Micron and Sandisk led a tech selloff on June 23, falling 13%+ each; Western Digital fell 8%, Marvell 9%, Seagate 5%, Qualcomm 8%, and Applied Materials 8%
One analyst says investors are merely taking a ‘breather’ from buzzy tech stocks, following sharp year-to-date gains and ahead …
Context & Ripple Effects
The selloff extends a volatile 2026 pattern in chip equities: Micron, SanDisk and Western Digital had already fallen sharply in late March, while Qualcomm and other semiconductor names pulled back in May after an AI-driven run-up.
The related coverage also shows why memory is a focal point. Micron’s shares were pressured after Google Research described a compression technique, and the sector has prior experience adjusting production when memory demand weakens.
First-order effects
- Micron and SanDisk face the most immediate investor repricing, while Western Digital, Seagate and other semiconductor names are pulled lower in the same risk-off trade.
- The move broadens beyond memory into connectivity and chip-equipment exposure, affecting Qualcomm, Marvell and Applied Materials’ market valuations at once.
Second-order effects
- A synchronized decline makes it harder for investors to distinguish company-specific execution from a sector-wide reassessment, increasing the likelihood that peers trade on shared AI and memory-demand expectations.
- The inclusion of Applied Materials links the selloff from chip makers to their equipment suppliers: weaker confidence in memory economics can quickly feed into scrutiny of future manufacturing-spending plans.
Third-order effects
- If repeated AI-led rallies continue to be followed by sharp, cross-industry reversals, semiconductor valuations may become more dependent on proof of durable demand and capital-spending discipline than on broad thematic exposure.
- The March compression-algorithm reaction suggests that software efficiency advances can become a structural variable for memory investors, because they may alter assumptions about how much hardware capacity future workloads require.
The trend: This is another instance of the semiconductor market testing whether AI-related demand expectations can support elevated valuations across memory, compute, networking and manufacturing equipment simultaneously.