Chip Stocks Soaring Too Fast? Morgan Stanley Warns AI Spending May Start to Brake

Lisa Shalett, Chief Investment Officer at Morgan Stanley Wealth Management, has warned investors to remain cautious on semiconductor stocks as signs emerge that chipmakers' pricing power may be constrained. She suggests that market optimism around artificial intelligence (AI) spending may have driven chip stocks to overextended levels.

In a Friday interview with Bloomberg Television, Shalett noted that the technical architecture of AI data centers is being redesigned, with hyperscale cloud providers like Meta, Google, and Amazon developing proprietary, lower-cost chips to reduce reliance on external semiconductor suppliers. This trend could weaken the pricing power of certain chipmakers, particularly memory manufacturers.

She acknowledged that ample capital remains available for AI investments, but highlighted a recurring pattern in the semiconductor industry: when supply chain bottlenecks allow certain players to earn excess profits due to shortages, engineers begin seeking cheaper alternatives.

Shalett described semiconductor stocks as "clearly overbought" in her investment report this week, citing evidence from both semiconductor ETFs and the Philadelphia Semiconductor Index. Bloomberg data shows the index's price-to-earnings ratio has more than doubled since 2022, reflecting a significant valuation expansion for chip stocks.

Her warning comes as South Korea's SK Hynix officially listed on Nasdaq Friday, raising $26.5 billion in the largest-ever foreign IPO on U.S. soil. However, SK Hynix's shares in Korea have been highly volatile, falling 26% from last month's peak, underscoring rising market concerns over high valuations.

Shalett also views Meta's (META-US) recent shift in AI strategy as a signal that major tech firms are reevaluating massive capital expenditures. Meta CEO Mark Zuckerberg stated this week the company is exploring leasing parts of its AI infrastructure to external customers to enhance asset value.

She interprets this as evidence that tech giants are internally debating the pace, scale, and return on AI investments, and how to accelerate commercialization. Shalett predicts AI capital spending growth may already be entering an early slowdown phase. If spending slows and custom chips become more widespread, the high prices and valuations currently enjoyed by chipmakers could face increasing pressure.