Another Look at Some Information Order Selection Criteria in Autoregressive Processes

Abstract

In this paper, the selection procedure of the order of an autoregressive process of order P, AR(p) using the information order selection criteria is
considered. We compare the performance of three information criteria methods, the Akaike information criterion (AIC), Schwarz information criterion (BIC) and Hannan & Quinn information criterion (HQC). It is observed that the ability of a criterion to select the true order may depend on the coefficient(s) of the process. A special AR(2) process where these three criterion appears to select the true order with equal probability is also identified. A Monte-Carlo experiment is used to demonstrate the procedure.

Keywords: AIC; SIC; HQC