Do fat tails matter in GARCH estimation? Testing market efficiency in two transition economies

Harrison, B. ORCID: 0000-0003-2567-9446 and Paton, D., 2007. Do fat tails matter in GARCH estimation? Testing market efficiency in two transition economies. Economic Issues, 12 (2), pp. 15-26.

Preview

Text 200647_7045 Harrison Publisher.pdf Download (107kB) | Preview

Abstract

The use of the GARCH-class of models is commonplace when examining stockmarket returns. In this paper we use data on stock markets in two transitioneconomies, the Czech Republic and Romania, to demonstrate the importance ofusing the correct GARCH specification. When residuals are characterised by ‘fattails’ or kurtosis, the use of a GARCH-t specification is appropriate. Diagnostictests suggest that the GARCH-t specification is appropriate for modelling stockmarket returns in Romania, whilst the standard GARCH specification is adequatefor the Czech Republic. Using a standard GARCH specification leads torejection of the null hypothesis of market efficiency in Romania, whereas thisnull hypothesis cannot be rejected using the GARCH-t specification. The nullhypothesis of efficiency cannot be rejected in the Czech Republic using eitherspecification. Thus, we find that the presence of ‘fat tails’ can have importantimplications for inference in the analysis of stock market returns.

Item Type:	Journal article
Publication Title:	Economic Issues
Creators:	Harrison, B. and Paton, D.
Publisher:	Economic Issues Education Fund
Date:	2007
Volume:	12
Number:	2
Divisions:	Schools > Nottingham Business School
Record created by:	EPrints Services
Date Added:	09 Oct 2015 09:59
Last Modified:	09 Jun 2017 13:16
URI:	https://irep.ntu.ac.uk/id/eprint/5945

Actions (login required)

Edit View

Views

Downloads