Guest: Arthur Hayes, Co-Founder of BitMEX

Moderator: Kyle Chasse, CEO of Master Ventures

Podcast Source: Kyle Chasse Crypto

Original Title: Arthur Hayes: Bitcoin's Final Dump Before The Pump

Air Date: June 10, 2026

Key Takeaways

Arthur Hayes liquidated his largest crypto positions—HYPE, NEAR, Worldcoin, and Zcash—not due to crypto-specific concerns, but as part of a macro chain reaction spanning oil prices, the Iran conflict, Trump’s midterm election strategy, and the impending collapse of the AI bubble. He believes Trump may pivot to attack the AI sector in an effort to salvage his midterms, and once the AI bubble peaks, the crypto market will be dragged down with it. In his view, SpaceX’s $1.8 trillion valuation and 100x revenue multiple IPO represents a liquidity time bomb destined to detonate eventually.

Highlighting Insights

Why Liquidating Everything

Moderator Kyle Chasse: Arthur, welcome back. You’ve just sold off Zcash, HYPE, NEAR—everyone is calling you out for exiting scams, front-running, etc. Why sell everything? What’s going on?

I recently published an article titled “Reality Check,” around 5,000 words, laying out the argument I’ll summarize in this podcast in just a few minutes. If you want deeper insight into my reasoning, I strongly recommend reading it on my Substack. Fundamentally, the core lies in a reflexive interaction between oil prices and Trump’s midterm campaign rhetoric—he needs to help Republicans retain control of Congress in November. The issue is the ongoing Iran conflict: whether you like it or not, it exists, it’s here, it’s now.

Thus, Trump and Iran’s Revolutionary Guard must reach some kind of agreement to end the standoff. Both sides face real constraints: oil prices determine how angry different regions of the world are toward them. Trump must worry domestically—voters dislike high fuel prices and inflation driven by energy costs. Iran faces pressure from China and other developing nations: “What are you doing? We need these oils and goods through the Strait of Hormuz. I know the U.S. attacked you, but figure something out.” So the higher the oil price, the more eager people become for negotiations; when prices drop, suddenly no one wants to compromise. This back-and-forth has been going on for roughly three months—or as long as the war has lasted.

As this process unfolds, we’re gradually depleting commercial and national reserves of oil and other hydrocarbons. Pick any energy analyst—charts differ, conclusions don’t—they all agree: pre-war inventories were abundant, leading to belief in oversupply and relatively low prices. But we’re now consuming that surplus at an accelerating rate. At some point, we’ll hit a threshold—I don’t know how many billion barrels, each analyst has their own number and date. Once past that date, conditions will deteriorate rapidly and severely. The only way markets can rebalance is by sharply pushing oil prices up.

This is the worst-case scenario: Trump and Iran’s Revolutionary Guard fail to reach an agreement. By October, the Strait of Hormuz remains effectively blocked, with only 25% to 30% of volume passing through—far from enough. More likely, some agreement may emerge within one or two months, allowing partial restoration of shipping. But then everyone needs to rebuild inventories—you must refill national stockpiles, and you’ll stockpile even more than before because you’ve just experienced firsthand what it’s like to be entirely at the mercy of Trump and a group of Iranian generals deciding whether your country receives goods. So you think: “I need to hoard more oil, gas, helium, everything needed to run a modern economy.” This drives additional demand, which won’t push prices to catastrophic levels but still means oil, gas, and other commodity prices will be significantly higher in three or four months than they are today.

Oil, War, and Elections

Following this logic further, Trump and his Republican allies are highly likely to lose control of the House in the 2026 midterm elections (November). Check Polymarket right now—the odds of Democrats retaking the House have surged to 82%.

Why? Clearly, Trump is getting crushed on cost-of-living issues. People feel inflation is bad and worsening. To voters, it’s the Republicans in power in the White House, and this conflict was sparked by them—so the blame naturally falls on Republicans. That’s why everyone expects a decisive defeat.

