Even at $5 Trillion, Nvidia Is Underpriced
The Wall Street Journal’s “Heard On The Street” column on Friday made an unusual argument: Nvidia (NVDA), despite a $5 trillion market cap, is still underappreciated by the market. The argument isn’t that there’s no competition — AMD and Arm just demonstrated they can take real share at the CPU layer. The argument is that the addressable market is growing faster than competition can erode share.
Nvidia reported record quarterly sales and earnings on Wednesday, beating expectations across the board on agentic AI demand. The stock closed down 1.8% on Thursday despite the print. That’s a classic “sell-the-news” setup on great earnings — not a thesis change.
The Anthropic Disclosure
The most important data point in this week’s news flow wasn’t from Nvidia. It was from Anthropic. Buried in their Q2 update: Anthropic disclosed that in Q1, they spent 71 cents on compute for every dollar of revenue. That ratio is dropping — Q2 guidance is 56 cents per dollar — but the absolute dollar amount is exploding because revenue is more than doubling.
That money flows to Nvidia. Then to Taiwan Semiconductor (TSM), which manufactures the chips. Then to SK Hynix and Micron (MU) for the high-bandwidth memory. Then to Vertiv (VRT), Bloom Energy (BE), and the power complex.
AMD’s $10 Billion Bet
Advanced Micro Devices (AMD) made the strategic announcement this week: $10 billion-plus committed to Taiwan’s semiconductor sector to meet AI infrastructure demand. Their next-generation Venice CPU is already in production on TSMC’s most advanced 2nm node. The future plan is to manufacture at TSMC’s Arizona facility.
This is the AMD story we’ve been positioned for. As one analyst put it: “Nvidia is still in pole position in this AI trade, but Intel and AMD lead the way on CPUs for now.” Both can win. The pie is big enough.
What We’re Doing
Reinforcing NVDA, AMD, ARM, GFS, and TSM today. Adding IBM on the quantum-foundry catalyst. Adding Bloom Energy (BE) and reinforcing Enphase (ENPH) on the power-derivative breakouts. Watching Parker-Hannifin (PH) on the Circor aerospace deal.
The trade is not about owning any one of these names. It’s about owning the basket. The CW Aggressive Tier I/II/III/IV books are over-indexed to this complex because the data keeps validating it.