According to Woofun AI, Arthur Hayes, co-founder of virtual currency derivatives trading platform BitMEX and chief investment officer of investment company Maelstrom, updated his blog on the 5th to issue a severe warning that large-scale initial public offerings (IPOs) in the AI field could cause drastic stock price dilution, and indicated that this risk will directly affect crypto assets such as Bitcoin. Hayes' core argument is that as restricted shares of SpaceX (SPCX.US) are lifted one after another, and technology giants such as Anthropic and OpenAI are about to be listed, stocks held by insiders will pour into the market after the lockdown period ends, causing the stock price to fall below the issue price, which will have a profound impact on US dollar liquidity and the crypto market.
Hayes compared the current situation to Zcash (ZEC), a privacy coin launched on October 28, 2016. The initial supply of this token's genesis block was zero, and only the portion that was mined and sold by miners on the Poloniex exchange was extremely high in the early stages.
Although he acknowledged that ZEC's market capitalization was at its highest level in history, he pointed out that many investors who bought in the early stages of the transaction are still facing book losses. He emphasized that traders often focus on the price of each token rather than the total market capitalization, and this perception bias will be magnified in AI giants' IPOs. Take SpaceX, which was listed on the NASDAQ on June 12. Its issue price was $135. It opened at $150 on the first day of listing and closed at $160.95. Since then, it has broken through $190 and has now fallen back close to the issue price. Hayes expects an additional 8.6 billion shares to be sold by the end of June 2027, which is equivalent to an additional 50% increase in existing tradable shares, and the share price will drop further by about 50% as a result. Most of the new shares are due to the redemption of stock rights held by CEO Elon Musk. For Anthropic, which submitted S-1 registration documents to the US Securities and Exchange Commission on October 26, Hayes gave a more pessimistic forecast: these companies will face an 83% dilution of their shares by June 2027.
If we calculate the total market value of $1.5 trillion at the time of listing, even if the stock price remains the same, it would require about $1245 billion in additional capital. Since there was no mention of any dividend plan in the S-1 document, he deduced from a mathematical point of view that it was almost impossible for stock prices to rise during this period.
According to data compiled by Woofun AI, OpenAI secretly submitted S-1 on June 8, but the prospectus has not been disclosed; according to the Financial Times, the company has lowered its financing target from US$30 billion to US$20 billion, and plans to finance at a valuation of US$125 billion. Hayes believes that even if the company goes public in January, about 5% of the issued shares will be sold off, and employees and early investors will continue to reduce their stock holdings in the next six months; similar to Anthropic, maintaining stock prices will also require huge additional capital. He is particularly concerned about the reason these three companies are rushing to go public — to support the construction of huge data centers, they must obtain capital from the bond market, and listing is the path to this market.
If stock prices continue to be sluggish, bond investors will seek higher yields, leading to widening credit spreads on AI-related bonds. In an article entitled “D is for Dilution” (meaning “dilution”), he proposed “sideways trading is falling” and pointed out that even if the stock price only hovers around the issue price, it will break financing expectations. On the virtual currency side, its impact will be reflected through the liquidity of the US dollar. Hayes said that there is currently no trend of expanding the size of assets in the US federal reserve system, and the slight rise in virtual currency since August is actually due to the growth of the Bank of America's balance sheet;
However, the size of these banks' assets began to decline last week. Although he claims that this is only weekly data, he is still concerned about the deterioration of the credit market related to artificial intelligence. He believes that if the loan and securities financing business related to artificial intelligence contracts, the bank's balance sheet will remain stable or shrink. Maelstrom has now decided to suspend action and wait for the situation to unfold. He expects that once the AI bubble bursts and signs of a financial crisis appear, central banks and finance departments of major countries will step in to provide financial support; he believes this will be a favorable factor for holders of virtual currencies such as Bitcoin, but he also reminds everyone that “patience is needed.”
Looking at it from a macro perspective, Hayes's analysis revealed a tenuous balance between AI capital expansion and monetary liquidity. When tech giants draw huge amounts of capital through IPOs and the bond market, if there is no corresponding central bank liquidity injection, the widening of credit spreads will directly eat back on asset prices.
This “sideways trade is falling” logic not only applies to US stocks, but is also transmitted to the crypto market through dollar liquidity channels. As banks' balance sheets come under pressure due to shrinking AI credit, the market is at a critical turning point: whether it is falling into a spiral of exhaustion of liquidity or an inflection point where central banks are bailing out the market. For Bitcoin holders, the current fluctuations are not simply market shocks, but rather a stressful response from the global financial system in dealing with the risk of the AI bubble bursting. Patience and waiting for policy changes is probably the only strategy to get through this cycle.