The large-scale sell-off in artificial intelligence sector stocks is approaching its final stage — this is the conclusion reached by Tom Lee, managing partner and head of research at Fundstrat. The key arguments are bond yields falling below 5% and the accumulation of short positions in the tech sector, which creates the preconditions for a reversal.
The Dow index is trading more than 6% below its peak recorded at the beginning of the month. It is against this backdrop that Lee points to two fundamental shifts in market conditions.
What has changed in the market
The US Federal Reserve is transitioning from a tight policy to a looser one. However, the main signal was the pullback in bond yields below 5%, whereas investors had previously feared their rise to 6%. This reversal in yields traditionally acts as a trigger for a reassessment of risk assets.
Among the promising areas, Lee highlights Nvidia, memory chip makers, the seven largest technology giants (Magnificent 7), software companies, and cryptocurrencies. An additional argument is the expectation of broad cyclical growth: financial company stocks are lagging amid falling rates, which historically precedes the expansion of the rally beyond the technology sector.
A week earlier, Lee expressed similar theses, calling Friday's decline a convenient entry point for buying chips and crypto assets.
The debt market will determine the fate of the forecast
Lee notes two remaining risks — a partially paralyzed debt market and uncertainty surrounding the geopolitical situation. At the same time, he considers the cooling of financing to be beneficial, since it curbs speculative bubbles. In the Fundstrat report, AI spending is estimated at $30 trillion over 25 years, and these investments are backed by real capital.
"AI is an infrastructure shift, perhaps even more important than mobile technology," Lee emphasizes.
At the same time, Gartner analyst Will Sommer notes a shortfall in revenue: AI income will prove too small to cover investments under current plans. By his estimates, AI spending alone will reach almost $1 trillion in 2026, and from 2026 to 2030 it will grow by another $6 trillion.
Lee compares the current state of the AI field to 1997. At that time, Cisco shares fell by 50%, although they reached their historic peak only in March 2000, after which they collapsed by almost 90%.
A separate factor is stock market concentration: the 15 largest corporations hold a record 46.7% weight in the S&P 500 index versus 27.2% in 2000. The unblocking of the debt market amid falling rates will be decisive for the realization of the forecast.
Cryptalist analysis: Fundstrat's thesis is built on the classic link of "falling yields — growth of risk assets," and the current bond dynamics indeed create a foundation for the recovery of risk appetite. However, the record concentration of the S&P 500 and the divergence in AI monetization estimates between Fundstrat and Gartner point to the market's high sensitivity to any disappointments in corporate reports. For cryptocurrencies, this means the potential for synchronous movement with technology stocks while maintaining elevated volatility.