· 方法论
涉及标的： $AAOI$AEHR$AMKR$AXTI$COHR$INTC$LITE$MU$SNDK$TER$TSEM$TSM
对于绝大多数散户而言： 如果你想搭乘当前资本支出(capex)趋势的快车，我认为以下这些是必须持有的标的： 1. 存储(Memory) - $MU, 三星, 海力士, $SNDK 2. 光子学(Photonics) - $LITE, $COHR 3. CoWoS/晶圆代工(Foundry)/先进封装(Advanced Packaging) - $AMKR, $TSM, $INTC 随着向产业链上游深入，存在许多细微的瓶颈环节，例如我常提到的： 光子学的基板/原料层面的 $AXTI，共封装光学(CPO)中ELS领域的利基玩家如 $AAOI，良率相关的 $TER 或 $AEHR，硅光(SiPh)领域的 $TSEM。甚至是数据中心中从日月光(unimicron)到其他厂商的基板铜用量。 我的观点是，对于绝大多数(99%)的人来说，你可以选择“简单模式”生活，无需追踪 X 上的每日更新或 $AMKR 的资本支出流向。 只需关闭大脑中关于供应链映射/更新的部分，坚持持有像 $TSM 这样处于核心地位的标的，有时是更好的选择。 它很可能也会跑赢大部分上游玩家。
For the vast majority of retail: If you want to ride the capex trends happening right now, these are probably must have imo: 1. Memory - $MU, Samsung, Sk Hynix, $SNDK 2. Photonics - $LITE, $COHR 3. CoWoS/Foundry/Advanced Packaging - $AMKR, $TSM, $INTC There's a lot of nuanced bottlenecks as you go upstream I talk about like : $AXTI in the substrate/feedstock level for photonics, random niche players like $AAOI for ELS in CPO. $TER or $AEHR for yields, or $TSEM for SiPh. Or even copper usage in DCs to substrates from unimicron to others. My opinion is that for the vast 99% of people, you can live life on easy mode without tracking day-to-day updates on X or where the capex spend from $AMKR goes. Just turning your brain off from all the supply chain mapping / updates, then just sticking with things like $TSM which is the center of it all is sometimes the better thing to do. It probably outperforms a large percentage of the upstream players as well.