Big technology companies suffered a reality check in June that led to a sharp stock market decline. The so-called Magnificent Seven (Nvidia, Apple, Microsoft, Alphabet, Meta, Amazon, and Tesla) lost 10% of their value, marking their biggest correction since March 2025. In absolute terms, this represents a loss of $2.3 trillion in market capitalization, coinciding with the month of SpaceX’s IPO, a flagship public offering of an era and an all-time record both in terms of the amount raised ($75 billion) and the company’s valuation ($2.1 trillion).

Although the technology companies’ financial results remain exceptional, many investors have begun to express doubts about the enthusiasm surrounding artificial intelligence. Alfonso de Benito, chief investment officer at Dunas Capital — an asset manager overseeing more than $5 billion — explains that “the valuation of the U.S. stock market was unsustainable, mainly because of technology companies.” The fund manager notes that the share prices of AI-related firms “implied not only that revenues would continue growing in coming years at the current rapid pace, but that growth would actually accelerate.”

One trend visible in the sector is the rotation from hyperscalers — companies developing data centers to process AI workloads, such as Amazon, Meta, and Microsoft — toward firms specializing in manufacturing microchips, such as Nvidia. The latter has gained 4.5% year-to-date (despite the June correction), while Microsoft has fallen 24%. In the AI gold rush, increasing numbers of investors prefer to invest in the manufacturers of the “shovels and picks” rather than in the mining company that might discover the richest vein.

This year’s stock performance among chip and memory manufacturers has been striking. SanDisk has risen by roughly 760%, Intel has tripled in value, and South Korea’s SK Hynix has gained 300%. Demand for processing cards remains extremely high, and memory shortages are expected to continue through 2028.

Data centers

The investment that technology giants are making in data centers is unprecedented. According to data gathered by ING, the five hyperscalers — which include Alphabet and Oracle — have committed capital expenditures totaling $713 billion in 2026, $907 billion in 2027, $991 billion in 2028, and more than $1 trillion in 2029. Demand for chips is therefore effectively guaranteed. The question, however, is whether corporate AI investments will prove as profitable as expected.

Francisco Quintana, ING’s head of market strategy, believes they will. He argues that enthusiasm for AI is unlikely to fade, though he adds a caveat. “The IPOs of Anthropic and OpenAI [two major AI players behind applications such as ChatGPT and Claude] will test investors’ appetite, but we expect demand to be sufficient, because the trend in recent years has been almost the opposite: companies choosing to leave public stock markets,” the expert reflects.

One factor helping to explain stock market behavior in 2026 is capital flows. Despite geopolitical tensions in the Persian Gulf, money has continued to flow into equity funds. José García-Zárate, head of research at Morningstar, explains that “there have been very strong inflows this year into U.S. equity exchange-traded funds [ETFs], unlike last year, when Europe attracted greater attention.”

For many experts, investment through passive ETFs that track stock market indices represents “dumb money,” because it does not distinguish between expensive and cheap stocks. Any capital flowing into such funds must be used to purchase the companies included in the index. At the same time, increasing numbers of retail investors are buying shares directly. In the SpaceX IPO, 25% of the shares offered were reserved for retail investors, making millions of people shareholders in a company trading at extraordinary valuation multiples and facing significant business uncertainty. Some industry professionals consider it a “meme stock.”

One message increasingly communicated by private banks to their wealthy clients is the need to diversify portfolios and reduce dependence on the volatile investment appetite surrounding AI. Mark Haefele, chief investment officer at UBS Global Wealth Management, believes that “investors should consider more defensive areas within the AI ecosystem, such as data center operators and certain payment companies, as well as other structural themes such as energy.”

Another factor affecting technology stock prices is interest rates. Because these companies are expected to generate large profits in the future, their current valuation depends heavily on the discount rate applied to future cash flows. Quintana of ING recalls that “at the end of 2025, several interest-rate cuts were expected, which would have provided a boost for technology stocks. However, the war in the Middle East has changed the stance of central bankers, and markets are now expecting rate increases.” In fact, the European Central Bank (ECB) has already begun raising rates.

The strategist sums up the situation with a football analogy: “Technology companies started the World Cup with the pitch heavily tilted in their favor, but now the field has leveled out.”

Sign up for our weekly newsletter to get more English-language news coverage from EL PAÍS USA Edition