Image Source: Pexels

In the years leading up to the peak of the dot-com boom, the stock market stopped rewarding profits and started rewarding a clear connection to the future.

I remember it well.

Back then, companies tied to the internet surged, while companies with steady earnings but no internet angle were left behind.

The same pattern is back today, but this time it’s being driven by artificial intelligence.

And right now, the market is splitting even more sharply between the winners and everyone else.

If this chart gives you late-1990s déjà vu, you’re not imagining it.

As you can see, the strongest performers of 2025 weren’t the most profitable businesses. They were tech companies with no revenues, which were up more than 40% by November.

Right behind them sat the Magnificent Seven, up roughly 24% when this chart was created. By year’s end, the Mag 7 was up an average of 27.5%.

Unprofitable Nasdaq companies and unprofitable U.S. small and mid-caps also beat the broader market.

But once companies crossed into profitability, their returns dropped sharply.

Profitable Nasdaq stocks were barely up. Profitable small and mid-caps were almost flat. And the S&P 493 — everything in the S&P 500 outside the Mag 7 — was stuck well behind the leaders.

This represents a classic barbell market, where the middle has been crushed.

On one side, investors are paying aggressively to bet on early-stage tech tied to artificial intelligence, next-generation computing, energy infrastructure and other policy-accelerated themes.

For now, revenues don’t matter to these investors. What matters is exposure to technologies that were once considered long-term bets, but are now being built and adopted at full speed.

On the other side, capital is still crowding into the largest platforms in the world due to investors treating the Mag 7 as vital infrastructure for the future.

This chart also makes clear what the market doesn’t want.

It doesn’t want “solid” or “reasonably valued.” And it definitely doesn’t want companies whose main distinction is that they’re profitable today but could be strategically exposed tomorrow.

This barbell shape should look familiar to anyone who remembers the internet boom.

Back then, investors weren’t paying for earnings. They were paying for a claim on a future that felt inevitable.

Today, the story is similar.

No one wants to miss out on what could be the most transformative technologies in human history. Artificial intelligence, compute, energy infrastructure and next-generation technology continue to soak up capital because the market is paying for tomorrow, not today.

That’s why companies with no revenues are outperforming profitable ones.

And it’s why the market looks strong, even though only a few stocks are doing the heavy lifting.

HERE’S MY TAKE

Profitability wasn’t often rewarded in 2025.

And despite a short-term pop in energy stocks after the U.S. strike in Venezuela, I believe we’ll continue to see a barbell-shaped market in 2026.

Companies with little or no revenue will lead the market, while profitable businesses with no clear AI leverage will continue to be ignored.

And this divergence isn’t accidental. It’s the result of multiple super-cycles — AI, energy, compute and automation — colliding all at once.

What George Gilder and I call Convergence X.

Of course, it’s understandable if you’re concerned about this market’s similarities to the late 1990s. In 1999, many of the market leaders didn’t make money, and many never would.

But today, the biggest winners include massive companies with massive profits and control over critical infrastructure. And unlike the dot-com era, today’s surge is being reinforced by policy and a global demand for compute, power and automation.

That doesn’t mean every “no-revenue” stock will ultimately be a winner. Many won’t.

But it does explain why the market is behaving this way.

In periods like this, the market rewards positioning. And until that changes, the most dangerous place to be isn’t early or late.

It’s right in the middle.

More By This Author:

How AI Gets The Grid Through New Year’s Eve
Is Quantum Entering Its Infrastructure Phase?
JPMorgan Just Confirmed My Tokenization Thesis