Selecting the Forgetting Factor in Subset Autoregressive Modelling

21 Pages Posted: 14 Feb 2003

See all articles by Tim Brailsford

Tim Brailsford

Bond University

Jack H.W. Penm

Australian National University - School of Finance and Applied Statistics, Faculty of Economics and Commerce

R. Deane Terrell

Australian National University (ANU) - National Graduate School of Management

Abstract

Conventional methods to determine the forgetting factors in autoregressive (AR) models are mostly based on arbitrary or personal choices. In this paper, we present two procedures which can be used to select the forgetting factor in subset AR modelling. The first procedure uses the bootstrap to determine the value of a fixed forgetting factor. The second procedure starts from this base and applies the time-recursive maximum likelihood estimation to a variable forgetting factor. In one illustration using real exchange rates, we demonstrate the effect of the forgetting factor in subset AR modelling on forecasting of non-stationary time series. In a second illustration, these two procedures are applied to time-update forecasts for a stock market index. Subset AR models not including a forgetting factor act as a set of benchmarks for assessing ex ante forecasting performance, and consistently improved forecasting performance is demonstrated for these proposed procedures. ex ante

Suggested Citation

Brailsford, Timothy John and Penm, Jack and Terrell, R. Deane, Selecting the Forgetting Factor in Subset Autoregressive Modelling. Available at SSRN: https://ssrn.com/abstract=370485

Timothy John Brailsford (Contact Author)

Bond University ( email )

Gold Coast, QLD 4229
Australia

HOME PAGE: http://www.bond.edu.au

Jack Penm

Australian National University - School of Finance and Applied Statistics, Faculty of Economics and Commerce ( email )

Canberra, Australian Capital Territory 0200
Australia
+61 (02) 61250535 (Phone)
+61 (02) 61250087 (Fax)

R. Deane Terrell

Australian National University (ANU) - National Graduate School of Management ( email )

Sir Roland Wilson Building (120)
Canberra, Australian Capital Territory 0200
Australia

Here is the Coronavirus
related research on SSRN

Paper statistics

Downloads
12
Abstract Views
1,051
PlumX Metrics