Board governance index, economic performance and ...

Corporate Ownership & Control / Volume 8, Issue 1, Fall 2010

CORPORATE OWNERSHIP & CONTROL

КОРПОРАТИВНАЯ СОБСТВЕННОСТЬ И КОНТРОЛЬ

Postal Address:

Почтовый адрес редакции:

Postal Box 36 Sumy 40014 Ukraine

Почтовый ящик 36 г. Сумы, 40014 Украина

Tel: +380-542-611025 Fax: +380-542-611025 e-mail: [email protected] [email protected] www.virtusinterpress.org

Тел.: 38-542-611025 Факс: 38-542-611025 эл. почта: [email protected] [email protected] www.virtusinterpress.org

Journal Corporate Ownership & Control is published four times a year, in September-November, December-February, March-May and June-August, by Publishing House “Virtus Interpress”, Kirova Str. 146/1, office 20, Sumy, 40021, Ukraine.

Журнал "Корпоративная собственность и контроль" издается четыре раза в год в сентябре, декабре, марте, июне издательским домом Виртус Интерпресс, ул. Кирова 146/1, г. Сумы, 40021, Украина.

Information for subscribers: New orders requests should be addressed to the Editor by e-mail. See the section "Subscription details".

Информация для подписчиков: заказ на подписку следует адресовать Редактору журнала по электронной почте.

Back issues: Single issues are available from the Editor. Details, including prices, are available upon request.

Отдельные номера: заказ на приобретение отдельных номеров следует направлять Редактору журнала.

Advertising: For details, please, contact the Editor of the journal.

Размещение рекламы: обращайтесь к Редактору.

Copyright: All rights reserved. No part of this publication may be reproduced, stored or transmitted in any form or by any means without the prior permission in writing of the Publisher.

Права на копирование и распространение: копирование, хранение и распространение материалов журнала в любой форме возможно лишь с письменного разрешения Издательства.

Corporate Ownership & Control

Корпоративная собственность и контроль

ISSN 1727-9232 (printed version) 1810-0368 (CD version) 1810-3057 (online version)

ISSN 1727-9232 (печатная версия) 1810-0368 (версия на компакт-диске) 1810-3057 (электронная версия)

Certificate № 7881

Свидетельство КВ 7881 от 11.09.2003 г.

Virtus Interpress. All rights reserved.

Виртус Интерпресс. Права защищены.

за

информацией

3


EDITORIAL

Dear readers!

Summer of 2010 was a very hot season for regulators and practitioners attracting a strong interest of academics. The US initiatives in corporate governance regulation came first in the board practices which come as a primary field of research including Corporate Ownership and Control journal. Thus, U.S. regulators will let investors owning 3 percent of a company nominate directors on corporate ballots, a step that may help shareholders oust board members accused of overpaying executives and failing to boost shares. The Securities and Exchange Commission required that companies put investor board candidates on the proxy statements sent to stockholders before director elections. Investors or groups that meet the ownership threshold for three years will be eligible to offer nominees. The SEC acted in response to investor complaints that companyselected directors failed to rein in compensation and risk-taking that led to more than $1.79 trillion of financial- company writedowns during the credit crisis. Another point of view for proper research was the issue of the stakeholder participation in corporate governance. Earlier this year the European Commission published a green paper titled “Corporate governance in financial institutions and remuneration policies”. The main issues to discuss were the issues of the stakeholder involvement in corporate governance and the issue of remuneration policy. The recent issue of the journal contains some very interesting findings regarding those issues underlined by the European Commission too. The same issue, namely Director remuneration, has been stressed this Summer in the UK too. The Financial Services Authority has published a consultation paper in which it sets out proposed changes to its Remuneration Code in the light of the Capital Requirements Directive and the Financial Services Act (2010). The role of shareholders in the compensation setting has been underlined in the USA. The Wall Street Reform and Consumer Protection Act was signed in July by the President. The Act will bring about the most significant financial regulatory reform since the 1930s and also includes corporate governance measures. For example, Section 951 provides for a shareholder vote on compensation disclosure and Section 952 sets out requirements for compensation committee independence. As we see the issues of the board practices, director independence, remuneration and stakeholder participation in corporate governance have become the most important for regulators and practitioners. The recent issue of the journal Corporate Ownership and Control pays attention to those issues too, as well as to the issues of corporate ownership and control, financial reporting, etc. I hope that you will enjoy reading the journal and come back to us with your outstanding papers again!

4


CORPORATE OWNERSHIP & CONTROL Volume 8, Issue 1, Fall 2010

CONTENTS

Editorial

4

SECTION 1. ACADEMIC INVESTIGATIONS AND CONCEPTS SUBSTITUTE GOVERNANCE STRUCTURE AND THE EFFECT OF MONITORING: EVIDENCE FROM BANGLADESH

9

Mohammad Azim This paper investigates the role of monitoring mechanisms in a corporate governance structure, focusing on listed companies in a developing country, Bangladesh. Specifically, it examines whether different interrelated monitoring mechanisms - board of directors and committee, management and external auditors - affect firm performance. This research found the possibility of having a substitution or complementary links in monitoring mechanisms that explain why there is no consistent empirical evidence between individual monitoring mechanisms and firm performance. This study has policy implications for the Bangladeshi corporate environment. Progress of implementation of the guidelines appears to be reasonable. However, credibility of the reported figures and quality of implementation remain open to discussion. To what extent these status reports reflect improved governance or are largely a form of paper compliance is a debatable issue. This research also suggests that when considering any change in corporate monitoring, the Bangladeshi government should take into account the nation‟s business, social structure, culture and legal practices. CORPORATE OWNERSHIP CHOICE BASED ON ETHICAL CRITERIA: IS IT BIG ENOUGH TO NOTICE?

24

Charl de Villiers, Chris van Staden Ethical investors often exclude firms that participate in so-called controversial activities, such as tobacco, alcohol, firearms, gambling, the military, and nuclear operations, from their investment portfolios. Firms excluded in this way should experience an increase in their cost of capital and a reduction in their share prices. We use the KLD database to identify S&P 500 firms involved in controversial activities. Our results show no difference between controversial activity firms and other firms regarding relative share price and we find that the cost of capital of controversial activity firms is in fact lower. We conclude that ethical investing, of the type that excludes controversial activity firms, does not influence the capital markets in the expected way. CORPORATE GOVERNANCE AND THE USE OF EVA COMPENSATION

37

Ralph DeFeo, Ehsan Nikbakht, Andrew C. Spieler The purpose of this paper is to determine if companies that chose to incorporate Economic Value Added (EVA®) as part of their executive compensation package tend to have better corporate governance than similar firms who have not chosen to use EVA®. EVA® is an economic profit metric developed by Stern Stewart & Co., which is calculated by taking the Net Operating Profits after Taxes 5

Corporate Ownership & Control / Volume 8, Issue 1, Fall 2010 (NOPAT) and subtracting a capital charge from it. Through the use of binary logistic regression the strength of several key corporate governance measures were tested in order to ascertain whether they have a significant impact on influencing a firm to use EVA®. A major finding is that firms that employed EVA® as part of their compensation package tend to have a weaker corporate governance. THE VOLUNTARY CSR DISCLOSURE IN CORPORATE ANNUAL REPORTS: EVIDENCE FROM AUSTRALIA

49

Nicholas Andrew, Mark Wickham The relationship between credible Corporate Social Responsibility (CSR) performance and desirable firm outcomes is well established in corporate governance literature. Over the past two decades in particular, there has been an increased recognition of this relationship in the business community and a concomitant increase in the quantity and detail of CSR activities being voluntarily reported by corporations has been observed. The rationale for the increasing levels of voluntary CSR reporting has been attributed to two main corporate strategies: to conform to the expectations of the society and to socially legitimise their operations to their salient stakeholder groups. Whilst there has been extensive academic interest in the concept of CSR, it has focused almost exclusively on normative definitions of the concept, and/or the presentation of empirical evidence that details „why corporations should report their CSR activities‟ and „what CSR activities they should report‟. What is lacking the literature, however, is a focus on the question as to „how do corporations strategically report their CSR activities?‟ We find that there is evidence to support a „Core/Periphery Model‟ of strategic CSR disclosure, which we feel provides a framework for predicting how corporations will voluntarily disclose their CSR performance given the issues, events and/or crises that affect their industry environments. ENTERPRISE VALUE AND DISCLOSURE LEVEL: EVIDENCES IN THE BRAZILIAN MARKET

