SINGAPORE – The ongoing craze over artificial intelligence is sending technology-company valuations skyrocketing in the US and Asian markets. One country still offers some relative bargains: China.

A parallel universe of Chinese technology champions exists in stock markets: an e-commerce giant that isn’t Amazon, a search-engine and robotaxi company that isn’t Google and a chip maker that has a four-letter name other than TSMC.
While some smaller Chinese tech companies have experienced a rapid boom this year, shares of larger companies that are investing heavily in AI often have lower revenue or profit per dollar than their better-known counterparts overseas.
“This is an attractive opportunity,” Eva Li, head of Greater China equities at Swiss bank UBS, said in a recent note. The top companies in the AI sector are at “historically low valuations,” he said.
Beijing is supporting the AI industry with favorable policies and financial incentives, trying to create a tech ecosystem that is independent from Washington.
“They’re in a different cycle,” said Alvin So, equity strategist in Asia for Goldman Sachs. He observed that while the launch of ChatGPIT in 2022 fueled the US generator-AI boom, China’s market only found its footing early last year, when DeepSeek released a large language model that proved local companies could compete on a global scale.
Alibaba has integrated its Quon AI models into its e-commerce platform and is investing $50 billion in cloud infrastructure over the next few years — even designing its own AI chips, like Amazon and Google.
Alibaba trades at a forward price-to-earnings ratio of 17, meaning investors are paying $17 for every dollar of earnings expected next year. The figure for Amazon, which runs a similar collection of businesses, is more expensive at 27. “We see Alibaba as the global AI winner,” Morgan Stanley wrote in a recent note.
The risk is that Chinese companies remain largely trapped in their own market, which is in an economic recession. Despite being huge, China alone cannot match the global empire that companies like Amazon have built.
Unstable geopolitics is a challenge. The Pentagon this month updated its list of Chinese businesses it says work with Beijing’s military, designating nearly two dozen new companies, including Alibaba. This potentially limits their operations in the US. Alibaba and other companies said they are not linked to the Chinese military and are not on the list.
Buying Chinese AI stocks is not intuitive and sometimes even impossible for US investors. China’s Semiconductor Manufacturing International, known as SMIC, is on the US blacklist and Americans cannot buy its shares. Shares of SMIC’s larger rival, Taiwan Semiconductor Manufacturing, or TSMC, are available through American depositary receipts traded on the New York Stock Exchange.
Some Chinese companies are traded in Hong Kong, whose stock exchange is relatively open to global investors, and some, including Alibaba, are traded on the New York Stock Exchange. Foreign investors can also buy Shanghai- and Shenzhen-listed shares through a regulated channel involving brokers in Hong Kong.
More Chinese initial public offerings are coming that will provide investors with the opportunity to get in on the ground floor. Memory-chip maker CXMT won approval last month to raise about $4 billion in a Shanghai IPO, potentially valuing the company in the billions of dollars. It enters a market dominated by global giants like Samsung, SK Hynix and Micron – which have recently topped $1 trillion in market value due to the AI chip boom.
Goldman Sachs research shows global investors are less risk averse to China. According to Goldman, while the country accounts for 10% of global AI-related market capitalization, global mutual-fund managers allocate only 1.2% of their global tech portfolios to Chinese AI equities.
For investors who already have exposure to U.S.-focused AI leaders, Chinese stocks “offer different exposures,” Goldman’s So said.
Some Chinese stocks are already mimicking the US enthusiasm for AI, pushing their valuations well beyond bargain territory. Zipu, which is developing a larger language model to compete with companies like OpenAI and Anthropic, listed in Hong Kong in January and its share price is now nine times its initial level.
Shanghai-listed Cambricon Technologies designs AI chips, competing with Nvidia. Its shares have tripled in the past year and trade at a forward price-to-earnings ratio of 128, compared to Nvidia’s 23.
However, other companies with a keen interest in AI seem hardly enthused. Robotaxi leader Baidu trades at just 14 times expected earnings. Tencent, owner of the WeChat app, which has more than a billion users in China, is testing an AI agent for the app that could help users manage payments, food delivery and taxi-hailing more easily. Its shares trade at 13 times forward earnings.
Chinese battery giant CATL, which investment banks say is likely to be a beneficiary of power needs for AI data centres, is trading at 19 times earnings on the Shenzhen Stock Exchange.
Write to Rory Jones at Rory.Jones@wsj.com
