Title: Industry versus country effects in stock returns: the significance of value-weighted versus equal-weighted estimation

Authors: Jianguo Chen; Martin Young; Andrea Bennett

Addresses: Department of Economics and Finance, Massey University, Private Bag 11-222, Palmerston North, New Zealand ' Department of Economics and Finance, Massey University, Private Bag 11-222, Palmerston North, New Zealand ' Department of Economics and Finance, Massey University, Private Bag 11-222, Palmerston North, New Zealand

Abstract: Using European market data from 1980 to 2007, we re-examine the commonly used Heston and Rouwenhorst methodology for determining the relative effects of industry and country factors in portfolio performance. The whole estimation involves two sets of weighting method; one is in the restrictive conditions and the other in the regression process. We explain that the former does not change the nature of the regression; it is the latter that determines the nature of the regression result (value-weighted vs. equal-weighted estimation). When using the equal-weighted regression, our results show the country effect dominating until the late 1990s and from then the industry effect becomes more important. However, when using the value-weighted method the industry effect has dominated over the entire period. Previous studies claim a value weighted method and report the same as our equal-weight result. We believe this is caused by confusing the two weighting methods. This study clarifies this issue.

Keywords: industry effects; country effects; equal-weighted estimation; value-weighted estimation; stock returns; Heston and Rouwenhorst; portfolio performance.

DOI: 10.1504/IJPAM.2013.054411

International Journal of Portfolio Analysis and Management, 2013 Vol.1 No.3, pp.278 - 287

Received: 23 Feb 2012
Accepted: 04 Jan 2013

Published online: 05 Jul 2014 *

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