Sensex and Stylized Facts an Empirical

Abstract

Financial time series have been documented to exhibit various stylized facts such as non-normal return distribution, volatility clustering, leverage effect, regime switching etc. In this paper we attempt to characterize these stylized facts in the BSE Sensex. We observe that the Sensex is characterized by the presence of many of these effects. We estimate models like GARCH, Artificial Neural Networks and Mixture of Gaussians that accommodate these effects.

Keywords: Stylized facts, GARCH, Neural Network