Maelstrom Chief Investment Officer Arthur Hayes has restated his prediction that Bitcoin could reach $1 million by 2030, arguing that the asset's strongest advance could arrive in late 2027 or early 2028 as strain builds in debt-financed artificial intelligence infrastructure.

The forecast was relayed by Walter Bloomberg, the financial news account on X, which reported that Hayes expects Bitcoin's most powerful rally within that window. Under the outlook he outlined, a downturn in AI investment would push governments and central banks to pump money into the financial system — monetary support that forms the backbone of his $1 million target. Hayes, who co-founded the crypto derivatives exchange BitMEX in 2014, regularly frames his crypto outlook around monetary policy. Bitcoin's hard cap of 21 million coins is central to theses of this kind: central bank money creation does not dilute an asset whose issuance schedule is fixed at the protocol level.

ARTHUR HAYES SEES BITCOIN AT $1 MILLION BY 2030 Maelstrom CIO Arthur Hayes predicts Bitcoin could reach $1 million by 2030, with its strongest rally in late 2027 or early 2028. His thesis: an AI infrastructure bubble bursts as>September 30, 2026

A $1 million forecast built on an AI credit downturn

In an Aug. 5 report, crypto.news covered Hayes's AI credit crisis thesis, which framed much of the sector's infrastructure spending as debt-backed property development. His argument centered on land, buildings, electricity connections and cooling systems, along with processors that could lose value as newer equipment becomes cheaper and more efficient.

Drawing a comparison with earlier market crashes, Hayes characterized the boom as a "credit story like 2008 and not an earnings story like 2000." The distinction shapes how a downturn could spread: the dot-com bust of 2000 played out mainly in equity valuations as unprofitable companies failed, while the 2008 global financial crisis traveled through the banking system as losses on debt — most visibly mortgage-backed securities — hit the institutions that had financed the boom. A credit-driven episode, under that framing, risks reaching lenders, insurers and bondholders rather than stopping at shareholders.

In his view, the risk stretches well beyond falling technology shares. Banks, insurers, private lenders and infrastructure investors could absorb losses if projects fail to generate enough revenue to cover interest payments, leases and other obligations.

Hayes also argued that profitable technology companies could remain healthy even as weaker projects and their financiers struggle. His forecast therefore hinges on the debt supporting the buildout, rather than requiring every major AI company to suffer an earnings collapse.

A mismatch between hardware's useful life and longer financing schedules helps explain his focus on 2027 and 2028. Hayes expects equipment to age while borrowers remain tied to repayment terms arranged when revenue expectations were higher.

Late 2027 and 2028 feature in Hayes's spending outlook

In the August coverage, Hayes predicted that growth in announced AI capital spending would begin slowing during the second half of 2027, with the slowdown becoming clearer in 2028. He also expected investors to eventually favor companies that scale back construction plans. That timeline leaves observers with concrete markers to track: the pace of announced AI capex projects, signs of strain among project borrowers, and the reporting changes taking effect in U.S. insurance regulation described below.

While Hayes identified a possible window of stress, he acknowledged that he could not name the borrower that would trigger a crisis or pinpoint Bitcoin's exact bottom. His August scenario included Bitcoin trading between $60,000 and $70,000, with possible downside toward $50,000, before an eventual advance toward $1 million.

By Sep. 22, reporting on his AI debt argument had narrowed to a more specific concern: weaker demand for AI training and services could undermine the revenue assumptions behind data centers, chip purchases and related lending.

In his "Safety First" essay, Hayes argued that efforts to cut computing costs could hurt infrastructure investments financed on expectations of heavier spending. Debt obligations would remain, he wrote, even if customers purchased less computing capacity than lenders and developers had anticipated.

Apollo estimates AI financing will extend into private debt

Separate research from Apollo, one of the largest alternative asset managers in the United States, attaches figures to the financing requirements behind the buildout. In an Aug. 14 note, chief economist Torsten Slok estimated that the AI ecosystem could support more than $2 trillion of additional investment-grade debt.

Apollo said public investment-grade markets might absorb less than $1 trillion through 2030 because of limits related to issuer concentration and credit ratings. The firm expected more than $1 trillion of financing could shift into private placements, infrastructure lending, equipment financing and project-specific structures.

Using data through July, Apollo also found that AI-related borrowing already accounted for nearly 40% of longer-duration investment-grade corporate bond supply. Its research presented private financing as a means of meeting demand, with collateral and contractual protections available in some transactions.

As for a potential U.S. policy response, Hayes laid out two paths in "Safety First." Washington could purchase computing capacity to support the industry, becoming what he called a "compute buyer of last resort," or extend financial assistance to insurers facing losses on AI-linked debt. Interventions on that scale have modern precedent: the Federal Reserve and U.S. Treasury ran large-scale asset purchases and emergency lending programs during the 2008 financial crisis and again in 2020, measures that sharply expanded the monetary base.

In either case, Hayes expects such a response to expand the money supply and support Bitcoin prices. The Sep. 22 report noted that U.S. authorities had not announced either measure in response to an AI debt crisis.

U.S. insurance regulators tighten private credit reporting

The National Association of Insurance Commissioners — the standard-setting body whose members are the chief insurance regulators of the 50 states, the District of Columbia and U.S. territories — has flagged liquidity, pricing and transparency concerns in private credit, providing a direct U.S. connection to the lending risks discussed by Hayes. Insurers hold substantial portfolios of long-duration bonds and private credit, so assets of the kind Hayes warns about would fall directly under state-level solvency oversight.

According to its guidance, concerns about valuations, lending standards and sector exposure have contributed to withdrawal requests at some retail private credit funds. Some vehicles have imposed withdrawal limits, while software borrowers exposed to AI disruption have drawn closer scrutiny.

The association said those developments do not necessarily establish deterioration across private credit markets or insurers' holdings. State regulators and NAIC staff are monitoring credit quality, valuation practices and insurer investments.

Under amendments adopted in 2025, the NAIC requires private rating rationale reports within 90 days of an annual update or rating change. The reports must contain analytical substance, according to the association's explanation of the requirements.

For annual financial filings, the NAIC's Statutory Accounting Principles Working Group has adopted changes effective at year-end 2026 to improve reporting of insurers' private credit holdings.