MW AI-powered stock ETFs were hyped as superior investments. Then reality hit.
By Sam Wyatt and Gary Smith
It's tempting to think that computers can make better investment decisions than we can
Computers can remember things we forget and recite facts we never knew. They can make complicated calculations instantaneously and error-free. They can give us directions to unfamiliar places and tell us how to repair a leaky faucet. Computers can beat the best humans have to offer at chess, Go and Jeopardy. It's tempting to think that computers can make better investment decisions than we can. That temptation should be resisted.
In 2017, the word "AI" was selected by the Association of National Advertisers as the Marketing Word of the Year. Later that year, on Oct. 18, EquBot launched the first public AI-powered fund - the Amplify AI Powered Equity, ticker symbol AIEQ AIEQ.
EquBot said that AIEQ represents "the ground-breaking application of three forms of AI": genetic algorithms, fuzzy logic and adaptive tuning. Chida Khatua, chief executive and co-founder of EquBot, boasted that AI algorithms can outperform humans because they don't get tired and don't make arithmetic or mental mistakes: "EquBot AI Technology with Watson has the ability to mimic an army of equity research analysts working around the clock, 365 days a year, while removing human error and bias from the process." Skeptics (including us) might note that AI algorithms are not susceptible to human biases because they have absolutely no understanding of how the numbers they crunch relate to the real world. It is hard to be biased when you have no comprehension of things you might be biased about.
Two weeks later, Horizons launched the Active AI Global Equity ETF, (its ticker: MIND). Horizons heralded the ETF's investment strategy "entirely run by a proprietary and adaptive artificial intelligence system that analyzes data and extracts patterns ... The machine learning process underpinning MIND's investment strategy is known as Deep Neural Network Learning - which is a construct of artificial neural networks that enable the A.I. system to recognize patterns and make its own decisions, much like how the human brain works, but at hyper-fast speeds."
Steve Hawkins, Horizons' president and CEO, added, "Unlike today's portfolio managers who may be susceptible to investor biases such as overconfidence or cognitive dissonance, MIND is devoid of all emotion." Again, AI algorithms are devoid of emotion because they are devoid of intelligence.
Ticker Launch Date Termination Date Annual return * Percent S&P 500 Amplify AI Powered Equity ETF AIEQ 1/9/17 7.84 14.07 WisdomTree U.S. AI Enhanced Value Fund AIVL 1/3/22 5.44 8.22 WisdomTree International AI Enhanced Value Fund AIVI 1/18/22 4.66 9.23 Reality Shares Fundstrat DQM Long ETF DQML 11/2/18 10/22/19 7.97 11.80 Fount Subscription Economy ETF SUBS 10/29/21 5/25/23 -15.68 -4.80 Fount Metaverse ETF MTVR 10/29/21 7/25/23 -17.97 1.20 BTD Capital Fund ETF DIP 12/14/22 6/3/24 14.47 26.26 Active AI Global Equity Fund MIND 11/1/17 5/18/22 -2.72 11.58 Average 0.50 9.70 * Through 8/31/24
How did these trailblazers do? Through Aug. 30, AlEQ had a total cumulative return of 68%, compared to the S&P 500's SPX 147%. MIND was shut down in May 2022, having given investors a cumulative loss of 12%, compared to the S&P's 65% gain.
Several more AI-powered funds have since been created. Have they done any better? We looked at all Al-powered ETFs launched between Oct. 18, 2017, and Jan. 1, 2023. Ten of these funds advertise themselves as being completely AI-driven with minimal human intervention. The table shows the results for the U.S.-focused funds through Aug. 30. Every single one has done worse than the S&P 500. Three had negative returns. Five have closed down. The unweighted average annual return for the group was 0.50%, compared to 9.7% for the S&P 500.
In addition, two fully AI funds focused on stocks outside the United States: JAKOTA K-Pop and Korean Entertainment ETF KPOP and the AI Powered International Equity ETF (AIIQ). KPOP has posted a -18.28% annual return since inception compared to a 20.61% return for the S&P 500. Meanwhile, AIIQ had a 1.32% annualized return before it closed in 2022 (compared to an 11.59% S&P 500 annualized return over the same period).
The core problem is that current AI algorithms are really good at finding statistical patterns and correlations in historical data, including stock prices and other financial and nonfinancial data - but these algorithms have no way of assessing whether a discovered pattern or correlation is meaningful or just a temporary fleeting coincidence. We shouldn't expect their performance to be any better than monkeys throwing darts. Add in the costs of the computer jockeys, hardware, electricity and more, and they are, on average, almost certain to underperform a simple indexing strategy.
Perhaps AI funds that are not open to the public have done better than these publicly available funds. We are skeptical. However, no data exists to test that optimistic conjecture and, in any case, it isn't useful if we can't invest in them. So use computers to do the things they do best. Investing is not one of them.
Sam Wyatt studies economics and mathematics at Pomona College. His research interests lie in the intersection of economics and technology, particularly in the application of AI to investment strategies.
Gary Smith, Fletcher Jones Professor of Economics at Pomona College, is the author of dozens of research articles and 17 books, most recently, "The Power of Modern Value Investing: Beyond Indexing, Algos, and Alpha," co-authored with Margaret Smith (Palgrave Macmillan, 2023).
More: These investors are killing the game-changing tech startups of America
Plus: Big Tech's AI bubble is alive and unwell
-Sam Wyatt -Gary Smith
This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.
(END) Dow Jones Newswires
September 11, 2024 12:53 ET (16:53 GMT)
Copyright (c) 2024 Dow Jones & Company, Inc.
Comments