Application of GARCH Models for Modeling Stock Market Volatility : An Empirical Study

Return is the major attribute of an investment asset which can be construed as a random variable, and the ‘variability in return' can be interpreted as volatility. Forecasting volatility and modeling it are the most prolific areas for research. This paper empirically investigates the conditional variance (volatility) pattern in Indian stock market based on financial time series data that consists of daily closing prices of CNX Nifty 50 market index for 10 years from April 2006 to March 2016. For the purpose of estimating conditional variance (volatility) in the daily returns of the index, Autoregressive Conditional Heteroskedasticity (ARCH) models are employed. Both symmetric and asymmetric models are used to capture stylized facts about CNX Nifty 50 market index returns such as volatility clustering and leverage effect. The findings of the study show that the asymmetric models are a better fit than symmetric models, confirming the presence of volatility clustering and leverage effect