Thematic ETFs:
buying the story after the top.

READ10 min · UPDATED
Reviewed against primary sources cited at the bottom of this page.

AI, robotics, cannabis, clean energy, the metaverse, blockchain. There is an ETF for every story, and it almost always launches after the story has already peaked. The pattern is consistent enough to be a law: issuers package a hot narrative, you buy it near the top, and over the next five years it underperforms the broad market by roughly a third on a cumulative risk-adjusted basis (about −6% per year). The theme might be real. The entry price isn't.

Reading time: ~9 minutes · Parent: How Financial Products Are Marketed · See also: Index Funds, Passive Investing.

THE CORE IDEA

A thematic ETF is a bet on a narrative, sold to you by the people who profit from the bet. Issuers don't launch funds for themes nobody is talking about, there is no demand for a fund about a boring sector. They launch funds for themes that are already everywhere, which means the underlying stocks are already expensive. You are not buying the future of AI or clean energy; you are buying a basket of stocks that the rest of the market has already bid up, at the exact moment attention, and price, are near their peak. The academic record on what happens next is unusually clear, and unusually bleak.

Supply follows hype, not opportunity

Fund launches are a contrarian indicator. Issuers file for new thematic products when media coverage, search interest, and recent returns are all climbing, because that is when the funds will sell. The launch is timed for maximum attention, not maximum expected return. By the time a theme is popular enough to justify its own ETF, the easy gains are behind it.

The cycle is visible in the launch data. Following a record-breaking 2021, when markets peaked and roughly 670 new thematic funds were introduced globally, the category cratered. Global thematic assets had surged more than threefold to a peak near $892 billion before tumbling.[2] New launches peak with the hype and collapse with it, which is precisely backward from when a disciplined investor would want to buy.

THE TELL

If a brand-new ETF exists for the theme, the theme is already crowded. The opportunity to invest in a narrative arrives only after prices have neared a peak and valuations are stretched, because that is the only environment in which the product gets built and marketed to you.

The 30% finding

The definitive study is Ben-David, Franzoni, Kim & Moussawi, "Competition for Attention in the ETF Space," published in the Review of Financial Studies in 2023.[1] Examining ETFs from 2000 to 2019, they separate broad-based index funds from "specialized" ETFs, the narrow, single-theme and sector products that thematic funds belong to, and track what happens after launch.

THE HEADLINE RESULT

Over their first five years, specialized ETFs lose about 30% on a risk-adjusted basis ×DON'T TRUST, VERIFYClaim: Over their first five years, specialized (thematic/niche) ETFs lose about 30% in risk-adjusted terms versus the broad market.Verify at: NBER Working Paper 28369 ↗Direct quote from the paper's abstract and body: "over their first five years, specialized ETFs lose about 30% in risk-adjusted terms." Confirm the magnitude and the 2000–2019 sample.. That is the cumulative four-factor alpha, not a price drawdown, it is how badly the funds did relative to what their risk exposure should have earned.

Broken down annually, a portfolio of all specialized ETFs delivered risk-adjusted returns of about −3.1% per year after fees, with the underperformance concentrated in the years right after launch, the newly-launched cohort runs at roughly −6% per year in those first five years.[1] Broad-based ETFs, by contrast, sit near zero alpha, exactly what you'd expect from a cheap index tracker. The gap between specialized and broad-based ETFs is roughly −3% per year, and it is not noise.

Crucially, the authors show this cannot be explained by the higher fees alone, the gross-of-fee numbers look just as bad. The driver is overvaluation of the underlying stocks at the moment the fund is created. Issuers are catering to investors' extrapolative beliefs: they package the stocks people are already excited about, precisely when those stocks are most overpriced.

Survivorship: most of them die

Underperformance is only half the story. The other half is that the funds themselves frequently disappear, which makes the surviving track records look far better than the lived experience of investors. Morningstar's Global Thematic Fund Landscape quantifies the carnage.