The problem is, inflation is largely beyond policy control—there’s a long lag in monetary effects, and supply chains are only now digesting events from three to four months ago. I don’t believe Trump can reverse the inflation narrative much. People see and feel it at gas stations; there’s no magic trick he can use to convince you inflation doesn’t exist—it’s real, visible every other day you fill up. So what issue could shake the entire U.S. political spectrum? The answer is AI data centers—regulation, taxation, all of it. I believe Democrats are crafting a powerful campaign message: stop building data centers, tax AI giants, regulate AI. Because it’s not just the poor who fear job loss—rich people fear their jobs will be replaced too, at least that’s the widespread anxiety.

Trump Turning Against AI

If opposition parties can leverage this fear, they gain two potent messages: first, hyperinflation caused by Republican wars, and second, the actual AI construction boom backed by Republican politicians. So my theory is: if Trump wants to pull a rabbit out of a hat, the only issue he can flip his stance on is AI. Taking over the Democratic script, saying: “We’ll increase scrutiny on data centers, implement an AI national dividend, impose taxes on them.” That’s classic Trump rhetoric—he’ll say anything, and whether he actually does it after November is another matter. I believe this is their only viable path to win—to reposition themselves as protectors of Americans from AI harm. Then Americans will forget it was Republicans who funded all of this, because people are forgetful. That’s the main risk I see.

Trump’s willingness to attack AI depends purely on oil prices, which are a product of his reflexive relationship with Iran’s Revolutionary Guard. The longer the war drags on without resolution, the greater the future commodity pressure builds, and the more likely Trump becomes to target AI to win the election—at least to help Republicans keep the House. Obviously, the most devastating blow to the AI narrative would be taxation and regulation. We’ve already seen it in South Korea: a local politician proposes a national AI tax, and Cosby crashes the same day. So if this rhetoric starts being openly promoted by the ruling party—especially by Trump—I expect the AI bubble to peak, at least for the next several months until the election. And this will drag the crypto market down with it. That’s the core of my argument. I’m genuinely done thinking about this, so I liquidated my entire portfolio last week.

New Portfolio Allocation

Moderator Kyle Chasse: Where are your liquid assets now—cash or treasuries?

Treasuries and energy stocks.

Moderator Kyle Chasse: You still believe energy can hold up if the AI bubble bursts?

We still need oil, regardless of your personal feelings. People need oil—it powers civilization. And I’m not saying AI won’t grow further; the issue is our willingness to pay forward multiples for such growth will decline, causing asset prices to fall. That doesn’t mean companies won’t earn strong profits—just that we expected them to be even stronger, and they’re not. So we sell those stocks—this is the logic.

The Math Behind the AI Capital Expenditure Bubble

I trade based on gut feel and intuition, not deep analysis. I sense we’re in some phase of the AI bubble—I’m not sure which. I listened to Marco Papovich’s podcast over the weekend; he’s a strategist at BCA with a great YouTube channel called Geopolitical Cousins—I highly recommend subscribing. He made a crucial point in both the podcast and his article: when you invest in AI, you’re not betting on earnings—you’re betting on capital expenditure for data center construction. I often forget this: you’re investing in the second derivative, the acceleration or deceleration of trends. If momentum is accelerating, you’re willing to pay infinite multiples for future revenue; if it’s decelerating, you’re not, and the asset won’t grow as fast as you’d hoped.

He recently shared a chart showing the second derivative of capital expenditure growth—the higher the number, the harder it is to sustain acceleration. We’ve already hit $800 billion in 2026, and he forecasts the second derivative of AI capex will begin slowing from 2027 onward. You can’t justify a 100x revenue multiple for SpaceX or any AI company when both earnings and capex are slowing. Even if revenues keep growing, the key isn’t the absolute growth—it’s the speed of change and your perception of that rate. Mathematically, under the law of large numbers, capital expenditure growth cannot continue at the pace seen from 2023 to 2026—it’s physically impossible. When will the market discount this future and say: “I’m no longer willing to pay 50, 60, or 70x P/E for these AI stocks or supply chain firms”? When will the market realize that opposition parties worldwide are weaponizing this zeitgeist: “To hell with data center inflation, to hell with AI replacing my job”? Why are only Elon, Sam Altman, Zuckerberg, and about fifteen others becoming trillionaires, privatizing human knowledge, while my share is gone? This isn’t just a U.S. phenomenon—people globally ask the same question: if AI is trained on human interaction data, using all public and private data legally and illegally, why should they monopolize the profits? For those with sufficient capital involved in these equity stories, this is a legitimate concern.

