Estimation of the volatility diffusion coefficient for a stochastic volatility model

We study, in the stochastic volatility model introduced by Hull and White [6], the estimation of the diffusion coefficient for the volatility process. The model is discretely observed on a fixed length time interval and no ergodicity assumption is needed for the volatility process. We construct an estimator show its consistency and establish that its rate of convergence is N-1/4 (N is the number of observations). (C) 2000 Academie des sciences/Editions scientifiques et medicales Elsevier SAS.