Do foreign stocks substitute for international diversification?

• Published date: 01 November 2020
• Author: Vicente J. Bermejo,José M. Campa,Rodolfo G. Campos,Mohammed Zakriya
• Date: 01 November 2020
• DOI: http://doi.org/10.1111/eufm.12275

Eur Financ Manag. 2020;26:1191–1223. wileyonlinelibrary.com/journal/eufm © 2020 John Wiley & Sons Ltd.

|

1191

DOI: 10.1111/eufm.12275

ORIGINAL ARTICLE

Do foreign stocks substitute for international

diversification?

Vicente J. Bermejo

1

|José M. Campa

2

|Rodolfo G. Campos

3

|

Mohammed Zakriya

4

1

Universitat Ramon Llull, ESADE Business

School, Barcelona, Spain

2

IESE Business School, Madrid, Spain

3

Banco de España, Madrid, Spain

4

IESEG School of Management

(LEM‐CNRS 9221), Lille, France

Correspondence

Vicente J. Bermejo, Universitat Ramon

Llull, ESADE Business School, Av.

Torreblanca, 59, 08172 Sant Cugat,

Barcelona, Spain.

Email: vicente.bermejo@esade.edu

Funding information

International Mobility Program UC3M;

Spanish Ministry of Science, Innovation

and Universities, Grant/Award Number:

PGC2018‐099700‐A‐100; Government of

Catalonia, Grant/Award Numbers: 2017‐

FI_B‐00502, 2018‐FI_B‐00170, 2019‐FI_B2‐

00163

Abstract

Using a novel sample of foreign securities available

for trade in 42 countries during the last four decades

(1979–2018), we examine the rise in importance of

foreign stocks for investors in their host countries and

its implications for diversification across industries

and countries. The availability of foreign stocks

allows domestic investors to increase their interna-

tional diversification from home by investing in these

stocks. We conclude that including foreign stocks

in portfolio investments offers an effective substitute

for international diversification, and contributes sig-

nificantly towards increasing the integration of global

markets.

KEYWORDS

country/industry effects, foreign stocks, international capital

markets, international diversification

JEL CLASSIFICATION

F36; G11; G15

EUROPEAN

FINANCIAL MANAGEMENT

We have benefited from comments by the editor (John A. Doukas), an anonymous referee, Javier Díaz‐Giménez, Ilias

Filippou, Pedro Ángel García Ares, Gonzalo Gómez Bengoechea, María Gutiérrez Urtiaga, Pablo Ruiz Verdú,

Josep Tribó, Lucciano Villacorta, Daniel Wolfenzon, and participants at Universidad Carlos III and the FREE seminar

series. Vicente J. Bermejo acknowledges financial assistance from the International Mobility Program UC3M and

the Spanish Ministry of Science, Innovation and Universities under PGC2018‐099700‐A‐100. Mohammed Zakriya is

thankful for financial support from the Government of Catalonia (grants 2017‐FI_B‐00502, 2018‐FI_B‐00170, and

2019‐FI_B2‐00163). The views expressed in this article are those of the authors and do not necessarily represent the

views of Banco de España or the Eurosystem.

1|INTRODUCTION

Over the last decades, stocks that are listed and available for trade outside their home markets

have increased markedly (Doidge, Karolyi, & Stulz, 2009; Fernandes & Giannetti, 2014;

Sarkissian & Schill, 2016). In this article, we document the rise of these “foreign stocks”

1

over a

40‐year period and study how their presence affects the benefits of international portfolio

diversification. In particular, we focus on how the increasing availability of foreign stocks has

affected the relative value of diversifying across industries compared with countries.

The question as to whether it is more beneficial for investors to diversify across industries or

countries is far from settled. Initial studies by Heston and Rouwenhorst (1994), Griffin and Karolyi

(1998), and Rouwenhorst (1999) decomposed stock returns into industry and country effects and

found that stock returns are driven mainly by country effects. This implies that diversification across

countries is more valuable than diversification across industries. Since then, this result has been

confirmed by, for example, the studies of Phylaktis and Xia (2006), Baele and Inghelbrecht (2009), and

Bekaert, Hodrick, and Zhang (2009), while Baca, Garbe, and Weiss (2000), Cavaglia, Brightman, and

Aked (2000), and Eiling, Gerard, Hillion, and de Roon (2012), among others, found the opposite

result, that international equity returns are driven mainly by industry factors.

2

To our knowledge, the issue of how the presence of foreign stocks affects the benefits of industry

versus country diversification has not been studied directly to date. In this article, we first characterize

the rise of foreign stocks available for trade over the last four decades and show that they currently

amount to a significant fraction of investment opportunities available to investors. Whereas there

were roughly 750 foreign stocks listed internationally before 1989, this figure has risen to over 23,000

in the last few years. In relative terms, foreign stocks, which represented less than 7% of total listed

stocks in the 1980s, accounted for over 28% of total listed stocks in the last decade.

