Are AI ETFs worth the hype?

To say there has been hype around AI is an understatement, and partially for that reason, advisors caution investors against too much exposure to the roster of thematic ETFs focused on it.

Currently, it is a small but growing area of product development. There are more than a dozen U.S. exchange-traded funds that purport to focus on some aspects of artificial intelligence, representing about $5 billion in assets, according to data from Morningstar.

More are on the way. Late in December, Alger filed with the Securities and Exchange Commission for an artificial intelligence ETF. TCW, which has had a mutual fund focused on AI since 2017, recently filed to reorganize it in the form of an ETF.

“It’s kind of hard to figure out what exactly is an AI ETF. That’s one of the big problems for issuers,” said Bryan Armour, director of passive strategies research at Morningstar. “It’s hard to nail down what is AI and which companies are exposed to AI – and that’s led to variable portfolios.”

When the first ETFs in that category appeared beginning in 2016, the focus was generally on robotics, Armour noted. At $2.4 billion, the Global X Robotics and Artificial Intelligence ETF was not only the first, but it remains the biggest.

Overall, “they haven’t really taken off” as a category, he said. Last year, though, with so much attention to chatbots and generative AI – not to mention money backing them – investors showed more interest, with the ETF category raking in $2.2 billion. At least six ETFs in the category launched in 2018, and most started after 2020, according to data from Morningstar Direct.

There are significant differences in the AI ETFs on the market. While they differ considerably from wider tech-focused ETFs, some invest primarily in smaller companies and have higher potential for volatility, while others allocate some assets to very large, established companies like Microsoft and Nvidia, Armour said. The latter company, though, is a holding among 10 of the ETFs on the market, and many of the same stocks are included in half the AI ETFs on the market, he noted.

And, much as AI’s potential extends far beyond the tech world, the holdings in the ETFs also crosses various sectors and countries – most of them included international holdings, he said.

“In my opinion, the whole market is going to change,” Armour said. “It does feel like an opportunity. It’s just hard to put something like AI into an ETF.”

A similarity between many of the AI ETFs and broader tech-focused ETFs are holdings in the “Magnificent 7” stocks – Amazon, Apple, Google, Meta, Microsoft, Nvidia and Tesla, said Roxanna Islam, head of sector and industry research at VettaFi.

“But AI ETFs also typically hold smaller-cap companies in areas like health care robotics, autonomous vehicles, and industrial automation,” Islam said in an email, pointing to about 10 such ETFs. “Most of these have seen positive inflows in 2023 with a lot of investor interest related to generative AI and applications like ChatGPT. These can serve [to] complement or [as] a replacement for a broader technology ETF while providing a relatively high return.”

Last year, the Global X Artificial Intelligence ETF and the broader Technology Select Sector SPDR ETF were each up by about 55%, she noted.

Advisors didn't indicate they were eager to jump on the trend.

“The benefits of AI will be present among solid companies that already exist [across sectors],” said David Hunter, owner of First Light Wealth. “We don’t need something different what we’ve had [to invest in] over the last 20 to 30 years, but we also want to make sure we’re taking advantage of new technologies.”

That isn’t necessarily exciting, but if clients are intensely interested and have 1% to 3% of assets to take higher risks with, AI ETFs are one way to go, he said.

“If they’re in a diversified portfolio … they’re probably participating in this with the growth of the general economy,” Hunter said.

Another advisor, Eric Walters, founder of Summit Hill Wealth Management, said he isn’t a fan of thematic ETFs, including those focused on AI.

“They tend to be created and released during a period of hype and trade in equities that are overvalued. Investors put money in and then the bubble bursts,” Walters said in an email. “All we have to do is consider some recent theme ETFs that were popular with investors but have lost money: green energy, electric vehicles, bitcoin and crypto currencies, gold and gold miners.”