|
The United States could face "massive capital destruction" from its artificial intelligence (AI) investment boom as Chinese open-source AI models increasingly challenge the dominance and profitability of US hyperscalers, according to a report by Jefferies.
The brokerage said investors are beginning to question whether the hundreds of billions of dollars being spent on AI infrastructure by major US technology companies will generate adequate returns, particularly as cheaper Chinese large language models gain market share. "GREED & fear's base case has always been that the most likely longer-term outcome of the AI story in market terms will be massive capital destruction in the US with market share going to cheaper open-source Chinese large language models," Jefferies said in its latest report. The report noted that the four largest US hyperscalers are projected to spend about US$695 billion in capital expenditure (capex) in 2026, with spending expected to rise further to US$870 billion in 2027. Alphabet recently increased its 2026 capex guidance by another US$15 billion to US$195-205 billion, while investors now await guidance from Microsoft, Amazon and Meta. Jefferies believes the market is entering a phase where investors will increasingly scrutinise returns on AI investments, warning that the AI capex arms race is now more than three years old. The report said competition from China has intensified following the launch of advanced open-source AI models, including Moonshot AI's Kimi K3, while usage data indicates Chinese models are rapidly gaining global traction. According to Jefferies, "China has become a technological peer to the US in AI, as well as in so many other areas." The report cited OpenRouter data showing that leading Chinese AI models processed 36.39 trillion tokens in the week ended July 19, up sharply from 4.37 trillion tokens in late April, compared with 7.39 trillion tokens processed by the top US models. Beyond competitive pressure, Jefferies also highlighted rising financial risks linked to the AI spending cycle. It said hyperscalers have increasingly financed AI infrastructure through debt, with the sector now becoming the largest issuer of investment-grade debt in the United States. The report warned that profitability assumptions behind AI investments may prove overly optimistic amid falling AI token prices and increasing competition. "The reason credit risk has become more of an issue is that it can no longer be assumed... that large language models will ever be profitable given the related ongoing collapse in token pricing," the report said. (ANI) Jefferies also pointed to growing off-balance-sheet obligations and data centre lease commitments, arguing that investors may eventually reassess the sustainability of current AI spending. The brokerage said uncertainty remains over which, if any, of the US hyperscalers will successfully monetise their AI investments, while their traditional businesses could themselves face disruption from AI. It added that although AI remains a transformative long-term technology, the sector may first go through an extended period of weaker investor sentiment as markets reassess the economics of the current investment cycle. (ANI)
|
