Foreign giants reduce holdings in US tech stocks; AI investment profitability questioned

Recently, UBS Group and Goldman Sachs have respectively disclosed their Q4 2024 holdings reports, showing that the two major foreign investment giants significantly reduced their holdings in U.S. tech giants like Nvidia and Microsoft in the fourth quarter of last year, drawing market attention. Recently, aside from Nvidia, the “Big Seven” U.S. tech companies have all announced their latest earnings reports, which indicate that these giants plan to substantially increase capital expenditures. Since February, the stock prices of Amazon, Google, Meta, Microsoft, and others have experienced notable declines.

Industry insiders say that although there are currently doubts about the return on massive capital expenditures disclosed in earnings reports, this round of investment is essentially a strategic layout for future productivity, and the long-term industry trend of AI remains clear.

Multiple Tech Stocks Undergo Large-Scale Reductions

Recently, UBS Group and Goldman Sachs submitted their Q4 2024 holdings reports (13F) to the U.S. Securities and Exchange Commission (SEC), showing that both institutions have significantly reduced their holdings in several U.S. tech giants.

Specifically, in Q4 last year, UBS reduced its holdings of Nvidia by 10.042 million shares, an 11.47% decrease; Microsoft by 2.32 million shares, a 7.64% decrease; Apple by 5.267 million shares, a 10.57% decrease; Amazon by 1.658 million shares, a 4.57% decrease; and Google by 2 million shares, a 9.05% decrease. Additionally, UBS also reduced holdings in Micron Technology, Oracle, AMD, Western Digital, and other tech stocks to varying degrees.

According to Goldman Sachs’ latest disclosed holdings report, in Q4 last year, Goldman reduced its Microsoft holdings by 3.197 million shares, a 5.86% decrease; Tesla by 2.47 million shares, an 8.27% decrease; Broadcom by 3.433 million shares, a 9.33% decrease; and Meta by 2.414 million shares, a 13.51% decrease.

Doubts About AI Investment Realization

Recently, among the “Big Seven” U.S. tech giants, aside from Nvidia, the other six have all announced their latest earnings reports. According to these reports, the tech giants plan to significantly increase capital expenditures. Specifically, Meta stated that its total capital expenditure in 2026 could reach up to $135 billion, an increase of possibly 87%; Google plans to invest up to $185 billion this year; Alphabet announced a capital expenditure plan of $185 billion; Amazon also announced a capital expenditure plan of $200 billion for 2026.

Following the earnings disclosures, the stock prices of these companies mostly experienced some fluctuations. Data from Choice shows that as of February 11, Amazon, Google, Meta, and Microsoft have fallen by 14.72%, 8%, 6.67%, and 6% respectively since February.

A fund manager from a foreign public fund in Shanghai analyzed in an interview with the Shanghai Securities Journal that the core reason for the recent decline is that the market is beginning to doubt the return capability of the continuous AI investment by tech giants. Over the past two years, the market has tolerated high levels of AI-related capital spending; as long as companies announced increased investment, their stock prices often rose. However, investors are now paying more attention to clear ROI from AI investments and are worried about the sustainability of high spending without a clear closed-loop model.

“AI technology’s value ultimately needs to be realized through downstream applications that form a commercialized closed loop. Currently, except for a few vertical fields like programming, legal, and medical research, large-scale, highly penetrative ‘killer applications’ have not yet emerged, which raises doubts about the utilization rate and profitability prospects of upstream computing power and model investments,” the fund manager added.

However, some also believe that although the market has short-term doubts about the return on capital expenditures, this round of investment is fundamentally a strategic layout for future productivity, and the long-term industry trend of AI remains clear.

Sheng Jin, Chief Investment Officer of Wise Group, told the Shanghai Securities Journal that based on recent earnings reports from U.S. tech giants, AI is transitioning from a period of technological exploration to a new stage of large-scale application and infrastructure deepening. This stage shows clear trends such as continued increases in computing power investment, AI agents gradually becoming practical, and cloud services shifting focus from model operation to resource scheduling. The strategic and business directions of U.S. tech giants may evolve from “All in AI” to “AI in All.”

Dixinghua, a fund manager at Guohai Franklin Fund, told the Shanghai Securities Journal that recent earnings reports from tech giants reveal that the demand for computing resources in the tech industry continues to exceed market expectations. Whether large internet companies or independent cloud service providers, most report that computing power supply remains tight, and AI infrastructure is in a rapid demand expansion phase.

From a development perspective, the AI field currently exhibits two main characteristics: on one hand, leading companies continue to push model technological innovation, aiming to develop more powerful and cutting-edge large models; on the other hand, some companies are beginning to optimize costs and improve efficiency based on existing models, forming differentiated competitive strategies. Despite different paths, the entire tech industry is accelerating toward the goal of artificial general intelligence (AGI).

In Dixinghua’s view, the main bottleneck in AI development still lies at the computing power level, not only due to tight supply of core chips like GPUs but also involving supporting infrastructure such as power supply, storage capacity, and network bandwidth. He will continue to monitor new technological evolutions and breakthroughs, and explore growth potential investment opportunities along the industry chain.