A1563
Title: One-step estimation for general Gaussian processes
Authors: weilin Xiao - Zhejiang University (China) [presenting]
Abstract: The fractional Gaussian process has been widely applied in many fields, and statistical inference for it poses significant challenges due to the lack of the Markov property. A method for estimating unknown parameters in general Gaussian processes based on one-step estimation is considered, which corrects an initial guess estimator to an asymptotically efficient one. Under certain regularity conditions, the one-step estimator of general Gaussian processes is shown to be strongly consistent and asymptotically normal with the same asymptotic variance as the maximum likelihood estimator. Monte Carlo simulations show that the one-step estimator reduces computational time compared with the exact maximum likelihood estimator and outperforms method of moments estimators in terms of asymptotic variance. The estimation methods are applied to fit fractional Gaussian processes to logarithmic daily realized volatility series, where the series display the roughness phenomenon and the stationary property.
