The market capitalisation of the so-called Magnificent 7 group of major U.S. technology stocks has fallen by $2.3 trillion in June.

On June 30, CNBC reported that the market is reassessing whether big tech can recoup returns from large-scale artificial intelligence (AI) infrastructure investment, while still allocating funds to the semiconductor sector.

The Magnificent 7 consists of Microsoft (MS), Nvidia, Alphabet, Apple, Meta, Tesla and Amazon. The CNBC Magnificent 7 Index has fallen 10 percent in June. Among them, Amazon, MS, Alphabet and Meta are spending hundreds of billions of dollars to secure large quantities of chips and build data centres to expand AI services. Some investment is being financed through debt.

Market attention is focused on when this spending will translate into results. The second-quarter earnings season beginning in July is expected to be the first turning point. Dan Ives (댄 아이브스), managing director at Wedbush Securities, said technology stock investors are trying to re-verify expanded investment in the AI revolution ahead of the second-quarter earnings season, and anxiety over the cost burden could continue for some time.

Losses by stock were also significant. MS is down 20 percent in June, and Nvidia has fallen about 13 percent. Apple and Amazon also slipped about 8 percent each. Some in the market also say the investment narrative around the Magnificent 7 is changing.

Tom Lee (톰 리), head of research at Fundstrat Global Advisors, said the market is trying to understand a new narrative for the Magnificent 7, and these companies are shifting from asset-light cash-generating firms to ones with heavier balance-sheet burdens. He added that this spending is ultimately meant to replace human work with AI and, over time, investors will begin to view it as a moat, describing the current moment as a phase in which the narrative is shifting.

Semiconductor stocks, by contrast, were relatively strong over the same period. The Philadelphia Semiconductor Index, which includes TSMC, Micron and ASML, rose about 6 percent in June. On a year-to-date basis, it has risen more than 90 percent, contrasting with a 3.4 percent fall in the Magnificent 7 over the same period.

Behind this is big tech's expanded AI investment. As large platform companies bought semiconductors in large quantities, supply shortages persisted, and the impact spread across the supply chain from component suppliers to manufacturers. Memory was singled out as a key bottleneck, with prices surging due to shortages. The Roundhill Memory ETF, which tracks SK Hynix and Samsung Electronics, is up 166 percent this year.

Earnings also supported the semiconductor strength. Duncan Toms (덩컨 톰스), a multi-asset strategist at HSBC, said last week's Micron results "threw cold water" on AI scepticism and showed clear evidence that the demand base for AI is still alive and healthy. UBS also saw no sign that AI supply-chain bottlenecks are easing, and expected cloud revenue growth at major platforms to accelerate further over the rest of this year.

UBS said this trend shows that the fundamentals underpinning the AI growth narrative are solid. It added that even if, over the long term, exposure to AI-related stocks becomes a key factor separating stock market performance, diversified investment spanning both within and outside AI is also needed.

Ultimately, the market appears to be reallocating capital based not on denying big tech's AI spending itself, but on who can first show clear benefits in earnings. In the second-quarter earnings season in July, key variables are emerging, including signals of payback from investment by large platforms and how long semiconductor supply shortages persist.