Currently, the market is experiencing a persistent, low-volatility upward grind, with the Nasdaq noticeably outpacing both the S&P 500 and the Dow Jones. While mainstream analysis attributes this entirely to AI enthusiasm and earnings growth, looking closely at modern market plumbing reveals a structural mechanic at play: valuation overlap.

Due to a combination of regulatory reporting thresholds and modern index construction, the Nasdaq is experiencing a partial circularity in its valuation. Here is a breakdown of the mechanics driving this divergence, the algorithmic feedback loop amplifying it, and how we are positioning for it.

Investing in the ‘AI Mafia’: Why the Bricks Always Beat the Brains

Most retail investors are currently obsessed with “The Brains”—the semiconductor companies like Nvidia or AMD. They see the sky-high margins and assume that because AI needs chips, chips are the best way to play the trend.

But sophisticated investors know that a chip without power is just an expensive piece of silicon. The real leverage in the AI revolution isn’t found in a cleanroom in Taiwan; it’s found in the dirt, the diesel, and the copper. It’s found with the “AI Mafia”: Caterpillar (CAT) and EMCOR (EME).

I’ve always believed Google (Alphabet) would ultimately win the AI race — not because they’d have the flashiest model every week, but because of deep structural advantages that matter in a capital-intensive, multi-year war.

The market is now reflecting exactly that. Here’s the full thesis that has held up extremely well.

The Core Thesis (written more than a year ago)

1. They own their own compute (TPUs + custom silicon)
2. They don’t need to raise money — massive cash flow + ad revenue moat
3. They already have a huge enterprise business
4. No painful revenue-share drag with partners
5. Raw model quality matters less than integration, distribution, and scale

This was never a “best model on the leaderboard wins” game. It’s a game of attrition, economics, and distribution.