The Global Preference for Dividends in Declining Markets
DOI | http://doi.org/10.1111/fire.12078 |
Published date | 01 November 2015 |
Date | 01 November 2015 |
The Financial Review 50 (2015) 575–609
The Global Preference for Dividends
in Declining Markets
Michael A. Goldstein
Babson College
Abhinav Goyal∗
University of Liverpool
Brian M. Lucey
Trinity CollegeDublin, University of Ljubljana
Cal B. Muckley
University College Dublin
Abstract
Investors globally prefer dividend-paying stocks over nondividend-paying stocks more
in declining than in advancing markets, even accounting for firm-level growth opportunities,
size and risk effects. Dividend-paying stocks outperform nondividend-paying stocks, from
0.63% (China) to 3.79% (Canada) more per month in declining than in advancing markets. In
declining markets, dividend-paying firms outperform by more than any underperformance in
advancingmarkets. The results are robust across dividend taxation regimes, legal environments,
∗Corresponding author: G-05, Management School, Chatham Building, University of Liverpool, Liver-
pool, L69 7ZH, UK; Phone: +44-151-795-2150; Fax: +44-151-795-3001; E-mail: agoyal@liv.ac.uk.
The authors would like to thank Robert VanNess for suggestions, and the participants in Financial Man-
agement Association annual meeting 2013 in Chicago (October: 2013) and Auburn University (January:
2015). Cal Muckley is thankful to the Fulbright Commission for financial support and for the hospitality
shown by colleagues at the finance group in the YaleUniversity School of Management, where part of this
research was undertaken. We are grateful to the authors’ respectiveinstitutions for financial support. We
are responsible for all the remaining errors and the usual caveat applies.
C2015 The Eastern Finance Association 575
576 M. A. Goldstein et al./The Financial Review 50 (2015) 575–609
emerging and developed markets, periods prior to and after the 2008 global financial crisis,
the exclusion of the dividend declaration month and in respect to segmented or integrated
international capital markets.
Keywords: dividend policy, market movement, G-7, BRICS
JEL Classifications:F3,G35
1. Introduction
Around the globe, some corporations send returns to their shareholders via
dividends, while others do not. These corporations exist in different countries, with
different economic systems at different levels of development, different taxation
systems, different legal systems and protections, and so on. Investors (whether local
or global) in stock markets around the globe therefore have a choice of investing
in corporations from which they will receive a dividend and those from which they
will not. While the question of the role and importance of dividends has generated
hundreds of papers over the decades since Lintner’s (1956) seminal work, much
of this research has been on the corporate, institutional and legal determinants of
particular payout policies, the effects of changes in policies on future asset returns,
or on the managerial elements of payout policy setting. Much less work has been
undertaken on whether dividends are preferred under different market conditions. If
we take the Modigliani-Miller propositions as holding, then regardless of whether
markets are advancing or declining dividends should not matter to the returns of
stocks.1Recent research by Fuller and Goldstein (2011) suggests that dividends do
matter, and matter more in declining (as opposed to advancing) markets, although
those results were only examined in one large developed country (the United States),
and therefore only under one economic and legal system.
In this paper, we evaluate dividend and nondividend-paying stocks across the
market cycle around the globe, across a variety of developedand emerging economies
with different legal systems and tax regimes. We find substantial international evi-
dence that investors differentially prefer dividend-paying stocks over nondividend-
paying stocks, more in declining markets than in advancing markets. This finding
is robust across markets at different levels of economic development, different legal
environments, and different dividend taxation regimes. These results therefore have
implications for the relative importance that investors place on dividends under dif-
ferent market conditions regardless of the tax policies, legal systems, or economic
development of the markets.
1Even following for the incompleteness of the Modigliani-Miller propositions, as demonstrated by DeAn-
gelo and DeAngelo (2006), dividends matter if only for the ability of dividend-paying stocks to destroy
wealth via suboptimal payout policies.
M. A. Goldstein et al./The Financial Review 50 (2015) 575–609 577
We look at dividend-paying and nondividend-paying listed firms in seven de-
veloped (Canada, France, Germany,Italy, Japan, the United Kingdom and the United
States) and five developing markets (Brazil, China, India, Russia and South Africa),
from January 1995 to December 2011. This sample includes a variety of legal, eco-
nomic, and tax systems across periods that include boom and bust cycles, including
the recent global Great Recession and recovery period. Examining raw returns (with-
out adjustments for risk) across all 12 countries, we find that dividend-paying stocks
outperform nondividend-paying stocks.2When just examining raw returns when the
local market index is declining, we also find that dividend-paying stocks do better
than nondividend-paying stocks in declining markets, although this effect is stronger
in the developed markets. When just examining raw returns when the local market
index is advancing, we find that for nine of the 12 markets nondividend-payingstocks
statistically do better than dividend-paying stocks in advancing markets. However,
notably, for all 12 countries we find that the outperformance of dividend-paying
stocks over nondividend-paying stocks is larger in declining markets than in advanc-
ing markets. This effect is stronger in the developed marketsand in South Africa than
it is in Brazil, India, and China, but it is statistically significant in all countries.
Fama-MacBeth (1973) regressions controlling for (local market) beta, size,
and book value confirm these results, with the United States showing a 3.25% per
month outperformance of dividend-paying firms in declining markets overadvancing
markets, followed by Russia at 3.10% and Canada at 3.07%,3while the lowest results
were 1.40% for the United Kingdom and 1.48% for India. The outperformance for
dividend-paying stocks over nondividend-paying stocks in declining over advancing
markets held both for the 1995–2007 pre-Crisis period and the 2008–2011 Great
Recession and recovery period for both developed and emerging nations. The effects
were, interestingly, slightly weaker during the more recent period for the developed
nations, but not notably different for the emerging markets (except for Brazil, where
the effect was slightly stronger in the recent period). The results also hold if we
instead use the MSCI All Country WeightedIndex (MSCI ACWI) instead of the local
country index to determine advancing or declining markets. While we find the results
hold for both developed and emerging economies, we find that the outperformance
of dividend-paying stocks in declining versus advancing markets over nondividend-
paying stocks are even stronger in emergingeconomies than in the developed markets.
Likewise, while we find that the results hold for countries with Common Law legal
systems (Canada, India, South Africa, the United Kingdom, and the United States)
as compared with Civil Law legal systems (Brazil, China, France, Germany, Italy,
Japan, and Russia), the results are slightly larger in the Civil Law countries. Other
2Interestingly, the overall stock return for nondividend-paying stocks was negative for all developed
countries, except for the United States.
3For the Fama-MacBeth regressions, all markets (including the United States) showed outperformance of
dividend-paying stocks over nondividend-paying stocks in both advancingand declining markets.
To continue reading
Request your trial