Arthur Hayes, co-founder of BitMEX and Chief Investment Officer at Maelstrom, projects Bitcoin (BTC) will climb to $1 million. His latest prediction, detailed in a Substack essay titled “Situationship” published on August 4, 2026, hinges on an anticipated credit crisis within the artificial intelligence (AI) sector. He also expects Ethereum (ETH) to reach between $100,000 and $200,000 at the next market cycle’s peak.
Hayes contends that the substantial capital expenditure driving the AI industry, particularly in data centers, mirrors a 2008-style credit bubble rather than a dot-com era earnings bust. He believes this will inevitably lead to massive central bank intervention and money printing, acting as the primary catalyst for Bitcoin’s ascent to seven figures.
AI’s credit crunch, not earnings bust
Hayes differentiates the current AI boom from previous tech bubbles, arguing that data centers are more akin to real estate developments than pure technology ventures. These facilities are packed with graphics processing units (GPUs) that depreciate rapidly. He emphasizes this point, noting that “AI CAPEX is just another boring real estate play.”
He first articulated this view in May, calling AI spending the largest fiat credit bubble in history. That earlier analysis set an initial Bitcoin price target closer to $126,000. Now, he’s significantly raised that forecast.
Debt fueling data center growth
Hyperscalers, the major cloud providers, are increasingly funding new data centers through borrowed money, not just free cash flow. Hayes argues this shifts significant default risk onto banks and bondholders. He points to roughly $1.5 trillion in AI-related debt issued between November 2022 and mid-2026 as evidence of this credit expansion.
This scenario draws parallels to the 2006-2008 subprime mortgage crisis, where lending continued even as underlying asset growth stalled. Hayes suggests a similar breakdown could occur in the AI sector once construction spending significantly contracts.
GPU depreciation mismatch
A critical flaw in the AI funding model, according to Hayes, lies in the depreciation schedule of GPUs. These specialized chips become obsolete for cutting-edge AI workloads in about two years. However, the loans financing them are often amortized over five or six years.
This fundamental mismatch creates a significant risk. If the value and utility of the underlying collateral (GPUs) diminish faster than the debt repayment schedule, it could trigger widespread defaults. Hayes warns that if the market shifts towards cheaper Chinese AI models, these assumptions about cash flows become “spurious,” leading to a major credit event.
Central bank response and market liquidity
Hayes anticipates that the growth in AI capital spending will decelerate by 2027. This slowdown, he believes, will expose the weakest data center loans and trigger a broader credit crunch. He expects central banks and treasuries to intervene much like they did during the 2008 financial crisis and the 2020 pandemic.
Such intervention would likely involve emergency lending facilities and potentially direct equity purchases to prevent systemic defaults. This flood of new liquidity, driven by money printing, would dilute the dollar’s value, making hard assets like Bitcoin more attractive.
Historical parallels and monetary policy
The Federal Reserve’s recent decision to hold interest rates steady at 3.5% to 3.75% in late July, despite three Federal Open Market Committee (FOMC) members dissenting in favor of a hike, is seen by Hayes as confirmatory evidence. He interprets this as authorities prioritizing credit flow over tightening monetary policy.
He views continued bank lending to AI projects as further indication that officials are reluctant to let weak borrowers fail. This pattern of intervention, he argues, will become a familiar mechanism when the AI credit bubble bursts.
Bitcoin as a monetary policy signal
Hayes describes Bitcoin as “the capital market’s fire alarm,” reflecting currency debasement and capital-allocation failures before other assets. He argues that in a world awash with liquidity following a bailout, Bitcoin will become a primary beneficiary.
This isn’t the first time Hayes has predicted a $1 million Bitcoin. A similar forecast last year was tied to an expected Fed shift toward yield curve control. This time, however, the impetus is explicitly linked to AI credit stress and the ensuing government response.
Bitcoin’s path to seven figures
For Hayes, the impending AI credit bust sets the stage for Bitcoin to reach $1 million. He emphasizes that “once large-scale liquidity injections begin, Bitcoin could rise above $1 million.” This macro-economic backdrop, rather than specific crypto industry developments, forms the core of his bullish outlook.
At the time of writing, Bitcoin traded near $64,300, up roughly 1% over 24 hours. Hayes believes Bitcoin is currently building a strong base in the $60,000 to $70,000 range, with $50,000 acting as a significant support level.
Ethereum’s role in the forecast
Alongside his Bitcoin prediction, Hayes also reiterated a $5,000 Ethereum target by the end of 2026. He points to Ethereum’s growing importance as a settlement layer, particularly for tokenized real-world assets. This suggests he sees fundamental value in Ethereum’s evolving utility within the broader financial ecosystem.
Potential downside before rally
Hayes has previously suggested a Bitcoin bottom near $40,000 before any substantial rally toward his higher targets. This implies investors should be prepared for potential downside movement in the short term. The timing and severity of the AI credit cycle’s unwind remain key factors to watch in the coming quarters.
Investors will need to monitor hyperscaler earnings reports and bank loan books for early warning signs. The unfolding of the AI credit story, whether it aligns with Hayes’s 2027 timeline or not, will be a defining narrative for both traditional finance and the cryptocurrency markets.