ETFs that use humans and market indexes such as the S&P 500 to pick stocks are having a rough year. Will machines do any better?
This month, another ETF launched that uses artificial intelligence to pick stocks. The BTD Capital Fund uses a computer model, using more than 20 indicators, to scoop up shares of large-cap stocks deemed to be oversold and due for a bounce higher.
So what kinds of stocks is this machine buying? It owns 27 positions, and of those, eight are up this year. The top-performing holdings are Vertex Pharmaceuticals, General Mills and Nucor, up 33%, 26% and 20.5%, respectively.
Watching which stocks a machine would pick is intriguing in a year like 2022, when very few strategies are working. Even the traditional 60% stock and 40% bond portfolio is having its worst run in years.
But does a machine know how to pick stocks any better than a human?
The Kinds Of S&P 500 Stocks Algorithms Like
BTD Capital is far from being the only or first ETF to use AI to pick stocks. There are more than 10 other ETFs doing the same thing.
One of the largest AI-powered ETFs is SPDR S&P Kensho New Economies Composite ETF with nearly $1.6 billion in assets. The ETF uses AI to help spot companies pushing technology that's transforming the economy. Top holdings include tech-testing firm Teledyne Technologies, defense contractor Leidos Holdings and diagnostic system maker Bruker. So far it's not paying off. Shares of the ETF are down 32% this year, lagging the S&P 500's 19% drop.
Elementary My Dear Watson
But what if a more well-known AI system could have more freedom to choose stocks in any sector? There's an ETF for that, too, offering an insightful glimpse into what a machine would choose. The $97.8 million-in-assets AI Powered Equity ETF uses IBM's Watson AI to pick its portfolio using fast-changing algorithms.
The ETF has nearly 130 positions, but it's placing its biggest bets on six stocks, with weights in the portfolio of 3% or higher each. The number one investment bet is video-processing firm Ambarella at 8.5%, followed by financial firm American Express at nearly 4% and tech consultant DXC Technology at 3.7%.
"The only overlapping themes in its top 10 positions are energy and transportation, with the rest spanning everything from cybersecurity and drug manufacturing to video games and the semiconductor industry," said Nicholas Colas of DataTrek Research.
So far, the AI has its fair share of misses. Nearly three-quarters of the stocks it owns are down this year. The price of the ETF itself is also lagging the S&P 500: down 31%. Longer-term isn't much better, gaining just 11% in the past five years, while the S&P 500 is up 44% in that time.
Does AI Know Better Than The S&P 500?
So far, many of the promises of automation for stock picking haven't panned out. For one thing, they don't save investors much on fees. The average annual expense ratio of 10 major AI-powered ETFs is 0.77%. That's much more than the 0.09% you'll pay to own the SPDR S&P 500 ETF Trust.
What about performance? The AI-powered ETFs haven't exactly thrived this year either. In fact, most are lagging the S&P 500. The seven major AI-driven ETFs with trading history going back all year are down an average 24.7% this year. And the two with five-year track records are lagging too.
"I get the premise that finding market anomalies will be aided with computing power and taking emotions away," said Todd Rosenbluth, director of research at VettaFi. "But as with all active management, a live record to assess will matter most to investors who still often want to beat the broader benchmark rather than replicate it."
AI's Favorite Stocks
Top-performing current holdings of stocks in AI-Powered Equity ETF this year
| Company | Symbol | YTD % ch. | Sector |
|---|---|---|---|
| Texas Pacific Land | 105.9% | Energy | |
| Maxar Technologies | 74.4% | Industrials | |
| APA | 72.2% | Energy | |
| Steel Dynamics | 70.3% | Materials | |
| H&R Block | 69.3% | Consumer Discretionary | |
| Murphy Oil | 66.9% | Energy | |
| Halozyme Therapeutics | 45.0% | Health Care | |
| Merck | 44.6% | Health Care | |
| Alnylam Pharmaceuticals | 41.6% | Health Care | |
| Array Technologies | 34.5% | Industrials |
Sources: S&P Global Market Intelligence, IBD
Follow Matt Krantz on Twitter @mattkrantz