There will come a day when the market feels a backlash. Conflict between capital and labor is inevitable—whether voluntary or forced, a deal will eventually be reached. If you’re holding these assets when that deal happens, you’ll likely get crushed. These thoughts have been swirling in my mind. Eventually, I sat down, tried to understand what was really happening, and spent a morning liquidating.

Why Bitcoin Underperformed AI

Moderator Kyle Chasse: What do you expect the market to do from now until year-end?

To answer that, I’ve been asking myself: why hasn’t Bitcoin risen higher since November 2022? I’ve said this repeatedly on your show and elsewhere: everything comes down to liquidity. If more liquidity flows into the system, Bitcoin should rise. But clearly, that’s wrong now. Since ChatGPT’s commercialization on November 30, 2022, Bitcoin has indeed risen—but Nvidia and all those AI stocks have risen far more. Look at Bitcoin’s peak: $125,000 last October. All the liquidity created during that period—my model says trillions of dollars—so why hasn’t Bitcoin reached $500,000 or $1 million? Why did it underperform AI?

I usually don’t track where money flows, I just say “more money → Bitcoin should rise”—a lazy mindset that worked before but doesn’t now. So I went back to reassess my mental model: what did I miss? The answer is that we all believe AI might be one of history’s most transformative technologies, undergoing massive capital expenditure buildup—trillions. But how much debt has AI consumed during this period? Has AI essentially crowded out all other risk assets, absorbing excess liquidity ahead of schedule?

On a high level, I don’t usually rely on M2—it’s too crude—but let’s use it as an example. From ChatGPT to now, U.S. M2 has increased by at least $1.5 trillion. I asked Perplexity AI: how much debt has been issued to AI and AI-related companies? Estimated at around $1.5 trillion, with $1.3 trillion concentrated between 2025 and 2026. So even though we say this AI frenzy began late 2022, the capital markets’ debt pump only truly kicked in later—recently, not early.

My theory is: Bitcoin rebounded from its lows because genuine liquidity was created, and AI didn’t consume much of it before 2025. Bitcoin had open skies to absorb that liquidity. From 2022 to mid-2025, factors like declining repo rates helped. But if you look at charts of AI capex and loan disbursements, real scale only began in 2025, especially 2026. And that’s precisely when Bitcoin started struggling—peaked last October, now down 50–60%. So if all liquidity is flowing into AI, and there’s no sign of stopping, and if the AI bubble corrects or bursts, investors won’t suddenly flood Bitcoin with fresh cash. They’ll sell AI—and also sell Bitcoin, and everything else. When bubbles burst, all correlations go to 1. Everything drops together until dust settles, then specific assets start outperforming again.

Therefore, if I believe AI will undergo a major correction in the next six months due to rising oil prices and U.S. political factors, Bitcoin cannot escape. It should perform better afterward, but you must endure the drop first. That’s why I currently see no environment favorable to Bitcoin or other cryptos. And clearly, my positions in NEAR, HYPE, Worldcoin, and Zcash were very profitable—I exited in the green. I want to lock in those gains, sit on the sidelines, and watch. These assets may keep rising, but in my mental model, I’m uncomfortable with the risks present at this moment—those foreseeable unknowns and how they might evolve. That’s why I pulled the trigger.

