Jefferies warns AI boom may end in US capital destruction and shift to Chinese open models

In a quarterly report covered by Wallstreetcn on October 12, 2026, Chris Wood, Jefferies' global head of equity strategy and author of GREED & fear, compared the current AI boom to the 2008 subprime crisis. The structural risk he flagged is that hyperscale cloud providers are extending large amounts of unsecured credit to two AI labs that continue to lose money. Wood argued that the most likely long-term outcome of the AI story at the market level would be large-scale capital destruction in the United States, with market share shifting to cheaper Chinese open-source LLMs. He also said the AI story would face a major reset at some point in the not-too-distant future.

The evidence he cited comes from OpenRouter. According to the data Wood tracks, in the week ending September 27, top Chinese models processed 76.12 trillion tokens, up from just 4.37 trillion at the end of April this year, doubling within ten weeks. Over the same period, top US models processed 14.2 trillion tokens.

On quality, Wood noted that Moonshot's Kimi K3 is, in the industry's terms, competing at the frontier, with performance estimated at 95% of Anthropic Claude Fable 5, delivering that 95% level at a very low price. The report described Wood's warning as not new, but growing in scale.