Key points
- Burry placed a bet against Micron (MU) on July 2 with the stock at $1,052. It traded near $868 on July 27, about 17% lower.
- He has averaged down on Lululemon (LULU) since 2025, doubling to 100,000 shares in that year's third quarter. It is off about 44% this year.
- He deregistered Scion, so the 13F filings stopped. His trades now appear first on a newsletter with 300,000 subscribers.
- He posted "Sell" in January 2023, then "I was wrong to say sell" two months later.
Michael Burry stopped filing 13F reports. That single change is why most of what gets written about his portfolio is already out of date by the time you read it, and it is why I spent the past few weeks reading his newsletter instead of waiting on EDGAR.
Here is what happened. A 13F is the quarterly form large investors file listing their US stockholdings. For years it was the only real window into what Burry owned. Then he deregistered Scion Asset Management, and the filings stopped. No more forms, no more window.
What replaced it is a newsletter called Cassandra Unchained, named for the Trojan priestess who could see the future and was never believed. He crossed 300,000 subscribers on July 12. Every week or so he publishes what he calls a Trading Post, describing what he bought and sold. A 13F arrives up to 45 days after the quarter ends. These arrive in days. That is a better deal for a reader, with one catch I will come back to.
Putting them all in one place makes a point the reputation tends to flatten, which is that he is not always right. One of these is working very well right now. Another has been going against him for more than a year, and he keeps buying it.
What he owns right now
Everything below comes from a dated post of his own, priced at 11:49 a.m. ET on July 27, 2026.
| Company | The bet | What he did | Where it stands |
|---|---|---|---|
| Micron (MU) | Pays off if it falls | Opened July 2 at $1,052 | $868.38, moving his way |
| Oracle (ORCL) | Options, pays off if it falls, runs to January 2027 | Sold half on July 17 | $119.52, after closing at $114.99 on July 24 |
| Nvidia (NVDA) | Options, pays off if it falls, runs to December 2026 | Bought more July 17 | $196.64, down 4.93% on the day |
| Palantir (PLTR) | Options, pays off if it falls | Disclosed June 26 | $129.54, up 5.39% on the day |
| Lululemon (LULU) | Owns the shares | Averaging down since 2025, latest buys in the low to mid $110s | $117.51, down about 44% this year |
| DraftKings (DKNG) | Owns the shares | Averaging down, bought more near $25 on July 17 | $24.44, down about 31% this year |
| Tesla (TSLA), Caterpillar (CAT) | Pay off if they fall | Disclosed June 30 | $307.90 and $845.20 |
Read down that table and a shape appears. He owns two struggling consumer companies outright. Everything else is a bet that something expensive gets cheaper. His July 23 post sending readers to a Bank for International Settlements report on how AI deals get financed argues the same thing in longer form.
Micron is the one working best. He placed that bet near the highs in early July and the stock has since given back about a sixth of its value. We wrote it up the day he disclosed it. It moved further his way on July 27, when ChangXin Memory's Shanghai listing raised fresh questions about who will be making memory chips in 2027.
Oracle is the one I find more interesting, because he sold the wrong half. On July 17 he cashed in half of an options bet against the company, on a day Oracle touched what was then its lowest price in a year, roughly 63% below its peak. Reasonable enough. Then Oracle kept falling, all the way to $114.99 by July 24. The half he kept earned more than the half he sold. He also had to publicly correct reports that he had exited the whole thing, which tells you how closely people are reading him.
Palantir has not cooperated at all. It was one of his two big bets against AI companies, alongside Nvidia, and it rose more than 5% on July 27 by itself.
Now the part people skip
This ground has been covered to death elsewhere, so here it is in one table rather than another retelling.
| When | The call | What followed |
|---|---|---|
| Sept 2019 | Index fund money is distorting prices, to Bloomberg | Index funds kept climbing for years |
| Mar 2021 | Bitcoin compared to the housing market, later the biggest bubble in history | No collapse on that schedule |
| Jan 31, 2023 | One word, Sell | I was wrong to say sell, March 30. Revisited December 2025: Meme corrected |
What interests me more is the one still running, and it starts earlier than most write-ups let on. Lululemon was already his second largest disclosed holding in the third quarter of 2025, when he doubled it from 50,000 shares to 100,000 and it came to roughly 26% of the portfolio. The stock had fallen about 58% over the course of that year before he doubled. It has dropped another 44% or so in 2026, from around $210 in January to $117.51 on July 27, and he has carried on buying in the low to mid $110s.
The word for that is averaging down. You buy more as the price falls, which pulls your average cost lower and pushes your exposure higher at the same time. He is doing it with DraftKings too, down about 31% this year, where he bought more near $25 on July 17. The June 24 essay defending the Lululemon purchase is worth your time, mostly for how irritated he sounds about the people running the company. His case is that the brand outlasts them, and he calls it a patience trade. New chief executive Heidi O'Neill genuinely has not been there long enough to move any of the numbers. Patience is also the word everyone reaches for when a purchase has not worked.
The gap a newsletter leaves
None of this carries the obligations of a filing. He picks what goes in and when, most of it sits behind a paywall, and he never has to say how large any of these bets are next to the rest of his money. That last gap is the one I notice. A headline about him betting against Nvidia could be describing a serious commitment or a rounding error, and there is no way to tell from outside. The 13F was months stale by the time anyone read it, and it had a size column.
The dated thing I am watching now is Micron. Three weeks in and it has moved a long way his direction. He took his profit on Oracle after a comparable run, and then the stock kept falling without him. If he does the same thing here, we will know within days rather than at the end of a quarter, which is the one clear improvement in all of this.
Longer versions of that distinction: what a 13F does and does not show, and whether copying famous investors actually works.
Sources
- Position disclosures and dates from Michael Burry's newsletter, Cassandra Unchained, June 16 through July 25, 2026, as tracked in our own archive of his posts
- MU, ORCL, NVDA, PLTR, LULU, DKNG, TSLA and CAT prices at 11:49 a.m. ET on July 27, 2026, and settled July 24 closes, via Robinhood market data
- Index fund comments: Bloomberg, September 2019
- Sell call and retraction: Burry's posts of January 31, 2023 and March 30, 2023, revisited December 2025
- Related coverage: the Micron bet, the Oracle sale, and the Microsoft and Palantir positions
This article is for information only and is not investment advice. Markets can go down as well as up, and you can lose money. Always do your own research and consider speaking with a licensed financial professional before making any investment decision.
