AI investment enters the "fine cultivation era"! J.P. Morgan: All transactions become AI transactions, the future "how to invest" is more important than "whether to invest"

As the concept of artificial intelligence (AI) permeates every corner of the market, investors' understanding of the AI theme is shifting from the frenzy of "buying everything that goes up" to a more nuanced approach. David Lebovitz, global market strategist at J.P. Morgan Asset Management, stated that investors are becoming increasingly adept at distinguishing the potential risks and returns in different aspects of the AI field, which is a crucial lesson for Wall Street. "People are starting to make more distinctions between different sectors, identifying where there may be oversupply and where there is strong demand," Lebovitz said in an interview on Thursday. "It is no longer the mindset of 'we like AI, so we should buy everything related to AI'." Lebovitz specifically pointed out significant structural differences within the AI industry chain. In his view, the demand for data center construction and operation will have more "structural" significance than semiconductor and hardware production - chips may be at risk of oversupply. "My concern about supply risks mainly lies in the areas of chips and hardware," Lebovitz said. "This is clearly where investors are most enthusiastic, and historical experience shows that when enthusiasm is high, people tend to get carried away." This statement comes as the AI chip sector experiences severe volatility. Previously soaring chip stocks have recently seen significant selling, signaling a cooling of the market's enthusiasm for AI infrastructure investment. Recently, South Korean chip giant SK Hynix's stock price has dropped by over 20% from its historical high in June, although it has still accumulated over 200% since 2026. The company was scheduled to debut on the Nasdaq this week, but the sell-off of AI chips cast a shadow over this highly anticipated event. Even more shocking to the market is the case of Samsung Electronics. On July 7th, Samsung Electronics announced record-high guidance for the second quarter - with quarterly consolidated revenue of approximately 171 trillion Korean won and operating profit of approximately 89.4 trillion Korean won. However, this record-breaking performance not only failed to boost the stock price, but instead triggered a global chip sell-off. On that day, both Samsung Electronics and SK Hynix plummeted, with the Korean KOSPI index dropping more than 8% at one point, triggering circuit breakers. The market generally sees this as a typical case of "buy the rumor, sell the fact" - months of increases have already priced in all expectations for economic prospects, and when actual performance comes out, it becomes a window to take profits. In addition, the news that Meta is planning to sell excess AI computing power has also raised concerns in the market about possible excess demand for AI infrastructure. Earlier, in another research report, J.P. Morgan pointed out that in the past year, AI chip and storage manufacturers' stock prices have significantly outperformed those of mega-scale cloud service providers such as Amazon.com, Inc., Microsoft Corporation, and Alphabet Inc. Class C. The report suggested that the "income gap" has widened to unsustainable levels. It believes that the future of the AI market depends not on capital expenditure, but on commercialization and realization capabilities. Despite short-term pressure on the chip sector, J.P. Morgan remains optimistic about the long-term prospects of AI infrastructure. The firm predicts that the capital expenditures of the four major cloud service providers in the United States will grow by 80% year-on-year in 2026, surpassing $575 billion. J.P. Morgan even raised its forecast for the total global investment in AI infrastructure construction by 2030 to $5.5 trillion. Lebovitz pointed out that as the AI landscape rapidly expands, the key question for investors is no longer "whether to invest in AI" but "how to invest specifically." "I realize that everything is turning into AI trading," he said. "AI is everywhere. So the key is how you participate, not whether to participate. I believe this is the biggest issue that investors will face in the future."