Besides SoftBank, tech visionary Peter Thiel has also liquidated his Nvidia stake: Is AI market consolidation now imminent?

Strategic portfolio shifts in financial markets: An economic analysis of the AI ​​bubble debate and institutional investment decisions

Global financial markets are in a period of profound contradictions, characterized by a growing gap between public AI euphoria and the calculated actions of the most influential investors. While artificial intelligence, as the dominant trend, is channeling trillions of dollars into new infrastructure projects, leading figures in the industry are executing a remarkable strategic retreat: They are systematically reducing their positions in the very companies considered to be the primary beneficiaries of the AI ​​boom. This behavior raises a crucial question: Are we witnessing a long-overdue reality check that precedes a massive market correction?

The trend is particularly evident in the trading activity during the third quarter of 2025. Prominent figures like tech visionary Peter Thiel, who liquidated his entire Nvidia stake, and SoftBank, which also sold its complete position in the chip giant, are sending a strong signal. These moves are more than just simple profit-taking; they indicate a deliberate reassessment of risk in a market environment where warnings of an AI bubble are growing louder. While the masses of investors, driven by momentum effects, continue to flock to the market, the pioneers who recognized the rise early on now seem to believe the time has come to exit – a development that fundamentally calls into question the sustainability of the current optimism.

Suitable for:

• Signs of crisis or strategy? Softbank unexpectedly sells all Nvidia shares: Background and consequences

The quiet retreat of major investors: Between AI euphoria and reality check

Global financial markets are in a fascinating phase characterized by apparent contradictions. While artificial intelligence dominates public discourse as a trending topic and trillions of dollars are flowing into AI-related infrastructure, major investors at the forefront of the industry are pursuing a striking strategy: they are systematically reducing their positions in the companies considered to be the primary beneficiaries of the AI ​​boom.

This trend is particularly evident in the trading activity of hedge funds and venture capitalists in the third quarter of 2025. Peter Thiel, one of Silicon Valley's most influential technology investors, liquidated his hedge fund Thiel Macro LLC's entire stake in Nvidia. The 537,742 shares sold would have had a market value of approximately $100 million on September 30, 2025. While this amount seems minuscule compared to Nvidia's total market capitalization of roughly $4.5 trillion, the move signals something far more significant: a loss of confidence in a man whose investment instincts have been scrutinized for decades.

Thiel's decision came within a broader context, one markedly different from the confidence of the early AI era. The sales coincided with a period in which warnings about a potentially inflating AI bubble grew increasingly louder. Not only were speculative analysts raising concerns, but established financial institutions like the International Monetary Fund were also warning of high valuations and a growing risk of concentration in the stock market. A Bank of America survey of fund managers revealed that 54 percent considered an AI bubble the greatest risk facing global financial markets.

Even more telling is the parallel action by SoftBank, one of the world's largest and most aggressive technology investors. The Japanese conglomerate, led by Masayoshi Son, liquidated its entire Nvidia stake of 32.1 million shares for $5.83 billion in October 2025. This is particularly noteworthy because SoftBank had at one time been Nvidia's largest single shareholder. In 2019, Son had already made a historic mistake: he sold his then-current Nvidia shares for $3.6 billion. Had he held onto those shares, his stake would be worth over $150 billion today. The renewed exit from Nvidia in 2025 therefore suggests a deliberate balancing act between taking profits and offloading a potentially over-insured asset.

The official reason given for SoftBank's sale is liquidity. The company needed capital for planned financing, particularly for a proposed $30 billion investment in OpenAI and the $6.5 billion acquisition of chip designer Ampere Computing. Son emphasizes that the Nvidia sale has nothing to do with fundamental concerns about the company itself. However, this portrayal can only be considered part of the story. A manager who believes in AI and wants to make even bigger bets in AI could have used alternative sources of funding. The fact that Nvidia shares were used to raise liquidity suggests that Son considered the timing favorable.

Peter Thiel's portfolio realignment: A pattern of caution

Particularly revealing is the way Peter Thiel restructured his entire hedge fund portfolio in 2025. This wasn't simply a one-off sale, but a systematic repositioning program that should raise concerns among investors. In the first quarter of 2025, Thiel sold all of his Amazon holdings and established new positions in Microsoft, Vistra, and ASML. In the second quarter, he reduced these positions, selling both Microsoft and ASML. In the third quarter, he divested Vistra, returned to Microsoft, and established an entirely new position in Apple shares.

These movements reveal a portfolio management approach that is difficult to explain based solely on fundamental belief in the companies involved. Instead, the rebalancing suggests tactical positioning strategies, possibly as an attempt to profit from short-term market movements or to hedge against risks.

