Second-order properties of thresholded realized power variations of FJA additive processes (Q2330961)

The aim of the paper is to obtain sufficient conditions for the mean-squared-error (MSE) consistency of some threshould-based estimators for the jump component, the total number of jumpes and the \(k\)-th order variation of an additive process \(X\) with finite jump activity (FJA). The framework of the investigation as well as notations are introduced in the second section of the article. The finite-jump activity additive process and the FJA Lévy model are defined. The third section is devoted to the introduction of nonparametric estimators for the integrated volatility of a general Ito semimartingale. The authors introduce the bi-power variation estimator (BVP), the multi-power variation estimator (MPV) and the threshold power variation (TPV) estimator. Asymptotic properties of the estimators are shown in the next section. The authors examine and compare the rates of convergence of the bias, variance and mean-squared error for the TPV and MPV estimators. A CLT (central limit theorem) for TPV estimators is formulated followed by a short investigation of the performance of the jump component estimators. References are discussed all along the article. Proofs and supporting results are shown in the appendix.