56

Fabio Gallo Garcia, Elmo Tambosi Filho, Luiz Maurício Franco Moreira There is a strong tendency in global markets towards an enhanced level of corporate transparency regarding the activities of companies and, as a result, information on their performance. The Purpose of this study is to analyze the relationship between greater disclosure levels and shareholder value creation. Increasing levels of disclosure are required from companies‟ management before shareholders and the society in general. Obscure practices that fail to take into consideration the best interests of shareholders increase risks and cause shares to lose liquidity. The São Paulo Stock Exchange‟s “Novo Mercado” (“New Market”) emerged from the intent to improve the Brazilian stock market by adopting best practices in corporate governance, adding transparency to disclosed information, and heightening the respect for the interests of shareholders, whether they may be minority or not. The “Novo Mercado” intends to foster a differentiated environment in which companies committed to corporate governance are recognized and can benefit from better stock prices, resulting in lower placement costs and increased liquidity. Our research will assume that companies with American Depositary Receipts ADRs are committed to a higher level of disclosure as a result of the requirements of the Security Exchange Commission – SEC, and the Financial Accounting Standards Board - FASB; an empiric study about these firms will be performed. We will determine, through a Study Event concerned with cases where ADR have been issued, which consequences of the commitment to higher levels of disclosure as regards shareholder are responsible for value creation, and what are the reflections on the stock price quoted in the Brazilian market. INSIDE AND OUTSIDE SHAREHOLDERS AND MONITORING: EVIDENCE FROM DEVELOPING COUNTRY

62

Mazlina Mustapha, Ayoib Che Ahmad This paper tests the effect of managerial (inside) and block-holders (outside) ownership in relation to agency theory in Malaysian business environment. This study tests the agency relationship in different culture and social contact and provides evidence whether agency theory in non-western organizations have equal impact in Asian organizations. Consistent with agency theory and the convergence of interest hypothesis, managerial ownership (insiders) in Malaysia indicate a negative relationship with the demand for monitoring. This finding may be due to the fact that as the managers are also the owners, there is less conflict, less information asymmetry and less hierarchical organization structure 6

Corporate Ownership & Control / Volume 8, Issue 1, Fall 2010 in the companies, which lead to lower monitoring costs. However, another ownership structure, outside block-holders appear to demand more monitoring. This positive relationship may be explained by their effort to compensate their lack of involvement in the daily transactions and internal decisions of the company, especially in the concentrated business environment in the country. LINKAGES BETWEEN OWNERSHIP CONCENTRATION AND FINANCIAL RATIO COMMUNICATION

76

Norhani Aripin, Greg Tower, Grantley Taylor This study examines the relationship between ownership concentration and the extent of financial ratio disclosures (EFRD) in the 2007 annual reports of Australian listed firms. Using agency theory as theoretical background, it is suggested that firms with more concentrated ownership structures are less likely to provide voluntary disclosure of financial ratios information. The univariate tests demonstrate that profitable firms, those firms audited by Big4 auditors and firms belonging to financial services industry communicate more financial ratio information. OLS regressions show that more dispersed shareholding firms‟ are significantly associated with EFRD. Profitable and larger firms audited by independent and Big4 audit firms additionally reported more extensive financial ratio information.

SECTION 2. BOARD PRACTICES DIMENSION, STRUCTURE AND SKILL MIX IN EUROPEAN BOARDS: ARE THEY CONVERGING TOWARDS A COMMON MODEL OF CORPORATE GOVERNANCE?

87

Paola Schwize, Vincenzo Farina, Valeria Stefanelli Using a large sample of European listed Companies and their Directors our study supports the hypothesis that European corporate governance do not yet converging towards a common model. Our results put in evidence in fact the existence of three dominant corporate governance profiles in European listed companies, different each others. The determinants of these profiles depend on the typical governance system adopted by the same companies and on country‟s culture and history; as a consequence, each of them has a own attitude of the board, in terms of structure, organization and skill mix characteristics. Finally we suggest some important implications on European governance practices. DOES CHAIRMAN INDEPENDENCE MATTER?

108

Hafiza Aishah Hashim, S. Susela Devi This paper examines the relationship between the role of non-executive (independent) chairman and the quality of reported earnings. Recent corporate governance reforms recommend firms to appoint an independent leader to ensure the success of a split leadership structure (The Combined Code on corporate Governance, 2006; Higgs Report, 2003). Research on leadership structure to date has tended to focus solely on role duality and find weak or insignificance relationship between role duality and financial reporting quality. Although separating the roles of the chairman and the CEO seems appropriate, researcher argue that it would not necessarily lead to independence of the board if the chairman is not independent. Consistent with recent recommendations to strengthen board leadership by appointing an independent chair, this study evidences a positive and significant association between non-executive chairman and earnings quality in Malaysia. The study suggests that the nonexecutive status of the chairman is an important mechanism in enhancing the board‟s independence, thus improving earnings quality.

7


SECTION 3. CORPORATE GOVERNANCE: ITALY MIMETIC ISOMORPHISM IN THE GOVERNANCE OF IPO COMPANIES IN ITALY

117

Alberto Barbesta, Giancarlo Giudici, Stefano Lugo In order to comply with listing requirements and overcome information asymmetries, listing companies may be encouraged to adapt themselves with market standards („isomorphism‟) in the setting of governance devices in order to reduce the perceived uncertainty and obtain legitimacy towards investors. In this work we evaluate the isomorphism of IPO companies with respect to the board characteristics (i.e. board size and members‟ age). By analyzing a sample of 121 companies listed from 1999 to 2008 on the Italian Exchange, we find that mimetic strategies are frequent in IPO companies, and that the majority of them exhibit a reduction in the differences of board characteristics in the year after the flotation, compared to listed firms in the same sector. The percentage of mimicking companies is even larger if we consider only companies that introduce changes in the board composition. Multivariate analyses suggest that isomorphism strategies are targeted to signal the IPO firm‟s quality, and are an alternative to issuing underpriced shares. CORPORATE GOVERNANCE IN ITALIAN LISTED COMPANIES AND COMPLIANCE TO BEST PRACTICES: AN EMPIRICAL RESEARCH

123

Giulia Romano This study aims to: (1) monitor the evolution of corporate governance practices in Italian listed companies; (2) evaluate the formal and substantial compliance to the Corporate Governance Code; (3) define if listing segment/market (and consequently also capitalization) is significantly related with the adoption of best practices. The research arose from the analysis of data contained in an unpublished source of information made available by the public authority responsible for regulating the Italian securities market. We find that nearly every listed company adopts the Code and its most important best practices. However, we find some contradictory discrepancies. We also find that the adoption of the Code is significantly related to listing segment/market.

SUBSCRIPTION DETAILS

8

140


РАЗДЕЛ 1 НАУЧНЫЕ ИССЛЕДОВАНИЯ И КОНЦЕПЦИИ SECTION 1 ACADEMIC INVESTIGATIONS & CONCEPTS

SUBSTITUTE GOVERNANCE STRUCTURE AND THE EFFECT OF MONITORING: EVIDENCE FROM BANGLADESH Mohammad Azim* Abstract This paper investigates the role of monitoring mechanisms in a corporate governance structure, focusing on listed companies in a developing country, Bangladesh. Specifically, it examines whether different interrelated monitoring mechanisms - board of directors and committee, management and external auditors - affect firm performance. This research found the possibility of having a substitution or complementary links in monitoring mechanisms that explain why there is no consistent empirical evidence between individual monitoring mechanisms and firm performance. This study has policy implications for the Bangladeshi corporate environment. Progress of implementation of the guidelines appears to be reasonable. However, credibility of the reported figures and quality of implementation remain open to discussion. To what extent these status reports reflect improved governance or are largely a form of paper compliance is a debatable issue. This research also suggests that when considering any change in corporate monitoring, the Bangladeshi government should take into account the nation‟s business, social structure, culture and legal practices. Keywords: Bangladesh; Monitoring; Performance; Structural Equation Modelling; Corporate Governance Guideline *Accounting and Finance, Swinburne University of Technology, Faculty of Business and Enterprise, Hawthorn VIC 3122, AUSTRALIA Tel: +61 3 9214 4500 Fax: +61 3 9819 2117 Email: [email protected]

9

Corporate Ownership & Control / Volume 8, Issue 1, Fall 2010 Introduction The separation of ownership and control in publicly owned companies has the potential to create conflict between the interests of managers and shareholders. Such conflict can be reduced by devising effective monitoring mechanisms. However, corporate governance is the product of a complex set of cultural, economic, and social and legal structures; which differ from country to country. It is appropriate that corporate governance guidelines and practice codes be designed and adopted by each constituent country. Therefore, to reduce conflict between shareholders (principal) and management (agent), it is necessary to have culturally, economic and socially specific corporate governance practices that can be monitored by capital market regulators. Previous research has focused on examining different governance mechanisms, typically studying one or two governance variable(s), such as board of directors (Hermalin and Weisbach, 2001), management ownership (Shivdasani, 1993; Kaplan and Minton, 1994), audit committees (Ramsay, 2001), and external auditors (Watts and Zimmerman, 1990), in terms of their impact on firm performance. These studies are largely based on American, British, Japanese, German and Australian companies (Farrar, 2008). Very little research has been done on developing countries. Firstly, however, firm performance depends on the efficiency of a bundle of monitoring mechanisms in controlling the agency problem (Rediker and Seth, 1995, p. 87; Agrawal and Knoeber, 1996, p. 378). Secondly, the emerging capital markets in developing countries differ from developed ones in terms of legal, institutional, political and regulatory practices. The importance of corporate governance practices in emerging economies is increasingly evident to domestic as well as international bodies (Farooque, 2007). In recent years corporate governance has emerged as an important issue for Bangladesh due to the ongoing effects of globalisation, as the domestic economy integrates with the global economy and companies strive to gain international competitiveness (Farooque, 2007). Bangladesh, a developing country, has a very different capital market, board structures and governance objectives than the USA, UK, Australia, Germany and Japan. Conclusions reached in previous studies may not be applicable to Bangladesh. This study is motivated by a need for understanding how monitoring mechanisms work in Bangladesh, and the role of market-specific factors versus governance characteristics in determining the effectiveness of Bangladeshi corporate monitoring mechanisms, especially following the introduction of the Corporate Governance Guideline by the Securities and Exchange Commission in 2006. Little research has been done on Bangladesh in examining the implication of the corporate governance guidelines 10

