Morgan Stanley Warns AI Stocks Face a Major Spending Reckoning
The massive spending behind the artificial intelligence boom is facing growing scrutiny as investors demand clear evidence of returns. Morgan Stanley warns that semiconductor stocks could face further declines if major technology companies begin scaling back their AI infrastructure investments.
Mike Wilson, Morgan Stanley’s chief U.S. equity strategist, expects semiconductor stocks to fall another 10% to 15% before potentially reaching a bottom around mid-November. He sees the Philadelphia Semiconductor Index falling to approximately 9,500 points.
The concern centers on the widening gap between AI infrastructure spending and the cash flow it generates. Major cloud providers, including Amazon, Alphabet, Meta, and Microsoft, have continued investing heavily in AI data centers and hardware. However, Amazon reported negative trailing 12-month free cash flow of $42.4 billion, while Alphabet recorded roughly $9 billion in negative quarterly free cash flow.
Morgan Stanley also highlights the semiconductor industry's dependence on a small group of hyperscalers. If these companies reduce hardware purchases to protect cash flow, chipmakers could face simultaneous pressure.
Corporate adoption is another concern. According to figures cited in the article:
- The share of companies abandoning most AI initiatives increased from 17% to 42% in one year, according to S&P Global.
- IDC data indicates that 88% of AI proofs of concept never reach production.
- Only 21% of organizations have a mature AI governance model, according to Deloitte.
- Gartner reports that 63% of data management leaders say their organizations lack the necessary data practices for AI.
The potential exposure is substantial. Hyperscaler AI spending is projected to reach $1.58 trillion by 2027, according to Fortune. Nvidia is particularly exposed, with 91.7% of its projected 2027 revenue tied to AI infrastructure demand.
Concentration risk also affects semiconductor ETFs. Five major holdings, including Nvidia, Micron, Broadcom, and Intel, reportedly account for roughly 39% of the SOXX portfolio.
While historical data shows U.S. stocks have rallied after every midterm election since 1950, the article notes that geopolitical tensions involving Iran and local opposition to large AI data centers could complicate a broader market recovery.
The key question for investors is whether massive AI investments can translate into sustainable earnings and cash flow. As the market becomes more selective, companies may need to demonstrate measurable returns rather than relying on AI-related growth expectations.
