By Elsa Ohlen

Some would say that trading hasn't been easy amid the chaos and uncertainty that has characterized markets recently. But an artificial intelligence-driven stock picker has a clear plan: Dump Meta and Walmart and load up on stocks that have outperformed recently.

Tariffs, and their impact on the economic climate have weighed heavily on stocks -- and lifted them sharply as effects appeared not to be as bad as first expected.

SoftBank-backed Qraft Technologies is one company capitalizing on AI technology to pick stock for its exchange-traded funds. The Hong Kong-based business has three U.S.-listed ETFs with a combined asset value of roughly $50 million.

Their selling point? All decisions are 100% controlled by an AI model and based on different investment strategies and investors' risk appetite.

May was an interesting month, said Justin Tam, ETF Lead at Qraft. It showed how different their two biggest ETFs, the momentum fund AMOM and the more defensive option QRFT, are.

AMOM targets stocks on the rise on the assumption that they'll continue to outperform and is always looking for growth opportunities.

Nvidia is still the fund's largest holding, with about 9% allocated to the chip maker stock. But there were a few surprise removals this month as well. It dropped its entire holding in Meta, which had accounted for 6% of the fund previously, after the Facebook-parent traded largely flat in April. The removal indicates that the AI predicts the stock won't see gains anytime soon.

Walmart was also completely sold off after a solid month with a 10% gain, which Tam says could be profit-taking.

Chip maker Broadcom, drugmaker Eli Lilly, analytics software company Palantir, and streamer Netflix were notable additions to AMOM's portfolio in May. Those stocks all saw impressive gains between 9% and 40% in April.

One argument in favor of AI handling your investments is that an AI trader won't be influenced by emotions, fear or attachment to previous decisions. The counterargument is that it lacks intuition and the ability to imagine things can become different than they were.

The ETFs reallocate assets on the second trading day of every month. That means the last time they rebalanced was on April 2, or Trump's so-called Liberation Day. "It wasn't pretty," Tam recalls, referring to the heavy market losses in the weeks following the president's tariff announcement. "Things can change really, really quickly in this current condition. It's really about just finding the proper risk tolerance, I'd say, for the everyday investor."

While Liberation Day was painful for AMOM, the fund bounced back as most tariffs were put on a 90-day pause shortly after being announced, ending the month up nearly 4%. It's still down just over 10% so far this year, largely due to a "very ugly March," where tech, where AMOM is heavily overweight, took a big dive.

The momentum benchmark, the iShares MSCI momentum fund ( MTUM), however, had managed to rise 2% in 2025 by the end of April.

The company's more defensive QRFT ETF on the other hand, has been outperforming the broader market. It's down less than 2% this year, compared with the S&P 500's 4% drawdown. "We're crediting that to a more defensive allocation over the past few months," Tam says.

That fund is slightly overweight on information technology, financials, and healthcare compared with the S&P 500. However, among its 10 biggest holdings are all Magnificent 7 companies: Apple, Microsoft, Meta, Nvidia, Amazon, Alphabet, and Tesla. It also holds a decent chunk of Eli Lilly, Walmart, and Broadcom.

Diversification was the number one thing Qraft had in mind when designing QRFT. The model uses a multifactor approach and includes 350 large-cap U.S.-listed stocks. It makes no big bets on any particular name or industry, which Tam thinks is "very appropriate in the current market climate."

"Timing the market is very, very hard; some would say impossible. The more that you rebalance your portfolio, the greater the risk that you're wrong and you get hit by the whipsaw effect," he says.

While AIs are good at processing huge amounts of data, they lack human intuition. Now, given the president's proven ability to surprise the market, it's far from a given that human intuition will beat hard, cold -- and backward-looking -- data.

"Nobody can predict what's going to happen tomorrow, not even AI," says Tam.

Write to Elsa Ohlen at elsa.ohlen@barrons.com

This content was created by Barron's, which is operated by Dow Jones & Co. Barron's is published independently from Dow Jones Newswires and The Wall Street Journal.

(END) Dow Jones Newswires

May 12, 2025 01:00 ET (05:00 GMT)

Copyright (c) 2025 Dow Jones & Company, Inc.