(hereafter „guideline‟) on company performance. Therefore it is important to understand how effective and efficient such guidelines are in addressing the needs of the capital market. Market regulators are seriously enforcing the implementation of this guideline to reforming companies‟ corporate monitoring structures. The guideline includes areas such as board size, independent directors, chairperson and CEO, internal control and audit, function of company secretary, audit committee and appointment of an external auditor. The guideline is still on a „comply or explain‟ basis. This research develops a conceptual model to classify monitoring mechanisms and map their potential effects on firm performance. Profit-based firm performance measures are used as representative of shareholders‟ interests in the model. For testing the propositions this research uses the Structural Equation Modelling (SEM) technique. This specific statistical tool is used here because SEM is able to deal with multicollinearity and reveals potential complex interrelationships between monitoring mechanisms. There are three specific reasons why this research uses SEM. Firstly, in this research on corporate monitoring, it is not possible to observe most of the variables directly and these are termed latent variables. SEM is able to incorporate latent variables into the analysis. By using SEM with multiple indicator variables, it is possible to model important latent variables. Secondly, in all multivariate analyses it is assumed that there is no error in the variables but from practical and theoretical perspectives it is impossible to perfectly measure a latent concept without some degree of error. However, SEM makes it possible to account for error and improve the statistical estimate. Thirdly, SEM is a powerful tool with which to measure multicollinearity in sets of predictor variables. The SEM examines a series of dependent relationships simultaneously. This is particularly useful when one dependent variable becomes an independent variable in subsequent dependent relationships (Wooldridge, 2003). This paper is organised into six sections: section 2 discuss the theoretical basis for this research. Section 3 discusses capital market in Bangladesh, followed by a literature review for this research in section 4. Section 5 develops research propositions, and methodology. Analysis of the results is presented in section 6 and section 7 provides concluding comments and highlights possible future research. Theoretical Basis for this Research Agency theory is concerned with aligning the interests of owners and managers (Jensen and Meckling, 1976; Fama, 1980; Fama and Jensen, 1983a,b) and is based on the premise that there is an inherent conflict between the interests of owners and managers (Fama and Jensen, 1983a). According to contracting theory, the agreed agent‟s role is to act in the best interests of

Corporate Ownership & Control / Volume 8, Issue 1, Fall 2010 the principal. In reality, however, managers‟ motive may be to ensure their own job security, prestige and personal wealth. This motive gives rise to decisions and actions that conflict with the interests of shareholders. The clear implication for corporate governance, from an agency theory perspective, is that adequate monitoring or control mechanisms need to be established to protect shareholders from the agency risk in the modern corporation (Fama and Jensen, 1983a,b). In general, agency risk has a negative impact on firm performance (McColgan, 2001). Agency risk results in costs: monitoring costs, which refer to monitoring the agent‟s behaviour; bonding costs, which are the costs associated with inducing the management to work in the best interests of the principal; and residual loss, any remaining losses to shareholders when the agents‟ and principals‟ interests are at odds with each other (Jensen and Meckling, 1976). In contrast, stewardship theory assumes that managers behave as the stewards of a company‟s assets, not agents for shareholders (Donaldson, 1990; Donaldson and Davis, 1991; 1994). Stewardship theory deduces that superior corporate performance is linked to the existence of a majority of inside directors. These inside directors, it is argued, hold a steward‟s perspective, meaning they exercise their intimate understanding of the business, their commitment and access to operating information and technical expertise, all in the interests of the company. These factors give them an advantage over outside directors (Donaldson, 1990; Donaldson and Davis, 1991). The theory argues that the economic performance of a firm increases when power and authority is concentrated in a single executive, who is not distracted by external non-executive directors (Donaldson and Davis, 1991). In this respect, stewardship theory directly challenges agency theory where monitoring is believed to affect firm performance. In this paper, agency theory assumes the dominant perspective, i.e. assumptions about the management-shareholder conflict problem and the need to increase monitoring cost by implementing various corporate governance structures and mechanisms. This paper focuses on the complex interrelationships between corporate governance monitoring mechanisms and their substitution or complementary effects on key corporate performance measures that are important to shareholders. The literature has been begun to address the issue of lack of theory at a more micro-level to explain complex interrelationships between corporate governance mechanisms. To some extent, the vacuum of formal theory has been filled by empirical research (Hermalin and Weisbach, 2003). The empirical literature on the monitoring functions of management, board of directors and auditors is well developed, whereas a theory that can underpin the complexities

of these alternative monitoring mechanisms is still in its infancy. It is likely that subsequent development in theory will lead to more sophisticated design and interpretation for future empirical analysis (Hermalin and Weisbach, 2003). Capital Market in Bangladesh Bangladesh is one of the least developed countries (USAid, 2007) in South East Asia. The capital market is relatively underdeveloped with a market capitalisation of only 6% of gross domestic product (GDP) (“Roundtable discussion”, 2006). There are two stock exchanges in Bangladesh: firstly, Dhaka Stock Exchange (DSE) established in 1954; and secondly, Chittagong Stock Exchange (CSE) established in 1995. Both exchanges are private sector entities, self-regulated and have their own Security and Exchange Commission-approved operating rules. Currently there are 351 companies listed on DSE and 243 companies on CSE (SEC, 2009). As the regulatory authority of Bangladesh‟s capital market, the SEC was established on June 8, 1993 under the Securities and Exchange Commission Act, 1993 (Act XV of 1993). To govern the corporate environment in Bangladesh the following legal measures are operating: (i) Securities and Exchange Ordinance 1969, (ii) Bangladesh Bank Order 1972, (iii) Bank Companies Act 1991, (iv) Financial Institutions Act 1993, (v) Securities and Exchange Commission Act 1993, (vi) Companies Act 1994, and (vii) Bankruptcy Act 1997. There are a number of factors hindering the development of Bangladesh‟s capital market. Solaiman (2006, p. 195) found that „the existence of weak legal and regulatory frameworks, the absence of active market professionals, the predominance of individual investors, and a serious dearth of foreign and institutional investors‟, seriously hinder the development of a capital market. There have been some attempts at reform but most initiatives are far from ready. Nothing significant has been done to protect investors except the corporate governance guideline introduced in 2006. The neighbouring countries are well ahead vis à-vis Bangladesh in terms of depth of capital market. For example, in India, Pakistan and Sri Lanka, the market capitalisation is 56%, 30% and 18% of their GDP respectively. In Bangladesh, corporate governance guidelines are still on a “comply or explain” basis, providing some „breathing space‟ for the companies to implement on the basis of their abilities. Around 66.7% of the companies adopted corporate governance and 43.3% have a compliance policy so that they are consistent with national or international benchmarks.

11

Corporate Ownership & Control / Volume 8, Issue 1, Fall 2010 Corporate Governance Guidelines The introduction of the Guidelines in February 2006 is expected to provide effective monitoring and legal protection to investors. It also expects to enhance the confidence of investors. Under the Guidelines, all companies listed on Bangladesh‟s stock exchanges need to follow a “if not why not” approach. Again, directors should state in statutory declarations which conditions the company complies with and those that they are not, with explanations for non-compliance. These Guidelines contain three major monitoring groups: the (i) board of directors; (ii) chief financial officer, head of internal audit and company secretary and audit committee; and (iii) external auditors. Existing literature supports the premise that these monitoring mechanisms influence firm performance as shown in Shivdasani (1993), Kaplan and Minton (1994), Hermalin and Weisbach (2003) and Holderness (2003). [Figure 1 About Here] Board of Directors The first category in the Guidelines relates to the board of directors. Boards are responsible for ensuring that management‟s behaviour and actions are consistent with the interests of shareholders. Boards have the power to hire, fire and compensate executive managers and to ratify and monitor important decisions (Fama and Jensen 1983a; Jensen, 1993). The theoretical role of the board in monitoring and disciplining management is firmly grounded in the agency framework of Fama and Jensen (1983a,b). Empirical examination of board characteristics and firm performance focuses on: board size (e.g. Jensen, 1993; Yermack, 1996); independent directors (Dechow et al., 1996; Beasley, 1996, 2001); separate role of chairperson (Chair) and Chief Executive Officer (CEO) (Jensen, 1993); financial literacy (DeZoort, 1997); and board committees (Austin, 2002; Menon and Williams, 1994). Board size: In general larger boards are more likely to be vigilant for agency problems, simply because a greater number of people are reviewing management actions (Kiel and Nicholson, 2003). However, if the size of the board becomes too large (beyond the standard threshold), effectiveness of monitoring diminishes (Ryan and Wiggins, 2004). A large board increases problems of free-riding and it becomes difficult for directors to express their ideas and opinions in the limited time available during meetings (Golden and Zajac, 2001). It is also suggested that too large boards are relatively ineffective and difficult for the CEO to control (Lipton and Lorsch, 1992; Jensen, 1993). Kiel and Nicholson (2003) found an „inverted U‟ relationship between board size and performance, adding that directors can bring the board to an optimal 12

