Tax policies and agency costs

• Published date: 01 June 2023
• Author: Diogo Duarte,Brice Dupoyet,Sandrine Docgne,Florent Rouxelin
• Date: 01 June 2023
• DOI: http://doi.org/10.1111/jfir.12321

Received: 27 January 2021

|

Accepted: 15 January 2023

DOI: 10.1111/jfir.12321

ORIGINAL ARTICLE

Tax policies and agency costs

Diogo Duarte |Brice Dupoyet |Sandrine Docgne |

Florent Rouxelin

College of Business, Florida International

University, Miami, Florida, USA

Correspondence

Diogo Duarte, College of Business, Florida

International University, 11200 S.W. 8th St.,

236, Miami, FL 33199, USA.

Email: diogo.duarte@fiu.edu

Abstract

We show that when large corporations are subject to a

different tax system than smaller firms, the agency cost

of under‐and overinvestment is significantly altered. In

contrast to the findings in the literature, the gap between

the first‐and second‐best investment trigger prices do not

move in lockstep with variations in the corporate tax rate,

as in the case of a linear tax system. We show that the gap

can either widen or shrink, depending on the tax policy

design and regime. In addition, we find that the agency

cost under a progressive tax regime is considerably larger

than the agency cost under a regressive tax regime when

equityholders have to bear all the investment costs. These

results are reversed when managers have the ability to

issue additional debt to finance the firm's expansion and

transfer part of the investment costs to bondholders.

JEL CLASSIFICATION

G32, H25, H26

1|INTRODUCTION

The seminal studies of Jensen and Meckling (1976) and Myers (1977) show that agency costs arising from the

misalignment of interests between equityholders and bondholders ultimately affect the timing of investment and

the balance between debt and equity used to fund these investment opportunities. Quantitatively, agency cost is

typically measured as the difference between the (levered) firm value when managers follow the first‐best policy

(i.e., managers maximize the total value of the levered firm) and the second‐best policy (i.e., managers maximize

equity value instead of the total value of the firm).

J Financ Res. 2023;46:383–409. wileyonlinelibrary.com/journal/JFIR

|

383

© 2023 The Southern Finance Association and the Southwestern Finance Association.

The agency cost of underinvestment (see Mauer & Ott, 2000; Parrino & Weisbach, 1999,Pawlina,2010)

typically emerges when an equity‐maximizer manager delays investment decisions relative to a firm‐value‐

maximizer manager for bearing all the investment costs while sharing the benefits of investment with

bondholders. In contrast, the agency cost of overinvestment (see Childs et al., 2005;Leland,1998;Mauer&

Sarkar, 2005; Morellec, 2001) can emerge when an equity‐maximizer manager hastens her decision to invest

relative to a firm‐value‐maximizer manager for being able to shift part of the investment cost to bondholders

with additional debt issuance while preserving the upside potential gains of equity. In spite of the fact that

researchers have been devoting a lot of effort to investigate potential factors affecting the capital structure and

agency costs, such as debt restructuring (Goldstein et al., 2001), macroeconomic conditions (Hackbarth

et al., 2006), managerial traits (Hackbarth, 2008), asymmetric information (Morellec & Schürhoff, 2011), and lack

of coordination between the timing of investment and debt financing (Li & Mauer, 2016), it is still unclear

how different tax regimes and their design affect the agency cost of under‐and overinvestment of large

corporations.

We fill this gap by investigating the implications for investment decisions, capital structure, and the agency

costs of under‐and overinvestment when large firms are subject to a different corporate tax rate. Our model

builds on the real‐option dynamic framework of Mauer and Ott (2000) and Hackbarth and Mauer (2012)and

investigates the importance of tax heterogeneity and regimes to agency costs. Our theoretical model goes as

follows. At the beginning, a young firm is endowed with a growth option that can be exercised at any point in

time and its operating income is taxed at the rate of τ

1

. Depending on the level of its operating income, the firm

can either abandon its operations or mature (i.e., expand) by exercising the growth option. After expansion, if

thematurefirm'soperatingincomebecomessufficiently large (i.e., above an exogenous threshold level that we

label tax cutoff point

L

), a different marginal tax rate τ

2is levied upon it and the overall effective tax rate is thus a

weighted average between τ

1

and τ

2.

We consider two cases. First, the marginal corporate tax rate τ

2levied on a large operating income is larger than

the initial tax rate τ

1

(i.e., ττ<

1

2

). We call this system the progressive tax regime. One can interpret this case as large

firms drawing the attention of legislators that pass a targeted tax reform imposing a higher corporate tax rate on

large firms. In the second case, we assume that the marginal tax rate τ

2levied on large operating incomes is smaller

than the initial tax rate τ

1

(i.e., ττ<

2

1

). We label this system the regressive tax regime and interpret it as large firms

hiring high‐skilled accountants and lawyers that conduct aggressive tax avoidance. We benchmark our results

against the case where operating income is subject to a linear (flat) tax rate.

Our main findings are as follows. First, we show that the agency cost of underinvestment under a

progressive tax regime is considerably larger than under a regressive tax code when investment costs are

all‐equity financed. In our numerical analysis, the agency cost of underinvestment under a progressive tax code

can be 2.6 times larger than the agency cost of underinvestment under a regressive tax code. The result follows

fromthefactthatthetaxpolicyterms τττ L

(

,/,

)

112 affect the first‐and second‐best investment trigger prices

(i.e., the output price of the goods produced by the firm in which the total‐firm‐value‐maximizer and equity‐

maximizer managers invest, respectively) in opposite directions. Different from the findings of Mauer and Ott

(2000) that the first‐and second‐best investment trigger prices move in lockstep with variations in the linear tax

rate in a flat tax system, we show that in the presence of tax heterogeneity, the underinvestment problem is

aggravated under a progressive tax regime and alleviated under a regressive tax regime. These opposing effects

on the investment trigger prices caused by the granularity of the tax policy design (i.e., two tax brackets instead

of one) is carried out to other (partial) equilibrium quantities, such as the agency cost components (i.e., tax shield

of debt and bankruptcy costs), equity betas, credit spreads, and leverage ratios.

Second, we find that the first‐best policy investment trigger price is more sensitive to changes in the tax policy

terms than the second‐best investment trigger price when equityholders bear all the investment cost of the option

exercise. The main reason is that tax dispersion, defined as

τττ

Δ

=| −|

21

, significantly di storts the trade‐off between

tax shield and bankruptcy cost, resulting in large variations in the trigger price that maximizes the total value of the

384

|

JOURNAL OF FINANCIAL RESEARCH