In an engaging piece by David Booth, the founder and chair of Dimensional Fund Advisors, a critical question is raised – can artificial intelligence (AI) help pick stocks? With the advent of technology capable of processing vast amounts of data, many might be inclined to say yes. However, Booth offers an alternative perspective, suggesting that the market itself is the ultimate AI1.
Booth invites us to consider a simple scenario: pick a stock and check its price. That exact price is the result of an equal number of buyers and sellers deeming it a fair deal based on all available information, both public and private. In other words, the market is the world’s largest information processing machine, capable of creating a price for every publicly traded stock and bond1.
The stock market is an ever-evolving guessing game, where no one can predict what’s going to happen next. Thus, each stock or bond’s price represents humanity’s best current guess of its performance1.
Despite the promise of AI, Booth suggests that it’s better to accept market prices rather than those determined by algorithms. While large language models, like ChatGPT, can understand and generate human-like text, they are not designed to predict future outcomes. They can generate potential scenarios based on patterns they’ve learned but struggle to account for unknown factors or real-world changes that occur outside their training data. In contrast, markets encompass real human intelligence and the judgments of millions of market participants1.
AI and algorithmic trading can aid in executing trades, but Booth argues that there’s no reason to believe that AI should fundamentally influence how we perceive stock prices in the near future1.
Highlighting the complex nature of the market, Booth emphasizes that no one can pinpoint how a particular piece of information influences a price, given the multitude of simultaneous inputs. However, he asserts that the market ensures that a price is the most accurate current representation of a stock or bond’s value. This perspective aligns with the Efficient Market Hypothesis, a theory that has been gaining support over the past 50 years1.
Booth also raises a valid point about the feasibility of hiring a manager to use AI for stock picking. If they had an AI that could predict stock prices better than the market, why would they share this information? More often than not, after considering fees, the chances of consistently beating the market using AI are slim1.
In conclusion, Booth encourages investors to embrace the simplicity of the market’s wisdom. Based on nearly a century’s worth of data, the stock market has returned about 10% a year, which is 7% above inflation. This holds true regardless of advancements in technology, before and after computers, and before and after the advent of AI1.
If you’d like to read the article yourself, here is the link.