Since October 2022, the S&P 500's surge has been overwhelmingly powered by the "Magnificent Seven." Those seven stocks account for roughly 75% of the market's gains, driven almost entirely by one force: the relentless promise of AI.
AI spending remains an arms race with no finish line. And after months of nonstop enthusiasm, investors finally tapped the brakes in November, pushing the Nasdaq down 5% as valuations reset.
Then Nvidia reported earnings.
Another blowout quarter:
- Revenue: $57B (vs. ~$54.9B expected)
- Profit: $1.30/share (vs. ~$1.26 expected)
- Data-center revenue: $51.2B (vs. ~$49B consensus)
The scale is staggering. Nvidia is now on pace to earn more annual profit than Intel and AMD will generate in revenue—combined. Its sales are nearly 10x what they were three years ago. With more than 90% share of the AI accelerator market and a full-stack moat (chips, networking, software, services), Nvidia has become both the heartbeat of the AI boom and a barometer for the entire stock market. It now represents ~8% of the S&P 500.
But here's the part that matters:
Growth is slowing. Not collapsing—just normalizing.
Markets expect full-year revenue to reach ~$208B, up 59% year over year. Incredible, but well off the prior year's 114% pace. Wall Street accepts that—for now. But when a company has only ever crushed expectations, simply "crushing less" eventually becomes its own burden.
Another risk: concentration. Amazon, Alphabet, and Meta alone account for more than 40% of Nvidia's sales. Add in the circular investments between OpenAI, Microsoft, CoreWeave, Oracle, AMD, and Nvidia, and you have a tight cluster of firms driving the entire AI economy.
If that small circle ever pauses spending, the ripple effects would be enormous.
So where are we today? Still in overdrive. Demand is so intense that the industry physically can't keep up. CoreWeave's CEO captured it perfectly last week:
"The totality of demand overwhelms the capacity of the market to deliver… It's a systemic problem the industry will have to deal with for the foreseeable future."
I remember the Dot-Com era. This is the opposite. Back then, telecom companies buried the world in fiber-optic cable long before the internet boom actually arrived. Today, demand is screaming higher and supply—chips, data centers, power—is the choke point. The pipes aren't big enough.
That's a good problem, but not a harmless one. You can have explosive growth and still face fragility if the system can't scale fast enough. As Black Swan author Nassim Taleb reminds us, fragile systems don't need a specific catalyst to break. Predicting the trigger is pointless. Recognizing fragility is what matters.
This isn't a bearish AI call. It's a reminder. Moats can be strong. Elevated valuations can be fragile. Smart investing requires seeing both. Don't try guessing the catalyst that might turn markets. Focus on fragility, then adjust accordingly.
Find a way to ride the AI wave—there's far more to this theme than Nvidia alone—but don't fly too close to the sun. The goal isn't to avoid innovation; it's to participate without betting your entire plan on a single story.
In other words, find your balance.
Diversification doesn't guarantee against loss, but it does protect against the only real catastrophe in investing: losing everything at once.
That mindset is durable wisdom—in markets and in life.