第一财经·2025-11-15 05:28Core Viewpoint - The article discusses the recent sell-off of Nvidia shares by major investors, raising concerns about a potential AI bubble and market instability [3][4][5]. Group 1: Major Investor Actions - SoftBank Group sold its entire stake in Nvidia, cashing out $5.83 billion, which has sparked fears of an AI bubble bursting [3]. - Bridgewater, a leading hedge fund, significantly reduced its holdings in Nvidia by nearly two-thirds and also cut positions in Alphabet, Amazon, and Microsoft by over 50% [3]. - Citigroup disclosed similar reductions in its holdings of major tech stocks, including Nvidia, Microsoft, Apple, and Amazon, while raising Nvidia's target price from $210 to $220 [3]. Group 2: Concerns About AI Valuations - Industry leaders have expressed concerns about inflated market valuations, with DeepL's CEO warning of an impending bubble [4]. - Picsart's CEO noted that many AI companies are achieving high valuations without substantial revenue, raising red flags [5]. - Michael Burry, known for predicting the 2008 financial crisis, cautioned that major AI infrastructure providers may be underestimating chip depreciation costs, leading to potentially overstated profits [5]. Group 3: Counterarguments and Long-term Potential - Some industry professionals maintain a positive outlook on AI's long-term potential, arguing that current financial risks are justified by transformative technology [7]. - Lyft's CEO emphasized the importance of not being left behind in the technological wave brought by AI, despite current financial challenges [7]. Group 4: Future Projections and Market Dynamics - A report from Accel predicts that new AI data center capacity will reach 117 GW by 2030, necessitating nearly $4 trillion in capital expenditures and approximately $3.1 trillion in revenue to cover these costs [8]. - AMD's CEO forecasts that the data center chip and systems market could expand to $1 trillion by 2030, with annual revenues reaching $100 billion within five years [8]. - Nvidia's upcoming quarterly report is anticipated to be crucial in determining the trajectory of the AI supercycle and related expenditures [9].