Memory‑chip selloff hits semis
Memory chip stocks have slumped after a 10‑day selloff, dragging semiconductor ETFs down roughly 10% in March and pushing equipment names like Applied, ASML and Lam Research lower—analysts now call some names 'smart dip buys' for long‑term exposure. Investors are rotating to defensive tactics—dividends, covered calls and blue‑chip cores—while watching cycle risk and export‑control uncertainty. (cnbc.com) (247wallst.com) (247wallst.com)
Google’s TurboQuant paper claims it can cut the memory needed for LLM inference by a factor of six, a revelation markets tied directly to the price drops at SK Hynix, Samsung and Kioxia this week. (cnbc.com) The iShares SOXX basket has slid roughly 10% over the past month, a pullback Morgan Stanley flagged as evidence investors are trimming semiconductor exposure. (finance.yahoo.com) Major equipment suppliers including Applied Materials, ASML and Lam Research have retraced recent gains as traders lock in profits, a trend highlighted in recent 24/7 Wall St. coverage of the sector. (247wallst.com) BNP Paribas’ desk noted recent moves of roughly -6.4% for Applied and -8.1% for Lam in some sessions, warning that historic cycles could put a peak on equipment rallies by late 2026. (seekingalpha.com) Flow data and product picks show capital shifting into dividend-focused ETFs and income overlays, with 24/7 Wall St. and other outlets outlining dividend ETF beneficiaries of the rotation. (247wallst.com) Simultaneously, covered‑call funds and manager reports from Global X show rising interest in option-income strategies as volatility increased in March. (globalxetfs.com) Analysts are flagging two structural risks for the semiconductor trade: a potential cyclic peak in equipment spending later in 2026 and renewed ambiguity around U.S. export rules for AI-capable chips. (seekingalpha.com) Several sell‑side notes argue the latest memory pullback looks like profit‑taking rather than a demand collapse, citing the Jevons Paradox and notes that efficiency gains can lower cost‑per‑token and broaden adoption — a view echoed in Bloomberg and Morgan Stanley commentary. (bloomberg.com)