3

We show that the presence of foreign stocks raises the relative importance of industry to country

effectsinaninternationallydiversifiedportfolio. We estimate both the standard dummy‐variable

model for countries and industries by Heston and Rouwenhorst (1994) and a more general factor

model by Faias and Ferreira (2017) on a sample with and without foreign stocks, and find strong

evidence that industry effects exceed country effects once foreign stocks are taken into account. Using

the Faias and Ferreira (2017) model, we find that the relative decline in the importance of country

effects is not only at the expense of industry effects but also at the expense of a global factor.

We assess the performance of our model in explaining stock return comovements successfully by

comparing its root‐mean‐squared error with those of other standard factor models. We find that our

model outperforms these other models, confirming the finding by Bekaert et al. (2009) that meth-

odologies based on the Heston and Rouwenhorst model are appropriate for explaining variations in

international returns. To strengthen the point that foreign stocks substitute for international

1

Throughout this article, we use the term “foreign stocks”to refer to the equity stock of firms made available for trade in

markets outside their home country.

2

Brooks and Del Negro (2004) and Soriano and Climent (2006) argue that the predominance of industry effects may

have been only a temporary phenomenon associated with the stock‐market bubble at the beginning of the 21st century.

Other articles weighing in on the debate of country versus industry effects are those by Ferreira and Gama (2005),

Hargis and Mei (2006), Campa and Fernandes (2006), Bai and Green (2010), and Faias and Ferreira (2017). See the

recent review by Bekaert, Harvey, Kiguel, and Wang (2016) for further references on both sides of the debate.

3

This increase in foreign stocks occurred despite the increased cost of cross‐listing in many countries, such as the

United States (triggered by the introduction of Sarbanes Oxley Act in 2002), as shown in Doidge, Karolyi, and

Stulz (2010).

1192

|

EUROPEAN

FINANCIAL MANAGEMENT

BERMEJO ET AL.

diversification, we also perform mean–variance spanning tests in the spirit of Bae, Elkamhi, and

Simutin (2019), and show that including foreign stocks in a portfolio expands the efficient frontier.

Our results imply that locally traded foreign stocks act as a substitute for international di-

versification. This complements the findings in the extensive literature, starting with Errunza,

Hogan, and Hung (1999), who highlight that gains from international diversification can be

achieved by investing in locally traded securities.

4

An important difference between our approach

and that of such literature is that we do not restrict the analysis to the investors of any particular

country (usually the analysis is framed from the point of view of investors from the United States or

other developed economies). A second difference is that we focus on stocks instead of focusing on

different types of indexes or funds. Therefore, the evidence of a decline in the importance of country

effects that we uncover is disentangled cleanly from a decline related to the composition of indexes,

changes in the availability of mutual funds, or the strategies that fund managers use.

5

A number of articles following French and Poterba (1991) argue that home bias and in-

vestor's preferences for home stocks or differential information may limit the effective ad-

vantages of diversification (e.g., recently, Dumas, Lewis, & Osambela, 2017). In this same vein,

Portes and Rey (2005) show that investors prefer the stocks of foreign countries that are geo-

graphically closer, and Chan, Covrig, and Ng (2005) show that investors prefer the stocks of

foreign countries with equity markets more, not less, correlated with their own. We show that

the presence of foreign stocks in local markets can rationalize investing only domestically, given

that international diversification may be achieved through investing in these foreign stocks.

We structure the article as follows. In Section 2we describe the data and characterize the

evolution of foreign stocks available to investors; in Section 3we go over the methodology that

we use to identify the relative gains from international portfolio diversification; in Section 4we

present our empirical results; and in Section 5we conclude.

2|THE RISE OF FOREIGN STOCKS

To study the evolution of foreign stocks available for cross‐border trade, we use the information

available in the research lists of Datastream, which cover a broad set of markets with com-

parable data across countries. This provides the added benefit of comparability relative to

previous work. These research lists cover the entire menu of stocks in which an investor can

invest. The lists include all stocks listed in a particular country and do not a priori exclude

stocks based on certain criteria, such as market capitalization or country of origin.

We focus our analysis on firm‐level data and include all stocks listed in a particular

country.

6

The use of research lists allows the inclusion of firms with relatively small market

capitalization and, in some cases, from different stock exchanges within a country.

7

The use of

4

See Bae et al. (2019) and Lu and Vivian (2020) for two recent examples of this approach.

5

As stressed by Bai and Green (2010) and Bae et al. (2019), using individual stocks to assess diversification benefits is

preferable.

6

Bai and Green (2010) argue that the use of data on individual stocks is preferable to using indices, because indices have

several limitations: first, investment managers usually buy individual shares and not indices; and second, weighting and

composition of indices change over time in a manner that does not necessarily reflect underlying market trends. In

addition, Bae et al. (2019) suggest that relying on equity indices to assess the benefits of diversification understates

potential gains. We follow their lead and use individual stocks as our unit of analysis.

7

For example, in Germany we include stocks from the seven stock exchanges: Stuttgart, Frankfurt, Berlin, Munich,

Hanover, Hamburg, and Düsseldorf.

BERMEJO ET AL.EUROPEAN

FINANCIAL MANAGEMENT

|

1193