(Bloomberg) -- One of the oldest artificial intelligence exchange-traded funds can’t keep up with the technology’s latest hype. 

The $108 million AI Powered Equity ETF (ticker AIEQ), which uses a quantitative model running on IBM Corp.’s Watson platform to scan thousands of US companies, has rallied about 4% so far this year. That compares to a roughly 9% climb for the S&P 500.

The irony at the heart of AIEQ’s underperformance is that the current AI investment craze is fueling recent gains in the broader equity indexes. Investors have piled into stocks such as chipmaker Nvidia Corp. and Microsoft Corp. and Alphabet Inc., which are bulking up their AI efforts. While that boom has helped to push the benchmarks higher — Societe Generale SA estimated in mid-May that the S&P 500 would be 2% lower this year without it — AIEQ largely has missed cashing in on the craze. 

“It is ironic that an AI-powered algorithm has not capitalized on the rally in big tech stocks that’s been driven by its own disruptive technology,” said Jessica Rabe, co-founder of DataTrek Research. “AIEQ has previously tended to work best when it could catch momentum driven tech names in broad-based market rallies like during the pandemic crisis, but it’s clearly failed to do that this year.”


At one point in January, AIEQ was sitting on year-to-date gains of over 16% and handily outperforming the S&P 500. However, that rally has since faltered, with the ETF slipping behind the S&P 500 in mid-March. 

AIEQ’s portfolio of roughly 150 stocks is mainly skewed towards software and Internet shares, with Palantir Technologies — a company with big AI ambitions — ranking as its top holding. While Alphabet is among its top 10 holdings, it doesn’t own popular AI stocks such as Nvidia and Microsoft.

“We leverage a lot of unstructured data, so we’re analyzing millions of news articles, social media posts every single day and we extract the sentiment and intent,” Art Amador, the co-founder and chief operating officer of EquBot, which developed the model behind AIEQ, said on Bloomberg Television’s ETF IQ Monday. “Right now it looks like it’s riding the AI and cloud theme at the moment, but it does change over time.”

Coming into 2023, AIEQ’s allocations were far more mixed. The fund’s heaviest weightings included consumer discretionary stocks, with those such as home furnishing firm RH and Las Vegas Sands Corp. making up its top holdings. 

AIEQ’s best performance tends to come during bull markets when the models latch onto momentum trades, according to DataTrek’s Rabe. However, that process has stumbled in 2023 even as the AI craze sweeps up some of the market’s biggest tech stocks. 

“We find AIEQ is an interesting case study for asset allocators and stock pickers because its investment process differs from traditional approaches,” Rabe said. “But as with some of what we are all learning from ChatGPT, different does not always mean better.”

©2023 Bloomberg L.P.