Forecasting with Structural Change

by A. Thavaneswaran and M. Ghahramani.

Abstract: Many analyses of time series forecasting have been based on the assumptions of a constant, time-invariant, data generating process, that is stationary (no structural change), and coincides with the time series model used. A simple example is when the process has undergone a regime shift (structural change), forecasts based on past information need not be unbiased despite being the previous conditional expectation. In this paper we study improved estimates and forecasts for time series models with structural change in the mean as well as volatility.

Key Words: Structural Change, GARCH, Improved Estimate/Forecast

Authors:
A. Thavaneswaran, thavane@cc.umanitoba.ca
M. Ghahramani, umghahra@cc.umanitoba.ca

Editor: Shelton Peiris,shelton@maths.usyd.edu.au

READING THE ARTICLE: You can read the article in portable document (.pdf) format (133748 bytes.)

NOTE: The content of this article is the intellectual property of the authors, who retains all rights to future publication.

This page has been accessed 2638 times since July 24, 2006.


Return to the InterStat Home Page.