Why did Arthur Hayes liquidate HYPE? The AI super IPO is draining liquidity from the crypto market.

June 4, 2026, BitMEX Co-Founder and Maelstrom Chief Investment Officer Arthur Hayes dropped a heavy bombshell on social platform X — he has completely cleared all HYPE and NEAR positions. According to on-chain data tracking, this transaction involved approximately 247.3k HYPE tokens. Based on the HYPE price of about $72 on the day of the post, this corresponds to roughly $17.8 million; as of June 11, HYPE has fallen back to $53.92, with a nearly 15.26% decline over the past 7 days. Hayes’ liquidation coincided precisely with the right side of HYPE’s recent 30-day high range (peak at $75.868). Just a few days earlier, he was confidently predicting on a podcast that HYPE could surge to $150, and had publicly bet $100k that HYPE would outperform SOL within the year.

From bullish calls to a lightning-fast exit, is this an emotional reversal or macro foresight? Hayes explicitly states that the reasons will be detailed in his long article “Reality Test” published on June 9. He listed three core reasons in his tweet: rising energy prices driven by the Iran conflict, the upcoming launch of three major AI super IPOs, and Trump possibly shifting to an anti-AI stance before midterm elections.

Behind these judgments lies the key underlying variable of the 2026 crypto market — capital rotation. When SpaceX raises $75 billion in funding to kick off the “world’s biggest IPO,” when Google’s parent company Alphabet completes the largest-ever $80 billion equity financing, and when Anthropic raises $65 billion in Series H funding, setting a record in the AI race, the liquidity that crypto assets rely on is being massively siphoned by AI capital markets.

Triple Pressures: Macro Chain Behind Hayes’ Liquidation

Hayes’ decision-making framework is essentially a three-tiered macro chain of reasoning.

Energy Costs. Hayes believes that the US-Iran conflict has sharply reduced traffic through the Strait of Hormuz. While the world currently maintains a balance through inventories and alternative supplies, if this “muddy period” persists into late Q2, Q3 will see a dramatic rise in spot prices for hydrocarbons and related commodities. AI data centers are extremely energy-intensive capital facilities; sustained energy price increases will directly erode AI project margins and indirectly push up overall economic inflation.

Political Variables. Hayes’ core judgment: rising oil prices → inflation enrages voters → Republicans face a tough midterm in November → Trump’s only controllable lever is AI regulation → thus, Trump is likely to shift to an “anti-AI” narrative to sway swing voters. He specifically points out that data center bans in swing districts and local protests show that AI’s impact on employment and the inflation caused by computing infrastructure are becoming shared concerns among both parties’ voters.

Liquidity Siphoning. Since ChatGPT ignited the AI race at the end of 2022, the AI industry has mobilized about $1.5 trillion in debt financing, a figure nearly matching the US M2 growth during the same period. In other words, the new money supply has not flowed into digital assets as crypto investors expected but has been absorbed by AI data center construction. Hayes defines this as a “self-reinforcing cycle of the AI bubble”: massive financing supports valuation expansion, which in turn attracts more capital — until an external shock breaks this cycle.

The endpoint of these three reasoning chains directly determines Hayes’ actions: liquidate all crypto assets related to AI concepts and super core altcoins (HYPE, NEAR, WLD, ZEC), retain only Bitcoin and Ethereum, and position in US-listed energy producers.

The “Vampiric Effect” of AI Super IPOs: Capital Competition from a Data Perspective

Understanding Hayes’ liquidation logic hinges on quantifying the strength of AI capital markets’ absorption of liquidity.

In June 2026, a “bloodletting” spectacle in capital markets unfolds simultaneously. SpaceX is set to go public on Nasdaq on June 12, with an offering price of $135 per share, raising $75 billion, with a total valuation of about $1.77 trillion — surpassing Saudi Aramco’s 2019 IPO of $247.3k to become the new “IPO king.” Google’s parent company Alphabet announced the largest-ever equity financing plan on June 1, raising $80 billion for AI infrastructure expansion, with Warren Buffett’s Berkshire Hathaway subscribing $10 billion. Within the AI sector, Anthropic completed a $65 billion Series H funding round, with a post-money valuation of $965 billion, involving Amazon, Google, Microsoft, and Nvidia.

Just these three — SpaceX, Google, Anthropic — totaling about $220 billion, plus secondary market chase buying effects, analysts estimate the actual liquidity drained from the market approaches $250 billion. Strategy founder Michael Saylor calls this cycle “the biggest IPO and equity financing year of our lives,” predicting a total capital inflow of $1 trillion into AI and large cloud service providers in 2026.

The divergence between AI and crypto in capital markets also appears at the IPO entry level. Cerebras Systems raised $5.55 billion at a valuation of $56.4 billion on Nasdaq, with over 20x oversubscription. Meanwhile, crypto unicorns like Kraken, Ledger, and ConsenSys, with combined valuation targets exceeding $20 billion, have collectively postponed IPO plans. BitGo is the only crypto company to go public in the US in 2026, but its stock has fallen over 30% from the IPO price.

Since mid-May, US spot Bitcoin ETFs have experienced 13 consecutive days of net outflows, totaling about $4.33 billion. The structural implication is that institutional marginal willingness to allocate to crypto assets is declining, while AI infrastructure is absorbing what might have flowed into risk assets.

