Did the AI Bubble Burst?

Is it possible to identify a market bubble before it bursts? On this episode of The Active Share, Hugo welcomes James Mackintosh, senior market columnist at The Wall Street Journal, for a deep dive into the current state of the U.S. equity market. Together, they explore the impact of artificial intelligence (AI) and U.S. exceptionalism in today’s global economy and explore strategies for identifying compelling investment opportunities and navigating uncertainty in a rapidly evolving landscape.

Comments are edited excerpts from our podcast, which you can listen to in full below.

Is the U.S. equity market in a bubble?

James Mackintosh: The problem with bubbles is that you can never really know you were in one until after it’s burst, but there are some clear signs of bubbly behavior around AI-related or Big Tech stocks.

However, this is different from the last major bubble. And while everyone loves to compare today to the dot-com era, the dynamics aren’t quite the same.

Back in 2000, there was a clear bubble in dot-com companies. Businesses that had just been set up were shooting to enormous valuations; some companies would add “dot-com” to their name and suddenly double in price the next day. The bubbliness was obvious, but many companies weren’t making any profit at all.

This time around, we’re looking at a bunch of established, highly profitable companies. They may be pouring money into AI, but they can afford it. And if it turns out that AI can’t deliver the gains everyone is hoping for, Microsoft, for example, can still spend $80 billion on data centers next year without going bust.

I’ll also say it’s hard to justify some of these valuations based on growth. And the way the market’s been behaving, it seems like a lot of investors are buying in without much hesitation. That kind of one-sided weight is often a bad sign.

How sure are you about the near-term impact of AI on valuation and earnings?

James: It’s clear that AI could bring about huge change, but it’s not obvious where those earnings are going to come from. Take NVIDIA—it appears to be an easy AI play because it’s supplying chips, but there’s a lot of competition coming, which should give investors pause.

The bigger question is about who the winners will potentially be. Everyone seems to think they know, but it’s not that simple. Just having a big AI model doesn’t mean you’ll secure monopoly-like profits. These models are also increasingly easy to replicate. Meta, for example, has made its methodology public, and others can catch up quickly.

If these models become almost like public property, where anyone can access them if they supply the compute power, the potential for monopoly-like profits diminishes. It’s a highly competitive market for compute power, and it’s not clear that a company’s size alone will give it an edge.

And here’s a lot of uncertainty about how profitable AI will be, even if it works. Generative AI, for example, still has plenty of things about it that don’t work well.

The problem with bubbles is that you can never really know you were in one until after it’s burst.

Is there any data or patterns that suggest we're seeing signs of excess or overconfidence?

James: At the end of 2024, indicators of euphoria were very high across the board. Whether it was surveys, investor positioning, or cash holdings by mutual funds, it all pointed to investors being fully committed to AI.

But there were other themes at play such as broader market performance, the Trump trade, and U.S. exceptionalism.

Overall, sentiment was excessive, which doesn’t necessarily mean we’re in a bubble—though if there is one, it’s still in its early stages. We’re not in a full-blown March 2000 situation, but we are seeing a lot of bubble-like behavior. There are crowded trades, and contrarians are either being silenced or flipping to the bullish side.

Does U.S. exceptionalism drive earnings growth that powers the market higher? Or vice versa?

James: It’s a feedback loop; both things happen at once. The fact that the U.S. economy has been so strong and U.S. productivity has looked so good is not just because of AI. These two factors also contribute to the U.S. exceptionalism story as well as the U.S. market doing phenomenally well.

Over the past couple of years, the U.S. economy decoupled from most of the rest of the world and, not surprisingly, so did U.S. share prices.

Some argue that the key drivers of U.S. exceptionalism (such as access to capital, a deep talent pool, and the sheer size of the economy) gives the United States a lasting advantage. Do you think those factors explain the country’s strong earnings growth compared to places like Europe or China? And how sustainable is that advantage?

James: It’s clear these U.S. exceptionalism factors are strong, so I wouldn’t necessarily draw a direct comparison to China. If there are opportunities in Chinese stocks right now, it’s likely because they’re cheaper than they should be. The Chinese economy is struggling—not just with poor earnings, but with low productivity as well.

The United States, on the other hand, has been doing well, but people act like it’s a permanent trend. I’m old enough to remember when the United States was considered behind, especially in tech. Back then, Americans and their Motorola phones were laughed at by Europeans with their Nokia phones. The United States was seen as falling behind, and people feared it was over for American innovation.

Now, the United States still has a huge innovation machine, particularly in Silicon Valley, where capital, talent, and ideas come together. The United States does this better than anywhere else, but that doesn’t mean it has a monopoly. There’s always a risk that dominant companies become complacent and fall behind on innovation.

History shows that big companies struggle to stay innovative once they become huge. The past 10 to 15 years in the United States have been different, with large companies managing to remain innovative.

Maybe this time is different, but history suggests that eventually, they’ll get lazy and start extracting too much profit, which could hurt them in the long run.

