An exchange-traded fund driven by artificial intelligence booted Tesla and Amazon from its portfolio in June, instead choosing to load up on shares of companies including Qualcomm, Snap and GameStop.
The Qraft AI-Enhanced U.S. Large Cap Momentum ETF, trading as AMOM
on the New York Stock Exchange, removed major technology companies from its portfolio this month, as it shifted to favor retailers and other post-pandemic trades.
Electric-car maker Tesla
and online retailer Amazon
represented two of the fund’s three largest holdings in May, but were completely removed in the latest rebalancing on June 2, along with graphics microchip maker Nvidia
which was its sixth-largest holding. The artificial-intelligence program controlling the fund believes these stocks will see price declines across the coming month.
The standout among the stocks added in June was GameStop
the videogame retailer that epitomized the “meme stock” trading frenzy that began in late January.
This was when a flock of investors, largely organized on social media platform Reddit, helped squeeze hedge funds’ short positions on companies including GameStop earlier this year. The trading frenzy caused multibillion-dollar losses for hedge funds, unbelievable gains for individuals that timed it right, and ushered in a new era of internet-inspired trading.
GameStop’s stock price rose almost 10-fold from Jan. 15 to Jan. 27, from $35 per share to nearly $350. The stock is currently trading around $300 and makes up around 1% of AMOM. And now the AI calling the shots thinks it will move even higher in June, and the shares have already gained more than 6% since the stock was added to the fund for the first time.
“Few fund managers would take the risk of adding a meme stock to their portfolios, but Qraft’s AI model has no such prejudices,” said Geeseok Oh, a managing director at Qraft and the head of its Asia-Pacific business.
The top five stocks by portfolio weight added to AMOM in June include semiconductor group Qualcomm
Big Tobacco company Philip Morris
social-media player Snap
medical technology specialists Edwards Lifesciences
and orthodontics group Align Technology
After the fund was rebalanced, AMOM’s top five largest holdings by portfolio weight were tech giant Facebook
and Home Depot
software company Adobe
and semiconductor manufacturing company Texas Instruments
“This month, AMOM’s portfolio appears to be shifting towards post-pandemic trades, a bit more diversified of a portfolio from the previous month’s big tech-heavy strategy,” said Oh.
AMOM’s decision to remove Tesla from its portfolio came after a bullish bet failed to pay off. The fund bought around $1.4 million worth of shares in the electric-vehicle company in May after avoiding the stock for months, and shares in Tesla fell 7% before the AI ditched it from the fund. This mistake with Tesla was a rare occurrence for the robot controlling AMOM, which otherwise has a strong record of predicting moves in the company’s share price.
AMOM has been listed in New York since May 2019, and has delivered total returns of 11% so far in 2021 and 53% in the past year — outpacing its benchmark, the S&P 500 Momentum index, which has climbed a comparable 26% in the past year.
AMOM is an actively managed portfolio driven by artificial intelligence, tracking 50 large-cap U.S. stocks and reweighting its holdings each month. It is based on a momentum strategy, with the AI behind its stock picks capitalizing on the movements of existing market trends to inform the decision to add, remove, or reweight holdings. The artificial intelligence scans the market and uses its predictive power to analyze a wide set of patterns that show stock-market momentum.
The fund is a product of Qraft, a Seoul, South Korea-based fintech group leveraging AI across its investment products, which include three other AI-picked versions of major indexes: a U.S. large cap index
; a U.S. large cap dividend index
; and a U.S. value index
The entrance of AI-run funds onto Wall Street promised a new high-tech future for investing, though it hasn’t quite lived up to the hype yet. Theoretically, researchers have shown that AI investing strategies can beat the market by up to 40% on an annualized basis, when tested against historical data.
But Vasant Dhar, a professor at New York University’s Stern School of Business and the founder of machine-learning-based hedge fund SCT Capital Management, argued on MarketWatch in June 2020 that AI-run funds won’t “crack” the code of the stock market.
Advocating caution, Dhar said that it was difficult for funds underpinned by machine learning to maintain a sustainable edge over markets, which have “a nonstationary and adversarial nature.” He advised investors considering an AI system to ask tough questions, including how likely it is that the AI’s “edge” will persist into the future, and what the inherent uncertainties and range of performance outcomes for the fund are.