A shared confusion between China and the U.S.: Can silicon-carbon move from divergence to win-win?

Core conclusion: ① Stocks: extreme divergence between silicon-based and carbon-based. The divergence in the U.S. stock market happened earlier, and the divergence in A-shares is more pronounced; the root cause is that silicon-based profits are clearly stronger than carbon-based. ② Real economy: the AI dividend has not yet benefited carbon-based. AI capital expenditure has become a bright spot in the U.S.-China economies; in China, the number of Token calls surged 81 times, but the consumer confidence index fell to the historical 5th percentile. ③ Outlook: silicon and carbon will eventually move toward a win-win. When AI moves from large models to physical applications, China’s intelligent manufacturing will bring an engineering talent dividend and advantages in computing cost; at that time, integration and win-win outcomes between the new and old economies will be achieved.

The most notable feature of the global economy and stock markets in the first half of this year is the divergence between silicon-based and carbon-based. This article takes the leading countries in this AI technology revolution—the United States and China—as the subjects, analyzes the underlying reasons behind the silicon-carbon divergence, and discusses the path toward a future win-win.

1. Stock market performance: ice and fire—two extremes

Judging by the stock price gains and losses of silicon-based and carbon-based categories, the U.S. market diverged earlier, while A-shares diverged even more drastically. Using the CSI 300 and S&P 500 constituent stocks as samples, divided into two groups—silicon-based and carbon-based—the divergence in their trajectories can be clearly seen: U.S. silicon-based names took the lead first, continuously outperforming carbon-based starting in early 2024, with divergence intensifying after 2025. Meanwhile, A-share silicon-based companies surpassed carbon-based only beginning in the second half of 2025, but then rapidly widened the gap, especially in 26Q2—ice and fire, two extremes.

Note: A-share silicon-based companies include listed companies in CICC’s top-tier industries such as communications, electronics, computers, electrical equipment, and new energy. Carbon-based companies include listed companies in 15 industries such as steel, building materials, construction, and transportation. Industries like large finance and machinery are classified as “intermediate.” U.S. silicon-based companies include listed companies in SIC second-tier industries such as industrial and commercial machinery and computer equipment, measuring and analyzing and control equipment, electronic equipment and components, communication services, and business services. Carbon-based companies include listed companies in 40 industries such as food and related products, general merchandise stores and other retail (various types of retailers), and oil and natural gas extraction. Industries like large finance and various manufacturing are classified as “intermediate.”

Industry gain/loss patterns also show that the divergence in A-shares is greater than that in the U.S. In the first half of 2026, among 30 A-share industries, half closed higher; the two major silicon-based industries—electronics and communications—were the leading drivers. The top gainer, building materials, was mainly boosted by AI-demand-driven shipments of electronic components/bill of materials. Of the 15 declining industries, 8 carbon-based industries saw declines exceeding 10%, and at the individual stock level, the proportion of stocks falling accounted for as much as 71%. In contrast, among the 54 SIC second-tier industries within S&P 500 constituent stocks, 35 industries rose in the first half; although silicon-based industries did not post standout gains compared with A-shares, their overall driving force for the broader market was stronger, pushing more than half of all U.S. market stocks to close higher.

In terms of trading concentration, both A-shares and the U.S. market highly cluster toward silicon-based. Specifically, using CSI 300 and S&P 500 constituent stocks to define silicon-based and carbon-based groups, as of June 30, A-share silicon-based companies accounted for 37.7% of the number of companies but contributed 63.1% of trading value; for the U.S., silicon-based companies accounted for 33.3% of the number of companies and achieved 67.9% of trading value. Among them, the heat for A-share silicon-based surged more rapidly. In the U.S., silicon-based trading share has remained above 50% for a long time since 2021; it gradually rose after 2023. But in A-shares, before 2023 the share was still below 35%; starting in 2025 it rose markedly, and in the first half of this year it quickly broke through 60%.

The divergence in silicon-based versus carbon-based stock prices stems from profit divergence. From listed companies’ financial reports, profit divergence is clearly evident in A-shares: silicon-based industries all achieved positive year-over-year growth in attributable net profit in Q1, with half growing by more than 30%, while most carbon-based industries lack sufficient growth momentum. In the U.S., the distribution of profits is more balanced: while silicon-based led the gains, many carbon-based industries also delivered strong performances. From industrial enterprise data, in China, between January and May, industrial enterprise profits for the computer, communication, and other electronic equipment manufacturing industries grew year over year by 103.9%, far above the overall industry figure of 18.8%; but profits in 20 carbon-based-dominated industries shrank by 15.4%. Meanwhile, in the United States, while silicon-based profits led in Q1, the carbon-based segment performed relatively steadily.

2. The real economy: uneven heat and cold

In the real economy, silicon-based and carbon-based are experiencing uneven heat and cold. In China and the U.S. economies driven by AI computing investment, the internal structure is splitting sharply: incremental demand mainly comes from silicon-based, while carbon-based demand is being suppressed and even substituted.

