NVDA

 Investment Thesis: 

• The co makes the key semi chips for AI: GPU, dominating the field (90%+ market share)
• High Growth: the Latest 3Q25 saw a rev increase of 63% and EPS of 67%. 
• Growth Margin: 73%
• Operating Margin: 63%
• ROA: 66%, ROE: 91%, ROIC: 83%

Risks

• Sustaining these incredible high growth rate and high margin depend on ever increasing capex from hyper scalers (google, msft, meta, amzn). Msft spent $120B in year on AI, for instance, others are spending similar amount. Can this be sustainable? The slowdown is inevitable. The question is when. 
• Druckenmiller says invest on future, not on past nor present! (I skate to where the puck is going). 
• It is possible other players can step up the purchases when big techs slow down their capex?
• Competition: Google has TPU (Tensor PU), which is cheaper and seemingly more efficient for specific tasks. Gemini, which leads their peers, was trained on TPU. Google will make it available to others including META. Amazon and other have their own custom silicone as well. 

Technicals

• It is at the 80 day and 15 day moving average cross point. I lowered the position size from over 10% to 3%. 

Trading: 

• Initially, I got in at a decent level and position size.
• Subsequently, I bought high (on fake breakouts) and sell low (loss cutting:(, numerous times.