Cathie Wood: Memory Chip Cycle Is Treacherous, China's Capacity Expansion Will Amplify Volatility
nashnova research
ARK founder Cathie Wood told investors in Hong Kong that memory chips face a treacherous boom-bust cycle, with rising Chinese capacity threatening the sector's pricing — ARK avoids traditional memory plays entirely.
Why does ARK stay away from traditional memory chips?
Wood's core logic: once supply exceeds demand, memory chip prices can collapse fast — the cyclical swings are extreme.
This means → memory is not a "buy and wait" sector. Get the timing wrong and you're deep underwater. ARK opts out altogether.
Her words: "The trend is up, but the cycle is treacherous."
How does surging Chinese capacity change the picture?
High-bandwidth memory (HBM — high-speed memory built specifically for AI chips) remains tight, but Chinese producers are ramping fast.
CXMT (ChangXin Memory Technologies, a state-backed Chinese DRAM maker) has aggressively expanded capacity to grab share, pushing new supply into the existing pricing structure.
In plain terms = a market once locked up by three giants suddenly has a disruptor. Once supply tips into surplus, a price war spares no one.
How far have the incumbents already fallen?
SK Hynix and Samsung Electronics are down 51% and 36% from this year's highs; Micron is under pressure too.
This reflects a market already pricing in the "Chinese capacity shock + cycle peak" scenario — not just a short-term pullback.
What does Wood think about investing in Chinese tech?
ARK has limited interest in Chinese tech, arguing that "anti-involution" policy measures cannot fix deflation or pull the economy out of a deep property downturn.
This means → in ARK's view, China tech sits on an unstable macro foundation. Even if individual companies shine, systemic risk still suppresses allocation appetite.
Does she buy the AI-bubble argument?
Broadcom CEO Hock Tan recently warned that AI trades have turned speculative. Wood pushed back directly: AI bubble fears are overblown.
She expects AI to drive a major productivity lift, with long-term value still underappreciated.
Put simply = Wood's stance is this — she dodges the cyclical risk in memory, but bets on the long-term AI trend. The same person, looking at the same supply chain, reaches opposite conclusions for upstream and downstream.
市场有风险，内容仅供研究参考，不构成投资建议。