Token call volume surged, but consumer confidence is weak. As AI applications continue to deepen, real demand in 2026 has seen a blowout; currently, China’s Token call volume per week for large models is 19.8 trillion, up 81 times compared with the same period in 2025. In the United States, weekly call volume reaches 5.8 trillion, up about 3.9 times year over year. By comparison, consumer confidence in both China and the U.S. has fallen to historical lows: in May, China’s consumer confidence index was 89.9, at the historical 5th percentile; in the U.S., it was 44.8, a new low since the indicator was established.

The bright spots in GDP growth for both China and the U.S. come from AI capital expenditure. The U.S. has previously driven growth through consumption, but this year in Q1, AI investment’s contribution to GDP growth reached 85%, surpassing consumption. In China’s economic structure, external demand is strong while internal demand is weak. Meanwhile, export growth remained high mainly because it was driven by the U.S. AI capital expenditure; electronic products have averaged about an 11-percentage-point contribution to exports year over year, accounting for nearly 70% of the overall incremental export growth, becoming a core source of China’s high export growth as AI capital expenditure “spills over” outward.

The core of the silicon-carbon split is that the AI dividend is still circulating within silicon-based and has not yet spilled over into carbon-based. What AI reshapes first are enterprise production functions and capital returns; improvements in residents’ income and terminal demand lag behind. This leads to the current situation where silicon-based companies expand proactively, while carbon-based companies face pressure.

On one hand, the AI “arms race” itself requires continuous high-intensity spending and is still hard to spill over into the carbon-based economy. In profit distribution, technology companies are more inclined to put money into computing power, models, and data centers rather than immediately rewarding employees and shareholders. Taking the four major AI giants listed in the U.S. as an example, their expected capital expenditures in 2026 are projected to rise to about $604 billion, nearly two times compared with 2024, showing that capital spending in the AI era is clearly accelerating.

On the other hand, AI’s “substitution effect” keeps squeezing employment and income expectations, forming a “second blow” to carbon-based. The impact of the new economy on traditional sectors is not only reflected in enterprise profits and capital expenditures; it will also spread through employment. For example, in the U.S., employment in high-AI penetration industries such as technology and media, and manufacturing, has remained sluggish, and it is significantly negatively correlated with AI penetration rates. As employment and income expectations weaken, consumer confidence naturally cannot rise.

3. Outlook: How to turn divergence into win-win?

K-shaped divergence will not be the endpoint of the AI revolution. Looking back at prior technological revolutions, the new and old economies are not mutually exclusive; they shift from early conflict and substitution to eventual integration and win-win outcomes. Fundamentally, carbon-based represents people’s core needs, and the productivity boom brought by the silicon-based revolution ultimately must return to people’s needs. In the future, as AI costs decline and application scenarios become more diverse, the silicon-based dividend will expand demand broadly, pushing silicon and carbon from “divergence” toward “win-win.”

From a theoretical perspective, new technologies always first transform supply and then create demand. In the early stage, efficiency improvements shift the supply curve to the right; but as costs fall and scenarios expand, new demand gradually takes over, pushing the demand curve to the right and forming a new dynamic equilibrium with both quantity and price rising together. For example, in the early stage of the internet revolution, the U.S., as the source of the technology, benefited first; then infrastructure investment erupted, and the dividend spilled over along industry and trade chains. As the internet revolution accelerated globalization, after China joined the WTO in December 2001, it absorbed the dividend; China’s share of global GDP scale changed accordingly, and in 2003 China achieved a switch in momentum between the old and new drivers, with corporate ROE continuing to improve and China’s assets being comprehensively revalued.

When AI permeates from the digital world into the physical world, silicon-based and carbon-based will both win. At present, AI is still in the stage where computing power, models, and infrastructure lead. The U.S. is the origin and also the frontrunner. However, once AI shifts from large models to physical applications, with AI cars, AI phones, and robots accelerating in adoption, China’s ability to absorb the silicon-based revolution will become explicit.

First, China has a massive engineering talent dividend. In 2025, the average annual salary of engineers in China’s IT industry is about $34k, clearly lower than the U.S.’s roughly $180k. This means that under equal conditions, China can support larger engineering teams, more frequent product iterations, and faster manufacturing implementation.

Second, China has a huge advantage in computing cost. The price of mainstream AI large-model Tokens in China is only one-fifth or even lower than that of comparable models in the U.S. This means that when AI is deployed at scale, the call threshold is lower, which is more conducive to diffusion and widespread adoption. When competition shifts from pure computing power and algorithms to low-cost, high-frequency, large-scale applications, China’s advantage will be further amplified.

Intelligent manufacturing—represented by intelligent vehicles and robots—is expected to become the link between silicon-based and carbon-based, driving growth in residents’ income and consumption; at that time, incremental demand will spread toward carbon-based. Then, China’s economic transformation will move from quantitative change to qualitative change, and the scale of the new economy will exceed that of the old economy. The comparative analysis of the new and old economies has been covered in previous reports; see 《A comparison of the strength between the new and old economies: stock market and real economy dimensions》、《Logic of the PPI turning positive across the first and second halves and differences in stock market performance: drawing on 98-03》.

Sources: Xun Yugeng’s thinking

Risk notice and disclaimer terms

        The market involves risk; investments require caution. This article does not constitute personal investment advice, and it does not consider the special investment objectives, financial conditions, or needs of individual users. Users should consider whether any opinions, viewpoints, or conclusions in this article align with their specific circumstances. Invest accordingly at your own risk.