The first half of 2026 marks a historic inflection point for global technology capitalization, characterized by an unprecedented wave of initial public offerings. On June 12, SpaceX is scheduled to list on NASDAQ with a valuation of $1.77 trillion, potentially establishing the largest stock issuance in human history.

Concurrently, Anthropic has confidentially submitted its IPO application with a valuation of $965 billion, while Hong Kong-listed entities Zhispu AI and MiniMax have seen their market values surge, with Zhispu briefly exceeding JD.com's valuation by more than 200%. These valuations persist despite a lack of traditional profitability; Zhispu reported revenue of only 724 million yuan in 2025 with ongoing losses, SpaceX recorded a net loss of $4.9 billion against $18.7 billion in revenue, and Anthropic's annual recurring revenue, though exceeding $10 billion, supports a near-$1 trillion price tag. Data compiled by Woofun AI indicates that the market is pricing in a paradigm shift where AI represents the most significant platform transition since the Internet era, driven by a consensus that these firms will become trillion-dollar giants with unlimited industrial substitution potential.

The primary driver of these exorbitant valuations is a severe scarcity of tradable shares coupled with overwhelming demand. Zhispu issued just over 37 million H shares, creating an extremely thin float, while SpaceX's current offering represents only 4.2% of its total share capital. This supply-demand mismatch has pushed prices beyond the reach of traditional valuation models.

However, a rigorous analysis of substitutable output value reveals significant discrepancies. In the realm of AI coding, China's total enterprise R&D expenditure stands at approximately 3.4 trillion yuan. Even under optimistic assumptions where 30%-40% relates to software development and 70% is automated by AI, the maximum substitutable output value ranges from 700 billion to 1 trillion yuan. At a market value of roughly 460 billion yuan, Zhispu AI alone would account for 46%-66% of this theoretical ceiling. Similarly, in the SaaS sector, where US annual revenue is approximately $200 billion (1.4 to 1.5 trillion yuan), Zhispu's $64 billion valuation represents roughly 30% of the entire US market, despite China's SaaS ecosystem being less mature. Woofun AI notes that when combining these scenarios with multimodal content creation, which may only reduce production costs by 200 billion to 300 billion yuan, the total substitutable value across all sectors reaches approximately 2.4 to 2.8 trillion yuan. Even in this extreme case, Zhispu's valuation would constitute 16%-19% of the total, ignoring the competitive fragmentation of budget resources among numerous AI firms.

Across the Atlantic, the valuation logic shifts from arithmetic substitution to infrastructure faith. SpaceX's Starlink division reported 10.3 million users and $11.3 billion in revenue for 2025, with a profit margin of 39%, yet its $1.77 trillion valuation implies a price-to-sales ratio nearly 100 times its revenue. Industry analysts argue that the market is pricing the establishment of future space exploration infrastructure, encompassing logistics, resource valuation, and institutional frameworks, rather than immediate computing power. Anthropic's valuation surge is anchored in its explosive growth in annual recurring revenue, which jumped from $9 billion to $44 billion within two months, driven by dominance in the AI coding sector and a strategic position in safe AI. OpenAI, valued at $852 billion with 900 million active users, faces a different challenge; despite its comprehensive product portfolio, it currently lacks the specific 'imagination' premium driving higher valuations for its peers. The term 'faith' in this context is not blind optimism but a calculated bet on future causal chains, exemplified by SoftBank's strategic pivot.

SoftBank's resurgence illustrates the mechanics of this capital allocation. By June 2026, founder Masayoshi Son's wealth reached $100.7 billion, propelling SoftBank's market value to 49.3 trillion yen and overtaking Toyota Motor Corporation after two decades of dominance. Son's strategy involves a vertically integrated chain: investing $47.1 billion in ARM, which saw its stock price surge 268% in 2026 due to AI-driven server CPU demand; acquiring Ampere Computing for $6.5 billion to complete the data center CPU ecosystem; and investing over $64 billion in OpenAI for a 13% stake, generating $45 billion in returns by March 2026.

Furthermore, SoftBank committed $500 billion to the US Stargate project alongside OpenAI and Oracle, and initiated a €750 million French data center project targeting 3.1 gigawatts of computing power by 2031. Woofun AI analysis suggests that this vertical integration from silicon to application has created a self-reinforcing valuation loop, with SoftBank's stock rising over 140% in 2026 and net profit hitting a record 550.8 billion yen. Son asserts that the AI revolution's scale could be 50 times larger than the Internet era, comparing the current trajectory to the post-1929 industrial mechanization boom.

However, the fragility of these valuations is exposed by impending liquidity shocks. In July 2026, the lock-up period for 25.86 million shares held by Zhispu's cornerstone investors expired, representing 5.76% of total capital. Given Zhispu's initial liquidity rate of only 2.67%, this release tripled the tradable share volume overnight. The impact on MiniMax was even more severe; on July 8, nearly 50% of its shares became tradable as lock-ups for existing shareholders and cornerstone investors expired, causing the outstanding share volume to surge tenfold from less than 6% to nearly 56%. Historical precedents offer little comfort; Kuaishou's stock dropped 15.3% upon share release in 2021, and Xiaomi fell 7.5% in 2019 before hitting new lows. Unlike those cases where prices had already corrected, Zhispu and MiniMax were trading at 10x and 3x their issue prices respectively, creating massive profit-taking incentives for early investors. While inclusion in the Hang Seng Technology Index forced passive funds to buy approximately $1.25 billion to $1.75 billion of these stocks, and southbound funds were poised to inject HK$18.6 billion into Zhispu, the dilution of the 'scarcity premium' fundamentally alters the pricing dynamic.

The broader market faces a $200 billion fundraising wave from SpaceX, OpenAI, and Anthropic, equivalent to the total US IPO volume of the past three to four years, potentially realized within five months. SpaceX aims to raise $75 billion, while OpenAI and Anthropic target $60 billion to $90 billion each. This influx necessitates capital reallocation, forcing asset managers to trim holdings in other tech firms. While historical IPO booms often result in short-term panic rather than structural collapse, the current environment differs from the 2000 dot-com bubble. In 2000, VC investment peaked at $105 billion, with no single company losing more than $1.5 billion annually, leading to a 78% NASDAQ crash. Today's AI firms generate substantial, rapidly growing revenues and are backed by giants like Microsoft, Google, and Amazon. Nevertheless, Morgan Stanley analysts warn that the capital expenditure-to-revenue ratio for Meta, Microsoft, and Alphabet, ranging from 21% to 35%, has already exceeded AT&T's levels during the telecom bubble peak. In China, valuations continue to escalate, with DeepSeek approaching $60 billion and Jiebu Xingchen seeking a $2.5 billion Pre-IPO round. As the window for this 'golden age' remains open but finite, the market is engaged in a high-stakes race against time, balancing the potential for exponential growth against the looming risks of a liquidity-driven correction.