One notable constant is Tesla. It's the only stock to have appeared in each of the last four quarterly reports. Thiel's relationship with Elon Musk dates back to their time together at PayPal, where they worked in the late 1990s and early 2000s. However, the history of their relationship is complicated and fraught with well-known tensions, as Thiel temporarily replaced Musk as PayPal CEO during his honeymoon, an experience Musk later found painful. Despite these historical turbulences, Thiel later supported Musk's SpaceX company with a crucial $20 million investment that was central to the company's survival in 2008 after several failed rocket launches.

In the third quarter of 2025, Thiel reduced his Tesla position by 75 percent, after having quadrupled it in the second quarter. This volatile treatment of a supposedly strategic investment suggests that Thiel's portfolio is currently viewed as a means of short- to medium-term risk management, rather than as a long-term wealth accumulation strategy.

Our global industry and business expertise in business development, sales and marketing - Image: Xpert.Digital

Industry focus: B2B, digitalization (from AI to XR), mechanical engineering, logistics, renewable energies and industry

More about it here:

• Xpert Business Hub

A topic hub with insights and expertise:

• Knowledge platform on the global and regional economy, innovation and industry-specific trends
• Collection of analyses, impulses and background information from our focus areas
• A place for expertise and information on current developments in business and technology
• Topic hub for companies that want to learn about markets, digitalization and industry innovations

The macroeconomic irrationality of the current AI investment boom

The discrepancy between investments in AI infrastructure and the actual economic results achieved is reaching alarming proportions. Analysts predict that by 2030, approximately $7 trillion will be poured into the development and expansion of AI infrastructure. This sum exceeds the entire gross domestic product of many developed countries. Meta, led by Mark Zuckerberg, alone plans to invest around $600 billion in AI initiatives by 2028. OpenAI CEO Sam Altman is planning even more massive investments, and the Stargate data center project, involving Trump, Oracle, OpenAI, and the Emirate of Abu Dhabi, is slated to receive up to $500 billion in funding.

Despite these monumental investments, reliable evidence of a return on investment is still lacking. A detailed analysis by Forrester Research indicates that a major technology provider will reduce its AI infrastructure investments by 25 percent in 2025, driven by supply chain issues, unmet expectations, and investor pressure. This signals waning expectations regarding the immediate economic benefits of these massive capital expenditures.

The empirical data is also promising. Although billions of dollars flowed into AI infrastructure and generative AI technologies in 2023, only 20 percent of companies reported profit increases through AI in 2024. This suggests that the widespread penetration of AI technologies is falling far short of expectations, especially considering the enormous investments.

Another statistical element supports this cautious view. Between March 2023 and December 2024, the S&P 500 experienced a dramatic rally, with valuations of mega-cap technology companies multiplying. While the historical comparison to the dot-com bubble isn't entirely apt, as valuations back then were more extreme, today's valuations are also substantial. Tesla has a price-to-earnings ratio of approximately 120 projected for 2025, while Nvidia, with a ratio of around 45, is also significantly above its historical average. The S&P 500 as a whole is valued at a projected price-to-earnings ratio of 24, roughly double its long-term average.

Suitable for:

• The $57 billion miscalculation – NVIDIA of all companies warns: The AI ​​industry has backed the wrong horse

The systematic redistribution of risk and the signs of market maturity

Looking at the actions of large investors reveals a pattern that extends beyond specific investment decisions. It is a systemic redistribution of risk and concentration. The US stock market is becoming increasingly concentrated. Just 155 stocks in the S&P 500 account for roughly 70 percent of the total market capitalization, a number that stood at 274 companies about ten years ago. The ten largest companies make up about 35 percent of the S&P 500.

This concentration is not inherently pathological if justified by fundamental superiority. However, the sharp focus on AI narratives suggests that a significant portion of this valuation premium is based on collective optimism and momentum effects, not on hard fundamental realities. Historical experience shows that such periods of concentration typically do not last long, and that market rotations have recurred both in the past and will continue to do so in the future.

The cautious stance of renowned investors like Peter Thiel and Masayoshi Son towards Nvidia is not a criticism of the company's fundamental strength, which is indeed a market leader in its fields. Rather, it reflects a rational assessment of risks, valuations, and market dynamics. In an environment where 54 percent of professional fund managers see an AI bubble as the greatest systemic risk, a portfolio realignment at opportune times seems logical.

The fundamental puzzle of long-term AI profitability

A key economic puzzle remains unsolved: How will the enormous capital investments in AI infrastructure be transformed into proportional profits in the long term? The history of technological transformations, from the railways to electrification to the internet, reveals a familiar pattern. Technologies that enable fundamental productivity gains typically require decades to realize their full economic value. Furthermore, they need extensive complementary infrastructure, training, and organizational changes to achieve their full potential.

