Over the past twenty years, the financial markets have experienced several speculative booms. From the dot-com bubble to the SPAC craze, via cryptocurrencies and the era of zero interest rates, each cycle has been accompanied by a dominant narrative capable of attracting capital, generating extraordinary expectations and temporarily redefining the rules of the game. However, the potential wave of listings aimed at bringing SpaceX, OpenAI and Anthropic to the public markets in 2026 has characteristics that set it apart from any previous phenomenon. We are not merely facing a new phase of technological enthusiasm, but a potentially historic shift of capital towards the infrastructure that could underpin the artificial intelligence economy for the next decade. According to Reuters (2026a), SpaceX is reportedly considering a listing with a valuation exceeding $1.75 trillion, whilst Reuters (2026c), Fortune (2026) and Yahoo Finance (2026a) indicate that OpenAI and Anthropic could reach valuations close to $1 trillion (Anthropic was the first to apply for a listing, indicating a valuation of $965 billion, ed.). Overall, these transactions could absorb nearly $200 billion in new capital, a figure which, according to Bloomberg Television (2026) and Tom Tunguz (2026), is unprecedented in the recent history of financial markets.
At first glance, the phenomenon appears to stem from the classic ‘Fear of Missing Out’, or FOMO. Investors are watching companies that dominate the conversation on artificial intelligence, control strategic infrastructure and are recording extraordinary growth rates. The psychological consequence seems almost inevitable: the fear of being left out of the next major economic transformation. However, to stop at this interpretation would be to underestimate the nature of the phenomenon. Unlike the dot-com bubble of the late 1990s, the companies at the forefront of the current wave are not start-ups with no revenue or business models yet to be validated. SpaceX holds a dominant position in the commercial space sector and continues to expand the Starlink network; OpenAI and Anthropic are recording growth levels that Deloitte (2026), PitchBook (2026) and Stanford HAI (2026) identify as unprecedented in the history of enterprise software. Anthropic is said to have accelerated its annualised run rate to exceed $40 billion (47 to be precise, ed.), whilst Starlink is said to have surpassed ten million active subscribers during 2026.
For this reason, the trend emerging in institutional markets appears to be closer to what we might call ‘Fear of Losing Out’, or FOLO. In this case, the concern is not about the possibility of missing out on a short-term speculative opportunity, but rather the risk of not holding a stake in what could become the fundamental infrastructure of the future digital economy. If artificial intelligence does indeed become the dominant technological platform of the 21st century, many asset managers fear that a lack of exposure to the key players could result in a structural competitive disadvantage for their portfolios.
This perception is helping to drive a concentration of capital that is beginning to resemble some of the major speculative booms in financial history. According to Bank of America, the growing influence of large technology groups could push the concentration of the technology sector within the S&P 500 to levels higher than those seen during some of the most famous periods of financial euphoria in the past. Although the economic structure of the companies involved is vastly different from that of dot-com firms, the psychological mechanism driving the investment rush bears striking similarities.
As is often the case during periods of intense enthusiasm, there are, however, certain signs that warrant particular attention. Reuters (2026d) reports that SpaceX is reportedly considering unconventional share allocation structures, whilst the report highlights that hundreds of OpenAI employees and former employees are said to have cashed in a significant portion of their holdings through secondary market transactions ahead of a potential listing. Historically, when insiders begin to shift risk from private to public markets, investors should question not so much the quality of the company as the expectations embedded in valuations. This does not necessarily imply that valuations are excessive, but it suggests that the market may be in a phase where expectations are growing faster than economic fundamentals.
There is, however, an even more important issue that receives relatively little attention. Where will the capital needed to finance these operations come from? Markets often tend to assume that liquidity is an infinite resource, but the reality is quite different. Every dollar invested in a new IPO must necessarily be diverted from another investment opportunity. As discussed in my article The Great Capital Migration (Agostini, 2026), artificial intelligence is not simply creating new winners, but is redefining global priorities for capital allocation. According to Bloomberg, some international capital flows are already shifting away from emerging markets and traditional sectors towards companies perceived as leading the artificial intelligence revolution. This process risks creating a growing polarisation between companies considered strategic for the digital future and all other investment categories.
The real challenge, however, may not be financial. It may be physical. In recent years, the debate on artificial intelligence has focused mainly on models, algorithms and applications. Yet numerous studies published by the International Energy Agency (2026), the OECD (2026) and the US Department of Energy (2026) show that the growth of AI increasingly depends on the availability of energy, advanced semiconductors, computational capacity and digital infrastructure. In this context, the narrative that software is the main limiting factor risks becoming misleading. As I have argued in several articles on the relationship between AI, energy and data centres, the real bottleneck of the artificial intelligence revolution may lie in the deepest layers of the technology stack.
