A new model for periodically correlated process with conditional heteroscedasticity

Ramezani, Rohollah; Rezakhah VARNOUSEFADERNAI, Saeid; Farhadi, Majid

doi:10.22055/jamm.2018.20658.1377

A new model for periodically correlated process with conditional heteroscedasticity

Document Type : Original Paper

Authors

¹ Department of Statistic, &lrm;School of&lrm; &lrm;Mathematics and Computer Science&lrm;, &lrm;Damghan &lrm;University, Damghan, Semnan, Iran

² Faculty of Mathematics and Computer Science,. Amirkabir University of technology

10.22055/jamm.2018.20658.1377

Abstract

In this paper, we study LARCH processes with periodic structure as a new class of time series with periodic conditional heteroscedasticity and long memory property. We characterize the structure of inter and intra season correlations.
Under the proposed assumptions, the long memory property for each season is studied too. Finally, by simulation study the efficiency of the R/S estimator for estimating long memory parameter of each season is shown.

Keywords

Main Subjects