Goldman Sachs China AI In-Depth Report: Why Is China's AI an Independent Market? Where Is the Main Capital Flow Heading?

Goldman Sachs, in its latest series of reports, states that the potential economic benefits brought by China’s AI are far from being fully priced by the market and are forming an independent trend separate from global tech stocks.

According to ZhuaFeng Trading Platform, the report indicates that since the “DeepSeek Moment” in January 2025, Chinese AI stocks have risen an average of 50%, with related tech stocks adding over $3 trillion in market capitalization. Analysts believe that current valuations still underestimate the potential value creation of AI by 50% to 100%, meaning the structural opportunities are far from exhausted.

From an allocation perspective, global fund managers hold only 1.2% of their global tech portfolios in Chinese AI stocks, far below China’s approximately 10% market share and 16% revenue share in the global AI market. In the global competitive landscape, China has significant comparative advantages in power, infrastructure, and physical AI, becoming an important diversified source in global AI investment portfolios.

At the policy level, AI aligns with China’s core goal of technological self-reliance, and related stocks are expected to see earnings growth that significantly outperforms non-AI assets. Given the attractive current valuations, Goldman Sachs believes that allocating to AI has become a necessary strategy to hedge against risks of disruption in the real economy and traditional industries.

DeepSeek Sparks the Fire: AI Has Completely Changed the Game Over the Past Year

Goldman Sachs states that since the release of DeepSeek R1 in January 2025, China’s AI applications have entered a substantial acceleration phase. The significant reduction in model inference costs has accelerated technological penetration, and AI has rapidly become the absolute mainline of China’s stock market. During this period, Chinese AI-related stocks increased by an average of 50%, with 103 AI companies completing IPOs in Hong Kong and Mainland China. The tech sector added over $3 trillion in market cap, with AI stocks contributing about $3.4 trillion of that increase.

Meanwhile, China’s model capabilities have established a foothold in global competition. Large language models from companies like DeepSeek, Alibaba, and ByteDance have ranked highly in multiple global benchmarks, positioning China among the ranks of globally competitive AI model providers. The latest survey at Goldman Sachs Asia-Pacific Global Macro Conference shows that 68% of investors believe AI is the best investment theme for 2026, far ahead of consumer, global mobility, and dividend strategies.

Redefining China’s AI Value Chain: A $10 Trillion Ecosystem

Faced with rapidly evolving technological landscapes, Goldman Sachs has redefined the investment boundaries of Chinese AI stocks through three cross-validated methodologies.

First, based on mapping the global AI supply chain, Goldman Sachs classifies over 700 leading AI companies in the US, North Asia, and Europe into 29 industries and five major themes—power, semiconductors, infrastructure, models, and applications—and maps Chinese listed companies with AI-related revenue into corresponding industries. This results in a global AI stock universe covering 3,715 companies with a total market cap of $36 trillion, accounting for about 25% of the global total.

Building on this, Goldman Sachs introduces a revenue classification framework based on 2024 financial data to objectively quantify each company’s actual AI exposure. The results show that over 3,000 Chinese listed companies have traceable AI-related revenues, totaling about $10 trillion in market cap, with roughly half directly related to the AI value chain. Industry-wise, AI revenue exposure is highest in software at 84%, electronics manufacturing at 60%, while healthcare, mining, and other sectors are below 5%.

Finally, through bottom-up industry insights, Goldman Sachs analysts have identified 29 AI-enabled and enabling industries, and project that by 2035, the total potential market size of Chinese AI companies could expand to $16 trillion.

Four Investment Insights on Chinese AI

Goldman Sachs points out in the report that Chinese AI stocks are facing significant structural misallocations. China accounts for 10% of the global AI market cap, contributes 16% of related revenue, and nearly 20% of R&D investment, yet global funds allocate only 1.7% of their holdings to Chinese stocks, with only 1.2% of their global tech allocations in Chinese AI tech stocks. This huge gap suggests that once global investors begin correcting this misallocation, the potential capital inflow could be substantial.

From the perspective of global value chain division of labor, both the US and China have their advantages. The US dominates in semiconductors, AI models, and digital applications, while China’s comparative advantages are concentrated in power, infrastructure, and physical AI, accounting for 38%, 26%, and 27% of the global AI revenue pool respectively. Holding Chinese AI stocks can provide global investors with differentiated exposure and effective diversification returns.

Heterogeneity in returns is also increasing. Since the DeepSeek release in January 2025, Chinese AI stocks have outperformed their US counterparts by 30%, and North Asian AI stocks by 21%. More importantly, the 52-week rolling return correlation between Chinese AI stocks and US/global tech stocks is only 23%, far below the 69% level of the US and other regions, indicating that Chinese AI has formed an investment theme independent of US market trends.

