Arthur Hayes, the co-founder of BitMEX and the chief investment officer of Maelstrom, has a new theory about why Bitcoin (BTC) hasn't been behaving as his models predicted. In a recent podcast interview, Hayes revealed that he believes the answer lies in the surge of artificial intelligence (AI) spending and debt issuance. Hayes estimates that between 2022 and 2026, approximately $1.5 trillion in AI-related debt was issued, primarily in 2025. This massive influx of funds into AI has left little liquidity for Bitcoin to absorb, according to Hayes.
Hayes' liquidity thesis, which he revisited after Bitcoin's sharp decline despite continued money creation, suggests that the cryptocurrency's price should have been more resilient. However, the reality is that Bitcoin has fallen roughly 50% from its October 2022 highs. Hayes realized that his mental model was flawed because he hadn't been tracking the destination of the newly printed fiat money. He now believes that the money went to AI, which has been vacuuming up liquidity, rather than flowing into Bitcoin.
The AI buildout, fueled by hyperscaler capital expenditure, has left Bitcoin in a challenging position. Hayes argues that the cryptocurrency was able to rally off its FTX-era lows because AI hadn't yet started absorbing liquidity. However, as AI spending and lending accelerated in 2025, the dynamic reversed, leaving Bitcoin with limited liquidity. Hayes compares Bitcoin to a piggy bank, where the money is being taken out to fund AI trades, leaving the piggy bank empty.
As a result of this liquidity drain, Hayes is now bearish on nearly every risk asset except large energy producers. He is particularly concerned about three upcoming mega-IPOs: SpaceX, Anthropic, and OpenAI. He believes that investors will need to sell other holdings to fund these listings, potentially triggering a broad selloff in risk assets, including Bitcoin. Hayes' concern is that Bitcoin could get dragged down in a correlation-driven selloff, as investors seek to free up cash.
Despite his bearish near-term view, Hayes remains optimistic about Bitcoin's long-term prospects. He believes that if the AI bubble bursts, central banks will likely return to money-printing mode. In this scenario, with investors no longer willing to pay high valuations for AI stocks, capital will need a new destination. Hayes thinks Bitcoin is well-positioned to be that destination, especially in an environment with freshly printed money. He sees Bitcoin as a safe haven asset that could benefit from the financial stress caused by the AI bubble bursting.
In conclusion, Arthur Hayes' new theory highlights the impact of AI spending and debt issuance on Bitcoin's liquidity. While he is currently bearish on risk assets, he remains optimistic about Bitcoin's long-term potential, especially if the AI bubble bursts. The cryptocurrency's ability to act as a safe haven asset in times of financial stress could be a key factor in its future performance.