skills/experience mix level. However, beyond that point the difficult dynamics of a large board prevail over the skills/expertise advantage that additional directors might bring. In general, board size will differ depending on country-specific factors, regulatory requirements, types of business and complexity of the business. In Bangladesh, the Guidelines have set maximum and minimum numbers of board members for listed companies at 20 and 5 respectively. The Guidelines also suggest that boards of banks and nonbank financial institutions, insurance companies and statutory bodies should be constituted as prescribed by their respective primary regulators. Independent directors: Independent directors are those board members who do not hold any executive position in the company or have any direct or indirect interest in the company (Suchard et al., 2001). It is generally argued that independent directors are more likely to protect shareholders‟ interests and reduce agency problems, add value to firms by providing expert knowledge and monitoring services (Fama, 1980; Fama and Jensen, 1983a). Hermalin and Weisbach (2001) report a number of findings: smaller boards and the greater proportion of independent directors each appear to lead management teams to take actions that are more in line with shareholders‟ interests; while boards with a larger ratio of outsiders are more likely to remove a poorly performing manager. A greater proportion of independent directors will be able to monitor any selfinterested actions of managers and thus minimise agency costs (Fama, 1980; Fama and Jensen, 1983a). By considering the above issues, the Guidelines encourage all listed companies to have „independent, non-shareholder‟ directors. The term „independent, non-shareholder‟ means directors who hold less than 1% or no company shares and have no direct personal or business relationship with the company, its promoters or directors. SEC also directs that nonshareholder directors should be appointed by elected directors. The Guidelines suggest that one-tenth (subject to minimum of one) must be independent non-shareholder directors. The emphasis on independent, non-shareholder directors suggested in the Guidelines is consistent with the Cadbury Report (1995), which emphasises improved board monitoring by increasing its independence from management and independent directors working for the best interests of shareholders. Separate role of Chairperson and CEO: In business, the two most important business positions are the chairperson of the board and Chief Executive Officer (CEO). These positions should be filled by different individuals since their functions are necessarily separate (Cadbury, 1995). The position of chairperson significantly influences the outcome of board decisions as the person controls board meetings, sets its agenda, makes committee assignments and influences the selection of new directors. The position


of CEO is also influential as he/she is responsible for any operating and financial decision-making. It has been argued that dual chair-CEO leadership role enables a CEO to have more opportunity to act in their self-interest (Jensen, 1993). Holding the position of both CEO and chair has been criticised as inappropriate in terms of influencing critical power relationships in the firm (Jensen, 1993). It is argued that where the two roles are combined in one person, it is more likely that the CEO will be able to control the board, reducing the board‟s independence from management, and making decisions in their self-interest at the expense of shareholders. To maintain independence, it is necessary that the board is independent from the CEO (Hermalin and Weisbach. 2001). In the Guidelines, SEC added the condition that the chair and CEO (or Managing Directors) are held by separate persons. In Bangladesh, companies have a CEO position but managing director is a more commonly used term. The Guidelines stated the chair of the board should be elected from the directors. This means an independent, non-shareholding director can be elected as chair. This might create some concern regarding confidentiality of information, as an independent, non-shareholding director can also be a director of other companies with similar operations and interests. Audit committee: The audit committee, a subcommittee of the board, is delegated specific financial oversight responsibilities (Menon and Williams, 1994). An audit committee is now being treated as a principal player in ensuring good corporate governance and rebuilding public confidence in financial reporting. The audit committee has the following functions: monitoring integrity of financial statements, reviewing internal financial controls, recommending appointment of external auditor and reviewing auditor independence and objectivity and audit effectiveness (Bosch, 1995; Klein, 1998). In reducing agency conflicts, audit committees function as a monitoring mechanism and their function has been emphasised by many researchers (e.g., Abbot and Parker, 2000; Chen et al., 2005). The Guidelines make it mandatory for all listed companies to have an audit committee. Furthermore the audit committee should consist of at least three members including one independent and nonshareholder director. The Guidelines also emphasise that the chair of the audit committee should have professional qualifications, knowledge, understanding or experience in accounting or finance. If members of the audit committee are financially literate, it is expected they will work more efficiently for the best interests of shareholders. Frequent audit committee meetings with independent and financially literate directors on the committee will enhance the monitoring ability of audit committees. The provisions set by the Security and Exchange Commission implicitly impose the condition that

there should be at least one member on the audit committee with a professional qualification or knowledge, understanding or experience in accounting and finance. The Guidelines also state that the audit committee reports to the board on its activities and any conflict of interest, fraud or irregularity, suspected infringement of laws or any other matter which they think necessary to disclose. Additionally the Guidelines empower the audit committee to report to the Security and Exchange Commission directly in case its proposals are ignored by the board and management without a valid reason. Audit committees should also be responsible to the shareholders and that report should be signed by the chair of the audit committee. Monitoring by Management and Audit Committee: The company must appoint a chief financial officer, head of internal audit and company secretary. The board of directors should clearly define these roles and it is mandatory for these persons to attend board meetings except where any agenda items relate to them. Chief financial officer: The CEO provides overall leadership and vision in developing the company‟s strategic direction, tactics and business plans necessary to realise revenue and earnings growth, and increase shareholder value. Head of internal audit: The head of internal audit is responsible for focusing and planning specific audits. The responsibilities of the internal audit division vary with the size, complexity and type of business. Some of the internal audit division functions are routine compliance auditing but may include duties in general accounting areas and even performance auditing. The function and duties of the head should be defined clearly. Company secretary: The company secretary is the chief administrator of the company. A company secretary‟s functions are connected with convening meetings of the board of directors. This requires the company secretary to be fully conversant with relevant law and meeting procedures. The Cadbury Committee on Corporate Governance (Cadbury, 1995) recognised the company secretary's unique position has a key role in ensuring that board procedures are followed and regularly reviewed. The chair and board look to the company secretary for guidance on their responsibilities. External Auditors Auditors serve to increase the quality of financial reporting. To maintain the quality of an audit, auditors need to be independent. Auditors do not directly monitor management, however, they provide an assurance service that improves the quality of information. The extent to which financial statements can reduce agency costs depends on the quality of the 13


audit. Although the preparation and audit of financial statements is required by the Companies Act 1994, there is significant variation in the quality and independence of audits. A major threat to audit quality and independence is the provision of both audit and non-audit services by accounting firms. Sharma and Sidhu (2001) find auditor independence is compromised when non-audit services are high in relation to audit fees. Lack of auditor independence will reduce audit quality through the auditor‟s reluctance to report any detected misstatements. Therefore, with respect to auditor monitoring, the general finding is that a high level of competence and independence is compulsory for a high quality audit. Empirical studies finds that Big 4 audit firms have brand names that are associated with higher quality audits (DeAngelo, 1981; Dye, 1993; Craswell et al.,1995). DeAngelo (1981) argues that large audit firms have stronger incentives to protect their reputations because they lose clients if they produce low quality audits. Dye (1993) argues that large audit firms face greater risk of litigation, and hence, large audit firms have stronger incentives to avoid litigation by supplying audits of high quality. Craswell et al. (1995) find large audit firms earn significantly higher fees and they attribute part of this premium to investments in expertise by large audit firms. From the above findings it is reasonable to assume that audit firm size is a good proxy for audit quality. A major threat to auditor independence, identified in the literature, is the joint provision of audit and non-audit services. This can both increase the competence and cost-effectiveness of audit firms and reduce the actual or perceived independence of the auditor (Arrunada, 1999). The revenue-based independence threat is suggested by a positive association between audit fees and non-audit service (e.g., Simunic, 1984; Palmrose 1986; Davis et al., 1993; Craswell et al., 1995; Butterworth and Houghton 1995). Publicly traded companies in Bangladesh * are required to have audits under the Corporations Act 1994. However, quality of audits, and subsequent ability to reduce agency costs, varies significantly (DeAngelo, 1981). DeAngelo (1981) defines audit quality as the joint probability that an auditor will: (i) detect a material misstatement in the financial report if one exists (auditor competence); and (ii) report the misstatement if it is detected (auditor independence). *

In Bangladesh Rahman Rahman Huq is the only audit firm that is affiliated with any Big 4 audit firm. RRH represents KPMG International Cooperative ("KPMG International"). Other audit firms having links with 3 other BIG 4 audit firms are: Howlader Younus and Co. and S.F. Ahmed and Co. (link with Ernst & Young); Hoda Vasi Chowdhury and Co. (link with Deloitte Touche Tohmatsu); and A. Quasem and Co. (link with PricewaterhouseCoopers).

