원문 보기
On the 28th of September 2026, Michael Burry closed every short he had against Nvidia, Palantir and six other names tied to the AI boom. And he did it all in one day. If you have gotten an index fund in a 401(k) or an IRA, there's a good chance the biggest single company you own is Nvidia. In an S&P 500 fund, about 8 cents in each dollar is sitting inside of it.
A global tracker in a British pension or ISA usually has it near the top too. Burry swapped every one of these for a different kind of bet, one with a date printed on it like a train ticket. He now thinks he knows roughly when the collapse happens and the very next day he posted a warning to everybody trying to copy him.
By the end of this video you'll understand exactly what he did and why. You'll also know how to place his Palantir trade yourself step by step in Interactive Brokers. I also took a $10,000 version of that trade and ran it through five possible futures. In one of them holding on to the end turns your $10,000 into about $7,000.
Following four simple rules however turns it into about $42,000. The only difference is what you decide on day one. So here is what's coming in eight short parts. First, what Burry actually did. Second, what a put is in plain English. Third, why he thinks this happens within a year, fourth what it means for your index funds, then the trade itself, the four rules for getting out, exactly how to place it in interactive brokers and finally the honest risks.
My name's Adam Feather, I research tax and residency options for high earners. I also pay to read Burry's posts every week, footnotes and all. I'm not a financial adviser and this isn't financial advice. You can copy this trade and I'll show you exactly how to, but a bet like this can lose every penny you put into it.
On the market's own odds, that's the most likely outcome here. Part one, what Burry actually did. He has been betting against Palantir for about a year, and he has changed the shape of that bet again and again. In 2025, he bought cheap puts a long way below the share price. It was fire insurance on a house he thought was built of matches.
In June he bought back half of his short near the low at $107. In August and September as the shares ran back up he added again. Then on the 28th of September he swapped the lot. Look at what stayed the same. He has been bearish on Palantir the whole time. What kept moving was the tool, the size and the date.
For Palantir his new bet is on the shares falling to the low hundreds of dollars by September 2027. They closed this Monday just gone at 187. By the time you watch this it could be different but this is the 28th of September that I'm talking about. So he needs a fall of roughly 40% within a year. He made the bet bigger to say so and that matters even if you never touch an option once because Palantir is in the same S&P 500 fund that you probably own.
People say Burry changes his mind every few weeks. What he changes is the date and you can watch it move along that timeline three times this year. So borrow his direction if you wish, which is well evidenced. Borrow his date as a promise though and you're standing on the platform with a ticket for a train that he may not be on.
Then the day after he posted the trade, so two days ago now, Burry posted this. In plain English, the moment he tells his readers what he has bought, big trading firms take the other side and he sees it most clearly in options. His readers pile in, the firms sell to them and his own position starts life as a loser.
It's like announcing on the radio which house you're about to bid on and then wondering why the price went up by lunchtime. He's now thinking of stopping the posts altogether. The man everybody is copying has noticed that being copied is costing him money. You could see it on the day. On the Monday he moved the September $115 put on Palantir traded about 1180 times.
Most of its neighbours traded fewer than 150. Every one of those contracts had a buyer and every one had a seller on the other side. What actually is a put? This is part two of our video. Let's say what a put is in plain English. Until Monday just gone, Burry was also shorting Palantir. If you short a stock, you borrow shares, sell them and buy them back later on, hopefully for less money.
There's no date on a short and if the price keeps rising, there's no ceiling on the bill. It's a taxi meter that keeps on running and you can't get out of the cab. A put is the right to sell a hundred shares at a fixed price until a fixed date. If you've ever brought travel insurance, you already understand it.
You pay a small known amount upfront. If the trip goes wrong, it pays out. If you have a lovely holiday, you don't get the premium back and you don't ring up to ask for it. Either way, you can never lose more than you paid. A few more words and you'll see them on every broker screen. The fixed price is the strike.
The fixed date is the expiry. and every option shows two prices. The bid, which is what buyers will pay, and the ask, which is what sellers want. Burry's new Palantir puts have strikes in the low hundreds and an expiry of September 2027. Next up, why Burry thinks this happens within a year? The collapse of Palantir's share price, to be specific.
Well, a man who has held the same view for a year has swapped a bet with no date for a bet with one. In his own words he now sees a timeline that will play out over the course of the next year. Some of the reasons are practical. Puts are cheap because the market's fear gauge has virtually been unusually calm especially lately and insurance is always cheapest when nobody's expecting any trouble.
