원문 보기
Michael Bur just made the most aggressive bet of his career, and it's not against housing or banks this time. It's against the two hottest stocks in artificial intelligence. The man who made $700 million predicting the 2008 crash just bet 80% of his entire portfolio, $1.1 billion in notional value, that Nvidia and Palunteer are about to collapse.
And 48 hours after his trade was disclosed, global markets are in freefall. Nvidia dropped 4%, Palunteer crashed 9% and $34 billion in market value evaporated overnight. But here's what nobody's telling you. This isn't actually a $1 billion bet. And understanding the real mechanics of what Bur did will teach you more about how the markets actually work than any finance course ever could.
Let me show you exactly what happened on November 3rd when Bur broke a 2-year silence on social media and posted three charts that sent Wall Street into panic mode. The first chart compared cloud revenue growth for Amazon, Google, and Microsoft from 2018 through 2022 versus 2023 through 2025. What it showed should terrify anyone holding tech stock.
Cloud growth has collapsed. During the earlier period, these companies were growing cloud revenue at 20 to 40% annually, now singledigit. But here's where it gets disturbing. The second chart showed capital expenditures, the money companies are spending on AI infrastructure. And that spending is accelerating to levels not seen since the.com bubble peak in 1999.
Let that sink in. Revenue growth is dying, but spending is exploding. That's not investment, that's desperation. The third chart Bur posted might be the most damning piece of evidence yet that we're in a bubble. It showed a complex web connecting Nvidia, Open AI, Oracle, Amazon, Microsoft, and every major tech company with arrows representing investments and partnerships flowing in circles.
Nvidia invests in AI startup. Those startups buy Nvidia chip. Those chips run on cloud platforms owned by companies that also invest in the startup. It's a circular economy where money chases itself with no actual end customer generating real profit. Bur captioned this post with a Star Wars reference.
These aren't the charts you are looking for. You can go about your business. Translation: The data is screaming danger, but everyone's pretending not to see it. Now, let's talk about what Bur actually did because the headlines are wildly misleading and understanding the truth is critical. His SEC 13F filing showed put options with a notional value of $912 million on Palunteer and $186 million on Nvidia, totaling 1.1 billion.
Every news outlet reported this as Bur betting $1 billion against AI. That's not how options work. A put option gives you the right to sell shares at a specific price by a certain date. You don't need to own the shares. You don't need to spend anywhere near their full value. You just pay a small premium upfront.
Think of it like insurance. If you buy insurance on a $300,000 house, you might pay $3,000 annually. You didn't spend $300,000. You spent $3,000 for the right to get paid if disaster strike. Bur's actual cash outlay was likely 1 to5% of that 1.1 billion. meaning he spent somewhere between 11 million and $55 million to control 1.1 billion in exposure.
This is why Bur's move is so brilliant and so dangerous for anyone on the other side of his trade. His downside is capped at whatever premium he paid. Let's say $50 million worst cake. His upside is theoretically unlimited. If Palunteer drops 50%, those puts could be worth $500 million or more. If it drops 70% like it did during COVID, they could be worth over a billion.
He risked 50 million to potentially make a billion. That's the asymmetric risk profile that's made Bur legendary. But here's what should really concern you. Bur doesn't make trades like this unless he's absolutely convinced something is structurally broken. He spent two years building his housing short in 2005 through 2007.
He watched his fund lose money, faced lawsuits from investors, got called by everyone on Wall Street, and he was right. Now he's doing it again with AI. Let's examine why Bur thinks these specific companies are doomed. Starting with Palunteer, because this is where his conviction is strongest. 912 million in puts on Palunteer makes it his largest position by far.
Palanteer's market cap just hit $450 billion. Their annual revenue 4.4 billion. That means investors are paying 100 times revenue for this company. To put that in perspective, during the dot bubble, Pets.com traded at absurd valuations and went bankrupt. Amazon, which survived and thrived, still fell 95% before recovering.
Palanteer would need to grow revenue by 2500% just to justify current prices at normal valuation multiple. And here's the problem. Their growth is already slowing. Third quarter revenue was 1.2 billion, up 63%, which sounds great until you realize that's decelerating from previous quarters. And they're spending massive amounts on sales and marketing just to achieve that growth.
