Over the past two years, the strongest logic in the AI industry chain has been simple: not enough computing power, not enough chips, not enough HBM, and not enough data centers. So capital has been continuously pouring into the upstream hardware sector. Companies like NVIDIA, TSMC, SK Hynix, and Micron have become the most direct and certain beneficiaries of the AI wave. But at the same time, downstream cloud vendors and application end have not enjoyed the same valuation premium. Microsoft, Google, Amazon, and Meta have been frantically investing in AI capital expenditure, building data centers, purchasing GPUs, and training models. However, the market's attitude towards them has become even more cautious: you're spending so much money...

最近纳指 100 进行季度调整，换入了 Astera Labs、CoreWeave、Nebius、Rocket Lab、Teradyne，换出了 Charter Communications、Cognizant、Insmed、Verisk Analytics、Zscaler。这次调整放在科技产业变迁的背景下看，其实很有代表性。它告诉我们：科技行业正在从上一轮互联网、SaaS、外包服务、传统通信的叙事...

价值投资之父本杰明·格雷厄姆在《聪明的投资者》里讲过一个非常经典的比喻：市场就像一个叫 “市场先生” 的合伙人。他每天都会跑来给你报价，问你要不要买、要不要卖。有时候他极度乐观，愿意给出一个高得离谱的价格；有时候他极度悲观，又会给出一个低得离谱的价格。问题在于，市场先生每天都在报价，但他的报价不一定代表企业真实价值，很多时候只是代表他当天的情绪。现在围绕美股七巨头和 AI 产业链的定价...

Recently, because I made some position reductions at high levels earlier, coupled with continuous capital injections, the cash position in my account has significantly increased compared to before. When there's more cash, it's easy to want to buy. But the more this happens, the more one must avoid making impulsive moves just because they have cash on hand. So I simply reorganized my operational thinking regarding the "Magnificent Seven" in the US stock market and the AI theme, to avoid emotional trading later when the market fluctuates. My overall judgment is: currently, the US stock market hasn't clearly turned bearish, and the AI theme isn't broken. Especially core tech companies like NVIDIA, Microsoft, Google, and Meta, whose fundamentals remain very strong. But the problem is...