But his short also had a trapdoor. He promised to buy it back if Palantir made new highs, only about 11% above last Monday's close. A put can't be taken off you like that. Once you've bought it, it's yours until the date on it, whatever the price does in the meantime, unless you choose to sell it of course.
But the main reason came from the people lending the money. You may not have heard of Ares Management, but it's one of the largest private credit firms in the world. In the autumn of 2026 it sent its investors a newsletter and Burry said it moved his timeline forward. The AI build-out needs more money than the tech companies, the banks and the venture funds have between them.
So much of it is sold as safe investment grade bonds to insurers who hold around nine trillion dollars of American savings. Insurers are built never to lose. These are the people who want to know the age of your boiler before they'll give you a quote, and they're now lending to projects run like venture bets.
Ares traced more than a hundred of these deals in a year. Nearly all of them lead back to eight names. Meta, Oracle, Microsoft, Amazon, Google, NVIDIA, OpenAI, and Anthropic. And those eight fund each other. NVIDIA puts money into OpenAI, which buys computing from Oracle, which buys chips from NVIDIA.
It's four friends at a dinner party passing the same $100 bill around the table, and every single one of them is counting it as income. Every single one of those deals rests on a single assumption, that being that AI spending continues to keep on growing. One flat quarter with a few boardrooms changing their minds on the same afternoon, and the guarantees across the whole web get tested at once.
Burry adds that rising interest rates make every link more expensive, which you'll recognise if you've ever watched a fixed rate mortgage come to an end. Palantir isn't on Ares's map because it isn't borrowing in order to build data centres. Its link is the price. It's valued as if AI demand compounds for years and years without interruption.
Burry's reading of its filings adds three worries. Its revenue behaves more like a consultancy's revenue than a software company's revenue. One customer accounts for about a quarter of the money Palantir is owed, and the time it takes to get paid has risen from about 35 days to 66. The obvious objection is that these are the biggest companies on earth.
True, but last quarter Alphabet and Amazon both spent more than their businesses brought in, and Meta's free cash flow fell nearly 91%. Being big isn't the same as having spare cash. Plenty of people in enormous houses are waiting for payday. and he could still be wrong, Burry of course, if the big companies raise their spending and the revenue arrives to match then the chain holds and Burry is wrong for now.
What it means for your index fund, this is the next part and even if you never buy an option this is relevant. If you own an S&P 500 fund you already own six out of those eight names so get a pen because this bit is about you and it just takes a few moments. Open the fact sheet for the biggest fund in your 401(k), IRA, pension or ISA.
The one whose statement goes straight back in its envelope. It's a two-page PDF nobody in history has ever printed out. Find its top 10 holdings. Write down every one that's meta, oracle, Microsoft, Amazon, Nvidia or Alphabet which is Google's parent company and add up the percentages. That total is how much of your retirement rests on the same assumption as the whole chain.
Whether it's too much is your call. Most people have simply never looked. But what about the trade itself? For instance, if you are convinced that Burry is correct and you want to copy him, well, there are two ways to copy it. Burry's own version is simple. September 2027 put at strikes in the low hundreds, around $105 to $115.
On the day he moved, those cost roughly $600 to $850 per contract. The second version spreads the bet across several strikes so that different sizes of fall each pay something. Here's how to build it with $10,000. That should be money that you could lose completely, by the way, and still book the same holiday next year.
Here's what you do. You buy September 2027 puts at seven different strike prices. One contract each at $150, $140, $120, $110, and $100. Two contracts at $130 and eight at $75. Here is why I have chosen to give you those numbers in order to build it that way. The top strikes cost the most and they're the first to pay on a moderate decline in Palantir stock price.
The lower strikes only pay on a larger drop of Palantir stock price. The 75s costs very little each and only pay if Palantir basically collapses, which is why you buy eight of them. At the prices on the day Burry moved, the whole spread costs about $10,000. Where you can hold it depends on where you live.
Most 401(k) plans won't let you buy options at all and only some IRAs will. In Britain they can't go inside an ISA so any gain is taxed as an ordinary capital gain so be aware of those facts. Before buying anything know the odds. The market publishes them in its prices. There's about a four in ten chance anything at all is left at the end.
There's about a one in four chance that you will get your money back. roughly the odds of a coin landing heads twice in a row. The model behind these odds understates crashes so the real chances perhaps are somewhat better but the trade only really makes sense if you believe with Burry that their odds are a lot better.
And know how much timing matters. If Palantir is at $120 in March 2027, six months early, the spread is worth about $21,900 hundred dollars compared to your initial ten thousand. At the same hundred and twenty on the last day in September it's worth about seven thousand. Same share price a third of the money because an option with six months left still carries the value of what might happen next.
