🐦 推特SerenityV1 · 实盘不做把握 8/10
北京时间·🕒 发布 22:02:20时区推断·⟳ 抓取 22:03:05·⚡ 决策 06-15 22:07:27 +4m22s·134.1 天前
KOL发布AI算力三大主题(Neoclouds/Photonics/Memory)投资回顾，部分标的已创新高或面临增发抛压
📄 原文Serenity↗ 原文
I’ve written a thesis on these 3 themes early on: Neoclouds, Photonics, and Memory. Now, it’s fun to sit back and watch…
不做 NBIS.US
KOL回顾帖，标的已创历史新高(ATH)，底层Neocloud逻辑已充分price-in，无新催化剂，追高期望值为负。
不做 IREN.US
KOL明确指出公司面临60亿美元ATM(增发)持续抛压导致股价停滞，存在明确的基本面稀释利空，回避。
不做 AAOI.US
光子学/光通信旧逻辑回顾，无新边际催化剂；KOL总结帖常处于行情中后段，缺乏独立做多edge。
不做 EWY.US
韩国ETF，存储(HBM)逻辑已发酵多时且为市场共识，作为ETF弹性有限，无新事件驱动。
展开详情(成交明细 / 研究档案 / 原文全文 / 决策记录链)
研究档案
这条决策没有研究档案。
原文全文
展开原文 · 798 字
I’ve written a thesis on these 3 themes early on:
Neoclouds, Photonics, and Memory.
Now, it’s fun to sit back and watch all my thesis ideas play out from $AAOI to $EWY to $NBIS.
Even got my warnings right too, $IREN is still stagnant due to the $6B of constant selling pressure from the ATM, while $NBIS reaches ATHs.
But the bagholders still don’t want to admit it.
Think a core part is knowing what theme comes next with markets, then comes picking the winner + heavy concentration in them.
If you went long on software and chose the ideal stock, you’d probably end up not as happy?
Photonics is still probably the earliest out of the three.
But I can see Nebius end up like AWS one day.
And $MU / SK Hynix / Samsung potentially end up like a mini $NVDA if memory demand is structural.
Neoclouds, Photonics, and Memory.
Now, it’s fun to sit back and watch all my thesis ideas play out from $AAOI to $EWY to $NBIS.
Even got my warnings right too, $IREN is still stagnant due to the $6B of constant selling pressure from the ATM, while $NBIS reaches ATHs.
But the bagholders still don’t want to admit it.
Think a core part is knowing what theme comes next with markets, then comes picking the winner + heavy concentration in them.
If you went long on software and chose the ideal stock, you’d probably end up not as happy?
Photonics is still probably the earliest out of the three.
But I can see Nebius end up like AWS one day.
And $MU / SK Hynix / Samsung potentially end up like a mini $NVDA if memory demand is structural.
决策记录链 · 只追加
每次领取产生一条记录,重新决策追加一条并以 supersedes 指向被替代的记录(ADR 0014)。
这条决策早于决策账本(2026-10-08 起逐条追加),没有追加记录。