Are AI Stocks a Good Investment? Jim Bianco: Two Stock Markets

Key Takeaways

The S&P 500 has split into two markets that no longer move together. Bianco says AI and AI-related stocks have decoupled from the rest of the index, with the correlation between them falling below zero, something that was not true even a couple of years ago.

A handful of AI names now rival the entire rest of the index. By his count, roughly 41 AI-related stocks make up about 45% of the S&P 500, down from a 49% peak in early June, leaving the other 459 stocks at 55%.

Which market to own depends on risk tolerance, and the quiet winner may surprise you. Bianco frames AI stocks as higher expected return but with the risk of losing a third or half along the way, while non-AI stocks are the "sleep at night" choice, and he notes they have outperformed AI year to date.

Every technology ends in a bubble, but Bianco doesn't think AI has peaked yet. He argues the classic bubble signal, massive overbuilding of unused capacity, is not present, because the market is still "compute constrained."

Key Moments

00:00 - AI is the biggest technology revolution ever Bianco ranks AI above the internet, the PC, even the railroads.

00:27 - There are now two stock markets The core thesis: AI and everything else have become separate markets.

02:15 - AI stocks vs. everything else The Monday when the index fell but the 459 non-AI names rose.

03:09 - Every technology ends in a bubble Why AI will end in a bubble, and the Global Crossing lesson.

04:14 - Why AI hasn't peaked yet The compute-constrained argument against calling a top now.

05:35 - What could end the AI boom? The efficiency breakthrough that could turn scarcity into glut overnight.

Are AI Stocks a Good Investment? Why Jim Bianco Says There Are Now Two Stock Markets

Concentration risk is what happens when a small number of stocks come to dominate an index, so that owning the "market" really means owning those few names. In a short conversation with Wealthion in early August 2026, strategist Jim Bianco argued that this has gone so far that the S&P 500 now behaves like two separate markets, AI and everything else, that increasingly move in opposite directions. Whether AI stocks are a good investment, he says, is really a question about how much risk an investor can stomach.

Are AI stocks a good investment right now?

Bianco's answer is that it depends entirely on risk tolerance, because the two markets offer very different trade-offs. If an investor wants the higher expected return, AI stocks are the place, but "you might lose half your money before you get there, or you might lose a third of your money before you get there," and he notes the AI names were already down about 15% from their June high. For a "more predictable, can sleep at night kind of life," he points to the non-AI stocks, and adds a detail many investors miss: those non-AI stocks have "outperformed the AI stocks year to date," even though AI has produced far bigger gains over the full year. As always on Wealthion, that is his framework for thinking about the choice, not a recommendation for any individual portfolio.

What does Bianco mean by "two stock markets"?

He means the two groups have genuinely decoupled. Bianco explains that if you build indexes of the S&P 500 excluding AI and of AI stocks only, the correlation between them has fallen below zero. His vivid example is a recent Monday when the headline index dropped and the financial press blamed worries about the economy, yet the 459 non-AI stocks rose almost 1%. In his words, "it's become two different stock markets right now," and "they're not even correlating with each other." A couple of years ago, he notes, you could argue there were two markets but they still rose and fell together; that is no longer true.

How concentrated is the S&P 500 in AI?

Extremely, by Bianco's numbers. Counting the Magnificent Seven, semiconductors, equipment makers, some pure plays, and the energy names that feed AI, he arrives at roughly 41 stocks, a figure he attributes to strategist Michael Cembalest at JP Morgan, that make up about 45% of the S&P 500. That share peaked near 49% in early June, which means the AI complex has grown to almost the same size as the other 459 companies in the index combined. For an investor in a broad index fund, that is the crux: nearly half the exposure now rides on a single theme.

Is AI a bubble, and has it peaked?

Bianco is direct that a bubble is coming, but does not think it has arrived. "Every technology ends in a bubble and this one will be no different," he says, while arguing the peak is probably still a couple of years off. His reasoning is that bubbles typically end with massive overbuilding of capacity that never gets used, the way Global Crossing laid far more fiber optic cable during the dot-com era than the world could use. AI, by contrast, is still "compute constrained": he estimates only about 2% of knowledge workers use AI at real productive capacity, and says tokens remain too expensive for heavy use, so there is no overcapacity yet and many more users still to come. Other Wealthion guests have pushed harder on the near-term risk, including why the smart money is selling AI stocks and signs the AI spending boom is cracking.

What could end the AI boom?

The risk Bianco flags is an efficiency breakthrough that flips scarcity into glut overnight. If a new advance, from China or the US, lets existing data centers and chips do far more work, the market could wake up to discover it has massive overcapacity rather than a shortage. He points to a newly released Chinese model that he said rivals the leading US models as exactly the kind of development that makes investors nervous about that scenario. For now he does not think the industry is there, but he is clear about the ride ahead: "before you get to that bubble peak, it's always going to be a thrill ride." This is the same concentration dynamic Bianco has described before in how AI is holding up a broken market, and a more bullish counterpoint runs through why some see AI as the biggest opportunity yet.

What Investors Should Watch

• The AI to non-AI correlation: Bianco's core signal that the two markets have decoupled, and whether it stays below zero.
• Index concentration: the roughly 45% AI share of the S&P 500, and whether it climbs back toward the June peak near 49%.
• Compute-constraint signals: token costs and the share of knowledge workers actually using AI, which he says show there is no overcapacity yet.
• Efficiency breakthroughs: new models that could suddenly turn "compute constrained" into overbuilt, the trigger he watches for the eventual bubble.

FAQ

Are AI stocks a good investment right now? Jim Bianco says it depends on risk tolerance. AI stocks offer higher expected returns but the risk of losing a third or half along the way, while non-AI stocks are the more predictable choice and have actually outperformed AI year to date. This is his framework, not advice.

What does Bianco mean by two stock markets? He means AI stocks and non-AI stocks have decoupled, with the correlation between them falling below zero. On a recent day the headline index fell while the 459 non-AI stocks rose almost 1%, which he says shows they are now effectively two separate markets.

How concentrated is the S&P 500 in AI? By Bianco's count, roughly 41 AI-related stocks make up about 45% of the S&P 500, down from a peak near 49% in early June, leaving the other 459 companies at 55%.

Is AI a bubble yet? Bianco expects AI to end in a bubble like every major technology, but does not think it has peaked. He argues the market is still compute constrained rather than overbuilt, so the peak is likely a couple of years away.

Which expert and interview does this article reference? This article draws on Wealthion's interview with strategist Jim Bianco, president of Bianco Research: "There Are Now Two Stock Markets."