ABSTRACT
This study investigates the sectoral impact of Foreign Direct Investment (FDI) on India's GDP growth by integrating econometric models with machine learning techniques to enhance predictive accuracy and identify key contributors. Results reveal a strong positive correlation (+0.78) between FDI and GDP, with regression confirming a significant positive effect (coefficient = 0.65, p = 0.001), while inflation and interest rate trends negatively influenced growth. Machine learning outperformed traditional econometric approaches, with XGBoost achieving the best results (R² = 0.88, RMSE = 170B USD), effectively capturing nonlinear, sectoral dependencies. Sectoral analysis highlighted services (0.35) and IT & telecom (0.28) as the most influential, with services FDI generating +$3.5B GDP per $1B invested. Comparative forecasts showed XGBoost predicting higher GDP (USD 4.48T by 2028) relative to ARIMA (USD 4.36T), reflecting sectoral investment linkages. GARCH volatility analysis indicated moderate fluctuations (6.5–7.2%) in FDI inflows, suggesting relative stability despite macroeconomic noise. Overall, the study demonstrates how combining econometric precision with machine learning flexibility advances understanding of sectoral FDI impacts, nonlinear spillovers, and economic growth forecasting in India.