Statements by Nvidia CEO Jensen Huang appear to support this interpretation. Reuters (2026e) reports that the market continues to operate under conditions of supply shortages for many critical components of AI infrastructure. Meanwhile, Reuters (2026b) estimates that spending on AI infrastructure could reach $600 billion, whilst Alphabet is said to have launched initiatives to raise up to $80 billion to expand its computational capabilities (Alphabet, 2026; Axios, 2026). If even hyperscalers with vast financial resources require additional resources to sustain the race for artificial intelligence, it is clear that the competition concerns not only software, but also access to energy, chips, data centres and network capacity.
This infrastructure pressure also comes against an increasingly complex geopolitical backdrop. Tensions in the Middle East, uncertainty over energy supplies and rising volatility in commodity markets are contributing to a less favourable environment for investments that require huge amounts of capital and increasing energy consumption. According to analyses by the Financial Times (2026), Pimco (2026) and other sources, rising energy prices and bond yields could exert significant pressure on the valuations of long-duration technology assets.
In light of these developments, the most important question is not whether SpaceX, OpenAI and Anthropic are extraordinary companies. Many indicators suggest that they indeed are. The real question, rather, concerns the markets’ ability to distinguish between structural value and collective enthusiasm. Current valuations appear to incorporate an extremely favourable scenario, characterised by sustained growth, adequate energy supply, falling computational costs and geopolitical stability. Economic history teaches us, however, that all these factors rarely evolve simultaneously in the desired direction.
To describe this phase as merely a tech bubble would probably be an oversimplification. At the same time, viewing it as inevitably destined for success could prove just as dangerous. The IPOs of SpaceX, OpenAI and Anthropic perhaps represent something more significant: the moment when some of the digital economy’s most important infrastructure shifts from private capital to the public market. For investors, entrepreneurs and policymakers, the challenge will not be deciding whether to participate in this transformation, but understanding which elements generate real value and which are instead the result of expectations that may prove difficult to sustain in the long term.
Ultimately, FOMO may well be the psychological driver fuelling this new wave of IPOs. However, behind the narrative of artificial intelligence lies a much deeper issue: the redefinition of the infrastructure, capital flows and economic balances that will underpin the next phase of global growth. And as with every major technological transformation, success will depend not only on innovation, but on the ability to align technology, energy, capital and governance within an economically sustainable system. (photo by Chenyu Guan on Unsplash)
References
Agostini, M. (2026, August 28). The great capital migration: How artificial intelligence is repricing venture capital, reshaping private credit, and redefining growth. Medium. https://medium.com/@tarifabeach/the-great-capital-migration-how-artificial-intelligence-is-repricing-venture-capital-reshaping-274de35221f1
Alphabet Inc. (2026). Alphabet announces proposed $80 billion equity capital raise to expand AI infrastructure and compute. https://abc.xyz/investor
Axios. (2026). Alphabet seeks $80 billion to fund AI buildout. https://www.axios.com
Bloomberg Television. (2026). 2026 IPO boom: SpaceX, OpenAI and Anthropic [Video]. https://www.youtube.com/watch?v=E-vnEbJnYz8
Deloitte. (2026). State of Generative AI in the Enterprise 2026. https://www2.deloitte.com
Financial Times. (2026). Commodities in ‘super-squeeze’ as Hormuz risks build, HSBC says. https://www.ft.com
Fortune. (2026). OpenAI’s potential IPO and the trillion-dollar valuation question. https://fortune.com
International Energy Agency. (2026). Energy and AI 2026. https://www.iea.org
OECD. (2026). Artificial intelligence, compute infrastructure and economic competitiveness. https://www.oecd.org
Pimco. (2026). Treasury yields driven by Fed bets, not AI, for now. https://www.pimco.com
PitchBook. (2026). Venture Monitor Q1 2026. https://pitchbook.com
Reuters. (2026a). SpaceX could seek IPO valuation of over $1.75 trillion. https://www.reuters.com
Reuters. (2026b). Big Tech investors gauge payoff as AI spending set to hit $600 billion. https://www.reuters.com
Reuters. (2026c). AI giant Anthropic confidentially files for U.S. IPO. https://www.reuters.com
Reuters. (2026d). SpaceX sets aside 5% of IPO shares for selected buyers, waives lock-up. https://www.reuters.com
Reuters. (2026e). Nvidia chief to Asia: We’re still supply constrained. https://www.reuters.com
Stanford Institute for Human-Centered Artificial Intelligence. (2026). AI Index Report 2026. https://hai.stanford.edu/ai-index
U.S. Department of Energy. (2026). Electricity demand implications of artificial intelligence infrastructure. https://www.energy.gov
Yahoo Finance. (2026). SpaceX, OpenAI and Anthropic: Here are the most anticipated IPOs in 2026. https://finance.yahoo.com
ALL RIGHTS RESERVED ©
