Can board diversity predict the risk of financial distress?

• Published date: 20 January 2021
• DOI: https://doi.org/10.1108/CG-06-2020-0252
• Date: 20 January 2021
• Pages: 663-684
• Subject Matter: Strategy,Corporate governance
• Author: Umair Bin Yousaf,Khalil Jebran,Man Wang

Can board diversity predict the risk of

f‌inancial distress?

Umair Bin Yousaf, Khalil Jebran and Man Wang

Abstract

Purpose –The purpose of this study is to explorewhether different board diversity attributes (corporate

governanceaspect) can be used to predict financial distress.This study also aims to identify what type of

predictionmodels are more applicable to capture board diversityalong with conventional predictors.

Design/methodology/approach –This study used Chinese A-listed companies during 2007–2016.

Board diversitydimensions of gender, age, education, expertiseand independence are categorized into

three broad categories; relation-oriented diversity (age and gender), task-oriented diversity (expertise

and education) and structuraldiversity (independence). The data is divided into testand validation sets.

Six statisticaland machine learning models that includedlogistic regression, dynamic hazard, K-nearest

neighbor, random forest (RF), bagging and boosting were compared on Type I errors, Type II errors,

accuracyand area under the curve.

Findings –The results indicate that board diversity attributes can significantly predict the financial

distress of firms. Overall, the machine learning models perform better and the best model in terms of

Type I error and accuracyis RF.

Practical implications –This study not only highlightssymptoms but also causes of financial distress,

which are deeply rooted in weak corporate governance. The result of the study can be used in future

credit risk assessment by incorporating board diversity attributes. The study has implications for

academicians,practitioners and nominationcommittees.

Originality/value –To the best of the authors’ knowledge, this study is the first to comprehensively

investigatehow different attributes of diversity canpredict financial distress in Chinese firms.Further, this

study alsoexplores, which financial distressprediction models can show betterpredictive power.

Keywords China, Machine learning, Financial distress, Board diversity, Relation-oriented diversity,

Task-oriented diversity, Structural diversity

Paper type Research paper

1. Introduction

A firm’s financial distress causes a substantial cost to investors and creditors including, but not

limited to, loss of sales and profits, reduced dividends, legal costs, high cost of further capital and

credit, tax avoidance, inability to issue new securities and the opportunity cost of positive NPV

projects (Bhattacharjee and Han, 2014;Habib et al., 2018;Zhou, 2019). Accordingly, much

attention has been paid to financial distress on both academic and practical ends. A vast majority

of financial distress prediction (FDP) literature can be classified into statistic al and machine

learning models that use accounting and market ratios as predictors (Altman, 1968;Barboza et al.,

2017;Campbell et al., 2008;Hillegeist et al., 2004;Lohmann and Ohliger, 2019;Ohlson, 1980;

Shumway, 2001;Taffler, 1983;Wang, 2017;Zmijweski, 1984). Several studies have explained the

relationship between corporate governance and financial distress (Abdullah, 2006;Chaganti et al.,

1985;Daily and Dalton, 1994;Darrat et al., 2014;Elloumi and Gueyie

´, 2001;Fich and Slezak,

2008;Lajili and Ze

´ghal, 2010;Lee and Yeh, 2004;Li et al., 2008;Muranda, 2006;Parker et al.,

2002;Shahwan, 2015;Udin et al., 2017). However, incorporating corporate governance measures

into FDP models have received less attention in the literature.

Umair Bin Yousaf is based

at the School of

Accounting, China Internal

Control Research Center,

Dongbei University of

Finance and Economics,

Dalian, China. Khalil Jebran

is based at the School of

Business Administration,

Dongbei University of

Finance and Economics,

Dalian, China. Man Wang is

based at the School of

Accounting, China Internal

Control Research Center,

Dongbei University of

Finance and Economics,

Dalian, China.