The Palunteer CEO's response to Bur's bet tells you everything about who's actually worried. Alex Karp went on CNBC and called Bur's move batshit crazy. He said the two companies he's shorting are the ones making all the money, which is super weird. The idea that chips and ontology is what you want to short is batshit crazy.
That's not the response of someone confident in their valuation. That's the response of someone who knows short sellers are shining a spotlight on unsustainable math. When CEOs start name calling investors who bet against them, it's usually because they can't defend the number. And the numbers for Palunteer are indefensible.
The company trades at 700 times earning. 700. For context, the S&P 500 average is about 20 times earning. Palanteer is 35 times more expensive than the market average. Now, let's talk about Nvidia because this is where most retail investors are exposed and probably don't even realize the danger they're in.
Nvidia just became the first company ever to hit a $5 trillion market cap. It's up 54% this year alone. The stock has been the single best performer in the entire market, making early investors extraordinarily wealthy. And Bur is betting 186 million that it's about to fall. Why? Because Nvidia's entire bullcase rests on the assumption that AI infrastructure spending will continue accelerating forever.
But Bur's charts show that's already breaking down. Cloud revenue growth is collapsing while companies keep spending on chips and data center. That's the definition of malinvestment. Spending money on capacity that won't generate return. Here's what most people don't understand about Nvidia's business model.
They don't sell to end consumers. They sell to other companies who are building AI capability. Microsoft buys Nvidia chips for Azure. Amazon buys them for AWS. Google buys them for cloud. But who's actually paying these companies for AI services that generate enough profit to justify this spending? Nobody yet.
Open AAI loses money. Anthropic loses money. Every AI startup burns cash. The only company consistently making money is Nvidia selling the picks and shovels. But what happens when the gold rush ends and companies realize they've spent hundreds of billions on infrastructure that doesn't generate returns?
They stop buying. And when Nvidia's revenue growth stalls, a stock trading at these valuations doesn't go down 20%, it goes down 60%. The geopolitical dimension makes Nvidia's situation even more precarious. China's internet regulator recently instructed Chinese tech companies not to buy Nvidia's H20 chip, preferring to build domestic capabilities instead.
That's a massive market effectively closed to Nvidia. Meanwhile, every major tech company is developing their own custom AI chip. Google has TPUs. Amazon has Tranium. Microsoft is working with AMD. Apple designs its own silicon. Nvidia's moat is eroding from multiple directions simultaneously. The competition that didn't exist 3 years ago is now spending billions to not need Nvidia.
And Bur sees this happening before the market does. Let's examine the historical parallels because they're eerily similar to previous bubbles Bur has traded. In 2006, everyone said housing only goes up. Subprime mortgages were being packaged and sold as AAA rated security. The math didn't work, but nobody cared because prices kept rising.
Bur spent months reading through mortgage documents, found the fraud, and shorted the entire housing market. In 2007, before Lehman Brothers collapsed, he posted warnings that went ignored. Fast forward to today. Everyone says AI will change everything. Valuations don't matter because growth is infinite.
And anyone questioning it gets called a lite. Bur is posting charts showing the math doesn't work and getting called crazy. This is the exact same pattern. On October 31st, 4 days before his 13th Filing became public, Bur posted a screenshot from the big short movie showing Christian Bale portraying him.
The caption read, "Sometimes we see bubbles. Sometimes there is something to do about it. Sometimes the only winning move is not to play." That's not a random movie reference. That's a direct warning that he sees a bubble forming and he's taking action. The timing is crucial. Bur went silent on social media in April 2023.
For two and a half years, nothing. Then suddenly, he returns with bubble warnings and immediately his filing shows massive shorts against AI. He didn't come back to social media because he was bored. He came back because he's seeing something so dangerous that he felt compelled to warn people. The market reaction to Bur's disclosure shows that smart money is paying attention, even if retail investors aren't.
November 4th, the day after his charts went viral, the NASDAQ dropped 2%. Palanteer fell 8%, Nvidia dropped 4%, Reddit crashed 8.4%. Soft Bank, a major AI investor, dropped 10%. That's not coincidence. That's institutional money reading Bur Signals and reducing exposure. Global markets followed. Japan's Nikki fell 2.5%.