Window (to mid-2024)Survived & outperformed global equities
Trailing 1 year~18–22%
Trailing 3 years~5% (one in 20)
Trailing 15 years9%

Source: Morningstar Global Thematic Fund Landscape, data as of June 30, 2024.[3]

Over the 15 years leading up to mid-2024, 60% of thematic funds closed entirely ×DON'T TRUST, VERIFYClaim: Over the 15 years to mid-2024, 60% of thematic funds closed, and only 9% both survived and outperformed a global equity benchmark.Verify at: Morningstar ↗Figures from Morningstar's Global Thematic Fund Landscape report (data as of June 30, 2024). Confirm the closure rate and the 9% survive-and-outperform figure., and only 9% both survived and beat a plain global equity index. Over the trailing three years, only about one in twenty did. In the US, fund closures outnumbered launches for the first time since 2008 during the first half of 2024. Europe was worse still: 68% of the thematic funds that existed in mid-2009 had shuttered by mid-2024.[3]

Why survivorship matters: when a dog fund is quietly liquidated, it drops out of the category averages. Any chart of "thematic fund performance" you see is computed mostly from the funds that lived. The real expected outcome of buying a random thematic ETF includes a large chance the thing simply ceases to exist, and you eat the loss on the way out. The dot-com history is the template, after the bubble burst, the vast majority of internet-economy funds closed.

Higher fees, fee-insensitive buyers

Specialized ETFs charge more than broad index funds, and they get away with it because the people buying them aren't shopping on price. In the academic sample, specialized ETFs carried a median expense ratio of about 58 basis points versus 35 for broad-based funds, with the full range running from 4 bps to over 150 bps.[1] Morningstar finds both active and passive thematic funds charge fees that are many multiples higher than their non-thematic counterparts on an asset-weighted basis.[3]

This is the giveaway. A cost-conscious investor compares a 0.03% total-market fund against a 0.60% thematic fund and walks. The thematic buyer doesn't, because they aren't optimizing for cost, they're chasing a story, and the story drowns out the expense ratio. The marketing isn't about the math; it's about steering attention away from it. Financial product marketing works on exactly this principle: a salient narrative out front, the cost quietly buried.

Compounded over a multi-decade holding period, a fee gap of half a percent or more is a meaningful drag on its own, before the overvaluation problem even enters the picture. Against a near-zero-cost index, high relative fees compounded over many years are a direct, mechanical reason these funds lag.

The mechanism: peak hype, reflexive inflows, mean reversion

Put the pieces together and the failure is structural, not bad luck:

  1. Launch at peak attention. The fund is created because the theme is hot, which means the underlying stocks are already richly valued, high market-to-book, high sentiment, the classic markers of overvaluation.
  2. Reflexive inflows. Investors extrapolate the recent run-up and pour money in after the gains. Money chases performance, not the other way around.
  3. Mean reversion. Overvalued stocks revert. The post-launch alpha is negative because the entry valuation was unsustainable, and it tends to be worst in the very first years.

Then a second, behavioral layer makes the lived outcome worse than even the fund's own bad returns. Because thematic funds are volatile and exotic, investors trade them at the worst times, buying after a surge, selling after a slump. Morningstar's dollar-weighted return analysis finds that over a recent five-year window, thematic funds returned about 7.3% annualized while the typical investor in them earned just 2.4% ×DON'T TRUST, VERIFYClaim: Over a recent five-year window, thematic funds returned ~7.3% annualized while investors in them captured only ~2.4%, a ~4.9 percentage-point behavior gap.Verify at: Morningstar Mind the Gap ↗Morningstar's "Mind the Gap" research on dollar-weighted vs. total returns for thematic funds. Confirm the gap figure and window., a roughly five-point gap they forfeited purely to bad timing. Sector and thematic funds show the widest such gaps of any category.[4]

This is behavioral finance in its purest form: the product is engineered to be bought at the wrong time, and its volatility ensures investors keep buying and selling at the wrong times after that.

The takeaway: own the haystack

If a theme is genuinely going to reshape the economy, the companies that win are already inside a broad total-market index fund, you own them, at a fair weight, for three basis points, with no narrative premium and no closure risk. You don't need to guess which clean-energy or AI names survive; the index holds the survivors automatically and drops the losers without you paying an exit toll.

