Trillions Wiped Off Chinese Markets: The AI Bubble Debate Reaches Beijing

Trillions of dollars have evaporated from Chinese markets in recent months. Shanghai's STAR 50 index, which tracks technology companies, has fallen roughly 40 percent from its June peak, while other indices covering large-cap firms have dropped around 20 percent. Behind the sell-off lies not a single bad earnings report but a more fundamental question: when will the massive capital flowing into artificial intelligence turn into profit?

What happened?

Investors have pulled out of China's AI story. Companies are spending heavily on model training and data centers, but the conversion of these investments into revenue is lagging. During the same period, Washington's restrictions on advanced chip exports are narrowing Chinese firms' access to the most powerful hardware. When the two pressures combined, the indices entered a sharp correction.

Why is the AI bubble no longer just America's problem?

The "AI bubble" debate that has dragged on for months was mostly centered on the United States. As the giant cloud companies wrote hundreds of billions of dollars in capital expenditure onto their balance sheets, investors increasingly asked, "where is the return on this money?" The decline in China shows that the same doubt is now surfacing in a different economic model.

The two countries' approaches are in fact opposites. In the US, the push is driven by private capital and publicly traded companies; the talk is of players like OpenAI and Anthropic with infrastructure commitments worth hundreds of billions of dollars spread over the coming years. In China, the direction is largely set by the state: strategic targets, subsidies, and priority for the domestic supply chain. But the outcome is similar — in both models, spending runs far ahead of profit.

Here is the crux: the market has begun to weigh both models on the same scale. State support does not cover unprofitability forever. Sooner or later, the investor asks about cash flow.

Two different bubbles, the same sustainability question

Capital intensity is extraordinary on both sides, but the sources of fragility differ.

On the Chinese side, the chip restriction is not merely a handicap but also a trigger. A company with narrowing access to hardware either runs on less efficient domestic chips or scales back. Both push costs up and efficiency down. That pushes the horizon of profitability even further away.

In the US, the problem is different: hardware is plentiful, but expectations are more than fully priced in. If a wave of investment is slow to turn a profit, the most inflated valuations pull back the hardest. The decline in China is a reminder that this scenario is on the table in the US as well.

What's next?

There are a few signals to watch. First, whether companies can show AI revenue in concrete terms — invoiced revenue, not demos. Second, how Beijing will respond to the decline: new incentives, or more aggressive investment in domestic chip production. Third, whether US indices will be psychologically affected by the sell-off in China; the same "where is the profit" question is being asked on both sides of the Atlantic.

This correction is not testing the value of the technology but the realism of its pricing. AI is here to stay; but so is the fact that not every capital move turns into profit. The losses in China show that even state support cannot indefinitely postpone the market's fundamental question — and that lesson is not unique to Beijing.