Value-at-Risk VaR estimation through volatility analysis is a regulatory requirement. Many asset management companies and bourses tend to use EWMA and GARCH based techniques towards this. This paper compares the predictive power of Stochastic Volatility Model SVM and Kalman Filter KF based approach vis-à-vis EWMA and GARCH based approaches with data from Indian security indices. A Quasi-Maximum Likelihood QML based on KF is used for estimation of parameters for the underlying state space SVM. It is found that, with a representative data set, VaR backtesting result from the SVM significantly outperforms the traditionally recommended EWMA based techniques.