Move Over, ChatGPT. AI-Powered ETF Outperforms This Year

AIEQ, which leverages IBM’s Watson supercomputer, is achieving outsized returns in 2023.

HeatherBell_green_bg
|
Reviewed by: Heather Bell
,
Edited by: Heather Bell

When the latest and greatest artificial intelligence ChatGPT tool was asked to create a market-beating portfolio, it reportedly told users the stock market is just too hard to predict. However, the $102 million AI Powered Equity ETF (AIEQ), issued by ETF Managers Group, claims to have been doing just that this year.  

AIEQ, which launched in 2017, leverages the power of IBM’s Watson supercomputer to select its portfolio holdings.  

It’s up 13.5% year to date as of Jan. 27, according to Morningstar data. Meanwhile, the Vanguard Total Stock Market ETF (VTI), which is a total market fund and provides a benchmark, is up 6.7% over the same period, essentially half the performance of the actively managed fund.  

AIEQ is a very different fund than VTI. AIEQ takes concentrated bets on the U.S. stock market, and has only 114 securities in its portfolio versus VTI’s nearly 4,000 holdings. Of course, as an actively managed fund, AIEQ is more expensive, charging 0.75% versus VTI’s rock-bottom 0.03%.  

It also looks like the two funds have very different factor exposures. As a broad market fund, VTI has little in the way of significant factor exposures, but AIEQ has a factor exposure to low size of 1.09, according to the MSCI data on ETF.com’s site, plus a -0.74 exposure to the low volatility factor. 

AIEQ actually only shares two of its top 10 holdings with VTI: JPMorgan Chase & Co. and UnitedHealth Group Inc. 

Beyond Standard Market Data 

Equbot Chief Investment Officer Chris Natividad said that AIEQ doesn’t just rely on standard market data like that which can be downloaded from Bloomberg or S&P. It also pulls data from what he terms “unstructured data” such as tweets, earnings calls and keyword data. 

“We're focused on investment related data, looking at how these different types of signals impact security prices across different time horizons,” he noted.  

Looking back over longer-term performance, AIEQ has mostly trailed VTI notwithstanding this year. Over the three-year period, AIEQ has returned 4.8% on an annualized basis relative to VTI’s 9.1%; over the five-year period, AIEQ has returned 5.5% versus VTI’s 8.6%, according to Morningstar. This year, however, AIEQ has been off to a great start. Given that it uses machine learning, it is possible the fund is simply getting better at picking stocks.  

“We continue to see improvement in the fund for a variety of different reasons. The systems continue to learn from every trade that is executing right now,” explained Natividad. “The best days of the fund are still ahead of it. And just as you'll see ChatGPT’s responses change and evolve with time and data, so will our fund.” 

 

Contact Heather Bell at [email protected] 

Heather Bell is a former managing editor of etf.com. She has also held editorial positions at Dow Jones Indexes and Lehman Brothers. Bell is a graduate of Dartmouth college and resides in the Denver area with her two dogs.