The euro area bias and the role of financial centres

• Author: Maurizio Michael Habib,Vincent Arthur Floreani
• Published date: 01 July 2018
• DOI: http://doi.org/10.1002/ijfe.1614
• Date: 01 July 2018

Received: 27 September 2017 Accepted: 12 December 2017

DOI: 10.1002/ijfe.1614

RESEARCH ARTICLE

The euro area bias and the role of financial centres

Vincent Arthur Floreani1Maurizio Michael Habib2

1World Bank Group,1818 H Street NW

20433 Washington, DC, USA

2European Central Bank,

Sonnemannstrasse 20 Frankfurt am Main,

60314, Germany

Correspondence

Maurizio Michael Habib, European

Central Bank, Sonnemannstrasse 20,

Frankfurt am Main 60314, Germany.

Email: maurizio.habib@ecb.int

JEL Classification: F3, G1

Abstract

This paper investigates the euro area bias—a stronger preference for euro area

securities in the cross-border portfolio of euro area residents with respect to

the prediction of an international finance gravity model. Compared with previ-

ous studies, we disentangle different partitions of the euro area and, crucially,

isolate euro area financial centres, finding some novel results. First, the euro

area bias is not always present and remarkably weak for securities issued by

euro area high-rating economies, those that would be expected to provide better

insurance during crisis periods. Second, the extent of the euro area bias could

be overstated, because euro area financial centres may conceal round-tripping

flows from euro area low-ratingeconomies. Finally, in the case of equity, the euro

area bias did not decline during the crisis period but, in the case of debt issued

by euro area low-rating economies, unsurprisingly, weakened. The decline in

exposure towards euro area de bt securities was common among all cross-border

investors, not a specific retrenchment of euro area investors.Overall, our results

call for a prudent treatment of cross-border financial statistics in the presence of

a financial centre bias.

KEYWORDS

cross-border portfolio, euro area bias, financial centres, gravitymodel, international finance

1INTRODUCTION

International finance gravitymodels have been used exten-

sively to study the impact of the introduction of the euro on

the financial integration of euro area economies. Although

the inclusion of gravity variables provides a good picture

of cross-border financial positions on a global scale, it falls

short to explain the “exceptionally” high degree of inte-

gration of financial markets among euro area economies.

Even accounting for the greater degree of trade integration

and lower distance among euro area economies, there is

still an euro area bias—a stronger preference for euro area

securities relative to other securities—in the portfolio of

euro area residents, which has been tentatively explained

by the decline in default risk and transaction costs or the

elimination of currency risk among Economic and Mon-

etary Union (EMU) economies. Do euro area investors

really prefer to hold securities of other euro area countries

beyond what a gravity model would predict? In this paper,

we offer an alternative, more pragmatic, explanation to

this puzzle: A measurement error driven by the presence

of important financial centres within the euro area. In

particular, we study deviations from a notional portfolio

allocation to the equity and debt issued by the euro area

countries in a reduced-form gravity model, accounting for

institutional features, their distance from the euro area,

and allowing for non-linearity in covariates. Compared

with previous studies, we partition the euro area into three

groups of issuers and investors: high-rating economies

(“core” in the literature), low-rating economies (“periph-

ery” in the literature) and, crucially,financial centres. Cru-

cially, isolating euro area financial centres, we are able to

Int J Fin Econ. 2018;23 233–256. wileyonlinelibrary.com/journal/ijfe Copyright© 2018 John Wiley & Sons, Ltd. 233:

234 FLOREANI AND HABIB

identify new patterns of intra euro area cross-border port-

folio holdings. We find that the euroarea bias is not always

present across different partitions of the euro area. Some-

what surprisingly,the euro area bias is remarkably weak in

thecaseofequityissued by high-rating economies. Singling

out financial centres, we find that some aspects of euro

area financial integration are less benign than previously

thought. In particular, we show that cross-border portfo-

lio statistics artificially inflate measures of equity financial

integration—a desirable type of integration through the

cross-border holdings of state-contingent risky capital—at

the expenses of debt—the one that raises more concerns

because debt is procyclical and volatile (ECB, 2012, 2016).

