Citadel warns: Tech stock sell-off is just the first warning sign, AI trading offers "almost no room for error"
The sell-off in tech stocks reflects rising doubts about AI profitability, with Citadel stating that there is very little margin for error in AI trading. Grayscale noted that demand is parabolic, and capital expenditure presents a two-sided risk.
According to information from Zhitong Caijing APP, the sell-off in technology stocks indicates that investor demands for massive AI expenditures to ultimately generate sufficient profits are growing increasingly stringent, as stated by Citadel’s Head of Fixed Income Sales for EMEA, Noshaad Shaha.
Shaha wrote: "Given the current high proportion of both the economy and the stock market hinging on AI trades, there is virtually no margin for error once other risks begin to emerge."
The debate about the pros, cons, and dangers of AI trades is intensifying among prominent investors: Ken Griffin of Citadel warned about bubble risks, while Blackstone Group (BX.US) President Jon Gray says there’s no need for concern, as demand is growing "almost parabolically."
Pablo Salame, Citadel’s Co-Chief Investment Officer, stated in a rare interview that the huge capital demands underpinning the AI revolution are colliding with the caution of lending institutions unwilling to take on all the risk.
Last week, a sharp unwinding of AI bets caused technology stocks to plummet, while risk-off signals in the credit markets climbed to their highest levels in more than a year, as the market worried that rising borrowing costs, combined with a range of global risks, would further complicate financing.

Rising artificial intelligence credit risk
Against this backdrop, Anthropic PBC filed for IPO on Monday, becoming the first such company whose core business is developing AI models to go public.
Shaha noted that tariffs, fiscal stimulus, and high energy prices are making both the financing and operational costs of AI investments more expensive, while public opposition to data center construction could slow project timelines and push costs higher.
Shaha stated that the United States has so far been able to absorb these pressures because profits and confidence remain strong. However, a continued ebbing of AI investment and financing could threaten the growth and household wealth that have previously helped the economy buffer higher costs.
Shaha echoed comments made by Griffin last week. Griffin said that US companies are struggling with how much to invest in technology and how much value they can extract from it.
Shaha wrote: "Capital expenditures themselves have become a two-way risk. Cutting spending raises doubts about the sustainability of the investment cycle... while further increases in spending bring questions about financing, execution, and ultimate returns."
Disclaimer: The content of this article solely reflects the author's opinion and does not represent the platform in any capacity. This article is not intended to serve as a reference for making investment decisions.

