Machines are coming. Artificial intelligence can write songs, talk and code, but can it replace investment managers? More and more AI-driven ETFs are trying to prove it can be done but many still struggle to outperform funds whose stock holdings are determined by the market value of a person or company.
Over the past few years, several AI-powered ETFs have been launched and touted. Like his mainstream ETFs, it is a stock-exchange-listed fund that offers investors exposure to a diversified portfolio of securities. But unlike their non-machine counterparts, they use machine learning, sentiment analysis, and natural language processing to identify patterns and trends that help guide asset selection.
Notable examples include South Korea’s Qraft Technologies’ three products: US Large Cap ETF (QRFT), US Large Cap Momentum ETF (AMOM), US Next Value ETF (NVQ), and Van Eck Social Sentiment ETF (BUZZ). is mentioned. EquBot’s AI Powered Equity ETF (AIEQ), Merlyn. AI’s Bull-Rider Bear-Fighter Index (WIZ) and its SectorSurfer Momentum ETF (DUDE), and WisdomTree International’s AI Enhanced Value Fund (AIVI).
However, the performance of AI-driven alternatives has been inconsistent so far.
For the three-year period ending April 19, 2023, the Traditional Index Tracking SPDR S&P 500 ETF (SPY) delivered a 14.8% return. QRFT trails closely with a 14.5% return. Other offerings AMOM contributed 12%, AIEQ managed 4.4%, WIZ 6.4% and AIVI 11.9%.
In the year to the same date the situation is more complicated, with both AMOM and AIVI outperforming SPY and the rest lagging behind.
“These ETFs use artificial intelligence to select stocks in hopes of outperforming the broader market, but decades of research have shown that active stocks of all kinds It’s hard to beat an index-based approach with low operating costs,” said Todd Rosenbluth. , Head of Research at his data provider VettaFi.
When picking stocks, experts say AI can struggle to track trends that don’t show up in historical data, company reports and news media that it analyzes using natural language processing.
“We live in a very complex world where human intuition adds value,” said Joseph Byram, chief data scientist at Principal Financial Group. “You can automate the reading of all corporate documents in the universe. But what you can’t train a model to do is look at how central banks behave — levels you don’t think you can model.” There is a unique risk of
As a result, AI-based systems still rely to some extent on human oversight and intervention. Wisdom Tree International’s offering of his AI ETF will allow his managers to review and reject trades in the firm’s portfolio, though this is “not happening very often,” the firm said. EquBot has operational controls to help you process false or inaccurate financial information and determine how often and what types of securities you trade.
But developers say AI models already eliminate inefficiencies in human decision-making.
Christina Bargeron, Client Portfolio Manager at Voya, who developed the unique model behind Wisdom Tree’s AIVI, said: “We don’t have HR. We work weekends and we don’t take days off.
AI ETFs also remove human bias and egos from the decision-making process. Francis Oh, his head of AI ETFs at Qraft, said the AMOM fund’s failure to include financial stocks in its portfolio at the beginning of the year helped avoid a crash following the collapse of Silicon Valley banks and the turmoil at Credit Suisse. It is said that it was useful for .
“As a portfolio manager, I have to decide whether to keep the strategy or change it,” says Oh. “Should we wait a few more months or change now? That decision could be painful for fund managers and investors.”
Human fund managers continue to find it difficult to prove their worth against indices and traditional ETFs that track indices. According to Morningstar data, only 48.7% of U.S. equity funds will outperform the index in 2022, and only 43.2% of global equity funds. The longer the timeline, the worse the situation. Between mid-2012 and mid-2022, only 12% of US equity funds and about 20% of global equity funds provided higher returns.
And the quality of competing AI technologies just keeps getting better. AI that can be used for stock picking is still in its relatively early stages, experts say, but it will improve its ability to make good decisions as companies like OpenAI and Google pour billions into its development. By adding more sources of data, AI models also provide a better foundation on which to base stock selection decisions.
EquBot Chief Investment Officer Chris Natividad said: “Our IBM partner said 90% of his data was created in the last few years, and we expect to see the exact same statement every two years.”
