Morgan Stanley's Chief Investment Officer for Wealth Management, Lisa Shalett, advises investors to be cautious on chip stocks as increasing signs indicate chipmakers' pricing power is coming under pressure. She believes market optimism over artificial intelligence spending may have already pushed related stocks to excessively high levels.
In an interview, Shalett stated:
"We are seeing a re-architecting of the AI datacenter tech stack, with more cost-effective, internally developed chips being integrated—chips that many of the hyperscale cloud providers are developing themselves."
Shalett issued this warning as SK Hynix officially listed on the Nasdaq, completing a $26.5 billion fundraising that set a record for a foreign company's U.S. IPO. However, SK Hynix has recently experienced significant volatility in its home South Korean market, with its stock falling 26% from last month's high.
Shalett noted:
"Broadly speaking, capital flows into this trade theme remain very robust."
But she pointed out that current industry developments are following a familiar pattern:
"When supply chain bottlenecks emerge, and some companies—like certain memory chip makers—take the opportunity to earn supernormal profits, engineers start looking for lower-cost alternatives."
Earlier this week, Shalett highlighted in an investment note that the semiconductor sector is showing signs of being "clearly overbought." She elaborated further, stating that multiple indicators, from semiconductor ETFs to the Philadelphia Semiconductor Index, support this view.
Compiled data shows that since 2022, the price-to-earnings ratio of the Philadelphia Semiconductor Index has more than tripled.
Shalett also mentioned that Meta Platforms' recent adjustments to its AI strategy serve as a notable signal, suggesting some tech giants may be starting to reassess their multi-hundred-billion-dollar capital expenditure plans.
Meta CEO Mark Zuckerberg stated in an interview this week that he is considering whether renting out parts of Meta's AI infrastructure to external customers could generate higher value.
Regarding this, Shalett commented:
"This, in a way, indicates companies are beginning to discuss the pace, speed, and return on investment of these expenditures, while also thinking about how to commercialize them sooner."
She concluded by saying:
"I believe we are in the early innings of a slowdown in the growth rate of AI capital expenditures."
Comments