
Arthur Hayes predicts that a collapse triggered by excessive investment in AI data centers will be followed by bailouts, with the liquidity released in the process flowing into the cryptocurrency market.
Astronomical overinvestment in AI data centers will ultimately lead to a market collapse and bailouts, and the massive liquidity supplied during the process will significantly drive up the prices of Bitcoin and other cryptocurrencies, according to a forecast. Arthur Hayes, co-founder of cryptocurrency exchange BitMEX, offered this analysis at an investment event in Singapore, pointing to historical patterns during periods of technological innovation.
He explained that every major technological innovation cycle in financial history was preceded by excessive infrastructure expansion, and that market collapses and bailouts by policymakers always followed. In today’s financial environment, Bitcoin and major cryptocurrencies have established themselves as key assets capable of absorbing excess liquidity released into the market. Accordingly, he expects crypto assets to deliver the strongest performance if large-scale bailouts gain momentum.
He specifically pointed out that leading companies such as SpaceX, OpenAI and Anthropic, which are driving enormous demand for computing resources, have yet to generate meaningful profits. The analysis suggests that financial pressure across the industry will peak around late 2027 to 2028, when data centers currently under large-scale construction are completed and infrastructure providers begin seriously demanding payment from these AI companies for computing resources committed in advance.
The assessment is that investors will ultimately benefit over the long term as liquidity-supply policies follow a market shock caused by the inability to sustain the pace of infrastructure expansion. Accordingly, rather than focusing on the short-term AI boom or short selling, investors were advised to allocate assets patiently in preparation for the liquidity-expansion phase expected to come.