14

This definition separates audit quality into two, namely competence and independence. The Guidelines prohibit the external/statutory auditors to provide selected non-audit services such as: (i) valuation services or fairness opinion, (ii) financial information systems design and implementation, (iii) bookkeeping or other services related to the accounting records or financial statements, (iv) broker-dealer services, (v) actuarial services, (vi) internal audit services, and (vii) any other service determined by the audit committee. Performance Measures In the absence of strong theoretical work on implementing any particular set of performance measures, this study uses both accounting and hybrid (mixture of accounting-market) performance measures to examine the effect of monitoring on performance. Performance measures are necessary for evaluating and comparing the effectiveness of monitoring and organisational control. This study measures the financial performance of firms on the basis of financial accounting information, which is the outcome of any company‟s accounting and reporting systems. This information provides quantitative data concerning the financial position and how a company has performed over a certain period. The financial statements supplied by the management are subject to external audits to verify their accuracy. ROE, ROA and EPS are used as accounting measures. It is expected that the different monitoring mechanisms will influence management to work in the company‟s best interests and this will eventually force them to report the correct accounting information. This will have an impact on the firm‟s performance. Market information is used to measure a company‟s hybrid performance, which is based on information from the capital market performance of the company. Monitoring enables management to have more influence in the market which could impact on market performance. PER, MBV and DY are used as hybrid (accounting and market) measures of performance. Proposition and Methodology To address the issue of multicollinearity, this study uses an interrelated structural setting. It sets all monitoring mechanisms in a structural equation model in order to establish within the structural setting how the individual monitoring mechanisms affect performance. This will be the first study on Bangladesh that analyses a large number of monitoring variables in a structural setting and determines their individual effect on firm performance. Based on SEM developed in this paper, the following null hypotheses are tested:


H1: In a joint monitoring context, firm performance is not a function of board monitoring. H2: In a joint monitoring context, firm performance is not a function of senior management monitoring. H3: In a joint monitoring context, firm performance is not a function of auditor monitoring. For measuring board of directors and audit committee monitoring this paper uses: (i) size of the board, (ii) proportion of independent directors, (iii) separate CEO, (iv) members on the committee, and (v) qualification for the chair of the audit committee and chair positions. Second, monitoring by management is measured by the presence of: (i) CEO, (ii) head of internal audit, and (iii) company secretary. Third, the external auditor is measured using audit in terms of: (i) quality, and (ii) independence (Table 1). [Table 1 About Here] In SEM there is no single statistical test that describes the goodness-of-fit of the model. Instead, researchers have developed a number of goodness-offit measures that assess the results (see Table 3 for details of results and cut-off value). For this type of study it is necessary to have enough data so important differences or relationships can be observed. SEM applications typically use 200-500 cases to fit models that derive from 8-15 observed variables. This research uses 281 listed companies on the Dhaka Stock Exchange listed companies for the year 2008, with reference to 8 monitoring mechanisms. Data and Firm Characteristics Total number of companies in January 2008 is 281 companies (with $10,822 million domestic market capitalisation). Out of 281 listed companies, 183 (65.12%) reported full compliance. Eighty-five companies (30.25%) reported partial compliance while 13 companies (4.62%) did not comply or respond (Table 2). SEC recommended board size from 5 to 20, subject to primary regulators provision not being inconsistent. This has been complied with by 253 (90%) companies out of a total of 281 listed companies. This provision is not difficult to comply with and it is not understood why 28 companies are still non-compliant. It is possible that some of these companies exist only on paper. [Table 2 About Here] Appointment of independent directors is a very important requirement. According to the guidelines, at least one tenth of directors should be independent directors who hold less than one per cent of the paid up shares or do not hold any share at all. They can not be connected with a company's promoters or directors.

It is a positive development that 202 companies have already complied with the provision. The guidelines provide that positions of chairperson and CEO should preferably be filled by different individuals. As reported, 243 companies have already complied with the provision of appointing separate chairperson and Chief Executive Officer. There should be a chief financial officer, a head of the internal audit and a company secretary with clearly defined roles and responsibilities. Chief Financial Officer (CFO) and company secretary are required to attend board meetings. According to the reports, 245 companies confirmed appointment of CFO, head of internal audit and company secretary. Provision of an audit committee is an important component of the guidelines. An audit committee is required to be formed with a minimum of three directors including at least one independent director. One member should be chairperson of the committee and he/she is required to have relevant professional qualifications or knowledge, understanding and experience in accounting and finance. The committee‟s most important responsibility is to ensure that the company‟s financial statements reflect a true and fair view of its affairs and report any conflict of interest, suspected irregularity or legal infringement, etc., to the board. It is indeed encouraging that 225 (80.07%) companies out of 281 have formed audit committees. Again, 261 (92.88%) companies reportedly ensured that their external auditors are not engaged in appraisal or valuation services, financial information system, accounting records, internal audit or other related activities. While these statistics look very impressive, they need to be considered cautiously because these are based on reporting by the companies themselves without verification by any independent authority.

Analysis of the Results The model of monitoring mechanisms and their effect on the firm performance fits the data well for all performance measurements (Table 2). Chi-square to degree of freedom index (CMIN/DF) is 1.89; Root Mean Square Error of Approximation (RMSEA) = .048; adjusted goodness-of fit index (AGFI) = .91; Normal fir index (NFI) = .91; and Comparative fit index (CFI) =.95. [Table 3 About Here] The results were generally consistent, when examining the impact of accounting and hybrid measures of firm performance. There are three variables in the model‟s monitoring mechanisms: 15


board of directors and audit committee; management; and external auditors. The relationships among them are significant for all the identified paths. In this research, a path diagram is developed that can present predictive relationships among constructs (i.e. the dependent-independent variable relationships), as well as associative relationships (correlations) between constructs and indicators.

MacAvoy and Millstein (1999) argue that one reason for not finding any relationship is because they have used “old” data – that is, data that preceded the board monitoring role in the current-year and performance. However, they found no difference in the result when they used the lagged year‟s performance. Hypothesis 2: Management and monitoring and firm performance

committee

[Table 4 About Here] Hypothesis 1: Director monitoring and firm performance Boards of directors are the most active monitors of management. Monitoring efficiency improves when there is a sufficient number of directors making up the board, high proportion of independent directors, and the CEO and Chairperson are separate people. Yet, whether monitoring by boards affects firm performance remains unresolved in the literature. The following results show that such monitoring has an inconsistent and statistically significant (and nonsignificant) relationship with firm performance. Accounting performance measures The data for board monitoring and accounting performance (Table 4) show that the impacts of board of directors on ROE, ROA or EPS are not significant. The relationship is positive for ROE and ROA but negative for EPS. The results are consistent with conclusions reached by Bhagat and Black (2000), who examined the influence of board composition on accounting performance. They failed to find any relationship between board composition and firm performance. In general, board of directors monitoring has been found to have more impact on market performance compared to accounting performance. Board monitoring may improve the transparency of reporting but this is not necessarily reflected in accounting numbers. Hybrid performance Results in Table 4 find a significant result concerning differences between monitoring by board and firm performance as measured by PER, MBV and DY. The relationship is positive for PER and MBV but negative for DY. The presence of a good board monitoring structure increases confidence among shareholders and this is reflected in market performance. Overall, this research finds significant results for hybrid performance models but no significant results for accounting performance. Therefore this research failed to reject hypothesis 1. These findings are consistent with Hermalin and Weisbach (1991), Mehran (1995), Klein (1998) and Bhagat and Black (2000), who examined the influence of board composition on firm performance and failed to find any relationships in accounting performance. 16

The results of management and committee monitoring and their impact on firm performance exhibit inconsistent but significant results in regard to accounting and hybrid performance. Accounting performance measures The data in Table 4 shows an association between monitoring by the management and committee and firm performance. A significant relationship is found when performance is measured by ROE and ROA. Nevertheless the results reveal that management monitoring does have some positive effect on accounting performance. In relation to EPS, this performance measure reflects how much has been earned during the financial year for each of the shares held. Earnings are an accounting number that reflect both the firm‟s economic results and management‟s accounting policy choices. The results in Table 4 show no significant relationship between management monitoring and EPS, suggesting that management does not cause a significantly higher or lower EPS. This result does not identify whether stronger management and committee monitoring does, in fact, achieve superior economic results for the firm in any one year, which are smoothed [smoothed over?] in the reported EPS due to the accepted accounting policy choice. Hybrid performance measures Similar to accounting performance measures, the data in Table 4 shows that there is a significant association between monitoring by the management and committee and hybrid performance measures. Results find a significant relationship between monitoring by management and firm performance as measured by PER, MBV and DY. A relatively higher PER could be a reflection of whether investors are willing to pay a market premium relative to current reported earnings due to the monitoring activities of management and committee. In this section, it is not surprising that a relationship is found between management and committee monitoring and PER. Table 4 shows a significant relationship between management monitoring and MBV Higher market capitalisation to book value of equity is deemed to reflect a stronger intellectual capital and intangibles that are not recorded in book value. The results infer that management and committee influence over


management leads to the development of greater unidentifiable goodwill by the firm. However, many factors affect MBV, so this is a very tentative inference. Table 4 provides results for management monitoring and performance as measured by the DY. It shows that in 2008 the relationship is significant. This suggests that management monitoring can occasionally affect management decisions concerning dividend payout relative to market price of shares. The result is evident that analysis failed to reject the proposition. There is a clear pattern in market performance, although, the results in Table 4 do not provide a clear pattern of relationship between management and committee monitoring and accounting firm performance. One possible reason for these inconsistent results is the presence of substitution or complementary effects among the range of governance mechanisms. These may encourage the firm to rely on various monitoring devices and structures.

thereby increasing the reported book value of net assets, which can lower MBV. In the final column of Table 4, there is a significantly positive relationship in the models for DY. Since DY is a reflection in part of sound cash management (i.e. ability to pay regular and increasing cash dividends), the quality of auditing and assurance services would benefit the firm‟s cash management. Overall, the results for relationships between auditor monitoring and firm performance are mixed. The results show both significant and insignificant; and both positive and negative outcome. Therefore, the result failed to reject this hypothesis. It is again posited that a key reason for not uncovering a consistent pattern of relationships is the presence of substitute effects among the governance variables. The possible interdependence among auditing and other monitoring mechanisms may explain the differences in relationships between individual monitoring mechanisms and performance measures over the period.