Added to this is the historical phenomenon of overinvestment. After technological surpluses, such as the railroad boom or the dot-com bubble, the crash typically leaves behind a massive oversupply of cheap infrastructure. While this later enabled innovative companies to utilize these resources at low cost, ultimately leading to productive applications, it meant massive capital losses for the shareholders who financed the initial enormous capital requirements.

Another critical factor is energy efficiency and the associated operating costs. AI data centers are extremely energy-intensive, and energy costs are becoming a major expense. As technology companies seek to secure supply contracts with renewable energy providers, massive expansion is creating competition for resources, which can drive up energy costs. This could ultimately significantly impact the profitability of AI services, especially if pricing for AI services comes under pressure.

Warning signs and institutional caution

Several institutional warning signs point to a growing awareness of inflated expectations. Jeff Bezos, the founder of Amazon, warned in October 2025 of a potential AI bubble, drawing explicit parallels to the dot-com crisis. This is noteworthy because Bezos is not a notorious pessimist and issues his warnings selectively. David Einhorn, a renowned hedge fund manager, has been quoted as saying that current investment figures are so extreme they are barely comprehensible. He sees the risk of massive capital destruction even within the current investment cycle.

Even at Nvidia, the company that has benefited most from the AI ​​boom, there are signs of internal caution. CEO Jensen Huang has repeatedly divested himself of stock holdings, selling Nvidia shares a total of 29 times in 2025. While these sales are partly automated and related to options, the fact that even the executives of the biggest winners of the AI ​​boom are reducing their positions remains a psychological signal to the market.

Suitable for:

• AI chip hype meets reality: The future of data centers – in-house development versus market saturation

The dynamics of institutional capital withdrawal

The fact that major hedge funds and institutional investors are systematically reducing their positions in AI darlings while the market as a whole continues to reach new highs suggests a two-stage phenomenon. First, smaller investors and retail investors, who are invested in mega-cap technology stocks through ETFs and other passive indexing instruments, benefit from the inflows of institutional money. This creates a technical boost that drives valuations higher. Second, these inflows are increasingly transforming into a momentum effect, where valuations drive themselves upward, regardless of improvements in fundamentals.

This is a classic bubble pattern. The most intelligent and best-informed capital allocators use their superior information to liquidate their positions at favorable prices. Meanwhile, less sophisticated investors, who operate by tracking an index, remain locked into these positions. This dynamic was observed during the dot-com bubble of the late 1990s, leading to extreme valuation ordinals and culminating in a dramatic crash.

Economic reality versus market psychology

The strategic decisions made by Peter Thiel, SoftBank, and other major investors in the third quarter of 2025 point to a deeper understanding of potential market overreaction. AI technology is real, and its long-term economic impact could indeed be transformative. However, current pricing of this future potential already reflects an enormously optimistic scenario. There simply isn't enough room for error if the expected returns from planned investments slow by a third or more.

Reducing positions in Nvidia and other high-flying stocks at prices that were on the verge of reaching new all-time highs is rational from an economic perspective. It doesn't necessarily reflect a misunderstanding of AI's potential, but rather a sober assessment of risks, valuations, and likely returns. In a market where valuations are twice as high as historical averages and concentration is at an unprecedented level, it seems wise to take this opportunity to realize profits and recalibrate the portfolio.

Independent AI platforms as a strategic alternative for European companies - Image: Xpert.Digital

Ki-Gamechanger: The most flexible AI platform-tailor-made solutions that reduce costs, improve their decisions and increase efficiency

Independent AI platform: Integrates all relevant company data sources

• Fast AI integration: tailor-made AI solutions for companies in hours or days instead of months
• Flexible infrastructure: cloud-based or hosting in your own data center (Germany, Europe, free choice of location)

• Highest data security: Use in law firms is the safe evidence
• Use across a wide variety of company data sources
• Choice of your own or various AI models (DE, EU, USA, CN)

More about it here:

• Independent AI platforms vs. hyperscalers: Which solution is right for you?

Konrad Wolfenstein

I would be happy to serve as your personal advisor.

contact me under Wolfenstein ∂ Xpert.digital

call me under +49 89 674 804 (Munich)

LinkedIn
 

Benefit from Xpert.Digital's extensive, fivefold expertise in a comprehensive service package | R&D, XR, PR & Digital Visibility Optimization - Image: Xpert.Digital

Xpert.Digital has in-depth knowledge of various industries. This allows us to develop tailor-made strategies that are tailored precisely to the requirements and challenges of your specific market segment. By continually analyzing market trends and following industry developments, we can act with foresight and offer innovative solutions. Through the combination of experience and knowledge, we generate added value and give our customers a decisive competitive advantage.

More about it here:

• Use the 5x expertise of Xpert.Digital in one package - starting at just €500/month