Arthur Hayes has never been shy about bold predictions, but his latest thesis is worth taking seriously even if the headline number makes you skeptical. The BitMEX co-founder is arguing that Bitcoin’s next major rally might not come from another wave of institutional adoption or a fresh surge in crypto demand. It could come from something far more disruptive: a credit crisis inside the artificial intelligence industry. His price target is $1 million. The mechanism behind it is more interesting than the number.
The argument is not necessarily a bet against AI itself. What it questions is how the industry’s infrastructure expansion is being financed, and whether the expected returns will materialize quickly enough to justify the debt being taken on to fund it. AI companies and their infrastructure partners are committing staggering amounts of capital to data centers, chips, power generation, and related build-out. The scale of that spending has encouraged increasingly creative financing structures. Reuters recently highlighted Nvidia’s roughly $500 billion financing arrangement involving private equity and credit firms for chip purchases, which gives some sense of how much capital is now being mobilized around AI infrastructure.
The concern is that investors may be treating these projects as if they carry the risk profile of the world’s most valuable technology companies, when much of the underlying infrastructure is closer to a leveraged real estate project. If AI demand grows as expected, the leverage may not matter. If capital spending slows sharply, the economics could change very quickly. Data centers still have debt obligations even when utilization, revenue, or expected returns disappoint. That is where the potential for a broader credit event lies.
The $1 Million Thesis Has a Big Catch
The Bitcoin forecast depends entirely on what happens after the credit bust, and this is where the thesis gets complicated. Hayes’ sequence runs roughly as follows: AI infrastructure spending becomes overextended, credit losses emerge, governments step in to prevent a disorderly collapse, liquidity expands, and investors eventually move into scarce assets like Bitcoin. That logic has historical precedent. During major financial crises, policymakers have repeatedly responded with aggressive monetary and fiscal measures, and Bitcoin’s fixed supply makes it an attractive destination for investors who believe large-scale monetary expansion weakens the purchasing power of fiat currencies.
The problem is the assumption that Bitcoin immediately benefits from a financial crisis, because it usually does not. When credit risk suddenly becomes a concern, the first reaction from investors is almost always to reduce exposure to volatile assets and raise cash. Bitcoin tends to fall alongside equities and other risk assets before monetary easing eventually becomes supportive. Some versions of this outlook place Bitcoin in a potential $50,000 area before any liquidity-driven recovery takes hold, and that distinction matters enormously for investors trying to position around this kind of macro thesis.
The trade is not simply that AI crashes and therefore Bitcoin goes up. It is closer to saying that AI credit stress could eventually create the monetary conditions that make Bitcoin substantially more valuable. Those are very different propositions, and the gap between them is where most investors who try to act on this kind of thesis get hurt.
What Could Prove the Thesis Right
There are several signals worth watching if you want to track whether this argument is moving from speculation toward something more concrete. The first is AI capital spending itself. If major technology companies continue increasing investment in computing infrastructure, the immediate case for an AI credit bust remains weak. A slowdown becomes more significant if it happens alongside falling utilization rates, weaker AI revenue growth, or deteriorating returns on infrastructure investment, not just a single quarter of reduced guidance.
The second is credit quality in the AI infrastructure ecosystem. Rising defaults, tighter lending standards, or increasing stress in private credit markets supporting data center construction would provide stronger evidence that the concern is becoming a real issue rather than a theoretical one. Credit spreads and the performance of private credit funds with AI infrastructure exposure are worth monitoring for early signs of stress.
The third, and arguably most important, is government intervention. If an AI downturn produces aggressive fiscal guarantees, central bank liquidity facilities, or other measures designed to prevent a wave of defaults, the resulting expansion in liquidity could become a powerful tailwind for Bitcoin. Without that policy response, the $1 million argument becomes considerably harder to defend. The credit bust alone does not get you there. The monetary response to it does.
What Investors Can Actually Use
Investors do not need to believe Bitcoin will reach $1 million to find something useful in this thesis. The more practical takeaway is that AI and Bitcoin may eventually become connected through liquidity rather than technology, and that connection creates a portfolio consideration worth thinking through now rather than after the fact.
An investor who owns AI beneficiaries is exposed to the first half of the scenario. Companies selling chips, data center equipment, networking technology, and power infrastructure could benefit enormously while AI spending remains strong. But those same assets could face serious pressure if capital spending suddenly slows. Bitcoin offers a different kind of exposure. It may struggle during the initial phase of a credit contraction but potentially benefit from the policy response that follows, which makes the two trades less contradictory than they initially appear. Watching the AI infrastructure cycle as a macro signal for Bitcoin, rather than treating the two themes as completely separate, is a more sophisticated way to think about both.
There is also a more immediate lesson buried in the thesis: liquidity matters more than price targets. Bitcoin has already demonstrated sensitivity to changes in global liquidity and risk appetite, reacting quickly to tighter funding conditions and declining available capital. That means the more important question for investors is not whether $1 million is the correct target but whether the conditions that could support such a valuation are actually developing. Right now they are not, but the AI infrastructure cycle is one of the more credible mechanisms by which they could.
The Bigger Risk Is Timing
The thesis could ultimately prove directionally correct while still being a very difficult trade. An AI credit bust could trigger a severe Bitcoin selloff before policymakers respond with meaningful stimulus, and investors who enter expecting an immediate rally could easily be forced out before the monetary support arrives. That timing problem is not a reason to dismiss the argument, but it is a reason to be honest about what the trade actually requires.
The biggest warning sign for Bitcoin investors would not necessarily be weaker AI spending on its own. It would be a combination of falling AI capital expenditure, deteriorating credit conditions, and tightening liquidity without a meaningful policy response. That scenario would deliver the pain of the crisis without the monetary payoff the thesis expects. Conversely, the strongest bullish signal would be evidence that AI credit stress is forcing policymakers into aggressive liquidity expansion.
Hayes’ $1 million forecast makes for a compelling headline. The real investment thesis underneath it is about liquidity cycles, credit dynamics, and policy responses, none of which move on a predictable schedule. For investors, that makes the AI credit cycle worth watching carefully, not because it guarantees any particular Bitcoin price, but because it could determine the direction of the next major liquidity wave, and Bitcoin has shown repeatedly that it knows how to move when that wave arrives.
Benzinga Disclaimer: This article is from an unpaid external contributor. It does not represent Benzinga’s reporting and has not been edited for content or accuracy.