Total botch jobs.
Stonks
There’s no denying that AI hype has been gangbusters for the stock market, even if the whole thing’s a bubble.
But how good is AI itself at actually selecting stocks? Two researchers, Gary N. Smith, a professor of Economics at Pomona College, and Sam Wyatt, a student at the college and a project lead at Pomona Consulting, conducted an analysis to find out.
Writing in a piece for Scientific American, they argue that “AI-powered investing is particularly interesting because it provides a quantifiable way to assess the abilities of the technology.”
And their findings, which are yet to be peer-reviewed, suggest that the abilities are shoddy.
Smith and Wyatt looked at every publicly available exchange-traded fund launched since October 2017 that either partly or fully depended on an AI system to make stock decisions — and the vast majority of them performed worse than the S&P 500, an index of 500 of the largest companies listed on the US stock exchange.
Tellingly, over half of the funds have since been shuttered.
Ill-Fated Investments
For some context, the S&P 500 is considered a benchmark of the stock market’s health. If it’s doing well and your fund isn’t, then you’re not cashing in on the gravy train.
Thus, you could argue that if out of 43 funds that partly used an AI system to inform stock decisions, only ten managed to do better than the S&P 500, there appear to be some serious shortcomings with the technology. Collectively, the partly AI-driven funds’ average annual rate of return was five percent worse than the index’s 12.4 percent.
But maybe in those cases, the humans in the mix cramped the AIs’ style. How did fully AI funds with no human intervention do?
Even worse, it turns out. All 11 lagged behind the S&P 500, and six even managed to lose money in what’s generally been a buoyant market. “Overall, the 11 fully AI funds lost 1.8 percent per year on average, while the S&P 500 gave investors an average annual return of 7.6 percent,” the researchers wrote. (Remember, anyone can invest in the S&P 500 for a lower-risk stock experience, rendering the case for these AI-managed funds extremely dim at present.)
Where the tech is going wrong, Smith and Wyatt argue, is that AI can correlate data, but it can’t understand it.
“The Achilles’ heel of AI systems is that while they are unparalleled at finding statistical patterns, they have no way of judging whether the patterns they find are plausible or pointless,” they wrote. “Until AI algorithms understand what words mean and how they relate to the real world, they will continue to be unreliable for important decisions, including but not limited to investing.”
More on AI: The AI Bubble Is Bursting, Experts Say