The SpaceX IPO Trap

Moderator Kyle Chasse: One more thing I’ve been pondering: the S&P 500 is rising, but most stocks are actually falling—the index is dragged up by a few tech giants. More importantly, we’re about to see new listings from OpenAI, Anthropic, and SpaceX—potentially over $4 trillion in new market cap entering the stock market. Do you think these will suck up all available liquidity for a while? What’s your take on these IPOs?

I think these will struggle to perform well because market expectations aren’t just “it trades normally”—they expect a 50% pop, a crazy surge, to prove the market still believes in AI, that they picked the right star company, and it will soar. With SpaceX’s ~$1.8 trillion issuance valuation, it will directly become the seventh-largest company globally. To grow 50% more, it would need to surpass Amazon. If you read its S-1 filing, SpaceX’s trading valuation approaches 100x revenue. This is absurd—it’s the seventh-largest company in the world yet has proven nothing.

Yes, the idea of space-based data centers is clever—addressing ground-level regulatory issues. I acknowledge the logic makes sense. I follow Semi Analysis on Substack—they do deep research on semiconductors and AI. They wrote an article comparing the full cost of space versus ground-based data centers, concluding: operating a data center in space is currently four times more expensive than on Earth. Not only that, you don’t have enough chips to fulfill Elon’s vision, and ground-based facilities still have capacity. Building on the ground may not be as easy as you’d hope, but it’s four times cheaper—so you choose ground until you can’t build anymore. According to the most optimistic estimates, space data centers won’t achieve cost parity with ground ones until sometime in the next decade.

So the current absurd reality is: global capital is willingly paying a 100x revenue multiple for a company whose product costs four times more than competitors, whose rockets explode, and won’t generate real profits for ten years.

And the worst part? It’s a classic crypto scam pattern: low float, high fully diluted valuation. Only 4–5% circulating, climbing to nearly 25% by September, insiders selling continuously from July to October. Yet it trades as the world’s seventh-largest company—with zero proof behind the data center thesis. Other businesses like satellite internet are executed brilliantly, but that’s not why you buy SpaceX.

So I think this will struggle to meet market expectations. I’m not saying it will crash—but even a 10% rise will trigger market disappointment: “That’s not good enough. I expected 50%, 60%, 70%.” Investors will start questioning: given that insiders at SpaceX can keep dumping shares, am I really going to bid aggressively for Anthropic or OpenAI’s IPO in September?

Setting the price this high creates almost impossible expectations. If it were a $100 billion company, a 2x or 3x move would be acceptable, and people would say AI still holds. SpaceX soared because they priced low. Now it’s maximizing extraction—$1.8 trillion. Outperforming Nvidia? Extremely difficult. Sorry Elon, you can’t beat Jensen. I don’t even know who Amazon’s current CEO is, but you can’t beat those companies. They have real revenue, operations, and proven track records. SpaceX’s everything remains mostly scribbled on napkins. Maybe it will prove itself over time, but would you really push a $1.8 trillion stock up another 50%? It’s too hard. That’s why I believe this will be a shock event—shaking faith in the AI narrative simply because it’s too big, almost impossible to grow.

Moderator Kyle Chasse: How will liquidity flow during these IPOs? Will there be mass capital shifting, or will it be like Elon—first to sell, best to sell?

First option. People will get excited and withdraw liquidity from other assets. If SpaceX disappoints, I believe Anthropic and OpenAI will face enormous pressure to lower their pricing. And if these two AI giants are forced to cut valuations or reduce fundraising size before listing, that sets a deadly precedent—like cutting their own flesh publicly, softly signaling to the world that the AI bubble is indeed inflated, and they’re lowering future expectations themselves. Suddenly, expectations drop, and investors hesitate: why did they lower prices after SpaceX? Why shrink the offering? All these changes could cool investor enthusiasm. So either capital gets pulled from other markets, or people simply sober up about the AI bull story and slowly exit, triggering a chain reaction of price declines.

Evidence for an Anti-AI Strategy

Moderator Kyle Chasse: Let’s return to Trump’s anti-AI narrative. Some might argue that all these AI titans helped him get elected—provided significant support or influence. We know he’s had numerous private dinners and discussions with them—they’re his major donors, and he’s publicly praised AI.

