Can artificial intelligence help pick stocks? More specifically, can investors use AI to determine the fair price of a stock or a bond? I bet a lot of people right now would say yes, given recent advances that allow for the processing of ever greater amounts of information.
I think my AI is better than all the other ones out there. My AI is the market.
For example, pick a stock. Check the price. Why is it that exact price? Because an equal number of buyers and sellers think they are getting a good deal when they sell or buy it at exactly that time. They make those judgments using every piece of information available to them, both public and private. The market is the world’s largest information processing machine, which creates a price for every publicly traded stock and bond.
These prices are set in an environment where no one knows what’s going to happen. So in that sense it is a giant model that is humanity’s best and constantly evolving guess of how each company stock or bond will perform.
Despite all the promise of AI, I prefer to accept market prices rather than prices from algorithms. Large language models, the types of AI that power tools such as ChatGPT, are intended to understand and generate text that seems as if it was made by humans, not predict future outcomes.
They can generate potential scenarios based on learnt patterns, but they struggle to account for unknown factors or real-world changes that come outside their training data. In that way, they are truly “artificial,” while markets are composed of real, human intelligence and the millions of judgments market participants make.
Sure, AI and algorithmic trading can help the execution of trades. But there’s no reason to think that AI should fundamentally influence the way people think about stock prices anytime soon.
The market is fantastically complex. So much so that no one knows exactly how much a particular piece of information impacts a price, because there are so many other simultaneous inputs. But the market ensures that a price is the most accurate current representation of the value of a stock or bond. It’s free and available to all. How great is that?
This isn’t just my opinion. There’s plenty of evidence to support it. In fact, it’s a 50-year-old theory that only gets more proven with each passing year. Google “efficient market hypothesis.” Better yet, ask ChatGPT to explain it.
Still don’t believe me? Then let me ask another question: Do you think you can hire a manager to implement the strategy of using AI to pick stocks that consistently beat the market? After fees, probably not. If they had some cool AI that actually did predict stock prices better than the market, why would they share the information with you?
What’s the takeaway? You can have a good experience without worrying about all that stuff. Based on nearly a century’s worth of data, the stock market has returned about 10 per cent a year, which is 7 per cent above inflation.1 That was true before and after computers, before and after the internet, and even before and after the second world war. It makes sense to me that it will continue to be the case after AI. Because our AI is “aggregate intelligence,” which includes artificial intelligence and betters it.
To be clear, I celebrate the innovation that this moment may represent. As I have witnessed over and over during the past 50 years of my career, many players will try to take advantage of the newest advance in technology to improve their company and also to build new ones. By buying the market, you can have a piece of all the publicly traded companies.
And if I still haven’t convinced you, I asked ChatGPT, “Is it safer to trust the market price mechanism than rely on an AI model to find mispricing in stocks and bonds?”
Here’s what I got back the day I asked: “It is generally safer to trust the market price mechanism than to rely on an AI model to find mispricings in stocks and bonds. The market price mechanism is based on the collective actions of all market participants and incorporates all available information into asset prices. As a result, it is difficult for any single investor or AI model to consistently outperform the market by identifying mispricings.”
So if you don’t trust me, trust the AI that is telling you not to trust the AI over the market.
- In US dollars. S&P 500 Index annual returns 1926–2021. S&P data © 2023 S&P Dow Jones Indices LLC, a division of S&P Global. All rights reserved. Indices are not available for direct investment; therefore, their performance does not reflect the expenses associated with the management of an actual portfolio.
The original version of this article has been reprinted with permission and was written by David Booth, Executive Chairman and Founder of Dimensional Funds Advisors.
This content is developed from sources believed to be providing accurate information and provided by Dimensional Fund Advisors. Please consult legal, finance or tax professionals for specific information regarding your individual situation. The opinions expressed and material provided are for general information and should not be considered a solicitation for the purchase or sale of any security.