Key points
- Michael Burry sold half his Oracle (ORCL) January 2027 puts on July 17, struck in the low-mid $100s, after the stock hit a 52-week low of $121.50, down 63% from its all-time high.
- He called reports that he'd closed the whole position "fake news" and says he still holds the other half.
- Same post: he added Nvidia (NVDA) December 2026 puts, bought more DraftKings (DKNG) around $25, and said Netflix (NFLX) would need to fall into the low $50s to interest him.
- Oracle's credit rating was cut to BBB-, one notch above junk, on July 9, with S&P citing OpenAI's roughly 50% share of Oracle's $638 billion order backlog.
I like checking back in on a trade once the market has had a few days to argue about it, and Michael Burry's Oracle (ORCL) short is a good example of why that's worth doing. Reports spread this week that he'd closed his bet against Oracle entirely, six months after he first disclosed it. He pushed back publicly, and the real story turns out to be more specific than either the original headline or the correction: he sold half, not all of it, and he did it because the trade had already worked.
What he actually sold
Burry's Substack, Cassandra Unchained, posted "Trading Post July 17, 2026" the same day Oracle touched a fresh 52-week low of $121.50, a level roughly 63% below the stock's all-time high of $345.72 set last year. In the post, he said he had sold half of his Oracle January 2027 put options, struck in the low-to-mid $100s, because the position had grown "too big." He kept the other half in place.
That distinction mattered because it got lost fast. Within a day, several outlets reported he had exited Oracle completely, treating it as the end of a six-month campaign against the stock. Burry called that read "fake news" and said he had "covered half because winning." He still has a real short position on Oracle. He just took some profit off a trade that had gone his way.
A thesis that's since picked up outside backing
Burry first disclosed Oracle puts back in January, shortly after he deregistered Scion Asset Management and moved to disclosing trades on Substack instead of through SEC filings. His case then centered on three things: Oracle's aggressive buildout of AI data center capacity, the debt he thought was financing too much of it, and how much of Oracle's growth depended on one customer, OpenAI. Oracle had already piled up roughly $95 billion in debt, making it the largest non-financial corporate borrower in Bloomberg's high-grade bond index.
That thesis stopped being just Burry's opinion on July 9, when S&P Global Ratings cut Oracle's long-term credit rating a notch to BBB-, one step above junk. S&P's reasoning tracked Burry's almost point for point: it flagged OpenAI's concentration risk, noting OpenAI now accounts for roughly half of Oracle's $638 billion order backlog, and it raised its own forecast for Oracle's free-cash-flow deficit in fiscal 2027 to $42 billion, nearly double its prior estimate, as projected capital spending climbed toward $90 to $95 billion. We've covered S&P Global (SPGI) before, when its own CEO bought a million dollars of the stock not long after taking the job.
Oracle stock actually rose the day of the downgrade, as investors focused on the record backlog rather than the credit warning. It kept sliding afterward anyway, and hit that $121.50 low eight trading days later, which is the point at which Burry decided the position had gotten big enough to start selling part of it.
The rest of what he disclosed that day
Halving the Oracle position wasn't the only thing in the July 17 post. Burry made three other moves at the same time:
- He added to his Nvidia (NVDA) short, buying more December 2026 put options struck in the low $100s, a bet he described as a small, "pocket change" addition rather than a major new position. He said Nvidia still looks expensive to him given margin and earnings risk over the next five years.
- He bought more DraftKings (DKNG) stock in the $25 range, adding to the stake he first built on July 8. He said there was "no change in the basic thesis."
- He weighed in on Netflix (NFLX), which had fallen to around $68 in the session after its earnings report, well off its June 2025 peak near $134. He didn't disclose a position either way, just an opinion.
The DraftKings stake traces back to a July 8 post, where Burry built a position split roughly 60% Flutter Entertainment (FLUT) and 40% DraftKings, buying Flutter around $107 and DraftKings in the low $26s. His reasoning was regulatory, not sports-related: he expects prediction markets, the newer betting-style products that have been pulling customers away from licensed sportsbooks, to eventually get pulled into the same regulation and taxes that DraftKings and Flutter already operate under. "I believe that the political climate will not tolerate this," he wrote. "Prediction markets exist in a loophole adjacent to a heavily regulated and taxed industry. In time, prediction markets will be subsumed into regulation and taxation."
His read on Netflix
Netflix wasn't a new trade, just commentary, so there's no position size or entry price attached to it. Burry questioned whether Netflix's content library holds up the way Disney's or Pixar's does, the kind of catalog people rewatch across generations, and suggested that gap matters more to Netflix's long-term value than the market currently prices in. Using the same valuation approach he laid out earlier in July, centered on a benchmark he calls IV15, he said Netflix would need to fall into the low $50s before it looked genuinely cheap to him. That's not a prediction the stock gets there, just where his own math says the price would need to go.
Where the trades stand right now
As of today, Oracle is trading around $123, down about 2.7% on the day and still sitting almost exactly where Burry cut his position, which tells you his remaining puts are close to the money rather than deep in it. Nvidia is around $205, meaning his new December puts, struck in the low $100s, are a long way from paying off and would need the stock to fall by roughly half to be worth anything, consistent with him calling it a small, low-cost addition rather than a core bet. DraftKings is trading just under $25, essentially flat against both of his entry prices, and Flutter is just under $107, almost exactly where he bought it. Netflix is around $67, down further since his post and still well above the low-$50s level he said he'd want to see.
None of that tells you whether any of these end up working. It's been a few days, not months, and Burry has been early on trades before.
The bigger picture
The honest way to read July 17 is that Burry didn't turn more negative on the market or more positive, he just managed the size of bets he already had. He took profit on a short that was working, added a little more to one that wasn't yet, kept buying a stake he'd already started, and offered an opinion on a stock he doesn't own. That's a more boring story than "Big Short investor exits Oracle," which is probably why the wrong version spread faster than the correction did.
It's also the second time in a row that Burry has cut a short in half rather than closing it outright. He did the same thing with his Palantir (PLTR) short in June, and he's added other AI-adjacent shorts since, including Tesla (TSLA) and Caterpillar (CAT) and a short against Micron (MU). Taking half off a winner instead of closing it is starting to look like a pattern for him, not a one-off.
Because Burry deregistered Scion, none of this shows up in a 13F. His last one is nearly a year stale and there won't be another (here's how those filings work if you want the background). We track his Substack posts directly, alongside every other fund's SEC filings, on his Michael Burry fund page.
This is general market commentary and opinion, 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.
