Indian Stock Market Volatility Analysis Based on a GARCH-type Model

Abstract

The Indian stock market, being an emerging market, exhibits significant volatility. This paper employs the returns of the top three under Nifty50 companies, i) Reliance India Ltd., ii) HDFC Bank, and iii) Bharati Airtel as per the current market capital to do an analysis utilising an Autoregressive Conditional Heteroskedasticity (ARCH) model. The Autoregressive Moving Average (ARMA) model with a t-distribution for the sample series to evaluate model performance across various distributions and orders. Conversely, it introduced threshold GARCH models to encapsulate the characteristics of the index. Furthermore, the accuracy and predictive outcomes of the models were assessed using mean squared error (MSE), mean absolute error (MAE), and root mean squared error (RMSE). The findings indicate that the ARMA model utilising Student’s t-distribution surpasses alternative models in forecasting the return series of Reliance ARMA (7,7), HDFC is ARMA (1,1), and Bharti Airtel is ARMA (1,1). The GARCH model exhibits superior predictive performance for the return series of the Reliance GARCH (2,1), the HDFC is GARCH (1,1), and the Bharti Airtel is GARCH (2,1), compared to all other models. This study may serve as a valuable informational resource for governmental macro decision-making, the operations of publicly traded enterprises, and investors' investment strategies.