Is bank capital sensitive to a tax allowance on marginal equity?

• DOI: http://doi.org/10.1111/eufm.12163
• Published date: 01 March 2019
• Date: 01 March 2019

325Eur Financ Manag. 2019;25:325–357. wileyonlinelibrary.com/journal/eufm © 2017 John Wiley & Sons, Ltd.

DOI: 10.1111/eufm.12163

ORIGINAL ARTICLE

Is bank capital sensitive to a tax allowance on

marginal equity?

Jose Martin-Flores

|

Christophe Moussu

ESCP Europe and LabEx-ReFi, 79

Avenue de la République 75543 Paris

Cedex 11, France

Emails: moussu@escpeurope.eu;

jose_maria.martin_flores@

edu.escpeurope.eu

Funding information

This work was achieved through the

Laboratory of Excellence on Financial

Regulation (Labex ReFi) supported by

PRES HeSam under the reference ANR-10-

LABX-0095. It benefitted from French

government support managed by the

National Research Agency (ANR) within

the project Investissements d'Avenir Paris

Nouveaux Mondes (investments for the

future Paris-New Worlds) under the

reference ANR-11-IDEX-0006-02.

Abstract

This paper studies how bank capital changes following the

implementation and removal of a tax incentive on equity.

We examine the impact of the introduction of a tax

allowance in Italy granted to banks (and other firms) that

increase their equity from a base year. Using a difference-in-

differences setting, we observe an 8.83% increase in bank

capital ratios following the implementation of this reform.

When this tax mechanism is phased out, we observe an

opposite effect on the equity ratio, showing the absence of a

hysteresis effect in bank capital. We document a heteroge-

neous effect for large and small banks.

KEYWORDS

tax, bank capital, debt-equity tax bias

JEL CLASSIFICATION

G21, G28, G32

1

|

INTRODUCTION

Bank capital is important but imposing higher capital requirements on banks affects the offer of credit

to the economy (Jiménez et al., 2016) and may result in a credit crunch with procyclical negative effects

The authors thank the Editor, John Doukas and the three referees. They also thank Gunther Capelle-Blancard, Thomas

David, JulienFouquau, Arthur Petit-Romec, Mark Roe, AnjanThakor, Michael Troege and Lei Zhao for helpful comments,

Vincenzo Cuciniello and GiacomoRicotti for providing information on the Italian ACE mechanism and participants in the

LabExReFi Workshop on Bank Taxation (2016), the PhD seminar at ESSEC Business School (2016), the IFABS

Conference in Barcelona (2016), the EFMA Conference in Athens (2017), the 2017 International Advisory Board

Workshop of the LabExReFi in Paris (2017) and the World Finance Conference in Cagliari (2017).

2

|

MARTIN-FLORES AND MOUSSU

M

326

in crisis periods (Cornett et al., 2011; Dermine, 2013). Another instrument available for policymakers

to strengthen bank capital is to use tax incentives. Since interest payments are in general deductible

from the corporate income tax base whereas equity returns are not, firms have an incentive to use higher

leverage. The effect of interest tax shields on firm leverage has been widely studied for non-financial

firms (e.g., Arena & Roper, 2010; Desai et al., 2004; Faccio & Xu, 2015; Feld et al., 2013; Graham,

1996; Graham & Tucker, 2006; Heider & Ljungqvist, 2015) and for banks (e.g., De Mooij & Keen,

2016; Hemmelgarn & Teichmann, 2014; Horváth, 2013; Milonas, 2016; Schandlbauer, 2017).

However, banks are, in principle, more systemic than non-financial firms so the effect of this tax bias is

particularly pernicious for them. Since bank capital provides some protection against failure in case of

crisis, a tax bias that incentivizes the use of debt undermines to some extent the safeness of the banking

system.

In this paper, we are interested in the impact of the implementation and removal of a tax reform

which reduces this debt-equity tax bias. The tax scheme introduced in Italy in 1997 has similar

characteristics to the Allowance for Corporate Equity (ACE) initially proposed by Devereux and

Freeman (1991). This ACE-like mechanism grants a tax shield on equity to banks that increase their

equity from a reference year. Banks (and other firms) apply a reduced tax rate (19% instead of 37%) on

a notional return computed on equity increases. Using a difference-in differences setting, with Italian

banks as treatment group and a matched sample of banks from other euro area states as control group,

we find an 8.83% increase in bank capital ratios relative to the control group following the

implementation of this tax reform. We also observe that this increase is driven by a higher level of

equity and not by a decrease in bank assets.

However, as the ACE tax incentive was withdrawn in 2002, we are able to provide new

evidence on whether the introduction and removal of the tax allowance mechanism lead to

symmetric effects. We find that once the ACE mechanism is no longer applicable, banks stop

increasing their equity and readjust downward their equity ratios by 4.6% on average relative to

banks in the control group. This finding reveals that there is no hysteresis in bank capital when a

tax incentive to increase equity is removed. This is in line with a static perspective of capital

structure. When an ACE is introduced, the debt tax bias is reduced and banks rebalance their

capital ratio upward. When the ACE is removed, the debt tax bias increases again, and bank

capital is adjusted downward accordingly. Additional robustness tests show that our results are

unlikely to be driven by changes in monetary policy or other economic events. Also, we show

that our findings are robust to different control groups and the introduction of various control

variables.

