Measuring Conditional Volatility using GARCH (2, 2) Model from Empirical Standpoint

Abstract

The article covers the broader aspect of time-variant volatility function using GARCH (2, 2) vo-latility function. The postulate of conditional heteroskedasticity and autoregressive in relation to residual distribution is examined for four listed daily stock prices from Indian markets namely Infosys, TCS, HDFC and Tata Motors. The basic aim of the paper is to understand how the GARCH (2, 2) model perform as against the unconditional volatility and whether any volatility clusters appear in the price series. The overall idea of using a GARCH (2, 2) is to observe the lag orders for 2 days across conditional mean (first order) and conditional volatility (second order) autocorrelation effects across four prominent stocks. Also, to observe the static parameters and their role in defining the conditional elasticity of errors in the univariate setup. The results clearly highlighted that volatility clustering was observed at different time period defining the role of macroeconomic information and price adjustments with mean-reversion effects.