Regression, ARIMA and ARIMAX Models to Study the Factors affecting Foreign Direct Investment in India

Abstract

The paper describes Regression, ARIMA and ARIMAX analyses on GDP, inflation, exchange rate, export, import, energy generation and trade balance to estimate the foreign direct investment (FDI) in India. Specifically, we compared the results achieved from the fitted models viz., ARIMA and ARIMA with explanatory variable(s). The main emphasis was to see whether the ARIMA model including other time series as input variables helps in improving the forecasting performance. For this empirical study, we found that ARIMA(1, 1, 0) model with GDP as explanatory variable outperformed the Regression/ARIMA models for estimating the value of FDI in india.