U.S.-China Relations
NextFin News - Global commerce did not shrink under the weight of the Trump administration's tariffs - it grew faster than expected. The World Trade Organization raised its forecast for world merchandise trade volume growth in 2025 to 2.4%, up sharply from the 0.9% projected just two months earlier, after trade volumes expanded 4.9% year-on-year in the first half of the year. The catch: the damage has not vanished. It has been postponed, with the 2026 outlook slashed to 0.5% and the full force of higher tariffs still waiting in the wings.
The story of 2025 is not that protectionism failed to bite. It is that three forces - panic-buying ahead of tariff deadlines, a boom in AI-related goods, and a shift of trade toward emerging economies - absorbed the blow and pushed the pain into next year. The question for investors and policymakers is whether that deferral is a soft landing or a bill coming due.
Layer 1: The Situation - Trade Grew, but the Growth Was Borrowed
On 7 October 2025, the WTO delivered a forecast that looked like a reversal. After warning in April that world merchandise trade would contract 0.2% in 2025, and then revising that to 0.9% growth in August, economists at the Geneva-based body lifted the full-year estimate to 2.4%. Volume growth in the first six months of 2025 ran at 4.9% year-on-year, and the value of world merchandise trade rose 6% over the same period.
The upward revision was not broad-based good news. It was concentrated, mechanical, and in important part reversible. US importers pulled forward shipments ahead of tariff hikes that took effect on 7 August 2025. American imports surged 11% year-on-year in volume terms in the first half of the year, including a 14% quarter-on-quarter jump in the first three months - followed by a 16% drop in the second quarter. That is not organic demand. It is demand relocated in time.
To measure the wall that trade ran into, consider the tariff structure itself. A baseline 10% duty took effect on 5 April 2025 on most US trading partners, and on 7 August 2025 country-specific reciprocal tariffs ranging from 10% to 41% hit imports from more than 60 economies, on top of existing steel, aluminum, and copper levies. By any normal standard, that is a large shock to the price of moving goods across borders. The surprise is not that trade kept growing. The surprise is how little the headline slowed.
At the same time, spending on AI-related products - semiconductors, servers, and telecommunications equipment - surged 20% year-on-year in value terms in the first half of 2025 and accounted for nearly half of the entire expansion in global trade during that period. Trade among emerging economies also accelerated: South-South trade grew 8% year-on-year in value terms in the first half of the year, faster than the 6% pace for world trade overall, and South-South trade excluding China ran even hotter at around 9%.
The WTO's own language captures the tension. Director-General Ngozi Okonjo-Iweala said "countries' measured response to tariff changes in general, the growth potential of AI, as well as increased trade among the rest of the world - particularly among emerging economies - helped ease trade setbacks in 2025." But she added:
Trade resilience in 2025 is thanks in no small part to the stability provided by the rules-based multilateral trading system. Yet complacency is not an option.
The baseline trajectory tells the deferral story plainly. World merchandise trade volume growth is expected to slow from 2.8% in 2024 to 2.4% in 2025 and then to 0.5% in 2026. Global GDP growth is projected at 2.7% in 2025 and 2.6% in 2026. Services export growth, not directly subject to tariffs but tied to goods trade, is set to ease from 6.8% in 2024 to 4.6% in 2025 and 4.4% in 2026. The tariff impact, in the WTO's own words, "has shifted into 2026."
Layer 2: The Analysis
Why Trade Did Not Break: Three Shock Absorbers
The first lesson of 2025 is that the tariff shock arrived in a form that markets and firms could prepare for, and preparation is the enemy of a clean natural experiment. The reciprocal tariff regime was announced on 2 April 2025, suspended, revised, and only took effect on 7 August 2025 - giving importers a four-month window to pull orders forward. The result was a textbook inventory cycle: North American inventories-to-sales ratios rose across machinery, motor vehicles, lumber, construction equipment, and non-durable goods in the first half of the year.
