By Li Gu and Jiaxing Li
SHANGHAI/HONG KONG (Reuters) -Investors are expected to remain hungry for China's Cambricon shares, with a looming reshuffle in a key tech index that could lead to over $1 billion worth of passive selling unlikely to shake their confidence in the AI bellwether.
Shares of Cambricon, an AI chip designer and dubbed China's answer to Nvidia, more than doubled in August, boosting its weight in the tech-focused STAR50 Index beyond the 10% cap for an individual stock.
A quarterly index
reweighting after the market close on Friday means Cambricon faces up to 8 billion yuan ($1.1 billion) in passive outflows from funds tracking the index, according to Reuters estimates.
Cambricon's shares fell 14% last week on profit taking and worries about rebalancing but have since bounced back, gaining 10% this week and flirting with record highs.
Cambricon is trading at 521 times trailing earnings, compared with a multiple of 50 for Nvidia, according to LSEG data.
But fund managers and analysts say the weighting adjustment is unlikely to derail the broader Chinese AI rally, betting that Beijing's tech self-sufficiency drive will turbocharge sector growth and replicate the U.S. AI boom.
"Maybe some investors will use it as a reason to take profit, but I don't think that will affect the long-term trend" or positioning by fundamental-based active managers, said Shihao Li, research analyst at CLSA.
The rally in AI stocks, a key pillar of China's stock market bull run, has been driven by Beijing's support for home-grown innovation, the DeepSeek breakthrough, and large AI investment by tech behemoths Alibaba, Tencent and Baidu.
The global AI frenzy has seen tech-heavy Nasdaq hit record highs through the year powered by Nvidia's 32% gain. Cambricon has surged 113% in 2025 with most of the gains coming last month after it posted a first-half profit.
"There is now more optimism that China's AI industry has passed a turning point, and entered a self-sustaining cycle of rising investment and higher profitability" said Tilly Zhang, analyst at Gavekal Dragonomics.
"The narrative that China can replicate at least part of the U.S. AI boom no longer seems so far-fetched."
Cambricon Technologies, a Beijing-based chipmaker focused on artificial intelligence, was founded in 2016 and listed on Shanghai’s STAR Market in 2020. The niche firm has been catapulted into the limelight due to its huge share surge.
CHINA BULLS
China's tech stock rally has extended into a twelfth month, with onshore equities largely catching up to offshore counterparts over the past few weeks and lifting the broader Shanghai stocks to levels not seen in a decade.
The CSI AI Index has jumped 60% so far this year, outperforming the 15% rise in the broader CSI300 index and Shanghai Composite index
The rapid surge in shares of Cambricon Technologies Corp has also brought valuation worries to the spotlight and led to the Beijing-based company issuing a risk warning last month.
The company reported last month an astonishing surge in first-half revenue to 2.9 billion yuan ($407.21 million) from 64.8 million yuan a year earlier, and swung to a profit of about 1 billion yuan. It expects to achieve a full-year operating revenue of 5 billion to 7 billion yuan.
"Investors are torn between hoping Cambricon can succeed in replacing foreign AI chips, and avoiding blowing up a speculative bubble," said Abraham Zhang, chairman of China Europe Capital.
The spotlight will be on the company's profitability and whether it is able to meet the demand for its AI chips.
Guangdong-based hedge fund manager Xu Qiongna said that although Cambricon is now much more expensive than Nvidia, potential rapid growth anchored by China's moves to replace foreign technologies could justify its valuation.
Tang Yibing, a portfolio manager of Bosera's STAR 50 Index Fund, said the tech sector's performance hinges more on industry trends and earnings growth.
The STAR50 Index has gained 5% this week.
($1 = 7.1216 Chinese yuan renminbi)
(Reporting by Li Gu and Samuel Shen in Shanghai, Jiaxing Li in Hong Kong; editing by Ankur Banerjee and Kim Coghill)
