Foreign currency borrowing surrounding the global financial crisis: Evidence from Korea

• Author: Taek Ho Kwon,Sung C. Bae,Hyeon Sook Kim
• Published date: 01 May 2020
• Date: 01 May 2020
• DOI: http://doi.org/10.1111/jbfa.12425

DOI: 10.1111/jbfa.12425

Foreign currency borrowing surrounding the

global financial crisis: Evidence from Korea

Sung C. Bae1Hyeon Sook Kim2Tae k H o Kw o n 3

1Department of Finance, College of Business,

Bowling Green State University,Bowling Green,

OH, USA

2Public Procurement Service, Government

Complex,Daejon, Korea

3School of Business, Chungnam National

University, Daejon, Korea

Correspondence

HyeonSook Kim, Public Procurement Service,

GovernmentComplex, Daejon, Korea.

Email:khsvega@korea.kr

Abstract

We present a complete profile of firms’ foreign currency borrow-

ing surrounding the 2007 global financial crisis. Employing extensive

data from Korean firms during 2002–2012, we find that foreign cur-

rency borrowing is significantly related to firm attributes of export

revenues, firm size, tangible assets and asset growth, as well as to

macro-level factors. These results offer two important implications.

First, macroeconomic factors alone cannot fully explain firms’ for-

eign currency borrowing. Second and more importantly, these firm

attributes are indicative of a lower default probability and larger col-

lateral value, which would not only facilitate borrowers’ access to

foreign currency debt marketsbut also offer lenders a better protec-

tive cushion from possible loan defaults in the face of exchangerate

changes and information asymmetry on borrowers’ credits. Period

wise, asset-related firm attributes have more pronounced effects in

the post- than pre-crisis period. We further show that banking regu-

lations following the crisis effectively limit the access to foreign cur-

rency borrowing byKorean firms, most significantly by those belong-

ing to large business groups.

KEYWORDS

banking regulations, firm attributes, foreign currency borrowing,

global financial crisis, Korean firms, macro-levelfactors

JEL CLASSIFICATION

F31, F34, G15

1INTRODUCTION

Over the past few decades, foreign currency (hereafter FC) borrowing has been an important source of debt cap-

ital for firms in emerging economies given their underdeveloped debt markets and limited access to global equity

markets. Since the 2007 global financial crisis (hereafter GFC), there has been a major resurgence in FC, especially

US-dollar-denominated bond issuance by emerging marketfirms. According to the Bank for International Settlements,

786 c

2019 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/jbfa JBus Fin Acc. 2020;47:786–817.

BAE ET AL.787

the issuance amount of FC bonds by non-financial firms in emerging markets quadrupled between 2007 and 2013,

reaching US$ 692 billion in 2013.1

The extant literature has advanced rationales for firms’ usage of FC borrowing, including to overcomesegmented

local capital markets (Campbell & Hamao, 1992; Jorion & Schwartz, 1986); to circumvent withholding taxes, cap-

ital controls, and other legal restrictions of local governments (Rhee, Chang, & Koveos, 1985; Shapiro, 1984); to

arbitrage differences in tax rates (DeMarzo & Duffie, 1995; Smith & Stulz, 1985); to take advantage of the interest

rate differential between local and foreign currencies (Allayannis, Brown, & Klapper, 2003; Basso, Calvo-Gonzalez,

& Jurgilas, 2011; Brown, Ongena, & Ye¸sin, 2011; Rosenberg & Tirpák, 2009); to hedge exchange rate exposures

resulting from exporting activities and foreign revenues (Allayannis & Ofek, 2001; Bae & Kwon, 2013; Graham &

Harvey, 2001; Kedia & Mozumdar, 2003; Keloharju & Niskanen, 2001; Mora, Neaime, & Aintablian, 2013); to ben-

efit from the trade-off between inflation and real exchange rate variability (Basso et al., 2011; Ize & Levy-Yeyati,

