Micah L. Bowens

I expect a large beat & raise when $NVDA results are released on 5/22 given more supply available from $TSM. Given the stock is currently trading 15% below its 5 year average forward PE but expectations are EXTREMELY high, I would expect a small positive reaction from the stock. AI landscape: If I compare the surge in spending on AI today to that of the internet in its early days: 1) From the launch of ChatGPT in late 2022 through Q1 of 2024 it has been roughly 5 quarters. Nvidia, the enabler of Generative AI, is expected to see its revenues increase over 4x from ~$6B (Oct qtr of 2022) to an estimated $25B for their soon to be reported Q1 (April qtr of 2024). 2) NetScape Navigator, the first mass internet web browser, was launched in December of 1994 and through Q1 of 1996 would be an equivalent 5 quarters. During this time, Cisco, the leading enabler of the internet infrastructure buildout, saw its revenues increase from $434M (Oct qtr of 1994) to $1,087M (April qtr of 1996) up 2.5x. 3) Cisco’s revenues peaked at up ~16x by the January of 2001 qtr at $6.7B compared to late 1994 and the stock increased 4000% to become the most valuable company in the world. Nvidia by comparison since the end of 2022 has seen revenues increase over 4x and their stock 533% to become the third most valuable company in the world. 4) From a valuation perspective, NVDA’s forward PE is in fact 15% below its average for the past 5 years. NVDA trades at ~35x forward twelve months PE for ~100% revenue growth in CY24. Cisco traded at a forward12 month peak PE of over 135x during the Tech Bubble for nearly 60% revenue growth so this is nowhere near a bubble for NVDA’s stock. 5) I believe that customers have been double and tripling ordering NVDA chips since the start of the AI boom but are unwilling to cancel orders as they await shipments of Nvidia’s new Blackwell chip in late 2024. While investors may anticipate an air pocket in demand before the Blackwell launch, I would expect a period of digestion in early 2025 once customers have received their initial Blackwell orders. 6) I expect increasing declines of 25-35% going forward for NVDA’s stock now that lead-times have come down from roughly one year to 3 months for their chips. As an analogy, Cisco’s stock saw declines of 26% in late 1995, 38% in 1997 and 34% in 1998 on its way to increasing ~4000% to its peak in 2000 from the end of 1994. I believe investors rightly will become more worried with Nvidia, much like they did with Cisco, if/when excess orders will get cancelled or spending will slow down now that supply is catching up with demand. 7) The spending on AI is being driven by some of the most valuable and cash flow positive companies in the world today such as $MSFT $GOOGL $AMZN $META. This compares favorably to much of the spending during the internet buildout which was more debt fueled with a mentality of build it and they will come and a focus on “eyeballs” and not profits to some extent. In summary: I believe the AI infrastructure buildout could be as powerful for NVDA as it was during the internet infrastructure buildout for Cisco when revs increased ~16x over 5+ years and the stock rose ~4000% but prepare for larger bumps and bruises along the way. I expect to see multiple 25-35%+ drawdowns over the next 3-4 years while Nvidia’s stock ultimately goes up 3-4x from current levels along with revenues.