Christopher Wood, the global head of equity strategy at Jefferies, has issued a cautionary note about the ongoing Artificial Intelligence (AI)-driven rally in U.S. stock markets. He warned that the surge in AI-related stocks risks ending in what he called a “massive overinvestment bust.”
AI Capex Arms Race Driving S&P 500 Gains
According to Wood, four major U.S. hyperscalers — alongside Nvidia — have been responsible for nearly 50% of the S&P 500’s gains since early 2023. This growth has been powered by an “AI capital expenditure arms race” rather than trade tensions or traditional factors.
The four largest hyperscalers are projected to spend an estimated $350 billion in 2025 on AI-related infrastructure, such as data centers, chips, and computing resources.
Retail Momentum and Feedback Loops
Wood also highlighted that this AI rally appears to be driven by retail investors. Many of them are reportedly purchasing stocks based on AI model recommendations themselves, creating what he described as a reflexive feedback loop — where AI-driven advice fuels further AI-related stock buying.
Valuations at All-Time Highs
The Jefferies strategist warned that U.S. market valuations have reached record highs on a price-to-sales basis. Despite these stretched valuations, investors continue to pour money into AI stocks, ignoring risks that could emerge if the hype fades.
DeepSeek’s Launch Signals AI Could Become a Commodity
Wood cited the recent launch of DeepSeek, an open-source large language model, as a reminder that AI technology may quickly become commoditized. If this happens, much of the current heavy spending on proprietary AI systems and infrastructure could prove wasteful.
Shift From Asset-Light to Asset-Heavy Models
Another concern Wood raised is that U.S. hyperscalers — traditionally known for their asset-light business models — are moving toward asset-heavy strategies by building massive AI infrastructure. If DeepSeek’s open-source approach becomes the norm, it may undermine the returns on these investments.
Equity Investors Overloaded on AI Stocks
Since Liberation Day, analysts have raised their earnings forecasts for the S&P 500 in 2026—but almost entirely because of the “Magnificent 7” tech giants. Other companies in the index haven’t seen similar optimism.
For the remaining 493 companies, earnings expectations have stayed low and show little sign of improvement. This means most of the market isn’t sharing in the growth hype.
In short, the S&P 500 is heavily concentrated in a few big AI-focused stocks, leaving equity investors very exposed if these companies stumble.
Sources: Bloomberg, Apollo Chief Economist
Be First to Comment