• Arthur Hayes, CIO of Maelstrom, predicts Bitcoin will reach $1 million by 2030, driven by an AI investment bust and massive government liquidity injections.
  • He expects an inflection point in late 2027 or early 2028, when AI capital expenditures slow and credit stress emerges, prompting bailouts.
  • The forecast is highly speculative; Hayes acknowledges Bitcoin could first drop to $50,000 or $60,000 before any recovery.

Arthur Hayes, the outspoken CIO of Maelstrom and co-founder of BitMEX, has doubled down on a bold prediction: Bitcoin will reach $1 million per coin by 2030. The call, which he first floated in late August, hinges on an unconventional macro thesis—that an eventual bust in the AI infrastructure buildout will trigger trillions of dollars in government bailouts, expanding liquidity and ultimately lifting scarce assets like Bitcoin.

In a series of public appearances, including at Korea Blockchain Week on September 30, Hayes characterized the $1 million figure as an aspirational target, not a high-probability forecast. He expects the rally to gain explosive momentum in late 2027 or early 2028, after an initial drawdown. Bitcoin was trading around $83,000 to $84,000 on September 30, down from a recent high of $87,400, leaving it roughly 12 times below his target.

The thesis rests on parallels between AI infrastructure financing and leveraged real estate. Hayes argues that data centers and power projects are being funded with large upfront costs, long-term contracts, and credit structures that depend on future utilization and cash flows. If AI demand disappoints or returns on investment falter, borrowers and lenders could face stress. Policymakers, fearing systemic fallout and loss of strategic advantage, would likely respond with public purchasing of computing capacity, expanded credit, and monetary accommodation—"printing trillions," as Hayes puts it.

That liquidity would first address the crisis, then flow into risk assets, with Bitcoin benefiting from its fixed supply. A move to $1 million would imply a market capitalization near $20 trillion, based on roughly 20 million coins in circulation, compared to today's approximately $1.7 trillion. Such a valuation would require not just adoption but a global repricing of Bitcoin alongside unprecedented monetary expansion.

Hayes' view is not without some grounding. AI investment is economically significant; Federal Reserve research estimates U.S. data-center investment could run at a $370 billion annualized pace by the second quarter of 2026. The Bank for International Settlements reports that hyperscalers issued over $100 billion in corporate bonds in 2025, increasingly using special-purpose vehicles and private-credit structures that link technology firms to banks and insurers. Goldman Sachs (GS) Research estimated nearly $500 billion in AI-related debt issuance in 2026, with roughly one-third of AI capex debt-funded.

Yet the leap from investment boom to systemic crisis remains uncertain. A capital-spending slowdown need not become a credit collapse if utilization and cash flows hold up. And Bitcoin is not a guaranteed short-term hedge; in a liquidity shock, it can trade like a high-volatility risk asset and fall alongside equities. Hayes himself has suggested an interim range of $60,000 to $70,000, with possible downside toward $50,000, before any recovery.

Maelstrom, Hayes' family office, focuses on digital assets, Web3, and decentralized infrastructure across liquid markets, venture capital, and private equity. It does not disclose audited financials, and Hayes' statements should be viewed as the views of an investor, not company guidance. No material leadership changes or restructuring have been reported.

Policy-wise, AI capacity is increasingly tied to national competitiveness and security, which could make governments reluctant to let strategic computing infrastructure fail. However, no current U.S. policy commits to bailing out AI investors. Crypto regulation is also evolving, with the SEC proposing tailored offering exemptions in August, though that does not assure higher prices.

The central debate is whether AI infrastructure resembles productive investment akin to earlier general-purpose technologies or a debt-supported overbuild like the telecom bust. The answer will depend on realized utilization, AI pricing power, power availability, and cash-flow durability—not merely on construction pace.

For now, Hayes' $1 million call remains a high-profile narrative linking AI-financing risk, state intervention, and Bitcoin's scarcity thesis. Investors would do well to distinguish a compelling macro story from a probability-weighted forecast.

Correction: An earlier version of this article misstated the date of Hayes' $1 million prediction. He first described it as possible in late August, not early September.