Burry Opened Six Shorts in Three Days. Then the Selloff Hit.
Michael Burry shorted Micron at $1,052, added five AI infrastructure longs, and watched every short turn profitable inside 48 hours as the June jobs report missed by half.
On Wednesday afternoon, Michael Burry shorted Micron at roughly $1,052 per share. By Thursday's close, MU sat at $975 — a 7.3% paper gain in under 24 hours, and the sixth consecutive profitable short in a basket he laid down over just three trading days.
He also added to five existing long positions in AI infrastructure, building a classic pair trade underneath the chaos.
The timing was surgical. The June jobs report, released Thursday at 8:30 AM, printed 57,000 nonfarm payrolls against expectations of 113,000 — the worst miss since early 2025. Combined with a Korean semiconductor rout that had already wiped 8% off the KOSPI, the data triggered the worst two-day momentum unwind on record. Burry was early by roughly 48 hours.
The Micron Short: Not His First MU Bet, But His Most Aggressive
Burry has traded Micron before — he held common stock as recently as last year — but this is the first time he's shorted the memory giant directly. His thesis, laid out in the July 2 Trading Post on Substack, is quantitative and specific:
- 200-day moving average extension: MU traded more than 100% above its 200DMA at the peak. Burry noted that historically, "the relatively few times 200DMA price extension has exceeded 100%, a 30% fall within 12 months is nearly certain."
- Peak-to-trough timing: Across prior cycles, crashes from this extension level lasted 2.6 months at the 25th percentile and 7.4 months at the 75th percentile.
- Why stock, not puts: "I shorted the stock directly as the puts seemed expensive. I will look to add puts should the stock settle down and bring volatility down."
Micron's numbers make the extension case stark. The stock hit an all-time high of roughly $1,255 earlier in the quarter and traded at $1,213 on June 25. By Wednesday, when Burry shorted, it had already fallen 13% — but was still more than 80% above its 200-day moving average. He fired into the selloff anyway.
| Date | MU Close | Event |
|---|---|---|
| Jun 25 | $1,213.56 | Near all-time high |
| Jun 30 | $1,154.29 | Burry discloses NVDA/TSLA/SOXX/CAT/AMAT shorts |
| Jul 1 | $1,032.28 | Burry shorts MU intraday ~$1,052; KOSPI drops 8% |
| Jul 2 | $975.56 | NFP miss (+57K vs +113K); Burry's MU short up 7.3% |
The Full Basket: Six Shorts, Three Days
Burry's June 30 Trading Post was titled, simply: "A little window-dressing, the $SOXX, and Caterpillar $CAT — It is only a matter of time now." He disclosed five shorts that day. Two days later, MU made six.
| Ticker | Direction | Entry Date | Jun 30 Close | Jul 2 Close | Total P&L |
|---|---|---|---|---|---|
| NVDA | SHORT | Jun 30 | $189.00 | $194.83 | −3.1% |
| TSLA | SHORT | Jun 30 | $420.60 | $393.45 | +6.5% |
| CAT | SHORT | Jun 30 | $1,065.00 | $963.53 | +9.5% |
| SOXX | SHORT | Jun 30 | ~$641 | $569.43 | +11.2% |
| AMAT | SHORT | Jun 30 | $729.00 | $603.04 | +17.3% |
| MU | SHORT | Jul 1 | $1,154.29* | $975.56 | +7.3%† |
*MU Jun 30 close shown for consistency; Burry entered at ~$1,052 on Jul 1 †From $1,052 entry
NVDA is the only short underwater — and barely. The stock's relative resilience (down only 1.4% on Thursday, with positive signed delta and accumulation signals in our distribution data) has made it the weakest link in the basket. Everything else is working.
The Five Long Additions: A Pair Trade in Plain Sight
Behind Burry's paywall, he also disclosed five additions to existing long positions. Cross-referencing Instagram summaries and reporting from TheStreet, the theme is unmistakable: AI infrastructure buildout plays.
- Four of five additions are tied to AI infrastructure — semiconductor test equipment, GPU cloud, and AI connectivity
- This is a classic pair trade: short the overhyped AI beneficiaries (chipmakers, the semi ETF) and long the picks-and-shovels that benefit from the buildout regardless of whether the AI monetization thesis holds
- Candidates include semiconductor test equipment names like FormFactor (FORM, down 15.7% Thursday), Teradyne (TER, down 12.2%), and Aehr Test Systems (AEHR, down 14.5%), plus GPU cloud (CoreWeave/CRWV) and connectivity (Arista/ANET)
The longs got hammered alongside the shorts on Thursday — it was a "sell everything AI" day, not a clean hedge — but the positioning tells you Burry is betting on a two-speed outcome: AI infrastructure spending keeps growing while chipmaker valuations contract.
The Bigger Picture: Burry's Dot-Com 2.0 Thesis
Thursday's events didn't happen in isolation. Burry has been building this argument for months:
- November 2025: Disclosed puts on NVDA and Palantir, then deregistered Scion Asset Management. He now operates exclusively through his paid Substack, Cassandra Unchained — there will be no Q2 2026 13F.
- February 2026: Warned CoreWeave had $29 billion in debt, speculative-grade credit, and a 42% implied default probability
- June 2, 2026: Published a "MEGA Trading Post" detailing what he calls the AI depreciation fraud — hyperscalers understating cumulative depreciation by an estimated $176 billion between 2026 and 2028 by booking GPUs on 4-to-6-year schedules when real useful life is shorter
- June 30, 2026: Laid down five shorts in a single day
- July 2, 2026: Added MU at $1,052 as the sixth
The core argument: AI hyperscalers are overbuilding compute capacity, the memory cycle is at a euphoric peak, and the bill is coming due.
What the Tape Shows
Our distribution algorithm flagged Thursday as a confirmed distribution session — warning triggered at 10:43 AM, confirmed at 1:47 PM. The tracked basket fell 1.83% with 54 names flagged. Among Burry's targets:
- AMAT was among the hardest-hit names, with heavy block selling and a distribution score above 60
- TSLA had the worst day in the basket, down 7.5%, with $74 million in volume and aggressive sweep selling
- MU closed at $975 after touching $950 intraday, with block sellers dominating $66 million to $45 million
- NVDA was the outlier — positive signed delta, positive OBV, and block flow tilted slightly toward accumulation
The VIX actually fell 0.55% on the day to 16.15. This wasn't panic. It was orderly liquidation of crowded positions, exactly the kind of unwind Burry's thesis predicts.
What to Watch
MU below $950. If MU breaks the June 23 panic low, the 30% crash thesis gains momentum — $850 becomes the next level on Burry's framework.
Burry's next Substack post. He explicitly said he would "add puts should the stock settle down and bring volatility down." A quiet bounce in MU followed by a vol compression is his signal to scale the bet.
No 13F to wait for. Scion Asset Management was deregistered in November 2025. Burry's Q3 2025 13F was his last — he now discloses trades exclusively through his Cassandra Unchained Substack. The only way to track his positioning is to watch for new Trading Posts.
Korean semis this weekend. The KOSPI and SK Hynix trade Sunday night Eastern time. If they stabilize, U.S. semis could bounce Tuesday. If they don't, Burry's basket gets another leg.
CAT earnings. Caterpillar is the odd one out — shorting it is unprecedented for Burry, who has been long the industrial giant for years. If CAT's late-July report shows infrastructure spending slowing, that's a macro signal far beyond semiconductors.
Sources: @michaeljburry Trading Post July 2 (tweet since deleted); @StockMKTNewz MU short report; TheStreet coverage; AI Stockwire; Bargo distribution data and signed flow.
More research at bargo.ai/research.