[source image]
Dow Jones NewsJul 2, 7:39 PM UTC
MW How hard is it to beat the market consistently? Even AI can't do it.
By Mark Hulbert
Many ETFs and stock pros use artificial intelligence to get an edge. But most still don't outperform their benchmark.
Artificial intelligence is an amazingly powerful tool for investing - but it's no match for human-run investment management.
Here's why: If one investment firm was the only one of its Wall Street peers utilizing AI, it's conceivable that it could beat the market. But when nearly all firms are using the technology, as is becoming the case, then the arithmetic of active management guarantees that the average manager will lag the market.
The number of firms employing AI has mushroomed, according to a survey conducted last year by fintech firm Broadridge Financial Solutions. Among surveyed firms' executives, 2 in 3 indicated that they "use [generative AI] most for investment or market research."
My reference to "The Arithmetic of Active Management" is to a famous article by that name by William Sharpe, the 1990 Nobel laureate in economics. In that article, he showed why active managers, on average, must lag broad market indexes - and that this conclusion depends "only on the laws of addition, subtraction, multiplication and division. Nothing else is required." That's because beating the market is a zero-sum game before transaction costs, and a negative-sum game after transaction costs.
The data support Sharpe's argument. Consider first the performance of the Eurekahedge AI Hedge Fund Index, which is "designed to provide a broad measure of the performance of underlying hedge-fund managers who utilize artificial intelligence and machine-learning theory in their trading processes." The index stopped being updated in January of this year - but at that point it was behind the S&P 500 SPX by 23 percentage points for trailing one-year return, 11 percentage points annualized for trailing five-year return, and 8 percentage points annualized over 10 years.
More data come from actively managed U.S. large-cap mutual funds and ETFs. If AI could help managers beat the market, we'd expect that the percentage of these funds doing so would have grown in recent years as Wall Street has increasingly embraced the technology.
Nevertheless, as you can see from the chart above, this expanded use of AI has not led the industry to perform any better. In 2024, for example, the percentage of funds beating the market was slightly below the average since 2001 - 35%, versus an average of 36%.
Still more data pointing toward the same conclusion come from the performance of the handful of funds that, in addition to using AI, bill themselves as basing their strategies on AI. Only a minority of these funds have beaten the S&P 500, as you can see from the table below.
This table is not an exhaustive list of U.S. equity funds that market themselves as basing their strategies on AI. But the list contains enough funds to make my point that AI is not a surefire path to beating the market.
Is AI creative?
The key to understanding AI's potential is whether it is capable of truly creative thought. In other words, can it come up with investment approaches that have never been thought of or even imagined?
In an interview, Lawrence Tint allowed for the possibility that some AI-informed managers will be able to beat the market in the future. Tint is the former U.S. chief executive of BGI, the organization that created iShares (now part of BlackRock (BLK)), and a colleague of Sharpe's when he wrote the article on the arithmetic of active management.
If AI is capable of being truly creative, Tint said, then its potential for beating the market actually will be limited. That's because most large Wall Street firms will discover the approaches which are temporarily working for this or that individual manager, and that manager's advantage eventually will be competed away.
At that point, the only hope of an AI-employing firm will be staying at least one step ahead of other AI-using firms. That could be a slim hope since, with the increasing speed and power of AI systems, any advantage wouldn't last very long. Tint said that one possible end result, assuming AI is capable of creativity, will be that it will recommend that we all invest in low-cost index funds.
If AI is not capable of creativity, Tint continued, but instead is essentially a powerful search tool for what is already known, then it is possible that - in the hands of a truly creative genius - AI could be used to beat the market. But ironically, in that event, it won't be AI beating the market but the creative genius who knows how to use AI "to better implement the concepts that the genius came up with, and to find ways to shield the concepts from other AI firms," Tint added.
Note that the question Tint is asking about creativity is different than the question of whether AI has human intelligence, or whether AI has surpassed human intelligence - as OpenAI founder Sam Altman suggests may have already occurred. The "intelligence" that some have in mind when comparing AI to human capacities is the efficient synthesis of what's already known. That's different than the genuine creativity needed to come up with something totally new.
The bottom line? Powerful as AI is, using it does not and cannot guarantee that you will beat the market. Keep your expectations in check - beating the market will always be difficult.
Mark Hulbert is a regular contributor to MarketWatch. His Hulbert Ratings tracks investment newsletters that pay a flat fee to be audited. He can be reached at mark@hulbertratings.com.
More: There's a 68% chance that stocks will rise by year-end. Do you feel lucky?
Also read: Stablecoins are setting the stage for the next financial meltdown
-Mark Hulbert
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
07-02-25 1539ET
Copyright © 2025 Dow Jones & Company, Inc.
