Thematic ETFs Are Popular Again. Yikes.

Investors are flocking to theme-based exchange-traded funds again. They’ll probably wish they hadn’t.

I last wrote about thematic ETFs back in January 2025. At the time, I’d noted these ETFs had seen poor dollar-weighted returns: The average dollar lost more than 7% per year over the three years ended Nov. 30, 2024—considerably worse than the ETFs’ 1% annual aggregate loss over that span.

Why the poor outcome? Investors binged and purged, buying thematic ETFs with abandon in 2020 and 2021—the ETFs hauled in $95 billion those years—and then fleeing from 2022 through late 2024. Those redemptions might have seemed prescient at first, as the average thematic ETF lost 36% in 2022. But they proved costly when the ETFs rallied hard the following years.

They’re Back

Fast forward to now: Thematic ETFs appear to have won back investors, gathering $76 billion in net inflows since 2024, with artificial intelligence, energy transition (think: smart-grid), and security (think: defense tech) leading the way.

Will things turn out differently this time? I doubt it. Consider the most popular ETF theme these days: AI, or in the parlance of our thematic-classification taxonomy, “Artificial Intelligence & Big Data.” Here are the rolling 12-month average returns of ETFs assigned to that theme.

These ETFs have been red-hot, routinely generating 20%-plus returns over recent 12-month periods. Investors have noticed and have piled in.

Ding, Ding, Ding!

That more or less fits the pattern we’ve seen more broadly for thematic ETFs: A narrative takes hold in ways everyday people can relate to and intuit, in this case AI’s potential to transform life as we know it; investors seek confirmation of the story, which they find in standout returns; and then they give chase.

Lest you doubt how often investors have chased performance in this fashion, here’s a plot that compares thematic ETFs’ rolling 12-month returns and flows over the decade ended Aug. 31, 2026.

Most of the time, inflows coincided with recent gains and outflows with losses. For instance, thematic ETFs raked in $65 billion in net new money over the year ended Jan. 31, 2021, during which the average thematic ETF rose 58%. Conversely, investors yanked $12 billion over the 12 months ended Feb. 28, 2023, when the average ETF lost around 21%.

That wouldn’t have been a problem if the performance trend had continued, but too often it reversed, wrong-footing investors. For example, the average thematic ETF lost 13% in the 12 months ended Jan. 31, 2022, right on the heels of the aforementioned $65 billion inflows. Similarly, thematic ETFs gained 14%, on average, in the year ended Feb. 29, 2024, following the $12 billion outflow.

You can see the relationship between flows and subsequent returns more clearly in this plot, which compares rolling 12-month flows against the average thematic ETF’s return over the subsequent one-year period. The upper-right (that is, inflows followed by gains) and bottom-left (that is, outflows followed by losses) regions are “good,” while the upper-left (that is, outflows followed by gains) and bottom-right (that is, inflows followed by losses) are “bad.”

There was a whole lot more “bad” than “good,” and that’s taken a toll on dollar-weighted returns. While my previous article focused on the three years ended Nov. 30, 2024, I’ve expanded the analysis to cover all trailing periods—ranging from one year to 10 years—ended Aug. 31, 2026.

The gap has been narrower over the shorter trailing periods primarily because the themes that have gotten the heaviest flows have kept chugging along. But as you extend the measurement period, the gap between the return of the average dollar and the ETFs’ aggregate total return widens dramatically. This reflects poorly timed purchases and sales in prior years as well as the effect of compounding those errors.

Investor Takeaways

Reject Good Stories

Thematic ETFs are predicated on the idea that you can tap into a burgeoning trend and ride a wave of popularity and adoption to big gains. A rule-of-thumb I use is that by the time I come across or make sense of something, it’s already been discovered and priced in by legions of other market participants who boast greater faculties and deeper resources than me.

Distrust Your Intuition

Sure, it sounds defeatist to say you shouldn’t count on your ability to make sense of an investment and play out its future. Heck, AI is already big, and it’s going to get bigger; if that’s the conclusion you’ve reached, I’m not here to tell you you’re wrong. But that isn’t a sturdy enough reason to buy something. Why? You need to be able to distinguish between the story you’ve constructed in your own mind and the story the market has effectively incorporated into the security’s price. When they differ, you can run into big problems.

Sideline Emotion

We seek patterns and extrapolate because it can instill a sense of calm and order in a world that otherwise might feel chaotic and random. Theme-based ETFs can seem to dispel complexity and make investing seem almost linear, where if something “gets big” or “breaks through,” you win. I’m not here to say you should reject your emotions, but they shouldn’t drive the decision. If buying feels exciting or selling brings relief, I’d revisit the decision.

Keep Perspective

It might not seem like it, but if you invest in a broadly diversified stock portfolio, then you probably have exposure to many of the leading themes. No, it’s not as fun and won’t win you any bragging rights. And, yes, it’s watered down compared with an allocation to an ETF that is focused like a laser on that theme. But let’s keep things in perspective: The average thematic ETF returned 10.5% annually over the decade ended Aug. 31, 2026, which was 5 percentage points per year less than the S&P 500’s gain over that span.

Find Another Way?

Full disclosure: Morningstar licenses thematic indexes that various ETFs track. So it would be pretty hypocritical for me to lecture anyone about launching theme-based ETFs. (I say that even as someone who is not involved in those commercial arrangements.) Nevertheless, as an unreconstructed believer in the idea that fund companies win over the long term when their investors succeed, I guess I’d hope that thematic ETFs’ poor dollar-weighted results would have fund companies doing at least a little bit of introspection? Maybe to ask whether we really need another drone-, photonics-, or “space-industry income blast”-themed ETF?

Switched On

Here are other things I’m writing, reading, and watching:

• Amy Arnott on the hocus-pocus at YieldMax Ultra Option Income Strategy
• Christine Benz with a needed reality check on bonds’ role in your portfolio
• Bryan Armour on untangling the semiliquid fund fee knot
• Michael Santoli’s “Market Memo” newsletter
• A thread on trends in ETF share splits, just because
• Daisy chain: A proposed ETF that would invest in swaps tied to futures linked to private firms
• What’s in your special purpose vehicle? The SEC wants to know.
• The Hugging Face hack was bigger and creepier than we thought
• “The Older I Get, the Less I Seem to Know” by Jonathan Eig
• “Aliens” gets the Rewatchables treatment

Don’t Be a Stranger

I love hearing from you. Have some feedback? An angle for an article? Email me at jeffrey.ptak@morningstar.com. If you’re so inclined, you can also follow me on Twitter/X at @syouth1, and I do some odds-and-ends writing on a Substack called Basis Pointing.