Industry rotation is accelerating. From a global perspective, AI leadership is shifting from semiconductors to power and infrastructure, reflecting market focus moving from computing power to supply bottlenecks. In China, infrastructure sectors have consistently performed strongly during both rounds of ChatGPT and DeepSeek rallies, reflecting the country’s competitive advantage in tech hardware manufacturing.

Valuation Reassessment: Why Chinese AI Is Not a Bubble

Goldman Sachs explicitly states in the report that Chinese AI stocks are far from entering bubble territory, and current valuations may underestimate the potential value and profit creation of AI by 50% to 100%. This conclusion is based on macroeconomic estimates, total addressable market (TAM), and corporate profitability analyses, pointing to a common theme: the economic value generated by AI far exceeds what current market valuations reflect.

On the macro level, generative AI is expected to boost China’s labor productivity by 8% cumulatively over ten years, translating into about $1.6 trillion in economic added value. Discounted to present value, the total economic benefit could reach $6 to $7 trillion, with approximately $3 trillion potentially accruing as capital income to Chinese companies.

At the industry level, by 2035, Chinese companies’ revenues in 21 specific AI-related industries could grow to $16 trillion. Discounted at a 15% net profit margin and 10% cost of equity, the potential profit pool’s present value is about $2.4 trillion.

On the corporate earnings front, widespread AI applications are expected to improve profitability through cost savings and new market development, increasing annual profits of Chinese companies by 3 percentage points over the next decade. For listed companies overall, this implies about 6 percentage points of incremental profit contribution, with a present value of roughly $800 billion.

Goldman Sachs believes that compared to these potential profit and value increments, the net growth of China AI market cap since the DeepSeek moment appears relatively moderate, further supporting the view that valuations are not yet overextended and the AI main theme still has room to rise.

The Cost of Missing Out: Four Risks of Not Investing in AI

Goldman Sachs’s latest report highlights that in the current era of AI reshaping global industries, not investing in Chinese AI itself is a risk that warrants careful assessment.

First is the risk of misallocation. China accounts for 10% of the global AI market cap, 16% of related revenue, and nearly 20% of R&D investment, yet global funds allocate only 1.2% of their tech holdings to Chinese AI stocks. This systemic underweighting means that once global capital begins correcting this bias, the cost of missing the rally could far exceed the valuation risks.

Second is the risk of structural misjudgment. The market tends to compare Chinese AI with the US, but their advantage tracks are fundamentally different. China’s strengths lie in power, infrastructure, and physical AI, capturing 26% to 38% of the global revenue pool. Using Silicon Valley-style stock-picking logic to select Chinese stocks could cause investors to miss the sectors where China truly holds advantages.

Third is the risk of growth divergence. Since DeepSeek’s release, Chinese AI stocks have outperformed US peers by 30%, but within the sector, returns are highly differentiated. Infrastructure sectors have remained strong across both rallies, while application sectors lag due to unclear monetization paths. Simple bets on “AI concepts” are unlikely to be effective.

Fourth is the risk of valuation lag. Goldman Sachs estimates that the potential economic value driven by AI’s efficiency gains and new profit streams exceeds current market expectations by 50% to 100%. Over the next decade, earnings growth of Chinese AI companies could outperform non-AI peers by 140 percentage points, yet the market’s pricing of this growth remains conservative.

How to Strategically Invest in Chinese AI?

Using TAM forecasts, Residual Income Model (RIM), and industry fundamentals, Goldman Sachs systematically evaluates the risk-return profile of various Chinese AI themes. The report notes that market implied growth expectations for power and infrastructure are relatively conservative, with model implied EPS compound annual growth rates of about 3% and 11%, respectively, versus TAM forecasts of 23% and 31%, indicating significant underestimation.

In contrast, application sectors—especially consumer services, healthcare, and autonomous driving—have embedded high growth expectations, with narrower valuation safety margins. The AI model sector’s current P/E ratio is only 17x, but as new IPOs enter the market, awareness is gradually increasing. The semiconductor sector remains globally competitive, though growth paths are relatively steady.

To hedge against AI disruption risks, Goldman Sachs recommends focusing on industries with the following features: high tangible asset ratios, clear AI revenue exposure, strong R&D and capital expenditure, and high entry barriers—such as sectors with high state-owned enterprise presence. Based on a comprehensive ranking, segments like wafer foundry, semiconductor equipment, and optical modules perform best on these dimensions.

This insightful content is from ZhuaFeng Trading Platform.

For more detailed analysis, real-time insights, and frontline research, please join 【**ZhuaFeng Trading Platform ▪ Annual Membership**】

![](https://img-cdn.gateio.im/social/moments-6d472b75c9-9b9860f41f-8b7abd-ceda62)

Risk Disclaimer and Terms of Use

Market risks exist; investments should be cautious. This article does not constitute personal investment advice and does not consider individual user’s specific investment goals, financial situation, or needs. Users should determine whether any opinions, viewpoints, or conclusions herein are suitable for their particular circumstances. Investment is at their own risk.