Q1-Q3 2026 customs data: 39.94 trillion yuan traded, 6.46 trillion surplus, and why the two growth rates differ by 15 points
39.94 trillion yuan. That is the number China’s General Administration of Customs put on the board on 14 October 2026 for the first three quarters of goods trade, up 18.8 percent year on year, the fastest pace for the period in five years. Exports came to 23.2 trillion yuan, up 16.4 percent; imports came to 16.74 trillion yuan, up 22.3 percent. The headline circulated globally within hours, and most coverage stopped where the presser stopped.
The ledger deserves one more pass, because the four published numbers do not all move together. 18.8 percent of total trade is not 18.8 percent of anything a trader or a policymaker actually pockets. The surplus, the number that dominates the political conversation, grew roughly 3.4 percent in yuan terms over the same period, computed from the customs totals themselves. Both figures are correct. They answer different questions, and the 15-point gap between their growth rates is where the actual story of 2026 sits.
The official claim, verbatim
At the State Council Information Office presser, customs deputy head Wang Jun stated: this year’s first three quarters, goods trade imports and exports 39.94 trillion yuan, up 18.8 percent year on year; exports 23.2 trillion yuan, up 16.4 percent; imports 16.74 trillion yuan, up 22.3 percent. He added that imports grew 5.9 percentage points faster than exports, that September alone reached 5.16 trillion yuan, the first month ever above the 5-trillion mark, up 27.5 percent, and that quarterly growth accelerated from 15.3 percent in Q1 to 18.3 percent in Q2 to 22.3 percent in Q3. All of that is quoted here as published. The task is to check whether the pieces cohere, which denominators each number silently assumes, and where two correct figures start to tell different stories.
Checking the arithmetic: the base year holds
Every growth rate customs published is year on year, in yuan terms, for the period January through September 2026. The 2025 base is therefore recoverable from the published figures themselves, by dividing rather than by hunting for last year’s presser. Divide 39.94 trillion by 1.188 and the implied base is 33.62 trillion. Divide 23.2 trillion by 1.164 for exports and the implied base is 19.93 trillion. Divide 16.74 trillion by 1.223 for imports and the implied base is 13.69 trillion. The three implied bases sum to 33.62 trillion, matching the total’s implied base to the second decimal. The ledger is internally consistent; no published rate contradicts another.
| Metric (Q1-Q3 2026) | Published value | Published YoY growth | Implied 2025 base (computed) | Consistency check |
| Total goods trade | 39.94 trillion yuan | 18.8% | 33.62 trillion yuan | equals sum of parts |
| Exports | 23.2 trillion yuan | 16.4% | 19.93 trillion yuan | sums to total base |
| Imports | 16.74 trillion yuan | 22.3% | 13.69 trillion yuan | sums to total base |
| Trade surplus (computed: exports minus imports) | 6.46 trillion yuan | about 3.4% (computed) | 6.24 trillion yuan (computed) | derived, not published |
| September single month | 5.16 trillion yuan | 27.5% | 4.05 trillion yuan | first month above 5 trillion |
Source: computed from figures in the customs press conference transcript 🔗, 14 October 2026
One computed row needs its denominator flagged before anyone quotes it. The surplus line is not a customs-published growth rate; it is the difference of two published totals in each year, so its 3.4 percent inherits rounding from both. Roughly is doing real work in that row. The direction, however, is robust: imports outgrowing exports by 5.9 percentage points mechanically compresses surplus growth far below trade growth, and no plausible rounding reverses a gap that wide.
Three denominators, three different stories
The same 18.8 percent total decomposes very differently by partner block, and the presser’s own numbers supply the pieces. Trade with free-trade-agreement partners reached 19.55 trillion yuan, up 30 percent. Divide that by 1.30 and the base is 15.04 trillion, an increment of 4.51 trillion yuan. The total increment of trade across the whole ledger is 39.94 minus 33.62, which is 6.32 trillion. The FTA-partner increment is 71.4 percent of the total, which is what the presser means by the phrase that FTA partners contributed over seventy percent of the trade increase. The claim checks out, and it is the most consequential sentence in the release: growth is concentrated, not broad.
| Partner bloc (Q1-Q3 2026) | Value | YoY growth | Implied increment | Share of total increment (computed) |
| FTA partners | 19.55 trillion yuan | 30% | 4.51 trillion yuan | about 71% |
| ASEAN | 6.84 trillion yuan | 22.7% | 1.27 trillion yuan | about 20% |
| APEC other economies | 24.07 trillion yuan | 23.9% | 4.64 trillion yuan | overlaps FTA figures |
| EU | 4.78 trillion yuan | 8.4% | 0.37 trillion yuan | about 6% |
| United States | 3.18 trillion yuan | 4.1% | 0.13 trillion yuan | about 2% |
Source: computed from the customs press conference figures 🔗 as republished by People’s Daily Overseas Edition
Read the last two rows against the first. The US line grew 4.1 percent in a year when total trade grew 18.8 percent, which means trade with the United States shrank as a share of the ledger for the fourth straight structural shift, and the phrase used in the official readout, that US trade “continued to recover”, is doing quiet work: 4.1 percent is recovery the way a tide recovering from low tide is high tide. The APEC row overlaps the FTA row, since several FTA partners are APEC members, so the two increments cannot be summed; the table computes each bloc’s share against the whole anyway, and the overlap is flagged in the row itself rather than hidden.