South Korea's Cosby dropped 2.85%. Europe's stocks 600 fell before recovering. This was a worldwide reassessment of AI valuations triggered by one hedge fund manager's filing. Here's where this gets even more interesting. Goldman Sachs CEO David Solomon warned the same week that a 10 to 20% market correction appears likely within the next year.
He's not talking about AI specifically. He's talking about overall market valuation. When the head of Goldman Sachs and Michael Bur are both warning about a correction at the same time, you should probably pay attention. These aren't perma bears who always predict doom. These are sophisticated investors who only sound the alarm when they see genuine systemic risk.
And right now, they're both flashing red. The concentration risk in tech stocks makes this even more dangerous than the housing bubble. In 2008, if you didn't own a house or work in real estate, the crisis didn't immediately affect you. But today, everyone owns tech stock. Your 401k is probably 60% tech.
Your index funds are dominated by the Magnificent 7. Your retirement account's performance is entirely dependent on whether Nvidia, Microsoft, Apple, Amazon, Google, Meta, and Tesla keep going up. In October 2025, Bank of America calculated that tech stocks contributed over 90% of the S&P 500's total return for the month.
90%, one sector. When that sector rolls over, there's nowhere to hide. Diversification doesn't save you when the entire market is actually just seven stocks in a trench coat pretending to be 500 company. Let me explain what happens next because this is where most investors make fatal mistakes. When Bur took his housing short in 2006, he was early, dead wrong for over a year.
His fund lost money. Investors sued him. He faced margin calls, but he had the capital and conviction to wait. Eventually, the math caught up and he made a fortune. The same thing is likely to happen here. Bur's puts have expiration date. If AI stocks keep rising for another 6 months, those puts could expire worthless and he loses whatever premium he paid.
But here's what retail investors miss. Bur can afford to be wrong about timing because his downside is cap. If you're long Nvidia or Palunteer, your downside is 100%. If these stocks drop 80% like many tech stocks did in 20202, you lose almost everything. Bur loses his premium, you lose your capital.
The options mechanics matter more than people realize. Bur bought puts that give him the right to sell shares at a specific strike price. Let's say hypothetically he bought puts with a $200 strike on Nvidia. If Nvidia falls to $150, he makes $50 per share. If it falls to 100, he makes 100 per share.
The further it falls, the more he makes. But this also means he needs a significant move to profit. If Nvidia only drops 10%, his puts might still be underwater because of time decay. He's not betting on a small correction, he's betting on a crash. That's a very specific prediction. It means Bur expects something to break in the AI narrative hard enough and fast enough that even with time working against him, he'll still profit massively.
The other positions in Bur's portfolio tell an important story about his overall market view. He added 50,000 shares of Lululemon, opened 125,000 shares in Molina Healthcare, 480,000 shares of SLM Corporation. These aren't AI plays, these aren't tech stock. Molina is a healthcare insurer. SLM is the student loan company.
Lululemon is Athleisure Apparel. These are defensive positions, companies that generate steady cash flows regardless of whether AI lives up to the hype. He also bought 2.5 million call options on Hallebertton and 6 million calls on Fizer. Energy and pharmaceutical, boring, stable old economy businesses.
The subtext is clear. Bur thinks tech is in a bubble, but he thinks other sectors are reasonably priced or even cheap. This isn't a bet that everything crashes. It's a bet that capital rotates out of speculative AI into real businesses with real profit. Let's talk about what you should actually do with this information because knowledge without action is worthless.
First, understand that you are not Michael Bur. You don't have his resources, his experience, or his ability to handle being wrong for extended periods. copying his trade by buying puts on Nvidia and Palunteer is probably a terrible idea unless you fully understand options and can afford to lose 100% of your investment.
What you can do is reassess your exposure to AI stock. If 50% of your portfolio is Nvidia, Palunteer, and other AI names, you're taking massive concentration risk. Consider rebalancing. Not selling everything, but diversifying into sectors that aren't priced for perfection. Second, understand how bubbles work.
They inflate longer than seems rational. The housing bubble was obvious by 2005, but didn't peak until 2007. The.com bubble was obvious by 1998, but peaked in March 2000. AI stocks could keep going up for another year, even if Bur's thesis is ultimately right. The question is whether you want to risk being caught when the music stop.
