Parameter estimation for fractional Ornstein-Uhlenbeck processes of general Hurst parameter (Q2417989)

The Ornstein-Uhlenbeck process is defined as the solution of the stochastic differential equation \[ X_t=X_0-\theta\int\limits_{0}^{t}X_sds+\sigma d{B_t}^ H, \] where $X_0\in R$ is the initial condition, $B^H$ is a fractional Brownian motion of Hurst parameter $H\in(0,1)$, and $\theta$ and $\sigma$ are positive parameters to be estimated. The purpose of the article is to estimate the parameter $\theta$ for any $H\in[0,1)$. A least squares estimator ${\hat {\theta}}_T$ and an estimator ${\tilde {\theta}}_T$ based on the ergodic theorem have already been considered in a previous paper by the first two authors [Stat. Probab. Lett. 80, No. 11--12, 1030--1038 (2010; Zbl 1187.62137)], where the almost sure convergence and a central limit theorem have been proved for $H\in [1/2,3/4)$. In the present article, a central limit theorem for the least squares estimator in the case $H\in(0,3/4]$ and a noncentral limit theorem in the case $H\in(3/4,1)$ are proved. Malliavin calculus is used. Next, an asymptotic law for the ergodic-type estimator is obtained. These two estimators are compared with the maximum likelihood estimator by computing their asymptotic variance. Simulations with an estimator defined for discrete data and using the form of the ergodic-type estimator are performed. Technical details used in proofs are contained in the Appendix section.