Hypothesis 3: Auditor-based construct and firm performance

Robustness Tests One set of robustness tests involved verifying the statistical inferences by testing the sample for the backward (for the year 2007) and forward (for 2009) years lagged model and seeing how the monitoring effects compared to one year lagged models. All lagged models conclude that one-year forward lagged models are more reflective of performance compared to one-year backward lagged models. Therefore it is assumed that results of current monitoring will be reflected in the year immediately after the financial information is published. There can be two probable explanations for the results of the robustness tests. Firstly, it suggests that the effect of monitoring is reflected on a concurrent basis. Secondly, as suggested in the literature, if the shareholders are not satisfied with the firm‟s performance they „exit‟ immediately rather than using their „voice‟.

As a monitor of the reported performance of any company, external auditors are responsible for safeguarding accounting information used for decision-making. Auditors do not directly monitor business but they do indirectly monitor business due to their audit function. Monitoring by auditors is reflected in accounting and hybrid performance measures. The following analysis reveals auditor monitoring on firm performance. Accounting performance measures Table 4 shows that there is no relationship between monitoring by auditors and firm performance when measured by ROE or ROA. This might be expected because auditor monitoring is concerned with ensuring accounting numbers are true and fair without being systematically biased in any year. There is a significant positive relationship between the monitoring by auditors and their effect on the EPS for 2008. Hybrid performance measures Similar to accounting measures, Table 4 shows that there is a relationship between monitoring by auditors and market performance when measured by PER. A further test using MBV shows that a significantly negative relationship exists in 2008. This relationship shows that when auditor monitoring increases, MBV value decreases significantly. These results again present a quite inconsistent picture. The curious finding is that auditor monitoring is negatively related to MBV. The explanation may be that higher quality auditors will generate financial statements that contain more up-to-date fair values of assets and more recognition of intangible assets,

Conclusion The concept of corporate governance is fairly new in Bangladesh and its current status is far from adequate. However, it is encouraging that in recent years this subject is being discussed in various forums among entrepreneurs, corporate managers, regulators and academics. Most Bangladeshi companies have concentrated ownership structures with a strong family orientation. The board of directors, dominated by sponsor shareholders often from the same family, control decision-making processes and annual general meetings are mostly ineffective. The board is often enthusiastically involved in management while the CEO‟s role is marginal. Independent directors - when there is any - can seldom act independently or play his/her role as an effective advocate for minority shareholders or as a useful deterrent to irregular 17


practices. Shareholder activism is still not a very popular concept in Bangladesh. Lack of auditor independence frequently gets in the way of transparent financial disclosures. In many companies, there is practically no accountability structure of the management to the board or shareholders. In the absence of any structured government mechanism, there is no central authority to enforce even minimum practice of corporate governance. With a few exceptions, all the models suggest that there is a limited discernable pattern of significant relationships between monitoring as carried out by board of directors, audit committee, management, auditors and performance. The most likely reason for there not being any consistent pattern is that most companies are family oriented. Such concentrated ownership structures affect the effectiveness of corporate monitoring mechanisms, which are weaknesses that cannot be rectified by laws and regulations. There is neither any value judgment nor any consequences for corporate governance practices. The current system in Bangladesh does not provide sufficient legal, institutional and economic motivation for stakeholders to encourage and enforce corporate monitoring practices. This result is consistent with Nandelstadh and Rosenberg (2003) who also find limited combinations of internal and external corporate governance mechanisms associated with firm performance. They based their conclusion on an analysis of data from Finnish publicly-listed companies during 1990–2000. This study has policy implications for the corporate environment in Bangladesh. When considering any change in corporate monitoring, the Bangladeshi government should take into account the nation‟s business and legal practices and culture. Limitations and Future Research Avenues There are a number of limitations that may influence the results of this study, and these need to be addressed in order to improve the integrity of future research in this area. Three accounting bases and three hybrid measurements were used in this study. Accounting measures of performance are subjected to accounting policy choice, while market measures of performance are affected by market inefficiencies. Consequently, this strategy may impede on some important performance features that could be obtained through other tools. Therefore, this study may not accurately report companies‟ intrinsic performance. This study did not consider other market and regulatory mechanisms that have been used in single country studies. Capital market, managerial labour market and legal systems are common to all firms and there is little scope in differentiating these factors (Agrawal and Knoeber, 1996; Denis and McConnell, 2003). Jensen (1993) states that the legal system which in itself is a corporate governance mechanism, 18

is too blunt to deal with agency problems between managers and shareholders. The same is true for the labour and capital market. Future research should focus on examining whether there are any substitutions or complementary effects existing between different monitoring mechanisms, specifically: ownership and board of directors monitoring; board of directors and auditor monitoring; and ownership and auditor monitoring and firm performance. Reference 1. 2.

3.

4. 5.

6. 7. 8. 9.

10.

11.

12. 13.

14.

Abbott, L.J. & Parker, S. 2000. Auditor selection and audit committee characteristics. Auditing: A Journal of Practice and Theory, 19: 47 - 66. Agrawal, A. & Knoeber, C.R. 1996. Firm performance and mechanisms to control agency problems between managers and shareholders. Journal of Financial and Quantitative Analysis, 31: 377 - 397. Arrunada, B. 1999. The provision of non-audit services by auditors let the market evolve and decide. International Review of Law and Economics, 19: 513 531. Austin D.V. 2002. Independence of the board committees. Hoosier Banker, 86: 8 – 10. Bhagat, S. & Black, B. 2002. The non-correlation between board independence and long-term firm performance. Journal of Corporate Law. 27(2): 231 273. Bosch, H. 1995. Corporate Practice and Conduct, 3rd edn, Business Council of Australia (Pittman Publishing). Butterworth, S. & Houghton, K.A. 1995. Auditor switching and the pricing of audit services. Journal of Business Finance and Accounting, 22: 323 - 344. Cadbury, A. 1995. Committee on the financial aspects of corporate governance: Compliance with the code of best practice. Gee Publishing, London. Chen, Y.M., Moroney, R., & Houghton, K. 2005. Audit committee composition and the use of an industry specialist audit firm. Accounting and Finance, 45: 217 - 239. Craswell, A.T., Francis, J.R., & Taylor, S.L. 1995. Auditor brand name reputations and industry specializations. Journal of Accounting and Economics, 20: 297 - 322. Davis, L.R., Ricchiute, D.N., and Trompeter, G. 1993. Audit effort, audit fees, and the provision of non-audit services to audit clients. The Accounting Review, 68: 135 - 150. DeAngelo, L.E. 1981. Auditor size and audit quality, Journal of Accounting and Economics, 3: 183 – 199. Dechow, P.M., Sloan, R.G., & Sweeney, A.P. 1996. Causes and consequences of earnings manipulation: an analysis of firms subject to enforcement actions by the SEC. Contemporary Accounting Research, 13: 1 - 36. Denis, D.K. & McConnell, J.J. 2003. International corporate governance. Journal of Financial and Quantitative Analysis, 38(1): 1 – 36.

