원문 보기
There's a number buried inside Michael Burry's famous AI short that almost nobody has said out loud. 65% That's how far Palantir has to crash from where it's trading right now before January or his entire bet expires worthless. Not a dip, not a correction, a 65% collapse. On a deadline. For 2 months, the internet has been telling you Burry is about to be right again.
Today, I'm going to use his own numbers to show you why he's almost certainly about to be wrong. Why the man who called one crash in 2008 has quietly called a dozen since and gotten almost all of them wrong. And why, if you sold AI stocks because of his headline, you may have just made the most expensive mistake of the decade.
Let's get into it. Let me start with the number everyone got wrong because it's the whole story. When Scion filed its final disclosure, every outlet ran the same line. Burry bet $912 million against Palantir. Burry corrected them himself. He spent 9.2 million in premium, not 912 million. Fine. The bears took that correction as proof of genius.
Look how cheap. Look how asymmetric. 27 to 1. But flip it around. $9.2 million is what a man running a $200 million fund spends when he is not actually sure. It's a lottery ticket. It's the smallest position that still buys a headline. And then, this is the part that should make you suspicious. He shut the fund down and stopped reporting.
Think about that. If you had the trade of the decade locked and loaded, would you blind the entire public to your moves? Or is that what you do when the trade isn't going your way and you'd rather nobody was watching? Here's the actual structure, because the math is everything. Burry bought 50,000 put contracts.
They give him the right to sell Palantir at $50 a share before January 2027. When he bought them, Palantir was trading north of 180. Today, it's around 140. So, yes, the stock has come down. The bears posted their money printer memes. But look at the strike. $50. From 140, that is a 65% collapse. And he's got roughly 9 months on the clock.
Those puts are deep out of the money. For Burry to actually win, not feel vindicated by a dip, but win, Palantir has to lose nearly 2/3 of its value by next winter. That is not a forecast of a correction. That is a forecast of a near-death event scheduled for a specific month. Markets do crash, but they almost never crash on a fund manager's expiration date.
We have seen this movie before. Back in 2000, you could have stood up and said the dot-com market was insanely overvalued, and you'd have been completely right. No earnings, no moats, companies selling dog food online at a loss. The valuations were absurd. But here's the trap. Being right about the valuation and being right about the trade are two entirely different things.
The people who shorted Amazon in 2001 because it was obviously overpriced were correct about the multiple, and they were obliterated by the outcome. Amazon fell 90% and then rose more than a thousandfold. Burry is making the year 2000 valuation argument. The problem is that the valuation argument has been, by a wide margin, the most expensive argument in the history of the stock market.
So, here's my thesis, and it comes down to one word: demand. A bubble is a price problem, Prices floating away from anything real, but AI is not a price problem. It's a demand problem. And demand is the one thing that is not slowing down. When Burry says the CapEx doesn't add up, that $665 billion in AI spending is reckless, he's assuming that spending is faith-based.
It isn't. Microsoft, Amazon, Google, and Meta are not pre-revenue startups lighting venture money on fire. They are the four most profitable enterprises in the history of capitalism, and they are spending out of their own operating cash flow because their cloud customers are demanding more capacity than they can physically build.
That is not the dot-com bubble. In the dot-com bubble, the demand was a story. Today, the demand is a backlog. And think about what a backlog actually means. It means customers have already raised their hands, signed the contracts, and are sitting in line waiting for compute that doesn't physically exist yet.
Nvidia's chips are sold out quarters ahead. Cloud providers are telling enterprise clients they'll have to wait for capacity. In a real bubble, you build the supply and then pray the demand shows up. Right now, the demand is showing up first and screaming for supply that can't be built fast enough. That is the exact inverse of a bubble, and it's the one detail Burry's spreadsheet conveniently leaves out.
And the receipts are public. Anthropic went from 9 billion to 30 billion in annualized revenue in a matter of months. Burry cited that himself to argue Palantir is losing, but read it again. A company added $21 billion of run-rate revenue almost overnight. That is not the signature of a bubble with no business underneath it.
That's the signature of demand outrunning supply. OpenAI did over 13 billion in revenue. Nvidia booked 68 billion dollars in a single quarter and beat estimates. Rewind to the year 2000. The mascot of the bubble was a sock puppet for a company with no revenue. In 2026, the picks and shovels company is doing 68 billion dollars a quarter and still can't make chips fast enough.
Those are not the same universe. They're not even the same galaxy. Now, Burry's other smoking gun is insider selling. Nearly 10 billion dollars of stock sold by Nvidia and Palantir insiders with almost no buying. Sounds damning. Here's what that framing leaves out. Executives sell for a hundred reasons that have nothing to do with the stock.
Taxes on shares that just vested, diversification when your entire net worth is one ticker, a house, a divorce, and pre-scheduled selling plans set up months or years in advance that execute automatically regardless of price. When your wealth is one stock that's up 2,000% trimming some isn't betrayal.
It's basic financial sanity. And notice what those insiders did not do. They did not quit. They did not stop building. They sold a slice and then committed to the most aggressive product road maps in their companies' histories. People who think the house is on fire do not pour the foundation for a bigger one.
