Arthur Hayes Backs Ethereum (ETH) Longs Over AI Shorts, Sees $1 Million Bitcoin

Hayes Calls for Bitcoin and Ethereum Longs

BitMEX co-founder Arthur Hayes told traders on Thursday to stop betting against the AI trade and to position for the money he says comes after it: long Bitcoin (BTC), long Ethereum (ETH). In a post on X, Hayes wrote that he does not know whether artificial intelligence will finally crash and will not bet either way; he wants exposure to assets that absorb the trillions of dollars in bailout spending he expects once a crack appears. The message arrived alongside a falling chart of CoreWeave (CRWV), the AI cloud provider whose 15% plunge one day after television host Jim Cramer named it the only AI cloud company he trusts has fed anxiety about the debt stacked into the sector. Hayes builds the case on a pattern he says every major technology follows: overinvestment, then the crash, then the rescue. He puts the wall in the AI data-center build-out somewhere in late 2027 or 2028. When the rescue arrives, authorities will create more money than they issued after the 2008 financial crisis, and that liquidity, in his reading, flows into Bitcoin and the wider crypto market, a path he believes carries Bitcoin above $1 million. Ethereum sits in the same trade as the second long, sized through the leverage markets his old exchange built around the perpetual futures contract. The call did not come out of nowhere. His September newsletter argued that large American insurers are already insolvent on a market-value basis under the weight of AI debt. Among crypto traders, he noted, the argument has shifted from whether AI stocks will fall to who pays the bill when they do. For the Ethereum price, the framing makes the coin a claim on the liquidity a rescue would unleash, rather than a hedge against an AI drawdown.

The 2020 Template and Its Loose Fit

The precedent behind the call is the pandemic rescue of 2020. As the Federal Reserve more than doubled its balance sheet, Bitcoin (BTC) climbed from under $5,000 to more than $60,000, and Hayes treats that run as the model for what a larger rescue would do to both assets. The current cycle, however, is straining the link between printed money and price. United States money supply set an all-time record in August, yet Bitcoin still trades roughly 35% below the high it marked in October 2025. The gap is not an American anomaly; liquidity records and crypto prices have drifted apart across major economies this year. His two chosen assets have also trailed most AI stocks through 2026, which means the rescue trade has so far rewarded the sector he refuses to short. Not everyone accepts the timeline. Tom Lee, co-founder of Fundstrat, argues that today's AI build-out resembles 1997 rather than the terminal year of 1999, and he places the bubble peak around 2033. If Lee is right, the rescue Hayes wants to trade sits years out, and the position has to survive a long stretch in which AI equities keep outperforming. That is the practical cost of the trade: the payoff depends on a policy response whose timing nobody can schedule, however certain the direction looks in Hayes's framing. How a rescue would reach these markets matters for anyone holding the position. A response on the scale Hayes describes would expand central-bank balance sheets the way 2020 did, and the money would arrive through familiar channels: spot buying, exchange-traded products, and the leveraged markets where his old exchange built its name. Within the pair itself, Bitcoin (BTC) plays the monetary story while Ethereum carries the utility story, with gas fees, staking yield and layer-2 activity giving holders reasons to stay paid through a long wait. It shapes where marginal money lands once direction returns.

A Duration Trade, Not a Timing Call

The trade, as we read it, is a duration bet rather than a call on this week's tape, and the flows do not yet back it. Ether had slipped below $2,400 in early Thursday trade, before the post landed, spot ETFs recorded a $129.5M outflow on October 14, and whale deposits pushed exchange balances to 14.93M, a one-month high. That combination reads as distribution, not accumulation. The scenario only pays once the money-supply link that powered 2020 reappears in the data, so we treat the $1 million target as a position to be staged over years, with the CoreWeave chart and the insurers' debt books as the triggers to watch.