AI High-tech Boosts May Exports, But Growth Quality Questionable

Overview

JPMorgan released a research report interpreting China's May trade data, pointing out that under the strong shipment drive of AI-related high-tech products, May export performance far exceeded market expectations. However, the report also emphasized that the current trade boom has obvious structural characteristics, mainly driven by price increases in memory chips, data center equipment, and new energy products, rather than a widespread recovery in demand. This mode of 'price rise, stable volume' and extremely high product concentration makes the actual transmission effect of the trade surplus on domestic production and economic growth more complex and uncertain. In addition, although global demand remains resilient supporting the export outlook, trade policy uncertainty from the US and EU remains the main downside risk.

Core views

Export side shows significant structural differentiation and price-driven characteristics. May exports grew 19.4% YoY, far exceeding market consensus (15%) and prior institutional forecasts (12.1%). This beat performance was mainly driven by high-tech and electromechanical products, among which high-tech product exports grew for the seventh consecutive month, rising 8.0% sequentially after seasonal adjustment; mobile phones, integrated circuits, and automatic data processing equipment surged 19.4%, 15.2%, and 11.0% respectively. The report believes this reflects price inflation of memory chips amid tight supply and device demand brought by AI data center construction. In contrast, exports of low-end consumer goods such as textiles, clothing, and toys fell slightly 0.3% sequentially, showing export momentum is highly concentrated in the AI and new energy industry chain. Import rebound reflects supply chain stocking rather than comprehensive domestic demand recovery. May imports grew 27.5% YoY, expanding for the sixth consecutive month, also exceeding market expectations. Structurally, high-tech imports rose 5.9% sequentially, mainly driven by price effects of Korean memory chip prices; commodity imports showed differentiation, with coal, natural gas, soybeans, and copper import volumes increasing, reflecting clear strategic stockpiling signs, while crude oil and refined product imports continued to decline, partly due to digesting previous inventory. The report indicates that the current import strength is more likely the result of AI supply chain pull and commodity restocking rather than a rebound in broad domestic consumption demand. The macro implications of trade data have become more complex and full of risks. Since about 60% of export growth concentrates on a few categories such as storage IC, AI modules, EVs, PV, and batteries, and semiconductor exports have shifted from 'volume increase' last year to 'price increase' recently, it makes the contribution of net exports to GDP and employment no longer clear, volatility increased, and more sensitive to relative price and exchange rate changes. Looking ahead, although global capital expenditure spreading from AI infrastructure to non-tech fields provides external demand support, expected full-year nominal export growth rate can reach high single digits or even over 10%, US trade policy focus is shifting to 301/232 clause investigations, EU defense posture against China trade is strengthening, these policy shocks constitute asymmetric downside risks, which may lead to slowdown in specific industry exports and supply chain restructuring.

Analysis framework

The report adopted a refined 'Volume-Price Split' and 'Structural Attribution' analysis approach. The institution did not stop at the total beat, but delved into breaking down the source of export growth, distinguishing between 'volume-driven' and 'price-driven', identifying that current growth is mainly dominated by the price cycle of specific products such as memory chips. Meanwhile, by classifying export products into 'High-tech/AI-related', 'New Three Items', and 'Low-end Consumer Goods', and cross-verifying with import side's 'Supply Chain Stocking' and 'Domestic Consumption Demand', the real mapping degree of trade data to the real economy is judged. In addition, the institution placed trade trends under a geopolitical framework assessment, focusing on the nonlinear impact of tariff tool switching (such as from IEEPA to Section 301) on specific industrial chains.

Methodology notes

• Industry/Industrial Analysis FrameworkVolume-Price Split

  Decompose the growth of total import and export amounts into volume effects and price effects to identify the real drivers of growth.

  In this research report, the institution determined that May exports exceeded expectations mainly due to 'price increases' rather than 'surges in physical volume' by observing price trends and export value changes of products such as memory chips. This method helps investors remove price noise and more accurately assess the actual intensity and sustainability of external demand.

• Macroeconomic frameworkCredit/debt cycle

  Analysis of the effectiveness of the trade transmission mechanism, i.e., the efficiency of trade surplus conversion into domestic production and GDP.

  The report points out that when trade growth is overly concentrated in a few high-priced products (such as AI chips), its pulling effect on broad manufacturing production and employment will weaken. This indicates that investors, when analyzing macro data, cannot simply apply 'export-GDP' elasticity coefficients from historical experience, and need to dynamically adjust expectations combined with industrial structure changes.

Key data

• May Export YoY Growth Rate19.4%Significantly higher than market consensus 15% and JPMorgan forecast 12.1%
• May Import YoY Growth Rate27.5%Expanded for the sixth consecutive month, higher than market consensus 26%
• May Trade Surplus1054 Hundred Million USDYTD cumulative 4530 Hundred Million USD, same period last year was 4700 Hundred Million USD
• High-tech Product Exports Sequentially+8.0% (Seasonally Adjusted)Grew for 7 consecutive months, Mobile +19.4%, Integrated Circuits +15.2%
• Automobile Exports YoY Growth Rate39.3%Maintained strong growth momentum

Impact & implications

For macroeconomy, the surface strength of trade data masks internal structural fragility, meaning the model relying solely on external demand to drive economic growth faces the challenge of diminishing marginal utility. For related industries, AI and semiconductor industry chains will still benefit from global capital expenditure expansion and price upward cycle in the short term, but need to be wary of precision strike risks brought by US-EU trade policy shifts; traditional low-end manufacturing faces continuous pressure from weak external demand. For asset allocation, the report suggests reducing linear extrapolation of total data, focusing more on structural opportunities and policy hedges, especially in the context where US-China-EU trade friction may intensify, the importance of supply chain security and domestic substitution logic further rises.

Risks

• US trade truce agreement not extended, and tariff focus shifted to 301/232 clause investigations
• EU stance on China trade hardened, potentially introducing stricter market access restrictions or countervailing measures
• Export growth overly dependent on a few high-tech products and price factors, weak anti-risk ability
• Ongoing Middle East conflict may disrupt regional oil markets and commodity supply chains

What to watch

• Officially upcoming detailed split data of import/export prices and quantities
• Specific progress of US 301/232 clause investigations and changes in tariff exemption lists for China
• Implementation status of EU trade defense measures against Chinese EVs and strategic industries
• Recovery strength of global non-tech capital expenditure and pace of inventory rebuilding