Long-range forecasting of the S&P stock market index using fractional integration techniques

Caporale, Guglielmo Maria and Gil-Alana, Luis A. (2007) Long-range forecasting of the S&P stock market index using fractional integration techniques. Centre for International Capital Markets discussion papers, 2007 (04). pp. 1-16. ISSN 1749-3412

[img]
Preview
Text
CentreForInternationalCapitalMarketsDiscussionPapers_2007-04_p01-16.pdf - Published Version

Download (229kB) | Preview
Official URL: http://www.londonmet.ac.uk/media/london-metropolit...

Abstract

In this paper we examine the stochastic behaviour of the S&P 5000 stock market index by means of fractional integration techniques. Specifically, we use a parametric method to test I(d) statistical models. Model selection criteria based on out-of-sample forecasting performance suggest that that best model specification is an I(d) process with d higher than 1, implying that the series under examination is nonstationary and non-mean-reverting.

Item Type: Article
Uncontrolled Keywords: Centre for International Capital Markets discussion papers; CICM discussion papers; fractional integration; long memory; long-range prediction
Subjects: 300 Social sciences > 330 Economics
Department: Guildhall School of Business and Law
Depositing User: Mary Burslem
Date Deposited: 23 Apr 2015 09:05
Last Modified: 23 Apr 2015 09:05
URI: http://repository.londonmet.ac.uk/id/eprint/489

Actions (login required)

View Item View Item