Corporate Ownership & Control / Volume 8, Issue 1, Fall 2010 15. DeZoort, F.T. 1997. An investigation of audit committees‟ oversight responsibilities. A Journal of Accounting Finance and Business Studies, 208 - 227. 16. Donaldson, L. 1990. The ethereal hand: organisational economics and management theory. Academy of Management Review, 15(3): 369 – 381. 17. Donaldson, L. & Davis, J.H. 1991. Stewardship theory or agency theory: CEO governance and shareholders returns. Australian Journal of Management, 16 (1): 49 – 64. 18. Donaldson, L. & Davis, J.H. 1994. Board and company performance – research challenges the conventional wisdom, Corporate Governance: An international Review, 2 (3): 151 – 160. 19. Dye, R. 1993. Auditing standards, legal liability, and auditor wealth, Journal of Political Economy, 101: 887 914. 20. Fama, E.F. 1980. Agency problems and the theory of the firm. Journal of Political Economy, 88: 288 - 307. 21. Fama, E.F. & Jensen, M.C. 1983a. Separation of control and ownership. Journal of Law and Economics, 26: 301 - 325. 22. Fama, E.F. & Jensen, M.C. 1983b. Agency problems and residual claims. Journal of Law and Economics, 26: 327 - 349. 23. Farooq, O. & Sheraz A. 2007. Do Governance Reforms Increase Information of Analysts‟ Stock Recommendations? Evidence From a Newly Emerging Market, Paper presented at the International Research Conference on Corporate Governance in Emerging Markets, Istanbul, 15-18 November. 24. Farrar, J.H. 2008. Corporate Governance: Theories, Principles and Practice. Melbourne: Oxford University Press. 3rd Edition. 25. Golden, B.R. & Zajac, E.J. 2001. When will boards influence strategy? Inclination x power = strategic change. Strategic Management Journal, 22: 1087 – 1111. 26. Hermalin, B.E. & Weisbach, M.S. 2003. Boards of directors as an endogenously determined institution: A survey of the economic literature. Economic Policy Review, 9: 7 - 26. 27. Hermalin B.E. & Weisbach, M.S. 1991. The effects of board composition and direct incentives on firm performance. Financial Management, 101 – 112. 28. Hermalin B.E. & Weisbach, M.S. 2001. Board of directors as an endogenously determined institution: a survey of the economic literature, FRBNY Economic Policy Review, 1 – 20. 29. Holderness, C.G. 2003. A survey of blockholders and corporate control, Federal Reserve Bank of New York Economic Policy Review, 9: 51- 63. 30. Jensen, M.C. 1993. The modern industrial revolution, exit and the failure of internal control systems. Journal of Finance, 48: 831 - 880. 31. Jensen, M.C. & Meckling, W. 1976. Theory of the firm: managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3: 305 - 360. 32. Kaplan, S.N. & Minton, B. 1994. Appointments of outsiders to Japanese boards: determinants and

33.

34. 35. 36.

37. 38. 39. 40.

41.

42.

43.

44. 45.

46. 47. 48.

implications for managers. Journal of Financial Economics, 36: 225 – 257. Kiel, G.C. & Nicholson, G.J. 2003. Board composition and corporate performance: how the Australian experience informs contrasting theories of corporate governance. Corporate Governance: An International Review, 11 (3): 189 - 205 Klein, A. 1998. Firm performance and board committee structure. Journal of Law and Economics, 41: 275 – 299. Lipton, M., & Lorsch, J. 1992. A modest proposal for improved corporate governance. Business Lawyer, 48: 59 - 77. MacAvoy, P. & Millstein, I.M. 1999. The active board of directors and its effect on performance of the large publicly trades corporation. Journal of Applied Corporate Finance, 11 (4): 8 – 20. Mehran, H. 1995. Executive compensation structure, ownership and firm performance. Journal of Financial Economics, 38: 163 – 184. Menon, K. & Williams, J.D. 1994. The use of audit committees for monitoring. Journal of Accounting and Public Policy, 13: 121 - 139. Palmrose, Z. 1986. The effect of non-audit services on the pricing of audit services: further evidence. Journal of Accounting Research, 24: 405 - 411. Ramsay, I. 2001. Independent of Australian company auditors: Review of current Australian requirements and proposals for reform. Report of University of Melbourne, Melbourne, Victoria. Rediker, K.J. Seth, A. 1995. Boards of directors and substitution effects of alternative governance mechanisms. Strategic Management Journal, 16: 85 99. Roundtable discussion on corporate governance guidelines of SEC and Its implementation practices in Bangladesh 2006. Dhaka Stock Exchange. Retrieved May 12, 2007 from: http://www.dsebd.org/pdf/roundtable.pdf. Ryan H.E & Wiggins R.A. 2004. Who is in whose pocket? Director compensation, board independence, and barriers to effective monitoring. Journal of Financial Economics, 73: 497 - 524. Security and Exchange Commission (SEC) 2007. Stock Exchange Information. Retrieved May 27, 2007, from http://www.secbd.org/stock.html Sharma, D.S. & Sidhu, J. 2001. Professionalism vs. Commercialism: The association between non-audit services (NAS) and audit independence. Journal of Business Finance and Accounting, 28 (5 & 6): 595–629. Shivdasani, A. 1993. Board composition, ownership structure, and hostile takeovers. Journal of Accounting and Economics, 16: 167 - 198. Simunic, D.A. 1984. Auditing, consulting and auditor independence. Journal of Accounting Research, 22: 679702. Solaiman S.M. 2006. Recent reforms and the development of the securities market in Bangladesh. Journal of Asian and African Studies, 41(3), 195–228.

19

Corporate Ownership & Control / Volume 8, Issue 1, Fall 2010 49. Suchard, J.A., Singh M., & Barr, R. 2001. The market effects of CEO turnover in Australian firms. Pacific-Basin Finance Journal, 9: 1–27. 50. USAid 2007, USAid Bangladesh program summary, web address: http://www.usaid.gov/bd/program.html, visited on 12 Feb, 2009.

51. Watts, R. L. & Zimmerman, J. L. 1990. Positive accounting theory: A ten-year perspective. The Accounting Review, 65: 131 – 156. 52. Wooldridge, M.J. 2003. Introductory Econometrics: A Modern Approach Thomson, Ohio.

Appendices Figure 1. Monitoring Model

Board Size

No. of meetings

Financially literate chairperson

Audit Committees

% of independent members

Composition & Structure

Boards of Directors

Firm Performance Quality

Independent Directors CEO/CH Duality

Chief financial officer

(BIG 4 Auditors)

Quantity (Audit/Non Audit Fees)

20

External Auditors

Management

Head of internal audit

Company secretary


Table 1. List of Variables

Variables: Management Monitoring: CFO= Chief financial officer HIA = Head of internal audit CS = Company secretary Board of Directors & Audit Committees: BSIZ = Number of directors on the board PBI = Proportion of independent directors on the board CHCE = CEO and Chairperson of the board PAF = Proportion of Financially Literate directors on the audit committee PAI = One tenth directors are independent on the AC QAC =Professional qualifications of the chair of audit committee External Auditors: BIG 4 = Affiliation/link with Big 4 audit firm PNAF = Proportion of audit/non-audit fees Control Variable: SIZE A: Size of the firm based on log of total assets SIZE B: Size of the firm based on log of total sales Performance Measures: ROE = Return on equity ROA =Return on asset EPS = Earning per share PER = Price earning ratio M/BV = Market to book value DY = Dividend Yield

21


Table 2. Descriptive statistics regarding monitoring measures (Sample Size: 281 Companies)

Monitoring Variables Chief financial officer (0,1) Head of internal audit (0,1) Company secretary (0,1) Size of the Boards (number) Proportion of Independent directors on the Boards Dual role of Chairperson and CEO (0,1) Proportion of Financially Literate directors on the audit committee (0-1) One tenth directors are Independent on the Audit Committee (0 -1) Professional qualification of the chairperson of audit committee (0,1) Big 4 Audit firms (0,1) Proportion of non-audit service fees (NAF/TAF) Return on Equity (ROE) Return on Assets (ROA) Earning per share (EPS) Price earning ratio (PER) Market to book value (MBV) Dividend Yield (DY) Size A (Log Assets) Size S (Log Sales)

Minimum 87.18% 86.83 87.54% 3 0 86.47% 63.23%

Maximum

Mean

Median

S.D

22 .95

8.4 0.675

6.1 0.81

6.245 .295

.23 0.356 0.37 645.74 16.34 79.501 45.76 9.568 10.356

.29 0.224 0.083 22.4 7.34 1.030 25.46 5.678 6.215

.26 0.184 0.081 10.45 8.52 0.705 22.54 6.246 5.456

.36 1.672 0.068 51.37 3.45 4.256 3.45 1.457 1.987

71.88% 69.07% 37% .12 -1.142 -0.45 -153.65 6.78 -20.654 5.45 1.921 -3.765

Table 3. Structural Equation Model Fitness Current year Model 2008 Model Return on Equity Return on Assets Earning Per Share Price Earning Ratio Market to Book Value Dividend Yield

χ2 (df) 247.367 214.206 275.162 257.438 266.517 273.258

∆χ2 (∆ df) 2.176 2.231 2.310 2.127 2.145 2.272

Model Return on Equity Return on Assets Earning Per Share Price Earning Ratio Market to Book Value Dividend Yield

χ2 (df) 293.620 281.807 287.249 282.503 297.032 295.637

Model

χ2 (df)

∆χ2 (∆ df)