Second, the product mix of world trade changed underneath the headline. AI-enabling goods are a small slice of total trade - roughly one-sixth - but they grew so fast that they carried 42% of global trade growth in 2025, with the value of trade in these products rising 21.9% year-on-year over the full year to $4.18 trillion from $3.43 trillion in 2024. Crucially, many of these products - chips, semiconductors, and data transmission equipment - are exempt from most of the new tariffs. A trade boom driven by a tariff-sheltered sector is not evidence that tariffs do not work. It is evidence that tariffs steer rather than stop.
Third, trade went where the tariffs were not. North America's goods exports, which the WTO had forecast in April to fall 12.6% in 2025, were later projected to decline just 3.1% - because the import collapse that was expected to hit the region in 2025 now lands in 2026 instead. Meanwhile, Asia's exports rose 4.9% in 2025, and emerging-market trade among themselves accelerated. The European Central Bank documented Chinese exports being redirected toward ASEAN countries, particularly intermediate goods for further processing - the signature of trade diversion, where goods still reach their final destination but through a longer, more expensive route.
The Mechanism: Borrowed Demand, Rerouted Flows, and a Bill That Matures in 2026
The transmission channel here is not mysterious. Tariffs raise the landed cost of targeted imports. Firms respond in three stages: first they pull forward purchases to beat the effective date; then they draw down the inventories they built; then they cut orders as higher costs pass through to demand. Stage one shows up as strong trade data. Stages two and three show up as weak trade data. The 2025 strength is stage one. The 2026 slowdown is stages two and three arriving on schedule.
This is why the same two-year period - 2025 plus 2026 - ends up looking similar in total whether tariffs bite early or late. The WTO noted that while the 2025 forecast was revised up from both the April and August projections, "the downgrade for 2026 results in a similar overall impact over the two-year horizon." The tariff did not disappear from the arithmetic. It moved from one cell of the spreadsheet to another.
The rerouting channel is different, and it is more durable. When Chinese goods face an effective tariff burden approaching half their declared value on some routes, but can enter the United States via Mexico or Canada at near-zero duty under USMCA rules, the incentive to transship is structural, not cyclical. An independent trade-policy monitor estimated that roughly $75 billion of Chinese goods were diverted to 17 partner economies between April 2025 and April 2026, with about one-fifth showing signs of being funneled back to the United States through intermediaries. That is not frontloading. That is a permanent lengthening of supply chains - more handling, more paperwork, more cost embedded in the system.
There is also a third-order effect that the headline trade numbers hide: inflation. The WTO warned that rising input prices and slowing shipments signal that inflation could increase in late 2025 as inventories shrink in tariff-exposed sectors. Frontloading masks the price effect in the short run - you pay the old price on the goods you rushed in - and reveals it later. That sets up a policy dilemma for central banks: the tariff-driven price impulse arrives just as growth is slowing, the classic stagflationary mix that makes rate decisions harder, not easier.
The Counter-Thesis: What If the Resilience Is Real, Not Borrowed?
The strongest case against the "deferred damage" reading is that it underweights adaptation. Firms did not just pull orders forward and hide behind intermediaries; they reconfigured. The WTO's Goods Trade Barometer stood at 101.7 in June 2026 - above its 100 baseline - with the electronic components index at 105.5 and container shipping at 102.4, both signalling continued expansion. Export orders at 100.5 were only slightly above trend, but they were not contracting. If tariffs were simply crushing commerce, these forward-looking indicators would be below baseline, not above it.
Nor is the slowdown as severe as the 0.5% figure implies in every forecast. The International Monetary Fund's July 2026 World Economic Outlook Update projected world trade volume growth of 3.5% in 2026 - far less dire than the WTO's baseline 1.9% for merchandise trade in its March 2026 report (a figure that already assumed a Middle East energy-price shock; the baseline excluding that shock was higher). The WTO itself acknowledged upside potential: if the Middle East conflict is short-lived and AI spending stays strong, merchandise trade growth in 2026 could reach 2.4%, boosted by 0.5 percentage points. And beneath the tariffs, the rules-based system still governs most commerce: the share of world trade conducted on a most-favoured-nation basis stood at 72% at the end of February 2026.
