One Bad Capex Number Could Crack the Whole AI Trade, Schwab's Kevin Gordon Warns
Schwab's Kevin Gordon says a single mega-cap miss on AI capital spending could disrupt S&P 500 earnings. The index sits 1% off its high while the average stock is already 14% down from its peak. That gap is the whole story.
The AI trade is a one-legged stool. And Kevin Gordon just pointed at the saw.
Schwab's senior investment strategist said the quiet part loud. A single mega-cap miss on AI capital spending could wreck the S&P 500's earnings math. That's it. One company. One guidance number. One CFO saying, hey, we're rethinking the data center buildout.
That's what passes for diversification these days.
The Numbers Don't Lie
Here's the setup. The S&P 500 sits about 1% below its all-time high. Looks great.
Now look underneath. The average S&P 500 member is down roughly 14% from its peak, peak to trough, since early August. Gordon's point is that this isn't today's loss. It's the drawdown the typical stock has already eaten while the index camped out near record territory.
Why the gap? Market-cap weighting. The biggest names drag the whole index around by the nose. And right now the biggest names are the AI capex spenders, Microsoft, Meta, Alphabet, Amazon. That hyperscaler crowd has been writing nine-figure checks on GPUs and data centers every single day, and the market keeps rewarding them for it.
JUST IN: that's not a market. That's a bet.
The Other Side of the Trade
Look, the bulls have a case, and it isn't stupid.
Hyperscaler capex is real money hitting real earnings. The cloud revenue it's building is real too. Amazon, Microsoft and Google aren't lighting cash on fire for vibes. They're expanding capacity because customers keep buying. Nvidia's order book isn't a rumor.
And that 14% stat cuts both ways. If the average stock is already 14% off its high while the index is 1% off, that's arguably rotation, not a bubble popping. Money left small caps and mid caps and parked in the megacaps. If capex holds up, that concentration keeps printing.
So what could go wrong?
A pause. Not a cut, a pause. If one of these giants says it's moderating spend in the back half, the earnings revision cycle turns on a dime. Analysts pull estimates forward, multiples compress, and the index-level EPS math everyone's been banking on gets a haircut nobody planned for.
And because the index is so top-heavy, the pain doesn't spread across 500 names. It lands on five.
My Verdict
Gordon's right, and the market isn't pricing it. Concentration is a feature until it's a bug.
Everyone knows the S&P 500 is a megacap fund wearing an index costume. What they're ignoring is how thin the earnings support really is underneath. One miss doesn't have to be a catastrophe to hurt. It just has to be one.
Mega-cap capex is a confidence game. The moment confidence cracks, the squeeze goes the other way, and it goes fast.
This rally is borrowing from tomorrow's earnings. That works right up until somebody decides not to pay.
What to Watch
Keep your eyes on the quarterly calls. Not the headline revenue numbers. The capex guidance lines. Watch Microsoft, Amazon, Alphabet and Meta specifically. If any of them drops language like efficiency, prioritization or moderating investment, that's your signal.
And watch breadth. If the average stock keeps bleeding while the index holds, the narrowness is getting worse, not better. That's the moment a single bad print flips from a headline into a haircut for everybody.
Traders are watching closely. You should be too.
