From a theme to a map
June was the month my research moved beyond collecting AI-related company names. I started following the spending chain: cloud investment, accelerators and servers, HBM, advanced packaging, networking, then power and cooling. This was a working research map, not a list of buy recommendations.
A more useful set of questions
A capital expenditure announcement is only the beginning. I want to understand when spending becomes an order, when an order becomes revenue, and whether the supplier has pricing power or simply needs more capital to keep up.
- Where is the capacity or yield constraint?
- Which part of the chain captures the economics?
- How much optimism is already in the valuation?
- What observation would make me change my mind?
The interesting question is where the next constraint lives.
Better research, quieter execution
My June review also exposed a gap: a deeper understanding of a theme did not automatically produce better trading. Repeated entries and overlapping exposures made the execution noisier than the research. I want the map to help me choose fewer, clearer expressions of an idea.