The export-composition story is the same shape. Equipment manufacturing contributed nearly ninety percent of the export increment. Integrated circuit exports grew 99.1 percent. Electric vehicles grew 72.3 percent, lithium batteries 36.4 percent, wind turbine sets 22.4 percent. Robot components, electronic parts, computer parts and electrical equipment together crossed 10 trillion yuan in combined imports and exports, up 58.9 percent, and the presser credits that basket with 11.3 percentage points of the total trade growth, meaning two thirds of the 18.8 percent came from one basket of goods. That is a concentration ratio, and the denominator, total trade growth, is the same one the headline uses.
The yuan-dollar seam
39.94 trillion yuan converts to “about 5.94 trillion US dollars” in Xinhua’s English dispatch. Divide 39,940 billion by 5,940 billion and the implied reference rate is 6.72 yuan per dollar. But customs also maintains a surplus series in dollars, and 938.11 billion dollars against a computed yuan surplus of 6.46 trillion implies 6.89. September’s single-month surplus of 133.53 billion dollars against a computed September surplus of about 0.90 trillion yuan implies 6.74. The range across the three implied rates runs from 6.72 to 6.89, roughly a 2.5 percent spread.
This is not an error, and calling it one would be the classic mistake of treating a denominator difference as a contradiction. Customs’ dollar-denominated series is converted at period-average market rates transaction by transaction, while a single-point conversion of a trillion-yuan aggregate picks one rate for nine months of flows. The seam matters anyway, because anyone who computes “the surplus in dollars” by multiplying the yuan figure by a spot rate will land somewhere between 930 and 960 billion dollars depending on the week they pick. The defensible statement is the customs series itself: about 938.1 billion dollars of surplus for the first three quarters. Anything more precise than that is a rate assumption wearing a number’s clothes.
What the ledger says against the plan targets
The customs release lands in the opening year of the 15th Five-Year Plan, and the plan’s headline numbers share the ledger’s weakness for compressed denominators. Three of them repay the same arithmetic.
About 3.8 percent. That is the 2026 annual target for cutting carbon dioxide emissions per unit of GDP, set in the government work report. Compound it: 0.962 to the fifth power is 0.824, so five consecutive years at 3.8 percent delivers a cumulative 17.6 percent reduction, comfortably clearing the plan’s binding 17 percent target for 2030 against 2025. The 3.8 percent itself is not arbitrary; the 14th plan period delivered about 17.7 percent cumulatively, which averages 3.8 percent a year, so the opening-year target extends proven pace rather than promising acceleration. The chain behind the 17 percent, from the 2022 carbon-intensity data through the Tsinghua estimate that 2025 landed near a 58 percent cumulative decline against 2005, leaves almost exactly the 17-point gap the plan commits to. The arithmetic is coherent at every link.
About 7 percent. That is the plan’s required average annual growth for nationwide research and development spending, with the plan’s growth target expressed only as a “reasonable range” and major analysts working from a 4.5 to 5.0 percent GDP assumption. R&D growing 7 percent against GDP growing 4.75 percent compounds to a relative gain of about 11 to 13 percent in R&D intensity over the plan period, depending on where GDP lands in the range. The range is wide because the denominator is a forecast, not a target; the plan deliberately left GDP unquantified, so intensity outcomes will inherit whatever growth actually happens. About 10 percent, separately, is the plan’s cut in energy consumption per unit of GDP, against the 17 percent carbon-intensity cut, and the gap between those two numbers is the plan’s implicit bet on grid decarbonization: carbon falls faster than energy only if the fuel mix changes.
Roughly 4 to 1. That is the ratio of trade growth to GDP growth in 2026, 18.8 percent against the 4.5 to 5.0 percent assumption. Trade has grown faster than GDP in most open economies, but a four-to-one multiple sustained for three consecutive quarters, with quarterly growth accelerating into the year rather than fading, raises the base-year question this article cannot resolve from the released figures: the 2025 comparison base would need its own full accounting to separate genuine acceleration from recovery arithmetic. The presser asserts the pace is the fastest for the period in five years, which implies the bases were not uniformly easy, but “fastest in five years” is a claim about rank, not about decomposition.
