Dividend commitment in firm bylaws and capital structure change

• Published date: 01 November 2021
• Author: Linyin Cheng,Guojun Wang,Yuetang Wang,Dan Yang
• Date: 01 November 2021
• DOI: http://doi.org/10.1111/eufm.12293

DOI: 10.1111/eufm.12293

ORIGINAL ARTICLE

Dividend commitment in firm bylaws and

capital structure change

Linyin Cheng

1

|Guojun Wang

2

|Yuetang Wang

1

|Dan Yang

3

1

Business School, Nanjing University,

Nanjing, China

2

School of Management and

Engineering, Nanjing University,

Nanjing, China

3

Business School, Beijing Normal

University, Beijing, China

Correspondence

Guojun Wang, School of Management

and Engineering, Nanjing University,

210093 Nanjing, China.

Email: wangguojun@nju.edu.cn

Funding information

National Natural Science Foundation of

China, Grant/Award Numbers:

71602010, 71602085, 71672082, 72072082

Abstract

Dividend policies can predict changes in capital

structure. We focus on a unique setting, namely, divi-

dend commitment emerging from a new dividend

policy in China, and explore its relation to leverage. We

find a commitment to increase dividends is associated

with a subsequent reduction in leverage, and this ne-

gative relation is enhanced in firms without pre-

liminary conditions specified for their commitments or

with greater dividend smoothing in previous periods.

Robustness tests support the main findings. Further

analysis addresses four internal mechanisms that play

essential roles in the above link, and we provide further

details about how leverage is reduced.

KEYWORDS

capital structure, China, dividend commitment, dividend policy,

leverage

JEL CLASSIFICATION

G32; G35

1|INTRODUCTION

Among capital structure theories,

1

a revised trade‐off theory recognizes that capital market

imperfections generate costs for capital structure changes (Frank & Goyal, 2005). Firm‐level

heterogeneity influencing such costs of change affects capital structure choices (Flannery &

Rangan, 2006; Lemmon et al., 2008). In this paper, we focus on a firm‐level factor, dividend

EUROPEAN

FINANCIAL MANAGEMENT

Eur Financ Manag. 2021;27:841–864. wileyonlinelibrary.com/journal/eufm © 2020 John Wiley & Sons Ltd.

|

841

We are grateful to John A. Doukas (the editor) and two anonymous referees for their detailed and constructive

comments. All errors and omissions are our own.

policy, and explore whether a firm's dividend policy is associated with a change in capital

structure. Dividend policy can affect capital providers' desire to offer their money, which is

linked to the cost of capital structure change. Accordingly, various dividend policies could

predict different changes in firm leverage. Unlike prior research that uses historical dividend

levels as a measure of dividend policy, we focus on a unique setting, that is, dividend com-

mitment emerging from a dividend policy recently implemented in the Chinese market. We

attempt to answer the following research question: is the quantitative level of a firm's dividend

commitment associated with a change in its leverage?

We focus on the unique setting of China for the following reasons. First, compared to

historical dividend levels, which represent an implicit commitment, dividend commitment in

China implies an explicit commitment to maintain a given dividend level and presents a new

type of dividend policy. Second, Chinese legislators have adopted a dividend commitment

policy to boost cash dividends, because Chinese listed firms normally underpay dividends

(Jiang et al., 2017). Such a policy provides us with a novel opportunity to explore whether a

dividend policy can act as a determinant of capital structure change. Third, differing from firms

in developed countries, where ownership is relatively dispersed (e.g., the Anglo‐American

counterparts), Chinese firms typically have highly concentrated ownership. In this sense,

agency problems arise between controlling and minority shareholders. Because the relation

between dividends and leverage can differ for different agency stories (Fama & French, 2002),

evidence based on developed countries cannot be generalized to the Chinese case.

Our research design employs the method of generalized difference‐in‐differences (DID),

because the commitment levels of Chinese listed firms making dividend commitments sti-

mulated by the regulation differ greatly. Specifically, the commitment percentage of some firms

is larger than the percentage of cash dividends they paid in previous periods, whereas the

commitment percentage of other firms is smaller. Thus, we identify firms with a larger divi-

dend commitment (larger than the average percentage of cash dividends they paid in the

previous 3 years) as the treatment group, and firms with a smaller dividend commitment

(smaller than the average percentage of cash dividends they paid in the previous 3 years) as the

control group. We then explore the differences in leverage changes between the treatment and

control groups after the dividend commitment is made.

2

To minimize any omitted variable bias,

we control for firm‐level factors identified in prior studies as affecting leverage. We also control

for firm and year fixed effects.

Using a sample of all Chinese firms listed between 2007 and 2019, we find that commitment

to increasing dividends is associated with a subsequent reduction in leverage, evidenced by the

treatment group presenting a lower level of leverage after their dividend commitment, com-

pared to the control group. Such a negative relation is enhanced in firms without preliminary

conditions in their commitments or in firms with greater dividend smoothing in previous

periods. We conduct several robustness tests to address possible endogeneity. First, considering

that firms with good financial performance and corporate governance could be more likely to

1

Capital structure theories include classic trade‐off theory (Jensen & Meckling, 1976; Myers, 1984), a revised trade‐off

theory (Frank & Goyal, 2005), pecking order theory (Myers & Majluf, 1984), and market timing theory (Baker &

Wurgler, 2002). Current capital structure research largely advocates the revised trade‐off theory.

2

Normally, in this case, firms that made a dividend commitment in response to the mandatory regulation should be

identified as the treatment group and firms that voluntarily made a dividend commitment before the regulation was

enacted should form the control group. However, due to the rather small number of firms that voluntarily made a

dividend commitment before the regulation was implemented (only 26 cases), this methodology could invalidate the

statistical results. Thus, we alternatively employ a generalized DID method.

842

|

EUROPEAN

FINANCIAL MANAGEMENT

CHENG ET AL.