THE DISCIPLINED MOVE

A broad, low-cost index fund beats narrative-chasing on every axis that matters: lower fees, no overvaluation premium at entry, no survivorship roulette, and no temptation to trade a volatile single theme at the wrong moments. The boring choice is the high-probability choice. See Index Funds and Passive Investing for the full case.

When you feel the pull of a themed ticker, ask the one question the marketing is built to suppress: am I buying this on its risk-adjusted, after-fee numbers, or on the story? If it's the story, you're almost certainly buying the top.

Quick answers.

On average, and over their first five years post-launch, the published evidence says yes: specialized ETFs lose about 30% on a risk-adjusted basis versus the broad market, driven by overvaluation of the underlying stocks at launch. Individual funds can beat that, but you are fighting both the entry valuation and a high closure rate. The expected outcome is underperformance.
The theme being real and the fund being a good investment are two different questions. Clean energy and AI are real; the funds tracking them still underperformed because investors bought in after the underlying stocks were already bid up. A correct forecast about technology does not rescue an overpriced entry. And if the winners do emerge, a broad index already owns them.
Issuers launch funds that will sell, and funds sell when a theme is already popular. So new launches cluster at peaks in attention and recent returns, the worst possible entry point. The 2021 record of roughly 670 new thematic funds, right before the category collapsed, is the textbook example. Supply follows hype, not opportunity.
Only partly. Fees are higher, a median near 58 basis points versus 35 for broad funds, and over decades that drag is real. But the academic study found the underperformance persists even before fees are deducted. The larger driver is that the stocks inside the fund are overvalued at launch and subsequently mean-revert.
A broad, low-cost total-market or global index fund. It already contains the companies behind every theme, at a fair weight and a fraction of the fee, with no overvaluation premium at entry and no risk of the fund being liquidated out from under you. This is educational content, not financial advice, but the evidence consistently favors the cheap, diversified haystack over the narrow themed bet.
Sources & Citations
  1. Ben-David, Franzoni, Kim & Moussawi, "Competition for Attention in the ETF Space," Review of Financial Studies 36(3), March 2023, pp. 987–1042 (sample 2000–2019). Key findings: specialized ETFs lose ~30% risk-adjusted over their first five years; ~−3.1%/yr after fees for the full specialized portfolio and ~−6%/yr for the newly-launched cohort; specialized vs. broad median fees 58 vs. 35 bps; underperformance driven by overvaluation at launch, not fees - NBER WP 28369 - nber.org/papers/w28369 · PDF - nber.org/…/w28369.pdf
  2. "Competition for Attention in the ETF Space," Review of Financial Studies (Oxford Academic), abstract and launch-cycle data; global thematic assets surged threefold to a peak near $892B before tumbling; ~670 new thematic funds launched in record 2021 - academic.oup.com/rfs/article-abstract/36/3/987 · Morningstar Global Thematic Fund Landscape 2024
  3. Morningstar, Global Thematic Fund Landscape (data as of June 30, 2024) and "The Thematic Fund Landscape in 7 Charts." 60% of thematic funds closed over the 15 years to mid-2024; only 9% survived and outperformed global equities over 15 years; ~5% (one in 20) over 3 years; 68% of European funds from mid-2009 closed by mid-2024; thematic fees many multiples higher than non-thematic on an asset-weighted basis - morningstar.com/funds/thematic-fund-landscape-7-charts
  4. Morningstar, "Mind the Gap" / "Understanding the Gap in Thematic Funds" - dollar-weighted vs. total returns. Over a recent five-year window thematic funds returned ~7.3% annualized while investors captured ~2.4% (a ~4.9 pp gap); sector/thematic funds show the widest return gaps of any category - morningstar.com/business/lp/thematic-mind-the-gap
NEW TO THIS TOPIC?

See the glossary for plain-English definitions of every term used here.

Last updated 2026-05-31. Educational content, not financial advice.

Subscribe via RSS for new articles.