I read his paper in details and very much like his modeling and analysis. In a nut shell, his finanical model is not fudamentally different from the one from Goldman Sachs. However, he goes in great details in deriving the data from much broader assumption such as industry segment, Nvidia's ability to penetrate the market, and the steady state margin and R&D expense, cost, etc.
Here is bad and good news
- Bad news is that his headline says NVDA is currently worth $87. I think that will catch the eyeball of everyone as 99.99% of people will not read his paper.
- The good news is that he (for some reason unknown to me) assumed an incorrect and much lower revenue base number for 2024 in his analysis. He used $96B as 2024 revenue, but NVDA aleady "made" $56B in the first 2 quarters, and is projecting to make $125B for the entire of 2024. There is no way NVDA can only bring $40B in the next 2 quarters.
- In terms of revenue growth, his model actually shows 40% revenue growth in 2025, inline to what wall street is expecting, and then 30%, 23%, and 18% growth from 2026 after. So, this is not like those "NVDA can only grow single digit" FUDs. I highly recommend any serious NVDA investors to read his article. You dont have to agree with all his numbers, but you can see how he estimate NVDA's future growth.
I modified his model with actual 2024 revenue number of $125B instead of $96B, and I came back with, suprisingly, $105 fair value today, and $120 fair value by end of 2024 (when NVDA will be priced on 2025 estimate). I kept all his margin estimate which is also lower compared to what NVDA is reporting + what wall street is projecting.
