The extraordinary rally in artificial intelligence chipmakers is showing signs of losing momentum as investors reassess lofty valuations and question whether the rapid pace of spending on AI infrastructure can be sustained, according to Reuters.
For much of the past two years, investors overwhelmingly favored semiconductor and infrastructure companies, betting that technology giants including Microsoft, Amazon, Alphabet and Meta would continue ramping up investments in AI data centers. That strategy helped propel chipmakers to record valuations as demand for AI hardware surged.
However, expectations are beginning to shift. According to UBS estimates cited by Reuters, capital expenditure by hyperscalers is projected to jump 76% this year to $673 billion, but growth is expected to slow sharply to 25% in 2027 and just 6% in 2028. While spending is still forecast to increase, the deceleration has prompted investors to reconsider whether semiconductor companies can continue delivering the revenue growth implied by their current valuations.
Fund Managers Rotate Into Hyperscalers
Some active fund managers have already begun trimming their exposure to semiconductor stocks and reallocating capital toward hyperscalers, software firms and sectors expected to benefit from AI adoption, including financials and healthcare, Reuters reported.
Portfolio managers argue that once cloud companies stop accelerating capital expenditure, it could ease pressure on their profitability while reducing growth prospects for semiconductor suppliers. As a result, several investors have shifted toward companies expected to gain from AI deployment rather than those supplying the infrastructure behind it.
Some managers have increased holdings in Amazon while favoring businesses involved in liquid cooling technologies, cybersecurity and enterprise software. Others have reduced exposure to memory chipmakers and semiconductor equipment manufacturers while simultaneously increasing investments in healthcare and major cloud providers.
Chip Stocks Remain a Crowded Trade
Despite recent weakness, semiconductor stocks have delivered exceptional returns over the past year.
The Philadelphia Semiconductor Index, whose largest constituents include Nvidia, Broadcom, Micron, ASML and TSMC, has more than doubled over the past 12 months, even after falling nearly 18% from its June peak. By comparison, the equal-weighted S&P 500 has gained about 11%, while Europe's STOXX 600 has advanced around 8%.
Bank of America's July global fund manager survey, cited by Reuters, showed that 82% of respondents considered semiconductors the most crowded trade in global markets, while none reported holding bearish positions against the sector.
Financing AI Expansion Faces Greater Scrutiny
Another concern emerging among investors is how hyperscalers will finance continued AI expansion.
After initially relying on internal cash flows to fund AI infrastructure, major technology companies have increasingly tapped debt markets to finance large-scale investments. Corporate bond issuance by Big Tech has reached billions of dollars this year.
However, demand for those bonds appears to be softening. Reuters cited Apollo's analysis showing that bond cover ratios, an indicator of investor demand relative to supply, have fallen below two times in July from nearly five times in February.
The Bank for International Settlements also warned in June that weaker-than-expected returns from AI investments could eventually reduce financing availability and turn today's spending boom into a prolonged downturn, according to Reuters.
Revenue Expectations May Need Adjustment
Research firms are increasingly highlighting a disconnect between slowing capital expenditure growth and optimistic revenue forecasts for AI hardware suppliers.
According to Reuters, Empirical Research believes either hyperscalers will have to increase investment plans again or analysts will eventually need to lower revenue expectations for companies supplying AI infrastructure.
Some investors believe earnings season will provide greater clarity. Portfolio managers at DWS still expect major cloud companies to signal continued investment in AI infrastructure and note that many institutional investors remain more optimistic than Wall Street analysts about spending beyond 2027.
Although DWS has booked some profits in semiconductor holdings following their strong rally, it continues to maintain an overweight position in the sector while adding exposure to industrial and electrical equipment companies after recent market pullbacks.
Data Center Expansion Faces New Challenges
Beyond financial concerns, AI infrastructure expansion is encountering growing political and community resistance.
Opposition to large-scale data center developments has been increasing across the United States due to concerns over electricity consumption, water usage and impacts on local communities. Estimates suggest that roughly 70% of planned data center projects face some degree of local opposition.
New York this week became the first U.S. state to impose a one-year moratorium on the construction of large new data centers, reflecting mounting concerns over the environmental and infrastructure costs associated with the AI boom.
Long-Term Optimism Remains Intact
Despite the recent volatility, many investors continue to view AI as a powerful long-term investment theme.
Morningstar data cited by Reuters show that chip-focused investment funds attracted a record $10 billion in net inflows through May, underscoring continued confidence in the sector.
Some market participants believe the recent correction resembles temporary pullbacks seen during previous technology booms, when leading companies experienced repeated sharp declines before resuming their upward trajectory.
At the same time, investors are increasingly seeking diversification by combining exposure to AI infrastructure providers with companies expected to benefit from AI adoption across industries such as finance, healthcare and enterprise software, reflecting a broader evolution of investment strategies as the AI market matures.