2003); and as a combined outcome of FC domestic deposits, cheap foreign funding, and supply-driven banks’ desire

for currency-matched portfolios (Brown & De Hass, 2012; Brown, Kirschenmann, & Ongena, 2014; Luca & Petrova,

2008).2

While existing studies offer a long list of rationalesfor FC borrowing, they mainly focus on market- and macro-level

factors and pay little attention to firm-level attributes that mayaffect firms’ borrowing in FC, relative to borrowing in

local currency (hereafter LC). Furthermore, evidence is lacking about whether and how the interventionsof the GFC

and subsequent government regulations on FC lending, as well as other influential macro-levelfactors, affect the firm-

level attributes of FC borrowing firms. Our study intends to fill this gap byexamining three unexplored research issues

on FC borrowing.

First, we offer a comprehensive profile of borrowing firms in FC debt markets by uncoveringfirm- and macro-level

factorsthat may differentiate firms borrowing in FC from those borrowing in LC. Our study is motivated by two streams

of research on FC borrowing. The first provides evidence, albeit limited, on the relationships between FC borrowing

and firm-level factors. Brown et al. (2011) report that firm-levelfactors, such as FC revenues, explain FC borrowing by

small firms in 26 transition European economies better than country and macroeconomic factors, such as interest rate

differentials and exchange rate volatility. Mora et al. (2013) find that while small Lebanese firms with more collateral

and higher net worth are likely toaccess US dollar debt, profitable firms are less likely to use such debt. Vonnák (2018)

notes the important role of the characteristics of FC borrowers such as credit status in assessing FC loan performance,

contrary to existingresearchers’ attention to those of FC lenders per se in Eastern European countries surrounding the

GFC. The second stream of research deals with FC borrowing by households, rather than bybusiness firms. Neverthe-

less, these studies offer insightful evidence on the roles of both household attributes and macroeconomic factors as

determinants of FC loans. For example, in a study of surveydata from Austrian households, Beer, Ongena, and Peter

(2010) report that risk-seeking, affluent, and married households are more likely to take a FC housing loan. Csajbók,

Hudecz, and Tamási(2009) show that the share of new FC loans by households across 10 Eastern European countries

during 1999–2008 is positively related to the interest rate differential. Pellényi and Bilek (2009) report similar evi-

dence regarding household FC loans in Hungary.

Our study extends the scope of existingresearch to firms in the major emerging economy of Korea. Korea has gone

through highly volatile changes in exchange rates since the 1997 Asian financial crisis, and Korean firms frequently

resort to FC borrowing for their local headquarters and overseas foreign subsidiaries. It is worth noting that currency

indebtedness is the outcome of the demand and supply sides of FC debt, which are influenced by both borrowers and

1In particular,growing Asian wealth and Asian investors’ desire for assets denominated in currencies other than their own have helped fuel the demand for

FCdebt, especially dollar bonds, in the region; during2017 alone, Asian firms and governments, excluding those in Japan, issued a record US$ 333.9 billion in

bonds(Bird, 2017).

2Severalother studies examine the effects of FC borrowing on firm performance and value with inconclusive evidence (see, e.g., Allayannis, Brown, & Klapper,

2003;Bae, Kim, & Kwon, 2016; Bleakley & Cowan, 2008; Clark & Judge, 2009; Endrész & Harasztosi, 2014; Harvey, Lins, & Roper,2004).

788 BAE ET AL.

lenders.3The nature of our data is, however,not suitable to disentangle these two effects because FC debt of Korean

firms comes from various sources not skewed to a few lenders, but still allows us to analyse what types of firm are

indebted in FC vs. LC.

