"Big Short" Investor Michael Burry Is Now Betting Against Micron, Nvidia, and Tesla. Should You Be Worried?

Michael Burry disclosed new short positions in Nvidia, Tesla, Applied Materials, Caterpillar, SOXX, and an additional short in Micron near $1,052/share, arguing AI-stock valuations are stretched. Micron reported fiscal Q3 revenue of $41.5B (+>300% YoY) and guided ~ $50B revenue for fiscal Q4, yet the article notes it still trades around 22x earnings; Tesla delivered 480,126 vehicles (+~25% YoY) but trades at >350x earnings and fell ~7.5% on the report date. Overall, the piece is more about valuation-cycle risk than immediate earnings deterioration, so impact is likely more sentiment/positioning than fundamentals in the near term.

Analysis

Burry’s setup is less a blanket “AI bubble” call than a dispersion trade across the capex stack. The most fragile leg is not the marquee GPU vendor; it’s the memory/equipment complex where earnings are most cyclical and incremental demand can reverse fastest if hyperscaler budgets normalize. That makes MU and AMAT the cleaner shorts versus NVDA, while SOXX is the best basket hedge for a broad de-rating if the market starts to question 2026 AI spending assumptions.

The second-order risk is that AI infrastructure is now interlocked with power, industrials, and data-center buildout, so a capex pause would ripple beyond semis into contractors, electrification, and heavy equipment proxies. CAT is less a direct AI beneficiary than a sentiment proxy for “build everything now” optimism; if that enthusiasm cools, CAT can underperform not on fundamentals but on multiple compression. Conversely, NVDA likely needs a demand or margin miss to break; valuation alone is not enough when earnings momentum is still compounding.

The contrarian miss is timing: crowded shorts can lose badly into one more quarter of upside revisions, especially if cloud capex and memory pricing stay tight through the next earnings season. For TSLA, the risk is a narrative gap rather than a cycle break; the stock can remain detached from fundamentals until a concrete robotaxi/software milestone slips. The thesis is most falsified if MU guidance stays elevated into the next 1-2 quarters and NVDA data-center growth reaccelerates without margin erosion.

Key Decisions for Investors

• Prefer relative-value shorts: initiate/maintain short MU vs long NVDA into the next quarterly print; MU has the highest earnings cyclicality and the most convex downside if memory pricing peaks.
• Use SOXX as the cleaner hedge for a broad AI-capex fade over 1-3 months; add on strength only if semiconductor breadth weakens and capex commentary turns cautious.
• Avoid outright shorting NVDA on valuation alone; wait for evidence of order digestion or margin compression. If you must express a bearish view, use put spreads into earnings rather than stock.
• For TSLA, consider a short-dated put spread only on rallies above the recent post-delivery range, with the thesis tied to a missed robotaxi milestone or another quarter of margin disappointment; otherwise the short is too timing-sensitive.
• Watch CAT and AMAT as secondary shorts if hyperscaler capex guidance rolls over; if they underperform SOXX by >5% on a single guidance reset, it would confirm the broader buildout trade is cracking.