Macroeconomic Determinants of Stock Market Performance in India and Selected Global Economies: An Econometric and Machine Learning Analysis

Abstract

Stock markets are increasingly influenced by macroeconomic fluctuations and global financial integration, particularly in emerging economies where financial systems are highly sensitive to economic shocks and external uncertainties. This study aims to examine the effects of macroeconomic determinants on stock market performance in India and selected global economies using econometric and machine learning approaches. Monthly and quarterly data covering the period 2010–2024 were collected from international financial and economic databases and analyzed using correlation analysis, multiple regression, Augmented Dickey–Fuller and Phillips–Perron unit root tests, Johansen cointegration, Vector Autoregression, Vector Error Correction Modeling, Granger causality analysis, Random Forest, XGBoost, Long Short-Term Memory, and a hybrid XGBoost–LSTM framework. The findings reveal that gross domestic product growth has a strong positive relationship with stock market performance (r=0.72), whereas inflation (r=–0.48), interest rates (r=–0.55), exchange rates, and unemployment exert negative effects on market returns. Regression results indicate that GDP growth is the most influential positive determinant (β=0.68, p<0.01), while interest rates and unemployment significantly reduce market performance. Among the forecasting models, the hybrid XGBoost–LSTM model achieved the highest predictive accuracy with an RMSE of 121.35, MAE of 98.42, and R² value of 0.93. The study concludes that macroeconomic stability plays a critical role in enhancing stock market performance and that hybrid machine learning frameworks provide superior forecasting capability compared with conventional econometric approaches, offering valuable implications for investors, portfolio managers, and policymakers.