China’s AI Bubble Bursts: $480 Billion Vanishes as Profitability Fails to Match Valuations

Key points

• The STAR 50 Index, tracking major Chinese tech firms, fell more than 40% from its July high, wiping out roughly $480 billion in market value over four months.
• CATL, a leading battery manufacturer, saw its stock drop 37% year-to-date despite 59% revenue growth in the third quarter, as net profit rose only 15% due to competition and tax changes.
• Specific AI and chip names suffered heavily; humanoid robot maker Unitree lost half its value since its August debut, while GPU developers MetaX and Moore Threads roughly halved in value.
• The STAR 50’s price-to-earnings ratio fell from near 250 to around 100, though it remains significantly higher than the Nasdaq 100’s multiple of approximately 30.
• Morgan Stanley forecasts that offshore Chinese stocks may fall another 10% to 15% before bottoming, citing continued uncertainty in monetization and domestic demand.

Background

This correction follows a period of intense speculation in Chinese tech stocks, which accelerated after the emergence of DeepSeek in early 2025. Recent global market volatility, including a broad tech sell-off on October 15, 2026, and concerns over AI infrastructure debt, have heightened scrutiny on the profitability of AI buildouts. Additionally, earlier reports indicated that Chinese memory chip maker CXMT’s advancements triggered global chip sell-offs, reflecting broader fears about competitive pressures and supply chain shifts.

How outlets are covering it

CNBC frames the sell-off as a necessary valuation correction, emphasizing that while technology remains an economic bright spot, expectations had outpaced fundamentals. Benjamin Cavender of China Market Research Group notes that the market is now critically examining whether earnings can justify AI excitement. In contrast, Yahoo Finance highlights the speculative nature of the prior rally, quoting Beijing-based fund manager Dong Baozhen, who described the market as a 'speculative crazy bull' that is now bursting. Yahoo Finance also draws parallels to U.S. AI concentration risks, noting that while the Nasdaq 100’s multiple is lower, rising concentration without rising earnings mirrors the pattern that preceded Shanghai’s crash. While CNBC points to undemanding valuations as a floor for downside, Yahoo Finance suggests that Morgan Stanley expects further declines, indicating a divergence in views on the immediate bottom of the correction.

Why it matters

The collapse of China’s AI bubble signals a global shift from growth-at-all-costs to profitability-focused investing. For U.S. investors, the sharp decline in the KraneShares China Technology & Semiconductor STAR 50 Index ETF (KSTR), which fell 42.63% from June to October, serves as a cautionary tale for concentrated AI holdings. The disconnect between rapid revenue growth and stagnant margins in key sectors like batteries and chips suggests that the next phase of the AI cycle will be defined by operational efficiency rather than market share expansion. Furthermore, the prolonged property downturn in China continues to suppress household confidence and consumption, limiting the ability of internet platforms to offset tech sector losses.

What to watch

Investors will closely watch whether Chinese tech companies can translate heavy AI spending into tangible earnings in the coming quarters. A sustained rebound depends on stabilizing margins and clearer returns on investment, as noted by analysts at JPMorgan and Union Bancaire Privée. Additionally, the performance of U.S. technology stocks may influence Chinese equities, as global sentiment on AI returns remains linked. If Morgan Stanley’s forecast of a further 10-15% decline in offshore Chinese stocks materializes, it could signal a deeper correction before any recovery. Conversely, if valuations stabilize and domestic consumption improves, the current lows may present entry opportunities for long-term investors.