Tom Lee Says AI Selloff Nearing End, Cites Cisco 1997 Parallel

Fundstrat Managing Partner Tom Lee stated on Friday that the recent decline in artificial intelligence stocks represents a buying opportunity rather than the beginning of a broader market unwind. Lee, speaking in a CNBC interview, argued that several pressures weighing on technology equities are beginning to ease, positioning the current dip as a moment to accumulate rather than retreat.

His remarks followed a sharp sell-off in the tech sector, where the Nasdaq Composite fell 2.6% on Thursday, marking its worst single-day decline since June. The weakness was particularly acute in semiconductor-focused instruments; the VanEck Semiconductor ETF (SMH) dropped 6% and the Philadelphia Semiconductor Index (SOX) fell 7% on the same day, both their steepest one-day losses since June. Despite these moves, Lee dismissed fears of a systemic tech decline, stating he would buy the dip.

Lee attributed his optimism partly to a shift in the interest-rate environment. He noted that Treasury yields have retreated below 5%, alleviating investor concerns that rates might climb toward 6%. Lee suggested the Federal Reserve could be "flipping potentially dovish," a stance that supports riskier assets. He maintained that the weakness in tech stocks does not indicate a breakdown in the underlying AI investment cycle, pointing instead to continued "astounding discoveries" in the field.

The strategist argued that current concerns are actually preventing the market from becoming a speculative bubble. He cited tighter debt markets as a restraint on excessive speculation, describing the situation as a "very healthy correction." In his view, constrained financing is preventing a "runaway speculative boom." Lee drew a specific parallel to the late-1990s technology market, highlighting Cisco Systems Inc.'s (CSCO) sharp decline in 1997. He noted that Cisco experienced a major drawdown as investors grappled with monetary policy and valuation sustainability, yet the broader technology investment cycle continued its advance afterward.

Lee sees similar opportunities emerging in companies addressing AI infrastructure bottlenecks, memory stocks, Nvidia Corp. (NVDA), and the "Magnificent Seven." He observed that these names are being sold off because they have become "easy to short." Beyond major hardware names, Lee expressed interest in software and cryptocurrency sectors, while also anticipating potential for a broader cyclical move. He estimated that AI infrastructure spending could reach $30 trillion over the next 25 years, potentially making AI a more significant infrastructure investment than previous major technology buildouts.

During Friday's after-hours session, broad market indicators showed slight gains. The SPDR S&P 500 ETF (SPY) rose 0.15%, the Invesco QQQ Trust ETF (QQQ) edged up 0.07%, and the SPDR Dow Jones Industrial Average ETF Trust (DIA) increased by 0.13%. Retail sentiment on Stocktwits toward the S&P 500 ETF remained in "extremely bearish" territory at the time of reporting.