Second, our study sheds new empirical light on how the intervention of the 2007 global financial crisis affected

firms’ FC borrowing behaviour.The current literature offers rich empirical evidence about FC borrowing surrounding

the Asian financial crisis and some evidence surrounding the 2007 GFC; the latter, however, is limited to European

firms, with little evidence for emerging market Asian firms surrounding the GFC (see, e.g., Brown & De Haas, 2012;

Brown et al., 2011, 2014; Vonnák, 2018; Zettelmeyer, Nagy, & Jeffrey, 2010).4As global capital markets, including

emerging markets,suffered a substantial decline i n economic activities with a lack of credit following the crisis (Dooley

& Hutchison, 2009), FC borrowing of emerging market firms may exhibit different characteristicsafter the GFC than

before the GFC. Indeed, FC borrowing peaked during the Asian financial crisis, declined after the crisis, and began to

increase once again, reaching its highest level during the GFC. Which firm- and macro-level factors have contributed

to the changes in FC borrowing around the GFC is important information for both policy makers and corporate

managers.

Third, our study examines whether and how banking regulations on FC lending surrounding the GFC affected the

levels and determinants of FC borrowing. Confronting the increasing FC borrowing and the volatilityof foreign capital

flow during the crisis, the Bank of Korea (BOK) enforced as a macro-prudential instrument a series of bank regulations

on FC lending, including FC loans’ ring fence in August 2007 and July 2010 and banks’ FC loan limit in July 2010.5

These bank regulations instituted by the Korean government were prompted by policy concerns about the high costs

of serving large amounts of FC debt and the potential costs of loan default by Korean firms, especially when the local

Korean won was rapidly depreciating against major currencies. While such bank regulatory changes are expected to

discourage firms’ borrowing and banks’ lending in FC, this study aims to offer further empirical insights into whether

and how the determinants of FC borrowing, as well as its levelsrelative to LC borrowing, are affected by the regulatory

changes.6

Employing extensive data from Korean non-financial firms during 2002–2012, we document that FC borrowing is

affected by different firm- and macro-level factors than LC borrowing. Firms with more exportrevenues, larger size,

more tangible assets, and higher asset growth are more likely to use FC debt over LCdebt. In addition, FC borrowing

is affected by macro-level factors such as interest rate differentials, inflation, and foreign exchangevolatility as well.

While these results are the consequence of borrowers’ heterogeneous demand for FC debt and of the fact that the

FC supply of the lenders also depends on the borrowers’ profile, our results offer two important implications, which

are undocumented in the literature. First, macro-level factors are influential determinants of firms’ FC borrowing, but

the significance of many firm-level factors in the presence of macro-level factors demonstrates that macroeconomic

factors alone cannot fully explainfirms’ FC borrowing behaviour. Second and more importantly, firms’ exportrevenues

provide a natural hedge against exchange rate fluctuations; thus, firms with a higher export ratioare less prone to

default in the event of an exchange rate shock. In addition, asset-related firm attributes such as firm size, tangible

assets, and asset growth indicate collateral value. Hence, firms with more export revenues and better asset-related

firm attributes are characterized by a lower default probability and larger collateral value. Accordingly, borrowing

firms with these attributes will be able to access FC debt markets with ease, compared to firms lacking such attributes.

Furthermore, faced with the uncertainty on exchange rate changes and a high degree of information asymmetry on

3Lendersof FC debt for Korean firms typically include foreign firms and institutions that offer international debt securities (e.g., bonds) and Korean banks and

localbranches of foreign banks that offer FC loans.

4A few other studies report that during the GFC, firms with more FC debt experienceda significant decline in their investments, compared to firms with less

FCdebt (Bleakley & Cowan, 2008; Endrész & Harasztosi, 2014).

5Seethe 2016 BOK report for details (pp. 72–75 & 264–265). TheDodd-Frank Wall Street Reform and Consumer Protection Act of 2010 was also one of the

mostwell-known and far-reaching new pieces of regulation that touched both US and foreign institutions.

6Zettelmeyer,Nagy, and Jeffrey (2010) suggest that regulation to the FC liability problem can be useful through limiting corporate and household foreign

exchangeexposures and correcting distortions that make FC borrowing too cheap.