Original Insight Summary

🧠 Semiconductor & Memory Cycle

• Memory looks like the highest-conviction semiconductor lane. The logic is simple: even the most aggressive sell-side numbers are now pointing the same way. UBS expects 2026 global semiconductor revenue at US$1.62T / 1.62万亿美金, with memory at US$960B / 9600亿美金. Gartner previously had 2026 total semiconductor revenue at US$1.32T / 1.32万亿美金, with memory at 6300万亿美金. Different forecasts, same direction: 2026 memory demand is being priced as the core growth engine. 1

📉 Tactical Trading Setups

• $BABA is fighting to hold the $100 psychological support. The tape is ugly: down 6 weeks in a row, and it has NEVER gone 7 red weeks in a row in its entire history. Price is also sitting on a blue trendline support, so the setup is not a clean long-term bull call, but a potential tradeable bounce watch. 2

• $RKLB is sitting on 3 confluences of notable supports. The read is straightforward: support stacking increases the odds of a near-term bounce, so it belongs on the watchlist rather than being chased after the move. 3

🧭 Long-Term Investing Framework

• Fundamentals beat charts, especially in U.S. equities. Staring at K-lines and worrying about one- or two-day moves is low-signal. The better process is to first judge whether the business is high quality, then read the financials, growth rate, and current valuation. That beats trading headlines and chart noise. 4

• Microsoft remains a long-term quality compounder. A few months, or even a few quarters, of weak price action does not mean the business is broken. Sometimes the market’s hot money is simply elsewhere. That is rotation, not a thesis kill. 5