If you want the whole picture pause on this next table. It shows you what the spread is worth at every price from 75 to $205 every two weeks until it expires. Now, the rules for getting out of the trade. The four rules for getting out. You write these down on the day that you buy the options and you don't renegotiate them once the price starts moving, because that's exactly when you will want to.
Rule one takes your stake back if Palantir falls to 150. Rule 2 gets you out if a fall turns into a bounce. Rule 3 banks some profit before results when option prices are at their most inflated. Rule 4 is there because a put still worth something on its final day is exercised automatically and that turns it into a short sale of 100 shares per contract.
You go to bed owning a put and wake up owning shares. There's another reason to have your own rules. Burry may stop telling anyone what he's doing and if he does you won't hear when he gets out. So why not just hold on until Palantir completely collapses? Well I ran the same position through five different futures as I said at the start of the video with the same rules applied on the same 10 dates.
Take future four. Palantir falls to $60 by February then rebounds to $120 by September 2027. If you hold to the end your $10,000 comes back as about 7 ,000. Sell by the rules on the way down and it comes back as about 42,300 which means that you were right either way but the difference is what you wrote down first and what rules you chose to follow which is why you have to know what you're doing when it comes to trading options.
The rules lose sometimes too. In a slow slide holding does a little better. In a collapse that never pauses is holding does far better and if nothing happens at all you lose almost everything either way. Nobody rings a bell at the bottom and that's the whole reason to write the rules down before you need them.
So you're probably thinking well this is interesting but how do I physically logistically step by step do this trade in interactive brokers which is what you promised me that you would show me how to do Adam. Well I will show you right now how to place the trade in eight steps. This is Interactive Brokers because it's one of the few brokers that lets people outside America buy American options easily.
Other brokers work the exact same way. Step one, wait. That's Burry's own advice. After he posts, give it a week or two before following him, and then compare the price of the put you want with its price the day before his post. If it's jumped, the crowd has already bid it up, and you'd be buying from the people selling to the crowd.
Step 2. In your account settings, under trading permissions, ask for US options. It'll ask about your experience, so answer honestly. If it says no, take that seriously. Step 3. Fund the account and convert to dollars first, using the currency conversion screen. If you skip this, interactive brokers can end up lending you the dollars and charging you the interest on them.
Step 4. Under market data subscriptions, add US options data. Otherwise every price you see will be 15 minutes old, which on a day like the 28th of September is a very long time. Step 5. In Trader Workstation type PLTR, right click it and open the option chain. Pick the September 17, 2027 tab and puts.
Step 6. Click the asking price on the strike you want and the order ticket fills in. Check it says buy put September 2027 and the right strike. Next set the quantity in contract. Each contract covers a hundred shares so a price of twelve means twelve hundred dollars. Step 7. Choose a limit order at the midpoint between the bid and the ask and set it today.
Submit it while the American market is open, ideally after the first half hour. That's after three in the afternoon in Britain. If it hasn't filled in 20 minutes, raise it five cents and wait again. Never use a market order on these. Step eight, repeat for each strike and then set two price alerts on Palantir at $150 and at $208 and put the four rules in your calendar.
To sell, right click the position, choose sell and use a limit order at the midpoint again and whatever happens be out by the week before the 17th of September 2027. Now finally to conclude, the honest summary. Burry is often early. He was right about American mortgages roughly two years before it paid and some of his own investors tried to take their money back while he waited.
Arres' own list of lessons says it plainly, cycles are easy, timing is hard. What's different this time is the source. For a year this was one famous short seller arguing with a share price. Now one of the largest lenders in the world describing the same fragility in writing is doing so to its own investors.
It's also a crowded trade with professionals leaning the other way every time Burry posts. And the date is still a guess and the market's odds still say the likeliest outcome for a position like this one is that it expires worthless. Only put in money that you could lose completely and not miss. So what changed on the Monday just gone was not Burry's opinion of these companies.
That had been in print for a year. What changed was who else was saying it and an option is simply a bet on when. Burry has bought his ticket and the date on it is September 2027. You can use his timetable without buying his ticket. It starts with a fund fact sheet and a pen and it just takes a few moments.
And if you do buy the ticket and choose to get these put options doing what Burry does, it starts with your exits written down on the day that you buy them. If you want to read a company the way Burry does, from its own filings rather than from headlines, my free book Cassandra's Playbook walks you through it. The link is in the description below. I will see you in the next one.
댓글 0