USE OF ARTIFICIAL INTELLIGENCE ALGORITHMS FOR ASSET PRICE PREDICTION IN G7 CAPITAL MARKETS

Abstract

This article presents the results of a quantitative and descriptive study that aimed to perform a comparative analysis of the performance of artificial intelligence (AI) algorithms applied to predicting the behavior of financial assets. The data analyzed were collected from capital markets that make up indices of developed countries that comprise the Group of Seven (G7) between the years 2001 and 2023. To generate the models in the research, the following AI algorithms were used: Random Forest (RF), Naive Bayes (NB), and K-Nearest Neighbors (KNN). The analysis of the study results was based on descriptive statistics and Shapiro-Wilk, Student's t-test, and Mann-Whitney tests. The results indicated the robustness and consistency of the RF model in predicting the behavior of the indices. Furthermore, the work confirms technical indicators as relevant inputs for AI models to predict the prices of financial assets in the capital market. Among the main theoretical contributions of this study, the indication that the Efficient Market Hypothesis (EMH) can be contingent on the time horizon and the type of data analyzed stands out. In empirical terms, this work shows that AI-based models can perform above the expected average based on the weak form of the EMH.