Received 21 June 2020

Revised 26 August 2020

5 October 2020

9 October 2020

17 November 2020

Accepted 8 December 2020

The authors are thankful to

Gagan Deep Sharma

(Associate Editor) and two

anonymous reviewers for many

insightful comments and

suggestions.

DOI 10.1108/CG-06-2020-0252 VOL. 21 NO. 4 2021, pp. 663-684, ©Emerald Publishing Limited, ISSN 1472-0701 jCORPORATE GOVERNANCE jPAGE 663

Moreover, it has been realized that the accuracy of standard accounting-based FDP

models has dropped significantly in recent times (Beaver et al., 2012). Therefore,

researchers argue for additional predictors to be incorporated into FDP models (Altman

et al.,2010

;Beaver et al.,2012).

Over the years, the diversity of the board of directors has captured a considerable debate

in corporate governance literature. Based on upper echelons (Hambrick and Mason,1984),

agency (Jensen and Meckling, 1976) and resource dependence theories (Pfeffer and

Salancik, 1978), we argue that diversity is an essential element for an organization’s

success and if we incorporate board diversity attributes in our prediction models, the FDP

ability improves significantly.The purpose of this study is not to prove a causal relationship,

but to assess the predictive power of board diversity in predicting financial distress.

Considering the limitations of some methods and to add robustness to our results, we

compare statistical models (static and dynamic) with machine learning models. Thus, this

study also explores, which FDP models better use corporate governance information along

with macro-economic, accounting, market and growth information to predict financial

distress (FD).

Prior studies have mostly focused on one or twofeatures of board diversity when examining

the relationship with FD (Kristanti et al.,2016;Manzaneque et al.,2015;Mittal and Lavina,

2018;Santen and Donker, 2009;Zhou, 2019). This study uses different facets of board

diversity, namely; gender, age, education, expertize and independence and categorizes

them into three broad dimensions of relation-oriented diversity, task-oriented diversity and

structural diversity to predictFD.

We used a sample of Chinese firms from 2007 to 2016. China provides a unique context to

explore the association between board diversity and FD. China has undergone an

enormous transition from a centrally planned to a market economy over the past three

decades. However, unnecessary protection of state-owned enterprises (SOEs), limited

access to financial resources that mainly concentrated on SOEs, poor protection of

creditors’ rights and weak legal infrastructure of bankruptcy make China a unique context

that cannot be generalized to Anglo-American studies (Bhat et al.,2019;Bhattacharjee and

Han, 2014;Sabbaghi, 2016;Wangand Deng, 2006).

Our study contributes to the literature in several dimensions. First, we integrate board

diversity with financial distress risk assessment and explore the predictive power of a wide

range of board diversity attributes in financial distress risk assessment. In the financial

distress risk assessment, the main focus is on predictive ability rather than causation.

However, our study allows us to predict moreaccurately as we explore causes (and not the

symptoms) of financial distress,which are deeply rooted in weak board diversification.

Second, most of the previous studies considered only one or two dimensions of

diversity, such as age or gender (Adams and Ferreira, 2009;Francoeur et al., 2008;

Mittal and Lavina, 2018;Talavera et al., 2018;Ullah et al.,2019;Zhou, 2019). However,

our study takes into account five key diversity facets and categorizes them into relation-

oriented (gender and age), task-oriented (education and expertize) and structural

(independence) diversity attributes. This is the first study to collectively incorporate

relation-oriented (gender and age), task-oriented (education and expertize) and

structural diversity (independence) attributes into FDP models.

Third, our study expands the FDP literature by considering the best predictors from

accounting, market, growth, macroeconomic and corporate governance variables by using

stepwise regression on feature selection. Finally, we further add to financial distress

literature by comparing popular static, dynamic and machine learning models. This study

uses a unique definition of financially distressed firms in the context of Chinawhere normally

special treatment (ST) stocksare used as a proxy of financially distressed firms.

PAGE 664 jCORPORATE GOVERNANCE jVOL. 21 NO. 4 2021