Nonparametric Estimation of Volatility and Its Parametric Analogs

Abstract

This paper suggests a nonparametric method for stochastic volatility estimation and its comparison with other widespread econometric algorithms. A major advantage of this approach is that the volatility can be estimated even in the case of its completely unknown probability distribution. As demonstrated below, the new method has better characteristics against the popular parametric algorithms based on the GARCH model and Kalman filter.