China Stocks Enter Bear Market As AI Shares Plunge

Chinese stocks have lost about 2.4 trillion dollars since May, but the main risk goes beyond investors’ portfolios. By dropping valuations and household confidence, the correction could deprive AI companies of the capital needed to compete with the United States.

In Brief

• The Chinese stock market has lost nearly 2.4 trillion dollars since May.
• Values related to artificial intelligence are among the hardest hit by this drop.
• The decline in stocks threatens the financing of Chinese tech companies.
• Stock market losses weaken individual investors and weigh on their confidence.
• Despite the correction, valuations remain high and slow the return of capital.

The stock market crash could cut future funding for Chinese AI

The CSI 300, which groups the main listed companies in mainland China, has fallen more than 20% since its June peak. It thus entered a bear market on October 22, while the STAR 50, more exposed to semiconductors and artificial intelligence, has lost 40% in less than four months.

Your 1st cryptos with SwissborgThis link uses an affiliate program.

This drop erased nearly 480 billion dollars on the STAR 50 alone. It comes as Beijing seeks to mobilize national savings to finance chips, robots, data centers, and the AI models needed for its technological autonomy.

Various indicators show why this decline threatens the sector’s financing :

• AI-related stocks represent 42% of Chinese capitalization ;
• They concentrate 57% of trades in the concerned markets ;
• Robot manufacturer Unitree has lost nearly half of its value ;
• Zhongji Innolight and Eoptolink have dropped about 60% since their peaks ;
• Many tech companies could postpone their IPOs.

A high valuation normally allows a company to raise more money by issuing fewer shares. When its stock price collapses, a new operation becomes more costly for existing shareholders, as their stake risks being further diluted.

Hong Hao, head of investments at Lotus Asset Management, estimates that the correction will “halt many companies that were seeking to get listed”. The crash therefore reduces the available resources at the very time Chinese players must finance heavy investments and catch up on advanced chips.

Households pay the price of a market that has become tech-dependent

The scale of the decline is also explained by the market’s concentration around a single theme. When AI-related values represent 42% of capitalization and more than half of trades, their correction mechanically drags the rest of the stock market down.

Individual investors hold an important place in Chinese markets. Many joined the movement thinking Beijing’s support for strategic sectors reduced their risk. Unitree’s IPO had thus received requests representing 8,000 times the number of shares offered.

State support for a technology does not guarantee the profitability of every company. China, an influential member of the BRICS alliance, can finance research, promote national chips, and encourage the adoption of robots without preventing listed companies from facing intense competition or declining margins.

Dong Baozhen, fund manager in Beijing, notes :

The stock market then evolved into a wild speculative rise. And now, we see the bubble burst.

American restrictions worsen this fragility. Washington limits China’s access to certain technologies and plans to exclude Chinese optical components from US data centers. The affected companies thus lose part of their potential markets at a time when domestic consumption remains weak.

Still too high valuations delay the return of capital

The 40% drop in the STAR 50 might give the impression that excesses have already been corrected. Yet, the index still trades around 100 times earnings over the past twelve months, compared to nearly 30 times for the US Nasdaq 100.

This level means investors are still paying a very high price for currently limited profits. At its peak, the STAR 50’s multiple approached 250. The correction thus reduced the excess without bringing valuations back to the standards of major international tech indices.

The problem does not necessarily come from a lack of innovation. DeepSeek has shown that a Chinese company could develop a performant model with costs lower than many Western competitors. Labs also claim they can train their models with Huawei chips.

The challenge now is to turn these advances into sustainable revenues. A price war between chip manufacturers, model developers, and robot makers could boost adoption while compressing profits.

The loss of 2.4 trillion dollars does not mean that Chinese AI has failed. It shows that Beijing cannot finance its tech strategy indefinitely through stock market enthusiasm. To attract household savings again, companies will need to demonstrate their ability to generate profits, not just their importance in the rivalry with the United States.