Jefferies' Global Head of Equity Strategy Christopher Wood has warned that cheaper artificial intelligence models emerging from China pose a material threat to US stock market valuations, as investors are set to increasingly question whether the enormous capital deployed into AI by American technology giants is generating adequate returns.

The Core Warning

In his latest weekly note, GREED & Fear, Wood identified the central risk for markets as not AI's growth trajectory itself, but the moment investors begin scrutinising return on investment. 'The key issue has always been one of timing in the sense of when the market will start to worry about the return on investment (made in AI),' he wrote. He added that the AI investment theme continues to favour companies supplying infrastructure and equipment to hyperscalers, rather than firms building AI applications.

Chinese Models Surge Past US Rivals in Token Processing

Wood pointed to striking data on the rapid global traction of Chinese large language models (LLMs). According to figures cited in his note, the top Chinese AI models processed 36.39 trillion tokens on the OpenRouter platform in the week ended 19 July, a sharp jump from 4.37 trillion tokens in late April. By contrast, leading US AI models processed just 7.39 trillion tokens over the same period — less than a fifth of their Chinese counterparts. 'There is also a growing realisation now that China has become a technological peer to the US in AI, as well as in so many other areas,' Wood said. He cautioned that financial markets have yet to fully price in China's rapid AI progress.

Debt-Fuelled Boom Raises Structural Concerns

Beyond the competitive threat, Wood flagged that the US AI boom is increasingly being financed through debt rather than internal cash generation — a structural vulnerability that amplifies downside risk if return-on-investment concerns crystallise. He reiterated his view that US equities have likely already peaked as a share of global stock market capitalisation, urging investors to closely monitor the relative and absolute performance of hyperscaler stocks.

Global Markets Feel the Pressure

Wood's remarks arrived as technology shares faced renewed selling pressure worldwide. South Korea's Kospi index slumped nearly 11 per cent on Tuesday, with trading temporarily halted during the session. Chipmakers Samsung Electronics and SK Hynix were among the biggest losers, reflecting broader concerns over the sustainability of the AI-driven rally. This comes amid a pattern of episodic risk-off moves tied to AI valuation anxiety — a dynamic that has intensified since the emergence of low-cost Chinese models earlier this year.

What to Watch

The immediate focus for markets will be quarterly earnings disclosures from major US hyperscalers, where capital expenditure guidance and AI monetisation commentary will be closely parsed. Any softening in revenue growth relative to infrastructure spend could accelerate the investor rethink Wood has flagged. The divergence between Chinese and US AI model usage metrics, if it widens further, is likely to become a recurring pressure point for tech valuations globally.