Michael Burry has been building short-semiconductor-book lately. I covered his Micron short, his Caterpillar bet, and his SOXX position. His pattern has been consistent.
He first identifies a narrative the market treats as infallible, finds the structural vulnerability underneath it, and waits.
Thursday, Aug. 6, he moved into two new targets. And the framing he used tells you exactly how he is thinking about the Artificial Intelligence (AI) infrastructure trade right now.
Burry disclosed a short position in Oracle (ORCL) at $144.63 per share and a larger short exposure in Nebius Group (NBIS) at $211.77, pointing out growing leverage risks tied to off-balance-sheet commitments across the tech and cloud sectors.
According to a Substack post, He described companies carrying long-term lease obligations that dwarf their current revenue metrics as “very fat, very large, easy to shoot” fish.
ORCL is down 25.68% year-to-date and 43.41% over the past year, according to Yahoo Finance.
NBIS is up 126.84% year-to-date and 244.67% over the past year. Burry is shorting the loser and the winner simultaneously. This is the interesting part of this disclosure.
Also Read: Michael Burry Latest News and Stories
The off-balance-sheet argument Burry is making
Burry’s bear case is not actually about revenue or product quality. It is about the liability structure.
Many cloud and AI infrastructure companies carry enormous lease obligations for data center space, power contracts, and compute capacity that do not appear prominently on the income statement but represent multi-year committed cash outflows.
When demand for AI infrastructure outpaces supply, those commitments look like competitive advantages. When the supply cycle turns, they become fixed costs against declining revenue.
More Manager Buy/Sells:
- Michael Burry increases his bet against popular chip giant
- Warren Buffett reveals he broke his own investing pattern
- Mark Cuban bets on MLB with Athletics minority stake
Burry’s thesis, as I read it across his recent disclosures, is that the AI infrastructure build-out has pulled forward years of demand into a compressed window.
The companies that signed 10-year data center leases and committed billions in GPU capacity at 2025 and 2026 prices to meet 2026 demand may find themselves over-leveraged when the cycle moderates.
He has made this argument across semiconductor equipment, memory chips, and now enterprise cloud. Oracle and Nebius represent two very different positions on the AI infrastructure spectrum, which is why both appearing in the same short book is the revealing detail.
Oracle — the record quarter Burry is still betting against
I just mentioned that Burry’s bear case is not about revenue or product quality. Because if it were, Oracle wouldn’t be one of his shorts.
Oracle’s most recent quarterly results were objectively strong.
- Q4 fiscal 2026 total revenue reached $19.2 billion, up 21% year-over-year.
- Cloud infrastructure revenue grew 93% year-over-year to $5.8 billion.
- Remaining performance obligations grew $85 billion in a single quarter, from $553 billion to $638 billion.
- Q1 fiscal 2027 guidance calls for total revenue growth of 27% – 29%.
Source: Oracle Q4 Fiscal 2026 Results
Burry actually made money on Oracle previously. Seeking Alpha reports that he disclosed earlier on August 5 that he closed short positions in January 2027 Oracle puts after a “substantial” profit, then re-entered.
“Should volatility come down, I may re-enter Oracle puts,” he said before proceeding to do exactly that with a direct equity short at $144.63.
What Burry is betting against is not Oracle’s cloud business. No. It is the valuation implied by the $638 billion remaining performance obligation figure, which represents contracted but not yet recognized revenue spread over many years.
If Oracle’s capital expenditure requirements to fulfill those obligations grow faster than revenue recognition, the earnings quality degrades. The $20 billion-plus in capital expenditure Oracle committed to for 2026 is the number Burry is watching.
Nebius: The high-growth, high-implied-volatility target
Nebius is the more aggressive of the two short positions. Burry disclosed he chose direct equity short rather than options specifically because put implied volatility exceeded 100%, making options “exorbitantly priced.”
At $211.77, he is short a stock that has returned 244% over the past year, according to according to Yahoo Finance.
Related: Nebius CEO doubles down on capex spending
Nebius reported Q1 2026 revenue of $399 million, up 684% year-over-year, with positive adjusted EBITDA of $129.5 million, according to the company’s May 13 earnings release. Management reaffirmed full-year 2026 revenue guidance of $3.0 billion to $3.4 billion.
Q2 earnings are due August 12. Consensus projects approximately $578 million in revenue and a loss of $0.52 per share, according to Zacks Equity Research.
The vulnerability Burry appears to be targeting is the gap between Nebius’s rapid revenue growth and its balance sheet commitments required to sustain it.
Burry’s long positions — the contrast matters
Burry’s long book provides useful context for understanding what he does and does not like right now.
He expanded his Flutter Entertainment (FLUT) position to a full size following earnings. He also added to Fiserv (FISV), Zoetis (ZTS), and MercadoLibre (MELI), according to Seeking Alpha.
The common thread in the long book is businesses with recurring revenue, finite capital requirements, and near-term profitability rather than speculative infrastructure commitments.
Flutter operates a platform business. Fiserv processes payments. Zoetis sells animal health products. Mercado Libre runs an e-commerce and payments ecosystem.
The pattern is consistent with his overall framework. Burry has long positions on businesses that generate cash today with manageable balance sheets, and short businesses that are committing enormous capital today against revenue that arrives years in the future.
Oracle and Nebius fit that short profile. Flutter and Fiserv fit the long profile.
Related: Michael Burry increases his bet against popular chip giant
