A growing number of China's most influential hedge fund managers are warning that the global artificial intelligence trade has become dangerously overheated. After a meteoric rise in semiconductor and AI development stocks from Silicon Valley to Seoul, these investors argue that valuations have decoupled from underlying business realities—and that early signs of a correction are already emerging.
Wealspring and Banxia Call AI a 'Super Bubble'
Two of China's best-known hedge funds—Wealspring Asset Management and Shanghai Banxia Investment Management—have issued unusually stark warnings, according to a Bloomberg report. Wealspring, founded by Yang Dong, who famously predicted China's 2007 market peak, told investors in a letter that global AI equities now resemble a 'super bubble' inflated by momentum rather than sustainable business models.
Wealspring, which oversees more than $1.4 billion, argues that many Chinese AI infrastructure companies lack durable competitive advantages and rely on relentless capital spending to sustain growth. The firm described current buying as 'brainless,' comparing it to the speculative frenzy of China's 2015 bull market. It cautioned that some of the most popular domestic AI names could ultimately fall more than 80% once sentiment shifts.
Global Concerns Extend Beyond China
Banxia, which manages about $294 million, contends that the warning signs are not limited to China's onshore markets. The firm points to slowing revenue momentum at Anthropic PBC, a closely watched U.S. AI startup, as evidence that the narrative of unstoppable growth is beginning to fray. Banxia believes Anthropic's annualized revenue run-rate may miss market expectations as enterprise clients resist rising token costs and competition from other model developers intensifies.
The fund also highlights a broader risk: the AI ecosystem's aggressive revenue assumptions are colliding with rising infrastructure costs and tightening corporate budgets. For Banxia, these pressures represent the early stages of a sentiment shift that could ripple across global AI valuations.
Market Performance and the Cost of Caution
Despite these warnings, AI stocks have delivered extraordinary gains in 2026. China's CSI Artificial Intelligence Index is up more than 35% year-to-date, far outpacing the 5% rise in the country's main benchmark. Overseas, the rally has been even more dramatic: South Korea's Kospi has surged nearly 100% this year, powered by SK Hynix and Samsung Electronics, both beneficiaries of unprecedented demand for high-bandwidth memory used in AI data centers.
Yet the very funds urging caution have paid a short-term price for staying on the sidelines. Wealspring's Zhiyuan fund and Banxia's low-volatility macro strategy have posted small losses in early 2026, even though both remain strongly profitable over longer horizons. Banxia founder Li Bei insists the discipline is worth it, advising investors to resist chasing parabolic AI trades and to prepare for a potential correction that could reset valuations across the sector.
For context, the broader market dynamics are also influenced by factors such as Nvidia lagging the chip rally as smart money rotates to memory and custom silicon, and the impact of China EV stocks shedding $100 billion as subsidies end and competition intensifies. These trends underscore the shifting landscape in technology and investment flows.
This article is for informational purposes only and does not constitute financial advice.
