May Foreign Trade Exceeds Expectations: AI and Chip Prices Drive Growth

Overview

Nomura released a report interpreting China's foreign trade data for May, stating that both import and export growth rates accelerated and exceeded market expectations. The main drivers were the boom in semiconductor trade under the global AI super cycle and the amplification of price effects. Although exports of traditional labor-intensive products remained weak, the substantial price increases in high-tech products, especially integrated circuits, supported nominal growth. Meanwhile, exports to the US saw a significant rebound due to tariff policy adjustments and a low base effect. The report emphasizes that the current high trade growth has distinct structural characteristics, namely 'rising prices with stable volumes' or even 'rising prices with falling volumes,' requiring a distinction between nominal prosperity and actual physical demand.

Core views

AI and semiconductors are the absolute core engines. The report calculates that AI-related exports (integrated circuits + automatic data processing equipment) collectively contributed 9.3 percentage points to total export growth in May, accounting for nearly half. This level has been maintained for two consecutive months, highlighting the huge拉动 role of the global AI super cycle on China's foreign trade. Specifically, the value of integrated circuit exports increased by 110.9% YoY, while volume grew by only 2.1%, meaning that more than 106 percentage points of growth were entirely driven by price increases. The same pattern applies to imports: IC import value grew by 68.0% while volume turned negative, with price contributions reaching 69.8 percentage points, reflecting the transmission of Q2 memory chip contract prices. The strong rebound in exports to the US stems from the叠加 of three factors. The YoY growth rate of exports to the US jumped from 11.1% in April to 37.3% in May, reaching a five-year high. The report attributes this to three points: first, the extremely low base caused by the tariff war in the same period last year; second, the substantive benefit brought by the decline in the actual effective tariff rate after the US Supreme Court overturned the IEEPA tariffs at the end of February; and third, the large amount of related demand generated by the local AI boom in the US. At the same time, imports from the US also rebounded due to a low base, but the trade surplus with the US further expanded to USD 26 billion. The import side shows a divergent pattern of 'strong chips, weak energy.' Although the value of crude oil imports maintained double-digit growth due to supply disruptions in the Strait of Hormuz and rising oil prices, the physical volume plummeted by 29% YoY, contributing only 1.9 percentage points to total import growth. In contrast, integrated circuits alone contributed 10.8 percentage points to import growth, confirming that semiconductor price shocks, rather than energy, were the primary force pushing up import data. Additionally, the growth rate of imports from South Korea further climbed to 84.3%, indirectly corroborating the strong momentum of technology supply chain imports. Traditional and emerging industries show a stark contrast. While electronic product exports soared, exports of labor-intensive products, despite marginal improvement, remained in negative territory. Apparel, footwear, luggage, and toys fell by 4.1%, 10.3%, 4.9%, and 7.0% YoY, respectively. Among transportation equipment, automobile exports maintained a high level of 39.3%, and ship exports rebounded significantly, but the growth rate of auto parts slowed down. This divergence indicates that current foreign trade growth is highly dependent on technology-intensive industries, while external demand for traditional manufacturing remains under pressure.

Analysis framework

The report adopted a typical 'volume-price decomposition' analysis method, not satisfied with the nominal value growth rates published by Customs, but decomposing exports and imports into two factors: 'volume' and 'price.' By comparing them, it was found that the current high growth is mainly contributed by prices (especially chip price hikes) rather than the expansion of physical volumes, thereby revealing the special structure of trade data under the AI cycle. In addition, the report used the 'growth contribution decomposition' method, weighting and decomposing total export/import growth rates by category (such as ICs, ADPs, crude oil, etc.) to quantify the pull points of each sub-item on the overall data. This method can precisely identify who the real growth engine is (e.g., AI-related categories contribute more than half), avoiding structural problems being masked by aggregate data. In analyzing US-China trade, it combined the 'base effect' and 'policy event-driven' frameworks. It not only looked at the numerical changes in YoY growth rates but also traced back the low base caused by the tariff war in the same period last year, and tracked changes in effective tax rates at specific policy nodes such as the US Supreme Court ruling and the implementation of Section 122 tariffs, thereby explaining the institutional reasons for short-term data fluctuations.

Methodology notes

• Industry/Sector Analysis FrameworkVolume-price decomposition

  Decomposing trade value growth into volume growth and price growth

  By comparing the YoY differences between value and volume, the report found that chip export value increased by 111% while volume increased by only 2%, thus judging that current growth is mainly driven by price hikes rather than an explosion in physical demand. This is a core method when analyzing inflation or cyclical industries, helping investors distinguish between 'real demand' and 'price bubbles.'

• Macroeconomic framework

  Contribution to Growth

  Decomposing overall growth rates into specific percentage point contributions based on the weight of each component. For example, calculating that ICs contributed 5.9pp to export growth, accounting for 30% of the total growth rate. This method allows readers to intuitively see which sector is the real 'locomotive,' avoiding being misled by averages.

• Event Gaming and Behavioral FinanceExpectation Gap/Expectation Management

  Analysis of deviation between actual data and market consensus expectations

  The report begins by comparing the actual value (exports 19.4%) with consensus expectations (15.0%), establishing an 'exceeding expectations' tone. In macro data commentary, expectations already priced in by the market are often more important than absolute values, and the magnitude of the surprise determines the direction of short-term asset price fluctuations.

Key data

• May Export YoY Growth (USD)19.4%Higher than market consensus of 15.0% and Nomura's expectation of 13.7%, further accelerating from 14.1% in April
• May Import YoY Growth (USD)27.4%Higher than market consensus of 26.0%, accelerating from 25.3% in April
• AI-Related Export Contribution9.3ppIntegrated Circuits (5.9pp) + ADP Equipment (3.4pp), accounting for about half of total export growth
• Integrated Circuit Export Volume-Price PerformanceValue +110.9% / Volume +2.1%Price contribution reached 106.5pp, with the degree of volume-price divergence further expanding compared to April
• Exports to US YoY Growth37.3%Highest since March 2021, compared to 11.1% in April
• May Trade SurplusUSD 105.4 billionSignificantly expanded from USD 84.8 billion in April

Impact & implications

The report believes that China's current strong foreign trade performance is highly bound to the global AI capital expenditure cycle. The increased pricing power of semiconductors and related equipment has allowed China to gain huge benefits in nominal trade, but it also means that once global chip prices correct or AI investment slows down, trade data will face the risk of a rapid decline. The phased warming of exports to the US is encouraging, but it stems more from the one-time bonus of tariff legal framework adjustments and the rush-to-export effect, rather than a signal of fundamental easing in US-China economic and trade relations. For the market, caution should be exercised regarding the weak physical volumes behind nominal high growth, paying attention to the potential drag of insufficient external demand for traditional manufacturing on employment and domestic demand.

Risks

• A slowdown in global AI investment or a correction in chip prices could lead to a cliff-like drop in trade growth rates
• US Section 122 tariffs are set to expire on July 24, and subsequent policy uncertainty may trigger export fluctuations in the second half of the year
• Continued supply disruptions in the Strait of Hormuz may further suppress the physical volume of crude oil imports
• Continuous negative growth in labor-intensive product exports reflects that traditional external demand remains weak

What to watch

• The衔接 of US tariff policies towards China after July 24 and the implementation of new rounds of Section 301 tariffs
• Results of Q3 contract price negotiations for DRAM/NAND and their transmission to import prices
• Progress in reciprocal tariff reductions for approximately USD 30 billion worth of non-sensitive goods mentioned by the Ministry of Commerce
• Whether there will be a new wave of rush-to-exports in the second half of the year due to expectations of tariff hikes