors should expect periodic shocks in both directions. Portfolios that require continuous improvement in either escalation or de-escalation are fragile. The pattern of the past three years is volatility around a slowly moving baseline.
How This Could Be Wrong
A Taiwan crisis is the largest risk. A Chinese military move on Taiwan, or a confrontation serious enough to trigger Russia-style sanctions, would force genuine decoupling regardless of either government’s preference. The economic damage would be enormous, but political dynamics could overwhelm economic logic.
A second risk is that trans-shipment enforcement actually works. If US customs starts rejecting goods routed through Vietnam, or if rules-of-origin tests get teeth, the substitution channel that currently absorbs the China decline could close. That would force genuine relocation rather than relabeling. Costs would rise across the consumer goods sector.
A third risk is that Vietnamese and Mexican capacity hits a ceiling. Vietnam’s economy is roughly $400 billion. The country already exports nearly half its GDP to the US. There is a physical limit to how much more it can absorb. If Chinese exporters run out of trans-shipment options, the substitution slows.
The framework in this article assumes none of these scenarios materializes. Investors should hold this view with appropriate humility. Geopolitical regimes can shift faster than economic logic suggests they should.
Conclusion
The decoupling story is comforting because it is simple. The data tells a more complicated one. Bilateral US-China goods trade is genuinely smaller than it was two years ago. That part is real. But the underlying supply chain dependence is largely intact. Chinese goods now reach American consumers through Vietnam, Mexico, and India rather than directly. The trans-shipment numbers match the bilateral decline closely enough that the pattern is hard to dismiss as coincidence.
Three specific sectors are genuinely separating: semiconductors, defense materials, and high-profile branded consumer electronics. Outside those, the structural picture has not changed. Both governments find the appearance of decoupling politically useful. Neither wants the reality of it.
For investors, the discipline is to see this clearly. Headlines will continue to suggest a fragmenting world. Government trade data will continue to suggest a managed one. Portfolio decisions made on the headline narrative will systematically overpay for resilience that turns out to be unnecessary. They will underweight the categories where dependence still matters.
The goal of reading the global economy is not to have the most dramatic view. It is to have the most accurate one. On US-China supply chains, the most accurate view in May 2026 is that decoupling is happening on customs paperwork and almost nowhere else.