Can artificial intelligence beat the stock market?

Abstract

PurposeAcademia and financial practitioners have mixed opinions about whether artificial intelligence (AI) can beat the stock market. The purpose of this paper is to investigate theoretically what would happen if AI has further evolved into a superior ability to predict the future more accurately than average investors.Design/methodology/approachA theoretical model in an endowment economy with two types of representative investors (traditional investors and AI investors) is proposed, and based on the model, a long-run survival analysis for both types of investors is implemented.FindingsThe model presented in this paper indicates that being equipped with a superior ability to predict the future more accurately than traditional investors cannot guarantee AI investors to always beat the stock market in the long run. Those investors may be extinct, all depending on the structure/parameters of the stock market.Originality/valueTo the best of the author’s knowledge, they are the first to set up a representative agent equilibrium model to explore the above question seriously.