Wall Street has spent 2026 pouring money into almost anything tied to artificial intelligence. One of the biggest banks now says part of that trade looks unhealthy.
JPMorgan technical strategist Jason Hunter told clients the AI rally is starting to resemble the market in the months before the dot-com bubble burst in 2000.
His concern is not that AI is fake. It is that money is flowing into one part of the AI trade while pulling out of another, and that kind of split has ended badly before.
For anyone holding a broad US stock index fund, like an S&P 500 or Nasdaq fund, the warning is worth understanding.
These funds automatically put more money into the largest companies, so the Magnificent Seven make up an outsized share even for investors who never chose those stocks individually.
Here is what the strategist flagged, why it matters, and a few practical steps to think through before the fall.
Why JPMorgan sees a 1999 warning sign in AI stocks
Hunter’s argument centers on a divergence, meaning two related groups of stocks moving in opposite directions.
On one side are the AI hardware makers.
Chip stocks have climbed for most of 2026, and the Philadelphia Semiconductor Index has jumped 87% this year.
It just logged its best-ever quarter, according to a JPMorgan note reported by Business Insider.
On the other side sit the hyperscalers, the tech giants spending billions to build AI.
Meta is down about 5% year to date, while Microsoft has fallen 18% to 20% and posted its worst monthly decline since 2000, according to TradingView.
Not every hyperscaler has been hit the same way, but the pattern is enough to worry JPMorgan.
More AI Stocks:
- Oppenheimer has a blunt Nvidia stock message ahead of earnings
- Morgan Stanley resets Nvidia stock forecast ahead of earnings
- Cathie Wood buys $16.2 million of popular semiconductor stock
In a healthy rally, the companies buying the chips and the companies selling the chips tend to rise together.
Hunter compared this to 1999, when communications equipment suppliers surged while companies making heavy capital investments crashed from peak valuations.
Dotcom stocks collapsed in early 2000, less than a year after that divergence appeared.
Courtney Olujobi, principal at Moon Pursuit Capital, sees the same concentration problem in how AI money is being spent.
In an interview I had with him, he said, “if we look at AI investments for the first, or call it second quarter of 2026, about 87 and a half percent of all venture dollars has gone into AI companies, and 43% of that has gone into two major names, OpenAI and Anthropic. What that is indicative of is concentration.”
The spending gap spooking investor nerves
The reason for the gap comes down to timing.
Chipmakers get paid immediately.
The companies buying those chips have to turn them into profit later, and that part is far from proven.
The numbers behind the buildout are large.
Meta (META), Microsoft (MSFT), Amazon (AMZN), and Alphabet (GOOG) are on track for combined capital expenditures of about $725 billion in 2026.
This is based on the companies’ own earnings disclosures.
What the AI spending numbers mean for investors
- Hardware sellers win first. Chipmakers collect revenue the moment a tech giant places an order.
- Big spenders carry the risk. Hyperscalers must convert that hardware into software and cloud profits, which takes time.
- Proof is now the demand. Investors have moved past the excitement stage and want to see the spending show up in earnings.
Olujobi frames the same idea around what survives once the hype fades.
“You have to be able to separate narrative story from price. Price is subject to customer sentiment. But at the end of the day, the market leader who has the clients is gonna be the one that remains.”
How concentrated the market has become
The bigger danger is in how much of the market now rides on a handful of names.
The Magnificent Seven, meaning Nvidia (NVDA), Microsoft, Apple (AAPL), Alphabet, Amazon, Meta, and Tesla (TSLA), make up roughly a third of the S&P 500’s market value.
Add the wider group of AI-linked tech and communications stocks, and that share climbs higher, with overall index concentration near multi-decade highs.
Related: Cathie Wood sells $5.5 million of surging tech stock
That matters for investors because of how index funds work.
When you buy an S&P 500 or Nasdaq fund, more of your money automatically goes into the largest stocks.
So if these few names drop sharply, the whole fund drops with them, even if hundreds of other companies are doing fine.
This played out in June 2026, when the Magnificent Seven lost roughly $2 trillion in market value over a few weeks and dragged the broader index down even as other stocks traded positively.
What tech-heavy investors can do now
Hunter’s warning is a caution, not a prediction of a crash.
Timing market peaks is close to impossible, and rallies can run longer than expected.
Still, the setup gives long-term investors a reason to check their exposure.
Before making any changes, make sure your basics are covered, including an emergency fund of three to six months of expenses.
If your portfolio leans heavily on tech, these steps are worth considering:
Steps to review before autumn
- Check how much you own in a few names. Many investors hold far more AI exposure than they think through index funds.
- Rebalance crowded positions. Trimming back to your target allocations takes profits without trying to call a peak.
- Broaden your holdings. Rotation out of AI names in June and July has partly moved into areas seen as more defensive. Non-Magnificent Seven companies still offer solid earnings growth.
- Watch the hyperscalers. If the big spenders can stabilize this summer, the risk of a sharp autumn pullback drops.
Olujobi points to history for why infrastructure booms can hurt even when the technology wins.
“I would liken this to railroads. The companies that built railroads, a large percentage went bankrupt. But the companies that use those railroads, they’ve gone on. People have used these things for hundreds of years, but the initial companies built the main infrastructure, and then market leaders emerged. There wasn’t several, there was a few.”
The bottom line for investors
Hunter is not saying AI is a scam or that these companies will disappear.
He is saying the market is treating one part of the AI trade very differently from another, and that pattern showed up right before the last tech crash.
The practical takeaway is straightforward. Know what you own, avoid betting everything on a few giant stocks, and watch whether the biggest AI spenders can prove the buildout is paying off.
If they can, the fear fades. If they cannot, the concentration that lifted these indexes on the way up could work just as hard on the way down.
Related: Nvidia just made a move Wall Street wasn’t ready for
