Corporate governance risk and the agency problem

• Date: 03 April 2018
• DOI: https://doi.org/10.1108/CG-08-2017-0195
• Pages: 254-269
• Published date: 03 April 2018
• Author: Walaa Wahid ElKelish
• Subject Matter: Strategy,Corporate governance

Corporate governance risk and the

agency problem

Walaa Wahid ElKelish

Abstract

Purpose –This study aims to investigate the relationship between corporate governance risk and

agencycosts across different countries.

Design/methodology/approach –Corporate governance risk indicators were obtained from the

Institutional Shareholder Services Europe (S.A.) for 4,135 firms across 27 countries. Agency costs and

other control variables were derived from companies’ annual financial reports using the DataStream

database. Ordinary least squares multiple regression analysis model was used to test the study

hypothesis.

Findings –Agency costs have a significant negative impact on corporate governance risk across

countries. The extent of corporate governance mechanisms used, however, varies across geographic

regions and industry types. The relationship between corporate governance risk and agency costs is

more obvious in the non-financialthan financial sector. These results were robustafter several statistical

checks.

Practical implications –The findings will help stakeholders, including corporate management,

regulators and investorsto improve corporate governance mechanisms and capital allocationdecisions

acrosscountries.

Originality/value –Evidence is provided on the role of agency costs in corporate governance risk

across geographic regions for financial and non-financial companies. The paper also overcomes

common problems in corporate governance research such as construct validity, limited data and

endogeneity.

Keywords Agency costs, Corporate governance, Financial and non-financial companies,

International variations

Paper type Research paper

1. Introduction

Previous empirical studies have often used the agencytheory (Jensen and Meckling, 1976)

to explain corporate governance practices (Aguilera and Jackson, 2010). This entails that a

conflict of interest between managers and shareholders creates agency problems, which

can be resolved using contractual agreements. Some scholars have argued, however, that

the basic principal–agent model provides limited solutions to agency problems arising with

other stakeholders, and undervalues international variations (Young et al., 2008). Some

empirical studies have, therefore, highlighted the importance of firm-specific characteristics

and/or national institutional factors such as legal, political, economic and cultural factors on

corporate governance mechanisms (Aggarwal et al., 2009;Ammann et al.,2013). The

results of these studies arevaried across countries (Doidge et al.,2007).

The main purpose of this paper is, therefore, to investigate the relationship between

corporate governance risk and agency costs across countries during 2014. The aim is to

highlight the heterogeneous nature of corporate governance models because of the

different agency problems in different geographic areas. This may provide useful

explanations for the mixed results in previous empirical studies. Corporate governance risk

Walaa Wahid ElKelish is

Assistant Professor at

Accounting Department,

College of Business

Administration (AACSB),

University of Sharjah,

Sharjah, United Arab

Emirates.

JEL classification –M40, M41

Received 28 August 2017

Revised 25 September 2017

Accepted 2 October 2017

PAGE 254 jCORPORATE GOVERNANCE jVOL. 18 NO. 2, 2018, PP. 254-269, © EMERALD PUBLISHING LIMITED, ISSN 1472-0701 DOI 10.1108/CG-08-2017-0195

is defined as the “strength of governance mechanisms”, with strong governance

mechanisms indicating low risk of wealth expropriation and vice versa. The corporate

governance index is measured at the firm level and includes four dimensions: board

structure, compensation/remuneration, shareholders’ rights and audit practices. This

study’s findings suggest that corporate governance risk is related to agency costs and

other control variables. More specifically, the strength of these relationships is dependent

upon the geographic region and industry type. This implies that different corporate

governance models are applied to suit localagency conditions across countries.

This paper contributes to the literatureby investigating the impact of a comprehensive list of

internal and external factors on corporate governance risk across countries. It uses a

composite index of corporate governance practices, rather than single mechanisms, to

reflect a comprehensive range of internal and external control systems. The paper includes

a detailed analysis of several corporate governance risk sub-dimensions to highlight

possible improvements. The empirical analysis used a large sample of firms across several

developed and developing countries, unlike previous studies, which focused mainly on

particular western countries, such as the USA. More importantly, the study addresses

several standard problems of statistical models, including construct validity, limited data

and endogeneity. In this paper, Section 2 deals with the background and hypothesis

development, Section 3 presents the data collection and methodology and Sections 4 and

5, respectively, contain the empiricalresults and conclusions.

2. Background

The agency theory (Jensen and Meckling, 1976) predicts a conflict of interests between

corporate management and shareholders because managers can pursue their own

objectives in defiance of the interests of other shareholders. This basic principal–agent

model can also create information asymmetry, which allows managers to withhold important

information to maximize personal interests (Godfrey et al.,2003). Companies with a high

ownership concentration may sufferfrom principal–principal agency problems, where major

shareholders expropriate company resources from minority shareholders,using techniques

like tunneling. Contractual arrangements between management and shareholders can be

used to address these agency problems by providing incentives or constraints for

managers. It is further argued that the original principal–agent model may provide only a

limited solution to the agency problem, and undervalue different formal and informal

institutions across countries. This implies that different corporate governance models may

be used to deal with agency costs across countries to suit local conditions.

Previous studies on corporate governance determinants show heterogeneous results

across countries because of the existence of different firm-specific and country

characteristics. Firm-specific factors include level growth opportunities, external financing

needs, firm size, ownership structure, firm risk and industry characteristics (Aggarwal et al.,

2009;Ammann et al., 2013;Barucci and Falini, 2005;Brick and Chidambaran, 2008;

Doidge et al.,2007;Durnev and Kim, 2005;Giroud and Mueller, 2011). For example,

Durnev and Kim (2005) found a positive relationship between the quality of governance,

disclosure and growth opportunities, need for external financing and concentrated cash

flow rights using data from 859 firms across 25 countries. Some studies have found a

different evidence for developed and developing countries. Doidge et al. (2007) showed

that firm characteristics explain corporate governance in developed countries more than

that in developing countries. Other studies focused on country-level characteristics,

including legal (La Porta et al.,1998), political and economic (Kim et al., 2017;Roe and

Siegel, 2011). National cultural values have also been shown to affect corporate

governance practices (Sarahand Catherine, 2017).

This led to the emergence of two main research themes. The first debates whether firm-

specific or national factors are more important in determining the level of corporate

VOL. 18 NO. 2, 2018 jCORPORATE GOVERNANCE jPAGE 255