China's STAR 50 Index Drops 40% from Peak as AI Stock Bubble Bursts, Wiping Out ~¥3.4 Trillion
nashnova research
Shanghai's STAR 50 index has fallen roughly 40% from its June peak, wiping out about $480 billion (¥3.4 trillion) in AI and semiconductor market cap; stretched valuations, thin margins, and limited policy firepower now put the board's commercial-profit thesis to a defining test.
How much has been lost — and where did it go?
The STAR 50 index is down roughly 40% from its June high; AI and semiconductor stocks alone have shed about $480 billion (~¥3.4 trillion).
Blue-chip benchmarks fell nearly 20% over the same period; Bloomberg estimates China's total equity market cap has shrunk by about $2.4 trillion since May.
This means → the STAR board is falling twice as fast as the broader market. AI stocks are not following the decline — they are leading it.
Why does one sector drag down the entire market?
Goldman Sachs data: AI-related stocks account for roughly 42% of China's (including Hong Kong) total market cap and 57% of all trading volume.
In plain terms = nearly half the market's value and more than half its turnover sit in a single theme — once that theme retreats, liquidity drains from the whole market.
The STAR 50 trades at roughly 100× earnings; the Nasdaq 100 sits at about 30×. This reflects a valuation built on future-profit expectations, not current earnings — leaving far more room to fall.
How badly have the marquee names been hit?
Unitree Robotics' retail IPO was oversubscribed roughly 8,000×, yet the stock has halved since its August listing.
Memory-chip leader CXMT has dropped more than 25% in the past month; several AI-chip firms that listed in 2025 with multi-fold first-day gains have lost over 50% from their peaks.
In optical modules, Innolight and Eoptolink are each down nearly 60% from mid-year highs — the market fears Washington may further restrict Chinese component suppliers' access to global data-center contracts.
Can these companies actually make money?
Investors note that while these firms carry state endorsement, their demand depends heavily on government agencies and research institutes — none has proven it can sustain profits in the commercial market.
This means → the earlier rally priced in a "state-will" premium, not orders and earnings. Once the market starts demanding real revenue, the valuations cannot hold.
Has Beijing stepped in — and is it working?
Beijing recently established a national AI investment fund of roughly ¥600 billion (~$90 billion) and encouraged households to move savings into equities.
Those measures have been offset by weak domestic demand, fierce corporate price wars, and disappointing earnings.
Dan Wang, China director at Eurasia Group, warned via *The Guardian* that thin margins, elevated global rates, and stagnant domestic demand have triggered deep anxiety over the long-term sustainability of China's state-led tech model.
What happens next?
Hong Hao, CIO at Grow Investment in Hong Kong, warned via Reuters that a continued slide could force tech firms to shelve IPO plans, severely narrowing equity-financing channels.
Beijing-based fund manager Dong Baozhen said bluntly that the prior rally was "speculative frenzy" — the bubble is bursting.
Some observers argue the correction helps purge inflated valuations; Asian funds have trimmed China positions over the past quarter but have not fully exited. This reflects a market caught between "too expensive" and "not ready to leave entirely" — whether the STAR board can rebuild support hinges on whether a commercial-profit thesis can be established.
市场有风险，内容仅供研究参考，不构成投资建议。