I haven’t quantified how many publicly oppose AI, but I know most people don’t have warm, fuzzy feelings toward it—so this could indeed be a smart pivot. But I’ve never heard anyone suggest it—a bold prediction. How confident are you? Any signs pointing toward this path?

I used Perplexity AI. I asked: “Polymarket says Republicans will lose—what paths remain?” First, grasp the underlying political logic: why does Trump desperately need to keep the House in the midterms? It’s not about ideology—it’s pure political self-preservation. If Democrats take the House, he and his family will face a torrent of congressional subpoenas for the next two years. Democrats can relentlessly pursue him. He’ll have zero chance to create any real “second Trump term” legacy. That’s why he must win.

I also believe Trump has no ideology—he only cares about winning. During the pandemic, he delivered the largest fiscal transfer in U.S. history since the New Deal—checks to everyone, no income thresholds, massive fraud, rich and poor alike received them. So believing he wouldn’t pivot to naked populism, directly appealing to public sentiment—especially since public sentiment dislikes AI—is unrealistic. AI generates negative emotions among both Republican and Democratic voters. So I asked AI: assuming these seat projections are accurate, remove the safely redistricted seats. Even then, they still need to win additional seats to keep the House.

So I asked: in all competitive, marginally close districts, are there any local laws banning data center construction or limiting their impact? Search all such districts. Result: if Trump turns anti-AI, he can flip enough seats to win the House—because these districts have already shown locally that residents absolutely don’t want data centers built in their communities and have taken local action.

And again, this is just rhetoric—Trump doesn’t need to do anything. He can call Jensen and other AI bigwigs: “Listen, I’ll hit you hard in the next four months—don’t panic. Nothing will happen by November.” That’s how he operates. He attacks them, stock prices drop, some lose money. Just look at his tariff moves—he made dozens of billions in losses for his hedge fund buddies when he tried to fundamentally rewrite U.S. trade infrastructure. He ultimately backed down at the last minute, but at least proved he was willing to try.

So I don’t see why Trump’s strategists wouldn’t pursue an anti-AI stance—even just rhetorical—if it brings enough votes. The only victims are the stock market, and those losing money are just wealthy people anyway. You don’t even need to do anything—just talk. No bills pass. By November, everything goes back to “we must win the AI race against China.” So in a situation where the inflation narrative is already locked in, this is a valuable path for Republicans to win elections. I don’t care if oil drops another 50%—gas prices might dip slightly, but too much stuff is already in the supply chain. By October, grocery shelf prices will be higher, and Trump will be powerless.

Fed, Wash, and Interest Rate Risk

Moderator Kyle Chasse: Let’s talk about Wash. I know there’s still uncertainty—his first FOMC meeting is next week. Based on his prior statements and the approaching midterm elections, what do you expect his policy direction to be?

I don’t recall his exact recent remarks, but one narrative is: you can see through wartime commodity inflation, then believe AI productivity miracles will bring growth without inflation—so rates can be cut. That’s the narrative the market wants to believe about Wash. Unfortunately, reality is higher oil prices that won’t fall soon. Two-year Treasury yields are currently about 60 basis points above the effective federal funds rate. The market is telling the Fed: you need to raise rates. That’s the signal the market is sending the Fed—whether they act or not, I don’t know.

I also believe Trump may privately soften his fixation on rate cuts, because if he wants to address cost-of-living issues, the last thing he should do is cut rates while inflation is at 3.5% to 4%. If he genuinely cares about winning over some voters on affordability, rate cuts would ruin his chances in the midterms. So considering market positioning, I believe Wash will find it extremely difficult to cut rates. My baseline judgment is inaction—only the wording matters: hawkish inaction or dovish inaction? If hawkish inaction, implying inflation pressures are building and the Fed may need to act later, the market will discount that: “They’ll hike at some point.” And the biggest fear of any bubble is rising interest rates—the increased cost of capital always pushes people out of the casino.