Who actually holds the ledger
The presser supplies one more decomposition that rarely makes it into the coverage: who the traders are. Private enterprises ran 22.73 trillion yuan of trade, up 19.1 percent, holding 56.9 percent of the total, which keeps them the largest trading entity in the country. Foreign-invested enterprises ran 11.69 trillion yuan, up 19 percent, and state-owned enterprises ran 5.46 trillion yuan, up 17.8 percent. Three ownership classes, three growth rates, and the spread between the fastest and the slowest is only 1.3 points, which is itself informative: the 18.8 percent total is not an artifact of one class surging while another shrinks. Roughly the same speed applies across all of them, so the concentration story told above is about products and partners, not about who owns the freight.
| Ownership class (Q1-Q3 2026) | Trade value | YoY growth | Share of total |
| Private enterprises | 22.73 trillion yuan | 19.1% | 56.9% |
| Foreign-invested enterprises | 11.69 trillion yuan | 19.0% | about 29% |
| State-owned enterprises | 5.46 trillion yuan | 17.8% | about 14% |
Source: computed shares from the customs press conference transcript 🔗, 14 October 2026
The geography row of the release splits the same way. Central region trade grew 20.5 percent and the western region 25.7 percent, both faster than the national total, which the presser attributes to the China-Europe freight trains and the New Western Land-Sea Corridor. A coastal-concentrated trade boom would look different in the interior lines than what was published; the interior is growing faster than the coast, or at least faster than the national average, which narrows one common objection, that the 18.8 percent is a port-volume illusion confined to a few coastal hubs.
Two more facts from the same release belong in any quotation of the headline. The first: exports of intermediate goods exceeded half of total export value, which the presser frames as the stabilizing anchor of global supply chains, and which is the structural reason the trade number moves with global manufacturing demand rather than with final consumer sentiment. The second: monthly growth stayed in double digits for six consecutive months through September, with the September single month at 5.16 trillion yuan, up 27.5 percent, and September exports alone at 3.03 trillion yuan, up 29.6 percent. Quarterly acceleration from 15.3 to 18.3 to 22.3 percent is therefore not a rounding artifact of one strong month; the September peak sits on top of a rising quarterly path. Any reading that treats the 18.8 percent as a fluke of the September print has to explain away five earlier months at the same order of magnitude.
Verdict: what holds and what does not
Everything customs published coheres. The implied bases sum correctly, the FTA increment share recomputes to 71.4 percent against the claimed “over seventy percent”, the monthly and quarterly acceleration figures stack, and the plan’s carbon arithmetic compounds to clear its binding target with margin. None of the four checked number groups contains a mathematical contradiction, and the one seam that exists, the yuan-dollar conversion spread, is a denominator choice rather than a mistake.
What the release does not advertise is composition. About 71 percent of the trade increment came from FTA partners. About two thirds of the growth came from one basket of tech-heavy goods, inside which integrated circuits doubled. The surplus, the number that dominates foreign-policy coverage, grew about 3.4 percent in yuan terms while the headline grew 18.8 percent, because imports outgrew exports by 5.9 points. A reader who takes only the headline will overestimate the surplus story and underestimate the concentration story, and both errors push in the same direction: toward treating a broad-based boom from a narrow-based one. The ledger supports the narrower reading. Whether the narrow base can keep compounding at 99.1 percent annual rates for integrated circuits is the question the fourth quarter will start answering.
Methodology
- Source hierarchy: customs presser and People’s Daily first, Xinhua second, plan analyses third.
- Denominators: all growth rates year on year, yuan terms, January-September 2026 unless stated.
- Derived values: 2025 base = 2026 value divided by (1 + growth); surplus = exports minus imports, yuan terms; increment shares against total increment.
- USD series treated as customs’ own dollar statistics, not conversions of the yuan series.
- Conversion checks: implied rate = yuan value divided by dollar value, across total, Q1-Q3 surplus and September surplus.
- Cross-validation: FTA increment share recomputed against the official “over seventy percent” claim; component bases summed against total base.
- Plan checks: the 3.8 percent carbon target compounded against the 17 percent binding cumulative target; R&D intensity growth computed across the GDP assumption range.
- Falsification: any recomputation changing a verdict, or a later base-period revision, overturns this audit.
Limitations
The 2025 base is recovered arithmetically and inherits rounding from four published figures. Surplus growth is computed, not published. “Fastest in five years” is accepted on rank authority. Plan-target arithmetic assumes annual rates persist uniformly.
Sources
- General Administration of Customs of China 🔗, Press conference transcript, imports and exports of the first three quarters of 2026, 14 October 2026 (in Chinese)
- People’s Daily Overseas Edition, via China Customs 🔗, Goods trade near 40 trillion yuan in the first three quarters, 16 October 2026 (in Chinese)
- Xinhua 🔗, China’s foreign trade maintains sound momentum in first three quarters, 14 October 2026
- Gov.cn 🔗, Reading the 2026 Government Work Report: main indicators of the 15th Five-Year Plan, March 2026 (in Chinese)
- Securities Times 🔗, Answering the 15th Five-Year Plan’s dual carbon-control exam paper, 13 November 2026 (in Chinese)
- Chinese Embassy in Singapore 🔗, National Day op-ed in Lianhe Zaobao on the 15th Five-Year Plan, 1 October 2026 (in Chinese)
About Global Infinity Labs
Global Infinity Labs is an AI innovation company specializing in MCP data services. This article was prepared by our expert team using insights and data provided through our MCP platforms.
Contact: support@globalinfinitylabs.com