Title: Measuring the forecasting accuracy of models: evidence from industrialised countries

Authors: Athanasios Koulakiotis, Apostolos Dasilas

Addresses: Department of International and European Studies, University of Macedonia, 156 Egnatia Str., P.O. Box 1591, Thessaloniki, 54006, Greece. ' Department of Accounting and Finance, University of Macedonia, 156 Egnatia Str., P.O. Box 1591, Thessaloniki, 54006, Greece

Abstract: This paper uses the approach suggested by Akrigay (1989), Tse and Tung (1992) and Dimson and Marsh (1990) to examine the forecasting accuracy of stock price index models for industrialised markets. The focus of this paper is to compare the Mean Absolute Percentage Error (MAPE) of three models, that is, the Random Walk model, the Single Exponential Smoothing model and the Conditional Heteroskedastic model with the MAPE of the benchmark Naive Forecast 1 case. We do not evidence that a single model to provide better forecasting accuracy results compared to other models.

Keywords: forecasting accuracy; stock price returns; industrialised countries; random walk; stock price index models; single exponential smoothing; conditional heteroskedastic.

DOI: 10.1504/IJMEF.2009.023065

International Journal of Monetary Economics and Finance, 2009 Vol.2 No.1, pp.44 - 57

Published online: 08 Feb 2009 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article