Semiparametric estimation of volatility functions of diffusion processes from discretely observed data

Abstract

This paper provides a semiparametric model of estimating states of the volatility defined as the squared diffusion coefficient of a stochastic differential equation. Without assuming any functional form of the volatility function, we consider the volatility state as an unobservable state in a state space model and estimate it by filtering. By using the estimates, we can draw the information on what kind of functions the volatility has behind the observable state of the process. We also prove the consistency of the model in the sense that estimated states converges to the true ones as the observation time interval goes to zero. In addition to that, from a numerical point of views, we carry out numerical experiments by examples of stochastic differential equations with linear/nonlinear volatility functions in order to check whether or not the model can actually estimate the volatility and capture the information on its functional form.