📈 Trading Calls & Execution

• $CIEN trade was closed for solid profit. The read is simple: payday taken, cash back in hand, no need to overstay the move. 1

• A prior add-on level has already delivered more than 10% upside. At this stage, trimming or taking profit makes more sense than chasing. 2

• The 820 level was framed as an add zone during the dump. Now that price has bounced, flipping bullish after the move looks like pure momentum chasing. 3

🧠 AI / Semiconductor Cycle

• SOX 2026 may look a lot like NDX 1999, but the setup is not a clean repeat of SOX 1999. The bubble analogy is useful, but too blunt if applied one-for-one. 4

• Rubin Ultra is facing serious implementation risk. A 5.5x reticle size version is already difficult in testing, and a 9.5x implementation is viewed as extremely hard. The key bottleneck appears to be advanced packaging/soldering execution into the first half of 2027. 5

• AMD–Anthropic could become a bigger deal than the OAI and Meta deals. If that expectation lands, the market may need to reprice AMD’s AI demand pipeline. 6

🤖 AI Infrastructure & Compute Demand

• $NBIS jumping more than 10% after $NVDA disclosed a ~9% passive stake is not just headline buying. Even if much of the stake reflects previously announced shares and prefunded warrants, the signal is stronger: Nvidia is getting deeper into the AI compute aggregation stack via Nebius. 7

📉 Single-Stock Sentiment Reads

• $NFLX sitting at $67.67 captures the recent tape well: dull, rangebound, and painfully average. No real leadership. No clean momentum. 8

• $ASTS ripping more than 10% after Jim Cramer told investors to wait for $40 creates a classic contrarian setup. The bear case is that the company is “losing a fortune” and still out of favor, but price action is refusing to cooperate with that caution. 9

• Calling for a new Nvidia add only at 160 without data support looks weak. If the thesis is bullish, the entry logic needs numbers, not vibes. 10