Cheaper AI models from China could undermine the US artificial intelligence rally by prompting investors to question whether heavy spending by American technology companies is delivering sufficient returns, according to Jefferies’ Global Head of Equity Strategy Christopher Wood.
In his latest GREED & Fear note, Wood said the biggest market risk is not AI adoption but the point at which investors begin evaluating the return on the massive capital investments made in the sector. He added that the AI investment story still favors companies supplying infrastructure and equipment to hyperscale data center operators over firms building AI applications.
Wood also highlighted the rapid rise of Chinese large language models (LLMs). Citing OpenRouter data, he said leading Chinese models processed 36.39 trillion tokens in the week ending July 19, up from 4.37 trillion in late April, compared with 7.39 trillion tokens processed by the top US models during the same period.
He said China has emerged as a technological peer to the US in AI and argued that financial markets have not yet fully reflected the country’s progress. Wood also warned that much of the US AI expansion is increasingly being financed through debt rather than internal cash flow.
Reiterating that US equities may have already reached their peak share of global market capitalization, Wood advised investors to closely watch the performance of hyperscaler stocks.
His comments came as technology stocks weakened globally, with South Korea’s Kospi index falling nearly 11% and major chipmakers Samsung Electronics and SK Hynix among the biggest decliners amid concerns over the durability of the AI-driven market rally.