苏36
09-01 16:42

My view: the memory supercycle still has room to run, but the easy money is probably behind us. The key is separating DRAM from NAND. DRAM remains structurally tight because AI accelerators and HBM are consuming enormous wafer capacity, while new fabs take years to build and ramp. NAND, however, faces a clearer risk of supply catching up with demand in 2027.

That makes me more comfortable with MU and SK hynix than pure NAND exposure. SK hynix’s ₩40T buyback is a strong signal that management remains confident in future cash generation, but buybacks are not a guaranteed price floor.

At ~85% gross margins, expectations are already extremely high. I would watch DRAM contract prices, HBM demand, hyperscaler capex and gross margins rather than stock charts. If DRAM prices roll over while AI capex slows, the cycle could turn quickly. Until then, I remain bullish—but I would buy selectively on pullbacks rather than chase momentum.

@TigerObserver [暗中观察]

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Report

Comment（1）

  • Top
  • Latest
  • BartonBecky
    ·09-01 18:22
    DRAM contract price growth already cooled for two straight months, so calling it structurally tight feels rich lol. HBM can't cover every weak end market forever
    Reply
    Report