Bears will push back. Fine, but Palantir specifically is doomed. Anthropic is eating its lunch. Government contracts are low margin. The Pentagon is cutting budgets. Let's take it seriously because there's truth in there. Anthropic is winning enterprise AI. And Palantir's multiple is genuinely extreme, around 180 times earnings is a number you respect, not wave away.
But expensive and going to zero are different claims. Palantir just printed 70% revenue growth. It's US commercial business grew 137%. It guided to 7.2 billion dollars next year. You do not write a company growing its top line 70% down to a $50 strike in 9 months. Not unless something breaks catastrophically.
Could the multiple compress? Absolutely. Could the stock drop 30%? Sure. But Burry doesn't need 30%. He needs 65. And a company with accelerating revenue is the single hardest thing in markets to crater that fast, that soon. Let me tell you about the most dangerous man in markets. It's not the bull.
It's the credible bear who was right exactly once. In 1996, the Fed chairman warned of irrational exuberance. The market then roughly tripled before it finally peaked in 2000. Imagine selling everything in '96 because the smartest, most credentialed, most respected man in finance told you the top was in.
And then watching four years of the greatest gains of your life happen completely without you. That's the permabear tax, and it is brutal. Burry's defenders love the line, being early isn't the same as being wrong. It's a great line, but here's the version they never quote. Being early for 15 years straight while the thing you shorted goes up tenfold is not early.
It's wrong with a publicist. Since 2008, Burry shorted Tesla and it doubled. He called the mother of all crashes in 2021 and the market climbed. He put 1.6 billion notional against the market in 2023, then closed it at a loss and admitted he was wrong. The 2008 call was real. It was also 17 years ago and this is the trap of the one-hit profit.
We remember the call that landed and quietly forget the dozen that didn't. It's survivorship bias dressed up as genius. If a man flips a coin and calls a recession every single year, eventually one arrives and he gets a documentary. But you don't get to run your portfolio on a broken clock just because it happens to be right twice a day.
The 2008 Burry made one specific, brilliantly timed call against a market everyone else refused to question. Today's Burry has been making the same bearish call in different costumes for the better part of two decades against a market that keeps handing patient investors the best returns of their lives.
But the Cape ratio hit 40, only the second time in 155 years. Yes, and here's the part that quietly dismantles the comparison. In 1999, the companies sitting at the top of the index had no earnings and no moats. Today, the top of the index is Apple, Microsoft, Nvidia, Google, Amazon. Companies with the fattest margins and the deepest moats ever assembled throwing off hundreds of billions in actual profit.
A high valuation built on real, growing earnings is not the same animal as a high valuation built on hope. Valuation tells you what to pay. It does not tell you when to sell. People have been calling this market overpriced at 3,000, at 4,000, at 5,000 on the S&P. Every one of them was right about the valuation and poorer for acting on it.
And here's what the doom narrative leaves out entirely. The other side of every Force Burry names. He says passive investing creates fragility. True, on the way down. But those exact same automatic flows are a relentless bid on the way up. Every paycheck, every retirement contribution, buying regardless of price.
He says corporate buybacks are collapsing because the cash is going into AI. But that cash going into AI is the demand that's driving the very revenue he claims doesn't exist. He's describing the fuel and calling it the fire. Every bear case here is a bull case read upside down. The only real question is which way the spring uncoils.
And a spring loaded with a decade of structural inflows, real earnings, and accelerating demand has a very clear bias. Now, let me be completely honest with you because you deserve both sides and I am not here to sell you a fantasy. Burry could be partly right. Valuations are stretched. Palantir at 180 times earnings is priced for perfection.
And if growth so much as hiccups, that stock can fall hard. A 20, 30, even 40% correction in AI names isn't just possible, it would probably be healthy. If you bought Nvidia at the very top with money you need next year, you are taking real risk and you should know it. Burry is not an idiot. He's a brilliant analyst pointing at something genuinely real, extreme prices.
Where he is dead wrong is the leap from this is expensive to this collapses 65% on my exact timeline. Diagnosing a fever is not the same as predicting the patient dies on Tuesday. So, what does this actually mean for you? It means stop trading on headlines from a man who deregistered specifically so you can't even see whether he still holds the trade.
And it means the lesson of every bubble is not sell everything the moment a famous bear opens his mouth. The lesson is the opposite. The people who compounded real life-changing wealth were the ones who stayed invested in genuinely productive technology through the noise, through the corrections, through every permabear victory lap.
AI is the most productive shift since the internet, and the internet had a savage crash right in the middle of it, and still minted more wealth than any technology in history for the people who simply did not flinch. Here's the pattern that actually matters. In every transformative era, a smart, credible bear stands up and says the prices are insane.
And he's usually right about the prices. And he is usually catastrophically wrong about what to do about it because he confuses this is expensive with this is over. The expensive part can last a decade. So, I'm not telling you to buy anything. I'm telling you not to let one terrifying headline scare you out of the most important technology of your lifetime at exactly the wrong moment.
The question isn't is Burry smart? He is. The real question is this. When the next scary headline drops, and it will, are you going to make a permanent decision based on a temporary fear? Because the people who build real wealth are not the ones who dodged every dip. They're the ones who knew the difference between a fever and a funeral before they sold, not after.
Comments 0