Machine Learning Forecasting of Stock Market Indices Based on Macroeconomic Indicators

Abstract

The accurate forecasting of stock market indices remains a significant challenge in financial research, given the complex interplay of economic variables and market sentiment. This study investigates the predictive power of macroeconomic indicators on major stock market indices using advanced machine learning (ML) techniques. The primary objective is to develop a robust forecasting model that leverages historical data on key macroeconomic variables—such as inflation rates, interest rates, GDP growth, and unemployment figures—to predict future index movements. Using a dataset spanning two decades, we trained and evaluated multiple ML algorithms, including Random Forest, Gradient Boosting, and Long Short-Term Memory (LSTM) networks. The results demonstrate that ensemble methods, particularly Gradient Boosting, achieve superior predictive accuracy compared to traditional time-series models. Notably, inflation and interest rates emerged as the most influential predictors. These findings underscore the potential of ML-driven approaches to enhance financial decision-making and risk management strategies for investors and policymakers.