In addition, we show that euro area financial centres may

conceal round-tripping flows from and back to low-rating

economies, implying a weaker ability of investors located

in these economies to diversify country-specific risks than

a superficial analysis of cross-border statistics would sug-

gest. Finally, we investigate the impact of the financial

crisis on the euro area bias. Even though weaker than

previously thought, the preference of euro area investors

for securities issued by other euro area countries has

been relatively resilient since the start of the global finan-

cial crisis, in particular through cross-border holdings of

equity, supported by the intermediation activity of euro

area financial centres. However, the appetite of euro area

investors for debt securities issued by other euro area

countries declined. Notably, this trend was driven by a

global retrenchment in the exposure towards debt secu-

rities issued by low-rating economies until 2012, not by

a specific retrenchment of euro area investors. The paper

is structured as follows. In the next section, we provide a

review of the related literature. Section 3 introduces the

basic concepts to measure financial exposure to the euro

area, develops the empirical model, and describes the data.

Section 4 discusses the results for equity and debt portfo-

lios. Section 5 investigates the resilience of the euro area

bias, in particular since the start of the global financial

crisis. Finally, Section 6 concludes.

2RELATED LITERATURE

Our paper follows the lead of two main strands of litera-

ture: (a) one related to international finance gravitymodels

with a specific role of informational asymmetries and insti-

tutions; and (b) a second one where these models have

been used to explain the impact of the EMU on financial

integration in the euro area.

First, international finance gravity models explain bilat-

eral cross-border financial positions through variables

proxying for informational frictions, such as physical dis-

tance (Portes & Rey, 2005), common language, legal ori-

gin, and, in particular, trade (Aviat & Coeurdacier, 2007;

Lane & Milesi-Ferretti, 2008). The development of these

empirical models finds its roots in the home bias litera-

ture, where informational asymmetries between domestic

and foreign investors have been identified as one of the

main determinants of cross-border asset holdings.1In par-

ticular, Martin and Rey (2004) and Okawa and van Win-

coop (2012) provide general theoretical frameworks that

produce a gravity form specification of bilateral financial

holdings. Additional factors—such as institutional quality

and institutional distance—could play a role in accounting

for bilateral financial asset holdings. Regulations, polit-

ical stability, government effectiveness, accounting stan-

dards, rule of law,and absence of corruption are all factors

that raise the transparency of financial information and

reduce informational asymmetries across countries.2The

relative level of institutional quality or financial develop-

ment across countries (“comparative advantage” hypoth-

esis) seems to matter for bilateral financial positions, as

shown by Forbes (2010) and other studies.3On the other

hand, a number of papers claim that “familiarity” (Huber-

man, 2001, Grinblatt & Keloharju, 2001, or Chan, Covrig,

& Ng, 2005) is the driving force of bilateral asset posi-

tions. A result that has been corroborated by cross-country

studies of regulatory differences such as Vlachos (2004)

and by specific applications on the role of governance

for the financial exposure to the United States (Abdioglu,

Khurshed, & Stathopoulos, 2013) or Korean firms (Kim,

Taeyoon, & Wei, 2011). Our paper will explicitly test the

relevance of financial development and institutions “at

home” as potential drivers of the exposure towards the

euro area.

Second, several studies used international gravity mod-

els to study euro area financial integration, finding a

degree of integration of financial markets among euro

area economies that goes beyond the level that an empir-

ical model would predict.4In particular, even considering

the greater degree of trade integration and lower distance

among euro area economies, there is a stronger prefer-

ence for euro area securities relative to other securities

in the portfolio of euro area residents: an euro area bias.

This euro area bias has been tentatively explained by the

decline in default risk and transaction costs (Balli, Basher,

& Ozer-Balli, 2010) or by the elimination of currency risk

among EMU economies (Kalemli-Ozcan et al. 2010).5The

link between euro area and global financial integration has

also attracted attention, following a particularly intriguing

narrative of the crisis. Hobza and Zeugner (2014) and Hale

and Obstfeld (2016) posit that the core euro area coun-

tries took on extra foreign leverage to expose themselves

to the peripheral economies; a downhill capital flow from

richer to poorer euro area economies, apparently resilient

to financial stress (Herrmann & Kleinert, 2014).