Efficient estimation of drift parameters in stochastic volatility models (Q2463719)

The aim of this paper is to estimate unknown parameters in the framework of the continuous time stochastic volatility models introduced by \textit{J. Hull} and \textit{A. White} [J. Finance, 42, 281--300 (1987)]. The authors construct the contrast function from the discrete observations. The idea is that from the high frequency observations it is known how to reconstruct the integrated volatility over intervals of size larger than some positive length. The main result states that, under suitable conditions on the quality of the reconstruction of the integrated volatility, the proposed estimator of the drift parameter is consistent and asymptotically Gaussian. The estimator has the same asymptotic behavior as the maximum likelihood estimator based on the direct observation of the volatility. Also, it is asymptotically efficient.