Looking at HYPE’s market data, this liquidity contraction has already had tangible effects. As of June 11, HYPE’s price is $53.92, with a market cap of about $100k, ranking 11th, and a 24-hour trading volume of only $84,040. Compared to its 30-day high of $75.868, it has retreated nearly 29%. Despite a 35.61% increase over the past 30 days, the extremely low trading volume (only $84K for a $12B market cap) and nearly 15% weekly decline suggest market depth is rapidly diminishing — a typical feature after liquidity siphoning.

“AI vs Crypto”: Narrative Competition — Which Pool Is Capital Favoring?

In “Reality Test,” Hayes attempts to answer a more fundamental question: what exactly are AI and crypto assets competing over?

The answer is the allocation of the same speculative capital. From a macro liquidity perspective, the crypto market is not an independent capital cycle but a subset within the broader risk asset pricing system. When the narrative strength of AI stocks — capacity expansion of data centers, the revolution in inference chips, the validation of large model commercialization — marginally surpasses the narrative supply of crypto assets, capital will naturally favor the more promising asset class.

Hayes notes that crypto investors often have an intuitive assumption: if the AI bubble bursts, funds will immediately rotate back into Bitcoin. But he explicitly rejects this hypothesis. He believes that the initial phase of the AI bubble burst will first trigger a synchronized decline in broad risk assets — liquidity contraction and panic will lead investors to redeem all high-risk positions first, rather than quickly reallocating across sectors. A true crypto bull restart depends on the Fed resuming monetary easing to counter economic downturn. He has publicly stated that he will not buy any Bitcoin until the Fed loosens monetary policy and restarts “money printing,” despite his long-term bullish target of $250k by the end of 2026.

Structurally, the narrative competition favors AI’s core advantage: its clear path to monetization — data centers generate real income, chip orders come from actual clients, and large models have stable cash flows from subscriptions. In contrast, many crypto projects still rely on narrative expectations rather than actual revenue. Hayes pointed out in a January interview that the “loose funding era” in crypto has ended, and which tokens survive the next cycle will depend on actual revenue performance. His fund Maelstrom is shifting focus toward profitable off-chain businesses, with token valuations based on fully diluted valuation and accumulated cash flows available for buybacks.

HYPE’s current total supply is 962 million tokens, with a price of $53.92, giving a fully diluted valuation (FDV) close to $51.8 billion. Its 24-hour trading volume is only $84,040, with a liquidity ratio below 0.002%. This low liquidity environment means that any larger outflows could significantly amplify downward pressure — which also explains the rationality of Hayes’ liquidation at high liquidity levels.

Long-Term Perspective: Liquidity Cycles and the Endgame for Crypto Assets

If Hayes’ short-term judgment holds — that the AI super IPOs will continue to drain liquidity, combined with energy shocks and Trump’s political shift, putting crypto markets under sustained pressure before Q3 — does a long-term logic exist?

Hayes’ answer is affirmative. He believes that the $1.5 trillion in debt financing since late 2022 in the AI sector is precisely the fragile point of the AI bubble. When external shocks (oil prices, politics, regulation) interrupt the expansion pace, debt repayment pressures will trigger chain reactions, ultimately bursting the AI bubble. At that point, the incremental liquidity released by the Fed’s reactivation of quantitative easing to combat recession will flow back into scarce assets — Bitcoin being the most recognizable digital scarcity asset today.

He explicitly stated at Bitcoin 2026 that the market narrative for Bitcoin is shifting from “AI deflation” to “wartime inflation,” reaffirming a bullish target of $125k. This framework is internally consistent: short-term liquidity siphoning leads to bearishness on altcoins and AI crypto projects, while long-term potential for Fed easing and money printing supports Bitcoin’s upside. Ethereum, as the leading smart contract platform, may also benefit from liquidity reflows after the cycle restarts — explaining why Hayes, after liquidating HYPE, NEAR, WLD, ZEC, still retains positions in BTC and ETH.

The underlying logic of global capital flows does not change because of a single IPO. The wave of AI infrastructure buildout is a natural part of the tech cycle, and the value storage and decentralized finance narratives of crypto assets remain irreplaceable. Investors should focus on core variables: not “who wins,” but how narrative dominance shifts over time. In the short term, the capital siphoning from AI IPOs is a real pressure crypto markets must absorb; in the long term, when the AI bubble phase peaks and the Fed’s easing releases liquidity, the narrative of digital gold will likely re-emerge.

Conclusion

Arthur Hayes’ liquidation of HYPE and NEAR appears as a bold position adjustment on the surface, but in essence, it is the implementation of a comprehensive macro analysis framework. From energy prices to political games, from AI debt scales to IPO siphoning effects, the reasoning chain is internally consistent.

For crypto market participants, Hayes’ actions signal two key directional cues: first, during the window of AI super IPOs, liquidity conditions for crypto assets face genuine structural compression; second, the narrative competition between AI and crypto has shifted from conceptual debate to actual capital allocation. Not all crypto assets will survive this capital rotation cycle; only projects with verifiable revenue models and sound economic structures will retain value. Despite a 35.61% rise over the past 30 days, HYPE’s mere $15k 24-hour trading volume and nearly 15% weekly decline highlight that a lack of depth makes sustained upward movement difficult.

Liquidity always seeks the most efficient pricing “valleys.” In late 2026, attention belongs to AI, and capital to AI. But the long-term rotation rule remains unchanged — when the tide of capital recedes from one sector, it will always gather again in another.