Debt and Taxes: Evidence from Foreign versus Domestic Subsidiaries in an Emerging Market

AuthorWoojin Kim,Hyo‐Jeong Lee
Date01 April 2015
Published date01 April 2015
DOIhttp://doi.org/10.1111/ajfs.12088
Debt and Taxes: Evidence from Foreign
versus Domestic Subsidiaries in an
Emerging Market*
Woojin Kim**
Seoul National University Business School
Hyo-Jeong Lee
Division of Business Administration, Kwangwoon University
Received 29 July 2014; Accepted 15 February 2015
Abstract
Foreign subsidiaries in Korea operate under a classical system where double taxation of cor-
porate and personal income provides interest tax shields, while domestic subsidiaries are
under an imputation system where preference for debt is largely eliminated. We find that for-
eign subsidiaries’ overall leverage is not higher than domestic subsidiaries, while the former’s
internal leverage is higher than the latter’s. Moreover, foreign subsidiaries with excess foreign
tax credit use more internal debt when home corporate tax rate is low. These findings suggest
that taxes have a first order impact on internal debt when parent companies hold both debt
and equity, but not on external debt.
Keywords Classical system; External debt; Foreign tax credit; Imputation system; Internal
debt
JEL Classification: G32, G38
1. Introduction
Whether taxes affect corporate debt policy is one of the still unresolved controversial
debates in corporate finance. While proponents of tax effect argue that the value of
tax shields from deductibility of interest payments amounts to up to 10% of firm
value (Graham, 2000), skeptics criticize this perspective, insisting that tax incentives
are at best of “third” order importance in corporate policies (Myers et al., 1998).
Although the argument for tax advantage of debt referred to as “interest tax shield”
*We would like to thank Hyeongsop Shim and other seminar participants at 2013 Allied
Korea Finance Associations Conference (Chonan, May 2013) for helpful comments. This
study was supported by the Institute of Management Research at Seoul National University.
**Corresponding author: Woojin Kim, Seoul National University Business School, Seoul,
151-916, Korea. Tel: +82-2-880-5831, Fax: +82-2-880-5831, email: woojinkim@snu.ac.kr.
Asia-Pacific Journal of Financial Studies (2015) 44, 246–280 doi:10.1111/ajfs.12088
246 ©2015 Korean Securities Association
in every single corporate finance textbook is innocuous (Modigliani and Miller,
1963), confounding effects from complexities in personal taxes and various measure-
ment issues raise serious empirical challenges in disentangling pure tax effect from a
host of “other” potential factors that may influence capital structure.
The first difficulty arises from the fact that it is not trivial to obtain adequate
levels of variations in tax treatments, especially when the sample firms are from a
single country. One may attempt to take advantage of different tax rates in various
countries in an international setting (Rajan and Zingales, 1995). However, this over-
looks an often ignored aspect of corporate taxation in countries other than the Uni-
ted States. As emphasized by Graham (2008), most countries outside the United
States, including the United Kingdom and France, adopt some variant of “imputa-
tion” tax system designed to remove or at least mitigate double taxation of dividend
income and accompanying distortion between debt and equity that plagues “classi-
cal” tax systems, such as that of the United States.
1
In a classical tax system, as adopted by the United States, dividend income is
taxed twice at the corporate level and personal level, while interest income is taxed
only once at the personal level after being deducted from corporate income. Such a
system, however, may distort capital structure decisions by favoring one type of
security over another in terms of after-tax cash flows available to investors. If the
tax system actually favors debt, although still somewhat controversial, the firms
may take advantage of the “interest tax shield”.
Under a typical imputation system, however, shareholder’s before-tax income is
“grossed up” based on corporate income taxes paid, and the shareholder is only
responsible for his or her personal income tax, after receiving tax credits for the
corporate income tax already paid at the firm-level. To the extent that the imputa-
tion system mitigates the tax advantage of debt and thus the distortion between
debt and equity, the effect of interest tax shield outside the United States may not
be as large as once expected (Graham 2003, 2008). That is, the interest tax shield
may be a United States specific phenomenon that may not exist in other countries.
Reflecting this difficulty with international studies, efforts have been made to
generate firm-level variations in tax status within the United States by relying on
some proxies for different tax brackets (such as profitability) or directly estimating
effective marginal tax rates through simulations. Nevertheless, this approach is still
subject to various endogeneity issues and measurement errors by construction. A
recent study by Desai et al. (2004) adopts a clever identification strategy that takes
advantage of cross-country variations in tax rates applied to subsidiaries of United
States multinational firms to circumvent limited variations in firm-level tax status
within the United States. Although this approach effectively generates variations in
tax rates for firms facing a classical tax system (i.e., double taxation of corporate
1
According to Gujarathi and Feldmann (2006), the only OECD country that still maintained
the classical tax system as of then was Ireland. Even the United States, since 2003, is classified
as a partial imputation system since dividend income is taxed at a lower rate.
Debt and Taxes
©2015 Korean Securities Association 247
and personal income), it does not provide a direct comparison between firms under
a classical system against firms under an imputation system. Besides, cross-country
variations reflect not only differences in tax systems, but also differences in various
other institutional arrangements, level of capital market development, and so on.
A typical approach in tax research ever since Miller (1977) is that relative bene-
fit of issuing debt versus equity is analyzed from a hypothetical “marginal” inves-
tor’s perspective who holds both equity and debt and is indifferent between the two
asset classes in terms of tax advantage or disadvantage.
2
But, as noted by Graham
(2003), the identity and tax status of the marginal investor is extremely difficult to
pin down, especially in widely-held publicly-traded firms, which are typically the
subject of extant studies based on United States data. This situation has forced
researchers to rely on some arbitrary assumptions about the identity of the marginal
investors. An example is Rajan and Zingales (1995), who conclude that the effect of
taxes depends on who is the marginal investor.
In this paper, we take advantage of an institutional setting that allows us to
overcome these empirical challenges raised above and directly compare firms under
the classical system and firms under the imputation system, holding other country
characteristics fixed. Specifically, we examine capital structure decisions of wholly-
owned firms in Korea where some firms are taxed under a “classical” tax system
while some firms are taxed under an “imputation” tax system. This provides an
ideal setting for a direct test of Graham’s (2003, 2008) predictions that tax incentive
to issue debt decreases with the degree of dividend imputation.
Most Korean firms are under the imputation system, and some investors are
likely to receive full tax credits, implying that there may not be much distortion
between debt and equity for at least a subset of firms, which renders the whole con-
cept of interest tax shield meaningless. However, there is another subset of firms
operating in Korea, the shareholders of which are not able to claim tax credits
through the imputation system. These are the subsidiaries of foreign parents. That
is, if the shareholder is a non-resident, he or she cannot claim tax credits for the
taxes paid at the corporate level, thus being de facto subject to a classical tax system
and accompanying double taxation of dividend income.
3
This setting allows us to effectively distinguish between firms that face a classical
tax system and those that face an imputation system operating under a unified
institutional framework. That is, we are able to compare the effect of differe nt tax
regimes holding all other institutional factors and capital market situations con-
stant. This approach complements previous research using cross-country data where
not only tax rates but other institutional factors also inevitably vary across coun-
tries. Moreover, our approach provides a more powerful test of the effect of taxes
2
A key assumption in Miller’s (1977) approach is that both debt and equity are riskless.
3
Non-resident shareholders may claim foreign tax credits at their home countries for taxes
paid overseas when they pay their domestic taxes. We provide an analysis of the implications
of foreign tax credits in a later section.
W. Kim and H.-J. Lee
248 ©2015 Korean Securities Association

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT