Chinese AI Stocks Have Been On a Tear—Why US Investors Have Been Left Out so Far
China’s AI rally has been concentrated in mainland-listed stocks, which are difficult for US investors to access.

Key Takeaways
- Beneath China’s lackluster stock market is a boom in AI and related names.
- Analysts say US trade restrictions have pushed China to develop a homegrown chip industry for its AI ecosystem.
- One strategy to access China’s AI prowess may be to go through its beaten-down internet giants, according to Morningstar analysts.
US investors looking to ride China’s artificial intelligence boom have been continually frustrated. Unlike in the US stock market, the largest Chinese stocks, such as Alibaba BABA and Baidu BIDU, have struggled despite being internet leaders. Meanwhile, the overall Chinese economy is in the doldrums. Over the past 12 months, the Morningstar China All Cap Target Market Exposure Index has lost 12.2%, and the MSCI China Index is down 12.3%.
Meanwhile, China’s tech and AI plays are up 25.5%, as measured by the SSE STAR 50 Index (little followed in the United States), which tracks science- and technology-focused equities. So why have China-focused investments done so poorly for US investors while the country’s high-tech stocks have thrived?
China’s Stock-Market Divergence
The divergence is partly due to China’s restrictions on foreign investors and the industries global investors are exposed to. China restricts foreign ownership of A shares (mainland-based companies that trade in renminbi on domestic exchanges). US investors can still purchase some A shares on the Hong Kong exchange, which has relationships with the Shanghai and Shenzhen exchanges. As a result, the MSCI China Index has just 20% in China and 80% in Hong Kong and US-listed shares. Meanwhile, Hong Kong is heavily exposed to Chinese internet companies like Tencent 00700 and Meituan 03690, as well as companies exposed to Chinese consumption.
China’s real estate bust, which began in 2020, when the government moved to curb speculation and excessive leverage, has weighed on household wealth and confidence, contributing to a dramatic slowdown in consumption. Consumer confidence in China peaked in February 2021 and has plunged to its lowest levels in 30 years. In July, retail sales grew by a weaker-than-expected 0.6% year over year, down from 1.0% in June.
“Consumption and internet names have dragged down the Morningstar China Index,” says Phelix Lee, who covers semiconductors for Morningstar in Hong Kong. “The two largest pillars are both selling off.”
Matthews Asia portfolio manager Tiffany Hsiao says, “There’s the same K-shaped economy in China as in the US. Most people haven’t seen real wage growth, and property markets aren’t doing well.”
How China’s Economic Woes Hurt Its Internet Giants
For China’s internet giants, the consumption bust has boosted competition and hurt advertising revenues, distracting leaders from developing AI strategies. In contrast, AI plays in the US have been built around enterprise solutions, including those provided by Anthropic and Palantir PLTR, notes KraneShares chief investment officer Brendan Ahern. KraneShares manages several China funds, including one focused on internet stocks and another focused on AI names.
Beijing directed domestic tech companies to curb purchases of foreign semiconductors to bolster Chinese industry. “China has to build the machines to build the semiconductors before building a robust AI model,” explains Matthews’ Hsiao. “But recreating the wheel is an exciting opportunity in itself.” That entails chip equipment, chip manufacturing, and building robust clouds. That decision was partly geopolitical; US export restrictions on Nvidia’s NVDA advanced AI chips helped spur China’s push for self-sufficiency.
You can see the enthusiasm around China’s homegrown AI investments in the so-called Star 50 Index. It’s part of the Star board, a subset of the Shanghai Stock Exchange. The Chinese government created the Star market in 2018 to boost domestic innovation, including chip design and manufacturing, software, healthcare technology, and automation.
The $395 million KraneShares China Technology & Semiconductors STAR50 ETF KSTR, which tracks the Star 50, has risen more than 30% over the past 12 months. Meanwhile, the much larger $4.8 billion KraneShares CSI China Internet ETF KWEB, which counts Tencent and Alibaba among its largest holdings, is down more than 30%.
The Star 50 Index has an array of chipmakers. The largest components include Cambricon Technologies 688256 at 10.1%, Hygon Information Technology 688041 at 9.2%, Semiconductor Manufacturing International 00981 at 9.1%, Montage Technology 688008 at 7.4%, and Advanced Micro-fabrication Equipment 688012 at 6.2%. Chip design company Cambricon surged 28.2% over the past 12 months, contributing 6.5% of the index’s gain. Chip equipment maker Advanced Micro jumped 143%, accounting for 6% of the index’s gain. When tech stocks globally plunged in July, the Chinese government reportedly stabilized stocks by buying ETFs, including those connected to the Star 50.
China’s most popular stocks are expected to join the Star 50 Index. One is chipmaker CXMT 688825, which went public in July and is one of China’s most valuable companies. Apple AAPL is seeking White House approval to use chips from CXMT in its devices sold in China. CXMT currently fetches a 147% premium to Morningstar’s fair value estimate. Another is Unitree Robotics, a humanoid robot maker whose shares surged 460% when it went public in August.
What about private AI labs DeepSeek and Moonshot? They’ve released powerful large-language models that require less computing power, which sank AI stocks in the US market. When DeepSeek and Moonshot go public, which is expected as soon as late 2026, they’re expected to join the STAR market too. In July, Moonshot launched its globally competitive model Kimi K3, raising concerns that US firms were overspending on AI and that Chinese AI developers and chipmakers would undercut semiconductor stocks like SK Hynix 000660 and Samsung Electronics 005930.
The tech companies in the STAR 50 are also benefiting from overseas demand, says KraneShares’ Ahern. “Why are Chinese exports so strong? A big part of the exports are these pick-and-shovel AI inputs. While Dell servers are made in the US, a lot of inputs come from China.”
For those wanting exposure to China’s AI prowess, the names on the Star 50 are “very overvalued,” says Morningstar’s Lee. He notes that customers will continue demanding lower AI prices, and Beijing “wants low-cost AI to proliferate, [which] will constrain pricing power.” At the same time, US trade restrictions on cutting-edge technology “force companies to pursue workarounds that are more expensive and lead to lower yields or lower production reliability.”
A Beaten-Down Opportunity in China’s AI Space
Ironically, the best way to access that AI prowess may be China’s internet giants, which have come to grips with the AI opportunity. Internet stocks such as Tencent, e-commerce giant Alibaba, and search engine marketing conglomerate Baidu have been under pressure since a crackdown on anticompetitive behavior began in 2020, when the government forced Alibaba affiliate Ant Group to scrap its Hong Kong and Shanghai dual listing. Today, they’re cheap—Tencent, Alibaba, Baidu and others are repurchasing shares.
Last year, Alibaba spent $1.6 billion to subsidize its Super Saturday food delivery promotions, notes Hsiao of Matthews, who estimates the food delivery business lost Alibaba about $5 billion in 2025. “Now, it’s putting money into next-generation semiconductors for AI infrastructure,” she adds. “They’re taking AI very seriously.”
Tencent is another likely winner, thanks to its user ecosystem. Tencent trades at a 45% discount to Morningstar’s fair value estimate of $99 per ADR. For AI, “monetization will be on the application side,” says Morningstar analyst Ivan Su. The spate of new LLMs means “people have no switching costs. What can ultimately capture attention is the ecosystem.” Tencent has integrated AI into its WeChat app, with which users manage their personal, social, and professional relationships, making it difficult to switch to other platforms. It’s using AI to better match ads to users and accelerate game updates, which all encourage users to boost spending.
The author or authors do not own shares in any securities mentioned in this article. Find out about Morningstar’s editorial policies.
