Resurrecting the size effect: Evidence from a panel nonlinear cointegration model for the G7 stock markets

• Published date: 01 January 2014
• Date: 01 January 2014
• Author: Nicholas Apergis,James E. Payne
• DOI: http://doi.org/10.1016/j.rfe.2013.08.003

Resurrecting the size effect: Evidence from a panel nonlinear

cointegration model for the G7 stock markets

Nicholas Apergis

a

, James E. Payne

b,

⁎

a

Departmentof Banking and FinancialManagement, Universityof Piraeus, Greece

b

Departmentof Economics and Finance,University of NewOrleans, 2011 AdministrationBuilding, 2000 LakeshoreDrive, New Orleans,LA 70148, United States

abstractarticle info

Articlehistory:

Received6 February 2013

Receivedin revised form 8 August 2013

Accepted13 August 2013

Availableonline 27 August 2013

JEL classification:

G10

C23

Keywords:

Size effect

Stockreturns

Panel thresholdcointegration

G7 stockmarkets

Firm size is known to be an important factor affecting stock returns. This study proposes a panel thresho ld

cointegration model to investigate the impact of the size effect on stock returns for the panel of G7 countries:

Canada, France, Germany, Italy, Japan, the U.K., and the U.S. over the period 1991:1–2012:12. The empirical analysis

is based upon the nonlinearcointegration framework using the asymmetric ARDL cointegration methodology

(Shin et al., 2011). This methodological approach permits a much richer degree of flexibility in the dynamic

adjustment processtoward equilibrium, than in the classical linearmodel. Our findings indicate the presence

of asymmetricadjustment around a unique long-runequilibrium. In particular, the empirical analysis provides

evidence of asymmetric effects between stock returns and the size effect, while controlling for the book-to-

market ratioand the price-to-earnings ratio.

© 2013 ElsevierInc. All rights reserved.

1. Introduction

The observation that small capitalized stocks tend to outperform

large capitalized stocks, known as the size effect (‘anomaly’), has been

well documented in the stock market literature (Banz, 1981; Berk,

1995; Fama and French, 1992). There are essentially two views with re-

spect to the size effect. One view argues, based on the concept of rational

asset pricing, that the stock market is characterized as an efficient market

in the pricing of assets with the differences in average returns attributed

primarily to differences in risk. The idea is that the higher returns on

small firms' stocks could be due to a greater exposure to an underlying

risk factor not incorporated in standard asset pricing models. This leads

firms to compute their cost of equity capital on the basis of an asset pric-

ing model that explicitly considers this source of risk (Fama and French,

2004). Another view argues, based on the behavioral finance literature,

that the market is characterized by the absence of rational investors

who usually drive prices to equilibrium. In that sense, markets are

dominated by naïve investors who just follow market trends or irratio-

nally extrapolate past information into the future (Lakonishok et al.,

1994). Furthermore, Whited (1992),Fazzari and Petersen (1993),and

Weinberg (1994) argue that extensive asymmetric information prob-

lems play a more substantial role for small firms due to liquidity con-

straints, bank loan access, cash flow management, dividend behavior,

and leverage volumes. As Hong et al. (2000) note, the asymmetric infor-

mation that restricts information flows for smaller firms contributes to

the size effect.

The empirical analysis surrounding the debate on the size effect has

primarily focused on the assumption that the size effect exhibits a linear

and symmetric adjustment process. However, Hong et al. (2007) present

evidence to suggest that large firms tend to have symmetric up and

down movements with the market, while small firms are riskier and

more sensitive to market downturns. Moreover, research on the nonlin-

earities in stock returns has emphasized whether such nonlinearities

actually exist without attempting to associate them with a number of

stylized facts within the finance literature, such as the size effect, the

calendar effect, and the contemporaneous volume/volatility effect

(Ammermann and Patterson, 2003; Hinich and Patterson, 1995;

Patterson and Ashley, 2000). Indeed, ignoring the asymmetric adjust-

ment of the size effect may lead to biased inferences and misleading

conclusions. This study addresses this underlying assumption by investi-

gating the nonlinear relationship between stock returns and the size ef-

fect for a panel of G7 countries through the use of the recently advanced

asymmetric ARDL panel cointegration model.

1

While the literature considers the sizeeffect more prevalent in de-

veloping markets, we focus our analysis on matur e markets to

strengthen the case fo r asymmetries with res pect to the size effect

Reviewof Financial Economics 23 (2014)46–53

⁎Correspondingauthor. Tel.:+ 44 11504 280 7199.

E-mailaddresses: napergis@unipi.gr(N. Apergis),jepayne@uno.edu (J.E.Payne).

1

The use of panelcointegration modelingcircumvents the endogeneitybias of relying

upon OLS estimation.

1058-3300/$–see front matter © 2013 ElsevierInc. All rights reserved.

http://dx.doi.org/10.1016/j.rfe.2013.08.003

Contents listsavailable at ScienceDirect

Review of Financial Economics

journal homepage: www.elsevier.com/locate/rfe