The United States still has a huge innovation machine, particularly in Silicon Valley, where capital, talent, and ideas come together.

With President Trump back in office, are we in a period of stability or change?

James: Honestly, I have no idea, and I don’t think anyone else does either. In his first term, Trump initiated some competition inquiries, but Biden was even more aggressive on that front.

And some in Trump’s camp have suggested they won’t be friendly to big companies, especially Big Tech, but I don’t know how that’ll play out.

I wouldn’t mind if Trump pushed for more competition in the United States, as that could benefit the economy and consumers in the long run. And he does seem to admire successful people, and many of these big companies are run by wealthy, successful leaders. That could swing him in their favor.

How do you think this could influence investing?

James: We’re in a period of uncertainty. Everyone I’ve spoken to—fund managers, strategists, investors—has said they don’t know what’s going to happen. The first Trump term showed that a lot of expected changes didn’t materialize.

Uncertainty matters just as much as change does. Investing in uncertain times means considering worst-case scenarios and dialing back your confidence. Right now, with a lot of volatility expected, especially with the new administration, I’d be cautious about potentially making large bets.

Diversification is still looked at as a great strategy. Some assets may offer cheap gains, others are overpriced, and many could see volatility in the short term.

Do you have any contrarian thoughts on the markets?

James: Right now, Europe is the most obvious contrarian bet. It’s cheap, hated, and facing challenges, but it could turn around. The bar is low, and if things improve even slightly, the market could react strongly. And most people are underweight Europe, so it’s not an extreme position.

For a more extreme contrarian take, I’d say China. But I’m not a contrarian for the sake of it. My concern with China is that too much depends on whether it launches major stimulus, of which I’m skeptical. It’s dealing with excesses from the housing bubble, and that could take years to work through.

Europe is the most obvious contrarian bet.

Can you talk about your journalistic process and how it’s evolved over time?

James: Being able to write to a word limit and meet deadlines is a basic requirement. But that alone doesn’t make a journalist. The real skill is spotting a story and explaining it simply and clearly. And that’s true whether you’re covering a murder inquiry or the markets.

As a columnist, my job is to spot what’s interesting and present it in a simplified yet accurate way. I think my task is harder than covering breaking news, but the fundamental principles of journalism are the same: spot the story, explain it clearly, and make sure the details are correct.

Has your job gotten easier with the internet, connectivity, and availability of data?

James: Being a journalist is both easier and harder today. The internet has made it much easier to find material, as I can quickly access earnings call transcripts and data releases and use AI to analyze them.

However, this also means the bar has been raised. In the past, just passing on information was valuable, but now people can get that information from countless sources.

So, to stand out, you need to dig out exclusives, offer better analysis, or present it in a more engaging way. That’s the real skill for journalists and columnists today.

Are people more or less inclined to speak to journalists these days?

James: I think people are still inclined to talk, especially when you're with trusted institutions like The Wall Street Journal or Financial Times. Most people understand journalists are trying to provide information, but many also want them to write about their perspective. It’s important to be aware of their point of view, but overall, many people are willing to speak.

Do you think trusted platforms like The Wall Street Journal will still dominate in 10 or 15 years, alongside individual creators on Substack? Or is there room for something in between?

James: I think The Wall Street Journal will absolutely thrive; it’s one of the most trusted U.S. news sources, so I have no doubt about its staying power.

Substack’s business model is interesting, but its biggest challenge is scale. With so many individual voices, even in niche areas, it can quickly become overwhelming.

I know some excellent writers there, some of whom are good friends, but even I can’t subscribe to all of them or keep up with everything they publish.

While a few big names will likely retain their audience, most writers struggle to stand out, and readers face the difficult balance of wanting variety without the cost and overload of subscribing to dozens of newsletters.

What we really need is a filtering mechanism, something that curates the best columns, highlights the standout pieces, and helps me avoid the ones that don’t hit the mark. That’s essentially what a traditional newspaper or website does.

I suspect Substack will evolve, and maybe move toward a more curated or bundled subscription model. But they haven’t quite figured that out yet.

To stand out, you need to dig out exclusives, offer better analysis, or present it in a more engaging way.

Can technology solve the trust and curation challenge in news media?

James: Human editors can curate content with an inherent level of trust in their judgment, but AI might one day take on that role.

However, AI isn’t fully capable of handling the complexities of editing and selecting relevant topics yet. The recent issues with Apple News shifting to AI show the challenges in replacing human curation.

Would you still become a journalist knowing what you know now?

James: When I graduated, I had the choice to go into early AI work, which I was interested in, but I’m glad I didn’t. In the early ‘90s, AI was a bubble, and I would’ve spent years banging my head against a wall because at the time, AI was about solving language with grammar, psychology, and linguistics.+

Looking back, I’m happy I went into journalism. It's been a fabulous career with great variety, and while I miss the times when journalism was a better place to work, I’ve had a wonderful time in the field.