This counter-thesis has real force. It correctly identifies that trade is a flowing system, not a static stock, and that firms adapt faster than models assume. But it conflates "trade did not collapse" with "tariffs did not matter." The data show the opposite: trade growth was preserved by exactly the three channels the tariffs were designed to punish - timing arbitrage, third-country routing, and tariff-exempt sectors. The system adapted at a cost, and that cost shows up as lower forecast growth ahead, higher logistics expenses, and a delayed inflation impulse. Resilience is not the same as immunity.
The falsifying signal is specific: if US and North American import volumes grow through 2026 without the unwind the WTO projected - if inventory-to-sales ratios normalize without a corresponding drop in new orders, and if merchandise trade volume growth in 2026 comes in above 2% rather than near the 0.5%-1.9% range - then the deferral thesis is wrong and the tariff impact has genuinely been absorbed rather than postponed.
Layer 3: Conclusion - Who Benefits, Who Is Exposed, and What to Watch
The practical implication is asymmetric. The beneficiaries of the 2025 pattern are clear: exporters of AI-enabling goods and the Asian supply chains that feed them; emerging-market economies integrated into South-South trade; and logistics, warehousing, and transshipment hubs in Southeast Asia, Mexico, and Central America that profit from longer routing. The exposed are import-dependent US manufacturers facing higher input costs, European exporters losing share as North American demand cools, and consumers who will meet the tariff-driven price impulse with a lag.
Stratified by time horizon, the picture differs at each layer. In the short term - sentiment and liquidity - trade data will likely stay firmer than the tariff narrative suggests, because frontloaded inventories take time to draw down and AI capital spending remains strong. In the medium term - fundamentals - the unwind arrives: weaker import volumes in North America, softer export growth in Europe, and a drag on global GDP that the WTO puts at 2.6% for 2026. In the long term - structure - the trading system does not break, but it thickens: more regional blocs, more intermediaries, a smaller share of trade on pure most-favoured-nation terms, and a permanently higher cost of moving goods across borders.
Three scenarios frame the path. The base case is the WTO's deferred-impact path: merchandise trade growth near 0.5%-1.9% in 2026 as frontloading unwinds, with a modest recovery to around 2.6% in 2027. The upside case requires two conditions holding together - a quick resolution of the Middle East conflict and sustained AI investment - which would lift 2026 merchandise growth toward 2.4%. The downside case stacks a protracted energy shock on top of broader tariff escalation, pushing trade-restrictive measures into more sectors and economies - the exact risk the WTO flags as its main downside - and could take growth well below 1%.
The rerouting channel carries its own political risk. The USMCA corridor - Mexico and Canada - offers the widest tariff gap of any route, with effective rates there sitting more than 30 percentage points below China's, so every dollar rerouted through North American partners avoids more duty than through any other channel. That gap is directly at stake in the July 2026 renegotiation of the US trade pact with Mexico and Canada, making the trade agreement a live lever on whether the diversion channel stays open or gets closed.
What to watch, concretely: monthly US import volumes and inventory-to-sales ratios through 2026; the WTO Goods Trade Barometer, especially the export-orders index; the most-favoured-nation share of world trade, which stood at 72% in February 2026 and would signal deeper fragmentation if it falls materially; the outcome of the USMCA renegotiation; and core inflation prints in late 2025 and early 2026, where the delayed tariff price impulse should appear if the deferral thesis holds.
The verdict: global trade's resilience under tariffs is real, but it is the resilience of a system that reroutes, retimes, and relabels - not one that escapes. The tariffs did not fail. They simply moved the pain to a date on the calendar that has not yet arrived.
Data as of late July 2026, based on official releases from the World Trade Organization, the International Monetary Fund, and the European Central Bank.
Explore more exclusive insights at nextfin.ai.
