Michael Burry expands short bets on Micron, Palantir and semiconductor stocks

Michael Burry increased short positions "in some size" in Micron, Palantir, Nebius and the iShares Semiconductor ETF (SOXX). The bearish positioning reflects his doubts about the durability of the AI and semiconductor boom, creating a negative sentiment signal for the targeted companies and broader chip sector.

Analysis

The disclosure is more relevant as a positioning signal than as a fundamental catalyst: MU, PLTR and NBIS have distinct earnings drivers, so a common short thesis is principally a bet on AI-capex normalization and valuation de-rating. MU is the cleaner cyclical expression because memory pricing and hyperscaler procurement can turn quickly; PLTR and NBIS carry greater duration risk, where a modest reduction in growth assumptions can drive disproportionate multiple compression. SOXX is less exposed to single-name execution but remains vulnerable if AI spending shifts from accelerated build-out to utilization and return-on-investment scrutiny.

Near term, public short advocacy can increase borrow demand and volatility but may also create squeeze risk in high-beta AI names if quarterly guidance remains intact. Over 1-3 months, the decisive data are hyperscaler capex commentary, HBM/DRAM pricing, lead times, and evidence that enterprise AI deployments are converting from pilots into paid production workloads. Over 6-18 months, the bearish case requires AI infrastructure returns to disappoint sufficiently to reduce orders rather than merely slow growth; absent that, shorts in structurally supply-constrained components can be costly.

The contrarian read is that broad AI skepticism may be too blunt. A capex digestion phase would hurt marginal infrastructure vendors and richly valued software narratives first, while suppliers with constrained high-end memory supply could retain pricing power. The key falsifier for a sector short is upward revisions to semiconductor revenue guidance alongside stable gross margins; that combination would indicate that demand is being funded by real workloads rather than speculative inventory accumulation.

Key Decisions for Investors

• Do not mirror the disclosed shorts solely on sentiment. Treat the filing as a volatility alert; require current valuation, borrow-cost and short-interest data before initiating any single-name short.
• For a 1-3 month downside hedge, prefer a modest long put spread on SOXX over outright shorts in MU, PLTR or NBIS. This isolates broad AI-capex de-rating while capping squeeze risk; invalidate the hedge if major hyperscalers raise capex plans and semiconductor guidance revisions turn positive.
• If memory pricing data weaken or MU guides below consensus, consider a tactical MU short versus long TSM as a relative-value expression: MU has greater spot-price and inventory sensitivity, while TSM is diversified across customers and end markets. Exit if HBM contract pricing tightens further or MU sustains gross-margin guidance.
• Avoid initiating a fresh PLTR or NBIS short ahead of earnings without defined-risk options. The appropriate trigger is decelerating remaining-deal-value or commercial-growth indicators paired with unchanged elevated valuation multiples; otherwise the risk/reward is dominated by upside guidance and short-covering.