Our paper contributes to the existing literature on the effects of ACE mechanisms on capital

structure. The most studied ACE mechanism in the finance and economics literature is the one

introduced by Belgium in 2006. Panier et al. (2013) and Princen (2012) document a positive effect

of the tax reform on equity ratios for non-financial firms, and Schepens (2016) for banks.

However, the Italian ACE scheme differs from the Belgian reform of 2006 in three important

respects. First, in Italy, the notional interest is computed only on new equity after the reform is in

place and not on the existing stock of equity as in the Belgian case. This is particularly interesting

as there are no ‘windfall rents’associated with reduced taxation on the existing equity. It therefore

provides a great opportunity to test the effect of a tax incentive to increase equity for banks, as

there are no tax-shields on existing equity in the Italian case. Our results confirm that a positive

impact on bank capital ratios is also observed for a ‘softer’version of the ACE. Second, the Italian

ACE scheme includes an anti-avoidance clause that targets transactions between related parties in

order to avoid abuses and tax planning (Zangari, 2014). In particular, as documented by Hebous

and Ruf (2017), multinational non-financial companies used the Belgian ACE mechanism to

i

d

t

p

p

s

i

m

H

A

d

b

T

e

a

e

b

l

r

p

2

i

o

d

i

r

b

c

a

(

D

c

a

o

f

a

l

s

r

1

2

r

c

U

MARTIN-FLORES AND MOUSSU

|

3

327

s

e

r

l

,

,

.

s

f

g

m

r

d

r

n

n

,

e

f

w

o

p

o

a

l

r

k

e

w

l

l

e

t

.

t

n

g

e

s

e

n

n

s

o

implement tax planning structures combining the benefits from the Belgian ACE with interest

deductions. Thus, our result on the introduction of the ACE in Italy is important because it shows

that a positive effect on bank capital is also found in an ACE system that limits the scope for tax

planning opportunities. Third, an interesting feature of the ACE experiment in Italy is that it was

phased out in 2002. This phasing-out allows us to provide new evidence on the existence of

symmetric reactions of bank capital to the introduction and removal of a tax shield on equity

increases.

Our paper also contributes to the bank capital structure literature. Previous work on this topic

mainly focuses on the determinants of bank capital structure (e.g., Berger et al., 2008; Gropp &

Heider, 2010) and their adjustments towards the optimal ratio (e.g., De Jonghe & Öztekin, 2015).

Another stream of the bank capital structure literature has focused on the extent to which tax rates

determine equity ratios. These papers show a negative (positive) relationship between tax rates and

bank capital (leverage) ratios for international (e.g. De Mooij & Keen, 2016; Hemmelgarn &

Teichmann, 2014; Horváth, 2013) and single-country samples (Bond et al., 2016; Gambacorta

et al., 2017; Milonas, 2016; Schandlbauer, 2017). Our paper differs from these papers in two

aspects. First, instead of looking at tax rate changes, we focus on a specific tax incentive targeting

equity. We show that an exogenous reduction of the debt-equity tax bias leads to better capitalized

banks. Second, we demonstrate that the effects of a tax incentive targeting equity are unlikely to

last once the incentive is removed. We provide evidence for a symmetric reaction of bank equity

ratios once the tax incentive is phased out. Given that accumulating higher capital in normal

periods determines the ability of a bank to withstand economic shocks (Berger & Bouwman,

2013), it is interesting to know whether a tax incentive on new equity leads to better capitalized

institutions and whether this effect remains after the incentive disappears. However, we are not the

only ones to document a symmetric reaction of bank capital to tax incentives. Bond et al. (2016)

document that bank capital ratios react symmetrically to corporate income tax cuts and tax

increases. In contrast, our paper provides new evidence about the effects of implementing and

removing an explicit policy tool addressing the debt-equity tax bias on the use of equity financing

by banks. This is, to some extent, valuable information for regulators and policymakers who may

consider a different taxation of banks that does not provide incentives that go against capital

adequacy regulations.

1

Finally, we observe that the introduction (removal) of the ACE incentive has a positive

(negative) and significant impact on bank capital only for small banks.

2

This result is in line with

De Mooij and Keen (2016), who reveal that the capital ratio of larger banks is not sensitive to tax

changes at local level. They attribute this effect to ‘too big to fail’considerations. However,

another potential explanation is the ability of multinational banks to exploit tax arbitrage

opportunities at group level. Gu et al. (2015) provide evidence that international banks shift debt

from one country to another to exploit differences in the tax codes between their home countries

and other countries where their subsidiaries operate. Our results can be consistent with the second

line of reasoning. However, they are less supportive of the ‘too big to fail’argument since medium

size banks' capital ratios do not change significantly when the ACE incentive is implemented or

removed.

1

See Roe and Troege (2017) for an extensive discussion on bank taxation.

2

The Italian banking sector has a large number of small cooperative banks that are subject to specific regulatory

requirements. We also test (in unreported tests) whether the effect on smaller banks persists when we remove those

cooperative banks. Our conclusions remain unchanged after applying this treatment.