Down Arrow Button Icon

It had to happen: Google searches for “short-selling ETFs” rocketed this week. Such exchange-traded funds exist—and amid GameStop madness, their prices have surged.

For any topic that’s hot or new, there’s probably an ETF. Last year, ETF marketers concocted COVID-19 vaccine ETFs, work-from-home ETFs, and SPAC (special purpose acquisition company) ETFs, each filled with stocks positioned to gain from those suddenly strong trends. When Bitcoin’s price peaked during the week of Jan. 3 this year, so did Google searches for “Bitcoin ETFs.” Regulators don’t let ETFs invest in Bitcoin, unfortunately for ETF marketers—but perhaps luckily for investors.

That’s the implication of a new analysis by researchers at Ohio State University, Villanova University, and the Swiss Finance Institute, which finds unequivocally that on the whole, sharply focused ETFs are not good for investors. “A portfolio of all specialized ETFs earns a negative risk-adjusted performance of 3.1% per year, after fees,” the researchers report. Newly launched ETFs perform especially badly: They “lose 5% per year in risk-adjusted terms.” (Those short ETFs, though riding a hot trend, are broadly based and don’t meet the researchers’ definition of a specialized ETF.)

So-called thematic ETFs are in some ways the opposite of the original ETFs, which first appeared in 1993. The early funds tracked broad indexes, especially the S&P 500, and competed by trying to charge the lowest fees; that’s still how broad-based ETFs compete. But specialized ETFs hold unique portfolios and charge much higher fees. As of December 2019 (the most recent period for which data was available), specialized ETFs accounted for 18% of ETF assets yet brought in 36% of fees.

The new analysis finds that these funds are just what they sound like: products created quickly to capitalize on high-interest trends of the moment. Examining the individual stocks in specialized ETFs, the researchers find that they are classic “glamour stocks that are likely to be overvalued.” Specifically, they have high market-to-book ratios and high short interest. They also attract lots of positive media attention—typically right up until the launch of the specialized ETF in which they’re included; then there’s a “quick reversal of the initial hype.” Specialized ETFs tend to get launched “near the peak of valuation for the underlying securities,” the researchers report. As a result, such ETFs tend to “start underperforming right after launch.”

Obvious question: Why does anyone buy them? The only logical explanation—that investors buy them as hedges against other investments—doesn’t pan out. After complex statistical analysis, the authors say they “do not find evidence consistent with an insurance motive.”

Another possible explanation, based on considerable previous research, is that some investors simply like gambling. Just as in a casino, they know the odds are against them, but they enjoy the fun and excitement of putting down money for an extremely remote chance of getting a giant return. And in fact the returns on stocks in specialized ETFs, when charted, tend to have long right-hand tails, indicating tiny probabilities of ultrarich payoffs. The researchers say that in specialized ETFs they “find evidence of catering to preferences for gambling.”

But the strongest explanation is the simplest and most obvious: Investors in specialized ETFs tend to be unsophisticated. They fit the profile of typical, hapless individual investors who buy at the top, “who chase past performance and neglect the risks arising from the under-diversified portfolio,” say the researchers. They find that owners of specialized ETFs are more likely to be retail investors than are owners of broad-based ETFs.

Fortune asked the three ETF giants—BlackRock, State Street, and Vanguard—to comment on the new paper. A BlackRock spokesman declined to comment other than to say, “We believe that there is a case to be made about thematic ETFs.” BlackRock offers 106 stock ETFs, many of which are thematic. State Street did not respond; it offers 85 stock ETFs, many of which are thematic. A Vanguard spokesman said the findings “seem to validate our product development philosophy, which focuses on broadly diversified strategies with enduring investment merit.” Vanguard offers 56 stock ETFs, most of which are broadly based; the firm offers 11 sector ETFs matching the 11 sectors of the S&P 500 and does not offer thematic ETFs.

The authors acknowledge that investors make poor investment decisions in countless ways, but specialized ETFs are different. What makes them especially worrisome, the authors say, is that ETFs have marketing budgets. The researchers’ harsh conclusion: “The marketing strategies of specialized ETFs attract unsophisticated investors to underperforming investment propositions.”