Chinese Hedge Funds Warn of AI ‘Super Bubble’

Shanghai-based Banxia Investment Management and Wealspring Asset Management warned investors that the global rally in AI-related stocks has become detached from fundamentals and entered what Wealspring called a “super bubble.” They pointed to signs that revenue growth forecasts for leading AI developers are moderating and that momentum at model maker Anthropic is slowing.

Banxia told investors the catalyst for a market correction may already be appearing. The firm highlighted early signs that revenue growth expectations for top AI developers are starting to ease and flagged weaker momentum at Anthropic as a possible signal that hyperscale cloud providers may take longer to monetise AI services. Banxia’s founder urged investors to exercise “very, very” high levels of caution when seeking AI exposure.

Wealspring described many AI infrastructure companies as being valued on aggressive growth assumptions despite lacking durable competitive advantages. The firm said a substantial portion of businesses across the AI supply chain remain heavily dependent on continuous capital injections to sustain expansion, raising questions about the sustainability of current valuations if investment or demand slows.

The warnings follow a sharp run-up in equities tied to AI, notably semiconductor and memory names such as Micron Technology and SK Hynix, which have rallied on expectations of sustained spending on data centres and high-performance computing. AI-focused indexes have outperformed broader benchmarks this year, while several Chinese hedge funds have moved to a more cautious stance on AI-related assets.

Some fund managers are questioning whether market prices fully reflect execution risks, including product delivery, margin pressure, competition and the pace at which hyperscalers can turn AI services into profitable revenue streams. Analysts and managers note that building and operating large AI systems requires ongoing capital for data centres, specialised chips and memory capacity. If revenue growth for AI developers or cloud providers slows, companies that depend on continuous funding could face higher financing costs or need to scale back capacity expansion, with potential effects on margins and cash flow.

The firms’ statements add to market discussion about how much capital the AI build-out will consume and whether current valuation levels match expected future cash flows.