Buffett has a famous quote, "You only find out who's swimming naked when the tide goes out. Right now, the tide is high. AI stocks look great. Everyone's making money. But when capital expenditures stop growing and companies realize they're not getting returns on AI investments, the tide will go out fast.
Third, pay attention to the actual business fundamentals, not the hype. Palunteer trades at 100 times revenue. That's not a valuation. That's a prayer. For that price to make sense, everything has to go right. Competition can't emerge. Margins can't compress. Growth can't slow. Regulation can't increase.
None of those assumptions are realistic. When any of them break, the stock doesn't decline gradually. it crashes. Same with Nvidia. The company is priced assuming every cloud provider and AI company will keep buying chips at current rates forever. But Microsoft just reported that despite massive AI spending, their cloud growth is slowing.
That's the first crack in the narrative. The Bank of England issued a warning before Bur's trade became public. That's worth understanding. They said the sharp increase in valuations of top AI tech companies creates risk of sudden correction in global market. OpenAI went from $157 billion valuation in October 2024 to 500 billion in 2025.
Anthropic went from 60 billion in March to 170 billion in September. These aren't graduations in value based on revenue growth. These are speculative explosions. The Bank of England specifically warned that a change in perceptions about AI progress could drive re-evaluation of currently high expected future earning.
translation. If AI doesn't deliver on promises, these valuations evaporate overnight. One detail everyone's missing is that this isn't Bur's first attempt to short Nvidia. In the first quarter of 2025, he had 97.5 million in Nvidia put. About half his portfolio at the time. Since then, Nvidia is up 50%.
If Bur held those puts, he lost money on them. And yet, here he is doubling down with an even larger position. That tells you two things. First, he's willing to be wrong about timing because he knows eventually valuation math wins. Second, the fact that Nvidia has run another 50% higher probably makes him more convinced, not less, that it's in bubble territory.
The higher it goes without fundamental justification, the harder it falls when reality asserts itself. Let's talk about what breaks the AI bubble because that's the critical question. It won't be Bur's warnings. It won't be charts or analysis. Bubbles pop when capital stops flowing in, and capital stops flowing when returns disappoint.
The catalyst could be a major AI company admitting they're not seeing ROI from their investment, or a recession that forces companies to cut speculative spending, or an Nvidia earnings report that shows demand softening, or just general market selling that forces tech heavy funds to liquidate positions for redemption.
The specific trigger doesn't matter. What matters is that at current valuations, there's no room for error. any disappointment, any growth slowdown, any increase in competition and the valuations collapse. The political and regulatory dimension is another underappreciated risk. If AI starts displacing jobs at scale, politicians will face enormous pressure to regulate or tax the technology.
If AI generates dangerous content or enables fraud, regulators will intervene. We've already seen initial moves in the EU with the AI act. If the US follows with regulation that limits AI applications or imposes costs on companies developing it, that changes the entire investment thesis. Bur might be betting not just on valuations being too high, but on political backlash making the technology less profitable than currently assumed.
Here's the final piece that ties everything together. Michael Bur doesn't trade to make headlines. He trades to make money. Everything he does is calculated based on deep analysis of data that others ignore. When he posts charts comparing AI capital expenditure growth to the 1999 tech bubble, he's not being provocative.
He's showing you the receipt. When he compares cloud revenue growth then versus now and shows it's collapsing, he's highlighting a fundamental disconnect between spending and return. And when he puts 80% of his portfolio into a bet that two of the hottest stocks in the market are going to crash, he's telling you he has high conviction this ends badly.
You have a choice. You can assume Bur is wrong for the first time in his career when he's made a major contrarian call. You can ignore the charts he posted showing AI spending matching the 2000 peak before the NASDAQ lost 78%. You can dismiss the fact that 95% of companies see no return on AI investment.
You can overlook that Palunteer trades at 700 times earnings and Nvidia at valuations that assume infinite growth. Or you can acknowledge that when the guy who called the biggest financial crisis in modern history puts his money where his mouth is, perhaps there's something to it. The people who ignored Bur's housing warnings in 2006 lost everything.
The people who paid attention saved themselves. History doesn't repeat, but it sure does rhyme. And right now it's rhyming with every bubble that came
Comments 0