So I believe the chance of rate cuts is extremely low—most likely, no change, then it comes down to phrasing. The Fed has few ways to support a bubble, because oil has already pushed two-year Treasury yields above the effective federal funds rate—oil spikes, yield curve steepens, yields climb. Right now, I see no room for Wash to cut rates. If rate cut expectations are one of the pillars supporting your optimism about the AI bubble and its sustainability, you need to seriously question that assumption.

Crypto Catalysts and Re-entry Timing

Moderator Kyle Chasse: Between now and the midterms, do you foresee any event that could give the market short-term breathing room or a rebound? Not manipulation, but a narrative or development that might spark a rally from now to year-end?

Maybe people believe MicroStrategy will somehow keep pumping—this could reignite some bullish sentiment. But I genuinely don’t see much money printing, and even if there is, it flows directly into AI construction. So I don’t see any major positive catalyst that could pull crypto out of this slump—or at least allow it to outperform AI. If we return to a perfect economic environment of high growth and low inflation, what would you buy? Nvidia or Bitcoin? You’d unhesitatingly pick Nvidia, Samsung, right? Because they’ve gone up 50x in two years. Would you buy Bitcoin? Absolutely not. That’s the problem—AI is performing too well. If the environment stays the same and these things keep doing well, why pick crypto? You’ll just keep betting on capex growing at 100% annually forever, buying these companies endlessly. Is that sustainable?

And that’s exactly the current market belief. But if I were an institutional investor and a client said: “Nasdaq is up 50%, why are you only up 10%?”—“Because I’m hedging, I bought volatility, etc.” The client would say: “Why give my money to a manager who only returns 10% when I can get 50% elsewhere?” That’s the logic devouring everyone: “I want maximum returns—why aren’t you in?” That’s the issue.

Moderator Kyle Chasse: When will you consider re-entering? What would convince you to come back?

If by autumn oil prices remain stable, no sharp rise, and Trump doesn’t turn on AI tycoons, I might re-enter the market to assess value. But all of this hinges on one extremely strict precondition: in the coming months, the epic IPOs of SpaceX, Anthropic, and OpenAI must succeed spectacularly—even set unprecedented historical highs—to match the scale of the largest IPO ever. When reality fails to meet expectation, we’re in trouble.

Moderator Kyle Chasse: Is there any way to gauge when the next crypto bull run might arrive?

We need to see more money printing—and that money must not flow entirely into AI. When will that happen? I don’t know, but I don’t believe it’s happening now. You often say: no matter what, governments’ only way out of crisis is printing money—that’s unavoidable. Is there a way to gauge the timeline or trigger for this? If the AI bubble truly bursts, some financial institutions collapse—then you’ll get a bailout. When? I don’t know. But that’s when crypto can outperform: AI will have credit exhaustion—not gone, but no longer soaring. Investors will need to trade something else. I hope that something else is crypto, and liquidity will flow back to crypto.

I absolutely believe the answer is always printing money—only the timing matters. Bitcoin has been the best-performing asset in human history over the past 15 years. Unfortunately, many didn’t enter at $1—most bought at higher prices. If you entered during the ETF era, on average, you’re down. Everything depends on path dependence and timing. Just because you entered six months ago doesn’t mean Bitcoin should rise for you. That’s a painful lesson many need to learn.

Quick Fire Q&A

Moderator Kyle Chasse: Final quick-fire questions. First: will Bitcoin be above or below $100,000 by year-end?

Below.

Moderator Kyle Chasse: When will the altseason come?

We just went through an altseason—four assets. People made a lot of money on HYPE and others. So I think we’ve just lived it—maybe it’ll come again, but I don’t know.

Moderator Kyle Chasse: Will you buy back HYPE before year-end?

Yes.

Moderator Kyle Chasse: If you had $1 million today to allocate across Bitcoin, HYPE, short-term treasuries, and gold—what would you pick?

Exxon Mobil.