Arthur Hayes says the roughly $1.5 trillion piled into AI debt could ignite a financial crisis bigger than 2008—one that forces massive money printing and ultimately drives Bitcoin toward $1 million.
On the Bankless podcast (June 22, 2026), the BitMEX co-founder and Maelstrom CIO argued that AI borrowing since late 2022 has soaked up nearly all growth in the U.S. M2 money supply. That, he said, has drained liquidity from Bitcoin while quietly building systemic risk. When the cycle turns, Hayes expects a credit unwind that could eclipse the subprime collapse, with Bitcoin as a key beneficiary.
This isn’t just a bold price call. Hayes frames it as a structural imbalance: too much capital has been funneled into AI infrastructure, setting up a large-scale misallocation that must eventually correct. When it does, Bitcoin stands to absorb the fallout-driven liquidity.
The mechanics are straightforward. Capital that might have gone into crypto has instead funded data centers and GPU clusters through multi-year debt. Hayes compares the boom to the railroad expansion of the 19th century—massive, transformative, and prone to overbuilding. The weak point is duration mismatch: loans run five to six years, while cutting-edge AI chips can become outdated in about two.
Competitive pressure adds another layer of risk. If U.S. AI firms are forced to match lower-cost Chinese models, projected revenues could drop quickly, undermining the loans tied to that infrastructure. Hayes sees this repricing as the trigger for a major credit event—potentially larger than the subprime crisis.
Evidence of the build-up is already visible. Bank for International Settlements data shows AI-linked private credit has surged from near zero to over $200 billion, about 8% of the total market. Meanwhile, major tech firms are shifting this debt off balance sheets using special-purpose vehicles and leases, creating hidden pathways for financial stress.
When the system cracks, Hayes expects a familiar response: policymakers step in with aggressive liquidity injections. In his words, authorities will flood markets with fiat to stabilize the damage caused by years of overinvestment.
Where that money goes is the key. Hayes argues investors burned by AI losses won’t rush back, leaving crypto—especially Bitcoin—as the primary destination. Because Bitcoin sits outside the traditional financial system, it becomes a natural recipient of crisis-era liquidity.
A $1 million Bitcoin implies a roughly $21 trillion market cap, requiring stimulus on a scale far beyond the COVID response. Hayes doesn’t pin down timing, noting the unwind could happen soon or take years. What matters is the trigger: a crisis big enough to force large-scale money creation.
For the thesis to hold, defaults must spread across AI-linked credit—particularly among smaller GPU lenders and highly leveraged data center operators—forcing central banks to expand balance sheets significantly. If institutions begin treating Bitcoin as a hedge against currency debasement, the rotation Hayes describes becomes plausible.
But there’s a counterpoint. In past crises, money first flows into Treasuries and gold. Bitcoin, which often behaves like a risk asset during market shocks—as seen in March 2020—could initially drop alongside AI stocks before benefiting later.
Hayes’ own positioning reflects that uncertainty. As of June 2026, he remains long Bitcoin but holds substantial cash in Treasury bills and has trimmed exposure to higher-risk tokens like NEAR and Hyperliquid. His approach prioritizes capital preservation, with the $1 million Bitcoin target framed as a potential cycle peak—not an immediate outcome.
