We're Still in the First Two Innings — Where the AI Infra Buildout Actually Sits
We're Still in the First Two Innings — Where the AI Infra Buildout Actually Sits
We're Still in the First Two Innings — Where the AI Infra Buildout Actually Sits
Micron nearly doubled from ~$430 to $818 in 30 days. The whole time it was happening, I was reading "you missed it" and "buying at all-time highs is idiotic" in the comments. Meanwhile, one Patreon member followed my Micron cash-secured-put alert from April 15 and rolled it eight times for $1,115 in profit.
That one snapshot captures the AI infrastructure trade in miniature. The market says the game is over. In baseball terms, we're around the first or second inning.
Where the $700B Actually Goes
In this capex cycle, Big Tech has committed roughly $700B to AI infrastructure. What makes that number interesting is where the money doesn't go — not into clicks, not into ad slots. It flows into the physical bottlenecks: data centers, power, optical networks, memory, GPUs.
Nvidia is clearly the kingmaker. But the real money is the factory and the real estate where the GPUs get plugged in. That's a multi-decade annuity stream, backed by the strongest credits on Earth as tenants.
Where Nvidia's Own Capital Has Gone
The clearest signal is where Nvidia itself has put capital.
- Iren (IREN): Nvidia received a $2.1B right to invest as Iren pivots from mining to high-performance compute.
- CoreWeave (CRWV): $2B direct equity investment from Nvidia, with a $99B revenue backlog.
- Coherent (COHR): $2B manufacturing partnership in optical networking.
- Nebius (NBIS): spending $4B to lock down its footprint, with growth rates that don't look real.
- Applied Digital (APLD): 1GW power pipeline secured.
These are footprints, not press releases. Even Nvidia doesn't think "sell the GPU and walk away" is the whole trade.
The Biggest Mispricing
The largest mispricing right now is that the market still treats these names like crypto miners. That's a legacy lens. What's actually happening is closer to "AI-era landlords" — Big Tech pays the rent, and that rent is locked in for years.
That doesn't mean any of them is buyable. Average leverage in this group is unhealthy, and the first tap on the brakes from the macro side will hit these names before anywhere else in the capex chain.
So How Do You Act on It
My summary:
- These five are not core holds. They are tactical positions.
- "First two innings" doesn't mean "buy anything." It means you have to pick the survivors.
- Debt and operating cash flow are the first two filters.
Risk and Counterpoint
Anyone calling the game over is mistaking inning one for inning nine. Capex cycles run on multi-year lag between announcement and actual deployment, and we're sitting at the front of that lag. The catch — if Big Tech trims even slightly in the next guidance cycle, this is exactly the group that bleeds first. The conclusion stays the same: infrastructure as a theme is multi-decade, but which names you carry inside that theme is a different question entirely.
More in this Category
CPI at 4.2%, Third Straight Rise — So Why Did Markets Exhale?
CPI at 4.2%, Third Straight Rise — So Why Did Markets Exhale?
US year-over-year inflation rose for a third consecutive month to 4.2%. A year ago it had fallen as low as 2.3%, near the Fed's target. Yet stocks popped slightly. The subtle but crucial detail: it came in line with expectations.
What Snapchat Taught Me About the SpaceX IPO
What Snapchat Taught Me About the SpaceX IPO
Across the last 15 years, 30 major IPOs posted an average one-year drawdown of roughly 55%. CoreWeave, up 300% in three months, is on that same list. Ahead of the SpaceX IPO, here's what tends to wait behind a glamorous debut — told through Snapchat.
The Largest IPO Ever: SpaceX Goes Public — and the Question of Fast-Tracking Unprofitable Giants Into the Nasdaq
The Largest IPO Ever: SpaceX Goes Public — and the Question of Fast-Tracking Unprofitable Giants Into the Nasdaq
SpaceX went public in the largest IPO in history, swinging more than 10% in its first five minutes of trading, with Nasdaq index inclusion fast-tracked within 15 days. Here's what rushing unprofitable mega-caps into the index really means.
Next Posts
SpaceX IPO at $1.75 Trillion: What Every Investor Needs to Know Before June 12
SpaceX IPO at $1.75 Trillion: What Every Investor Needs to Know Before June 12
SpaceX is set to become the largest IPO in stock market history at a $1.75 trillion valuation. With Starlink's $4.42 billion operating income doubling year-over-year, here's what the 120x revenue multiple really means.
OpenAI vs Anthropic: Inside the Trillion-Dollar AI IPO Race
OpenAI vs Anthropic: Inside the Trillion-Dollar AI IPO Race
OpenAI at $840 billion and Anthropic surging from $380 billion toward $1 trillion in months — two AI companies, two radically different strategies, and neither is profitable yet.
5 Principles to Separate Price from Value in the 2026 IPO Boom
5 Principles to Separate Price from Value in the 2026 IPO Boom
From SpaceX's $1.75 trillion valuation to Inspire Brands' $20 billion franchise empire, 2026's mega-IPOs demand discipline. Here are the principles that separate investors from speculators.
Previous Posts
Palantir Isn't a SaaS Company — It's Infrastructure
Palantir Isn't a SaaS Company — It's Infrastructure
Classify Palantir as SaaS and the valuation looks insane next to a 19% YTD drop. But +85% revenue, a 145% Rule of 40, and 150% net retention say this isn't software — it's industrial-grade infrastructure.
Palantir Just Printed Record Numbers — Why Did the Market Yawn?
Palantir Just Printed Record Numbers — Why Did the Market Yawn?
Palantir posted $1.63B in quarterly revenue (+85% YoY), 60% operating margin, 53% net margin, US business +104%, and a 145% Rule of 40 — all in one quarter. The stock is still down ~19% on the year.
Palantir Is Down 19% This Year — Should You Have Sold?
Palantir Is Down 19% This Year — Should You Have Sold?
Palantir is down ~19% YTD even as it printed its best quarter ever. Here's how I think through whether to hold, sell, or add — and why the real problem is usually anchor, not analysis.