Here’s a plot twist nobody saw coming: the companies that were supposed to dominate the AI revolution are getting left in the dust by the companies that actually *make* the chips. Yeah, you read that right.
Q2 was absolutely bonkers for the stock market. We’re talking best quarter since 2020, semiconductors had their best quarter *ever*, and everyone was basically printing money. But if you looked closer—like, actually paid attention instead of just scrolling past the headlines—you’d notice something weird happening. While chipmakers were absolutely crushing it, the Magnificent 7 (Microsoft, Meta, Google, Amazon, Apple, Tesla, and Nvidia) were… well, let’s call it “underperforming.”
The divergence is wild. Investors have basically decided they’d rather own the companies making the shovels than the companies digging for gold. And honestly? You can kind of see their point.
The hyperscalers—Meta, Microsoft, Google, Amazon—have been throwing absolutely *insane* amounts of money at AI infrastructure. We’re talking hundreds of billions in capital expenditure. The problem? Nobody’s entirely sure when (or if) that money is going to turn into actual profits. It’s like watching someone spend their entire paycheck on lottery tickets and hoping one of them hits big.
Meanwhile, chipmakers are sitting pretty. They’re selling the actual hardware that powers all this AI stuff, and they’re not taking on nearly as much risk. It’s the classic “picks and shovels” play, and Wall Street is eating it up.
But here’s where it gets interesting: the big brains on Wall Street think this divergence is totally unsustainable. Mike Wilson at Morgan Stanley basically said “yeah, this can’t keep going,” and he’s expecting semiconductors to correct while hyperscalers bounce back. Goldman Sachs is making similar noises—they think hyperscaler valuations are now attractive enough to justify another look.
The theory goes something like this: hyperscalers will eventually figure out how to monetize all this AI spending. They’ll generate revenue and earnings that justify the capex binge, and suddenly they’ll look like geniuses instead of reckless spenders. JPMorgan’s even laid out a scenario where this happens and everyone’s happy.
Sounds great, right? Except there’s a tiny problem: execution. It’s one thing to say “yeah, this will work out eventually.” It’s another thing to actually *make it work*.
The real test is coming in a couple of weeks when the Mag 7 report earnings again. That’s when we’ll find out if all this AI spending is actually paying off, or if investors are going to keep punishing these companies for their aggressive capex strategies.
So grab your popcorn. The next earnings season might just tell us whether the Magnificent 7 can close the gap, or if they’re destined to stay lagging.