RMSES

Return on Equity

237.456

1.931

Return on Assets

255.017

2.073

Earning Per Share

257.627

Price Earning Ratio

244.119

Market to Book Value Dividend Yield

RMSES .050 .053 .051 .047 .051 .052

AGFI .911 .920 .907 .912 .911 .909

CFI .960 .918 .955 .960 .930 .957

NFI .926 .924 .925 .912 .928 .921

CFI .962 .963 .961 .962 .961 .961

NFI .933 .937 .935 .938 .932 .933

AGFI

CFI

NFI

.045

.913

.920

.901

.048

.915

.942

.893

2.094

.049

.913

.982

.899

1.977

.046

.921

.928

.900

233.116

1.888

.044

.921

.931

.901

230.337

1.930

.045

.921

.940

.902

Backward Lagged Year Performance 2007 ∆χ2 (∆ df) 2.352 2.189 2.315 2.316 2.393 2.381

RMSES .054 .052 .053 .053 .054 .054

AGFI .903 .906 .904 .904 .902 .902

Forward Lagged Year Performance 2009

Here, χ2 (df) = Chi- Squire AGFI = Adjusted goodness of fit index (acceptable limit => .90) ∆χ2 (∆ df) = Normed Chi-Squire (Acceptable limit 1 – 5; 1 = best fit, 5 = reasonable fit) CFI = Comparative fit index (0 = no fit at all, 1 = perfect fit) RMSES = Root mean squire (.05 or less indicate a close fit) NFI = Normal fit index (0 = no fit at all, 1 = perfect fit) (Source: Hair et al., 2006)

22


Table 4. Monitoring and Performance

ROE

ROA

EPS

PER

MBV

DY

Shareholder monitoring and Performance (2008)

0.023 (0.053) *

0.012 (0.003) **

0.978 (0.197)

-0.926 (0.005) **

-0.099 (0.037)*

0.001 (0.011)**

Management Monitoring and Performance (2008)

0.017 (0.768)

0.058 (0.972)

-0.780 (0.551)

0.054 (0.008) **

0.035 (0.048) *

-0.003 (0.005) **

Auditor monitoring and Performance (2008)

0.239 (0.618)

0.121 (0.175)

5.186 (0.005) **

- 0.891 (0.861)

- 0.921 (0.000) **

0.04 (0.005) **

** Significant at the .01 level. * Significant at the .05 level. Note: Values in the bracket indicates “P value”.

23


CORPORATE OWNERSHIP CHOICE BASED ON ETHICAL CRITERIA: IS IT BIG ENOUGH TO NOTICE? Charl de Villiers*, Chris van Staden** Abstract Ethical investors often exclude firms that participate in so-called controversial activities, such as tobacco, alcohol, firearms, gambling, the military, and nuclear operations, from their investment portfolios. Firms excluded in this way should experience an increase in their cost of capital and a reduction in their share prices. We use the KLD database to identify S&P 500 firms involved in controversial activities. Our results show no difference between controversial activity firms and other firms regarding relative share price and we find that the cost of capital of controversial activity firms is in fact lower. We conclude that ethical investing, of the type that excludes controversial activity firms, does not influence the capital markets in the expected way. Keywords: Corporate ownership, Ethical investing, comparison of risk and return, controversial activity firms *Corresponding Author: Department of Accounting and Finance, The University of Auckland Business School, Private Bag 92019 Auckland, New Zealand Tel: +64 9 923 5196 Email: [email protected] Fax: +64 9 373 7406 **The University of Auckland Business School and University of Canterbury

Introduction Ethical investing has made great strides in recent times (Schueth 2003). Ethical funds attract individuals and groups who want their funds invested in socially responsible ways. Criteria differ, but most ethical funds screen out firms involved in controversial business activities from their portfolios (Social Investment Forum 2007). Screening is used to manage $2.1 trillion of the $2.7 trillion that is invested ethically in the US. The method of screening out firms involved in controversial business activities has been used since the 18th century, for example Entine (2003: 353) reports that during the 18th century some businessmen excluded firms involved in alcohol, tobacco and gambling, “so-called sinful behaviour”. The first screened U.S. investment fund (in 1928) excluded firms involved in alcohol and tobacco and during the 1960s firms involved in military contracting became controversial (Entine 2003). More recently it is reported that ethical funds exclude firms involved in controversial business activities, such as tobacco, alcohol, firearms, gambling, military, and nuclear operations from their portfolios (Renneboog et al. 2008). Kinder, Lydenberg and Domini (KLD), an independent ratings company, provide information on whether firms are involved in these controversial 24

business activities. We use the controversial business activities derived from the KLD database (i.e. alcohol, gambling, tobacco, firearms, military and nuclear power) as the controversial business activities in our study and motivate the use of the KLD database in a later section.2 Chatterji et al. (2009: 130) suggest that some ethical investors may avoid firms involved in controversial business activities in order to influence share prices “by raising the cost of capital for misbehaving firms and lowering it for socially responsible firms”. According to Merton (1987: 500), “...an increase in the relative size of the firm's investor 2

Although controversial business activity screens are historically widely used as ethical investment screens, we acknowledge that everyone will not necessarily regard all the controversial business issues as unethical. Investor‟s views on these issues differ, but they have the common trait of excluding any investments that might be seen to be supportive of human suffering or environmental degradation. We use the issues that have been identified over time as controversial and that are rated in the KLD database. In the rest of the paper we use the term “controversial business activities” (in scare quotes) to acknowledge that everyone will not regard these as unethical.


base will reduce the firm's cost of capital and increase the market value of the firm.” This is further explained by Angel and Rivoli (1997: 57) when they state that “[w]hen investors exclude certain firms from their portfolios, the result is a segmented market: The firm has access to one segment of the equity market but not another. Finance theory suggests that the effects of equity market segmentation will be to raise the cost of equity capital.” Bauer et al. (2005: 1752) also comment that “…socially responsible investors are able to influence the value of socially responsible companies by driving down their expected returns and cost of capital.” Herremans et al. (1993: 590) furthermore state that the popularity of ethical investing “is likely to have caused a relative increase in the demand for, and hence prices of, the securities of companies considered to be especially responsible.” This increase would be “relative” to the non-responsible companies. In summary, the accounting and finance literature suggests that firms that are screened out by some investors will experience an increase in their cost of capital and a reduction in their share prices. Such a lowering of share prices can be observed by referring to a measure of relative share price, such as return on equity (ROE) calculated using the market value of equity (Herremans et al. 1993). We therefore expect the ROE of “controversial activity” firms and their cost of capital to be higher than other firms if ethical investing is big enough to make a difference in capital markets.3 Using an analytical model for the reaction of firms (to investing sanctions), Heinkel et al. (2001) concluded that more than 20% ethical investing is needed before unethical firms would reform, because their cost of capital and thereby their share price would be adversely affected to the extent that spending money on reform makes economic sense. However, Entine (2003: 352) believes that “representations of the growing financial impact...of social investing are questionable”. If ethical investing is not big enough, then Entine (2003) may be correct and the ethical investment community do not influence capital markets. If ethical investing is too small, “controversial activity” firms can still exist and thrive at no real disadvantage compared to other firms. In this paper we split the S&P 500 into “controversial activity” and other firms using the KLD database measures, and compare their relative equity value and cost of capital for 2004, 2005 and 3

The reason we expect ROE (calculated using the market value of equity) to be higher for “controversial activity” firms, is that the literature shows that equity markets will reduce the share price of these firms (see for example Merton 1987, Bauer et al. 2005, Chatterji et al. 2009, and Renneboog et al. 2008), but this will not have an impact on the earnings (net profit) of these firms. The resulting ROE figure will therefore be higher than it would have been if the share price was not reduced in this way.

2006. We do not find significant differences between “controversial activity” and other firms on the relative value of equity measure and find, contrary to our expectation, that “controversial activity” firms have lower cost of capital. We, conclude, therefore, that ethical investing of the type that excludes “controversial activity” firms, does not involve large enough amounts of investment funds to influence the capital markets. Ethical investors will have to decide whether this information requires a change in their investment strategy. Our study contributes to the literature by revealing this important information to ethical investors who will be interested in whether the impact of their investment activities influences firms and the capital markets. Our study also complements the existing literature that compares ethical fund returns to others, by focusing on the firm level and comparing “controversial activity” firm returns with other firm returns. In the next section we give some background to ethical investing and then develop our hypothesis. We follow this with the method, results and conclusions. Background and development of hypothesis Background to ethical investing Ethical investing is growing and in the US involved $2.7 trillion in 2007 (Social Investment Forum 2007). It has its modern roots in the “impassioned political climate of the 1960s” (Schueth 2003). Ethical investors may simply invest the way they do in order to avoid feelings of culpability for bad firm behaviour or they may hope to ultimately change firm behaviour. If ethical investing keeps attracting more and more investment funds, individual firms that are screened out by ethical investment funds will find it increasingly difficult to attract investment capital and this will adversely affect their share price (Herremans et al. 1993, Merton 1987, Bauer et al. 2005, Chatterji et al. 2009). A time may come when they are forced to change their behaviour in order to ensure continued funding at competitive rates through the capital markets. Heinkel et al. (2001) explain that when ethical investors do not invest in, for example, polluting firms, this changes the risk sharing opportunities in the market. They use an analytical model to show that the lack of risk sharing leads to lower share prices for polluting firms and that this raises their cost of capital. In their model they found that if ethical investors constitute about 20% of the investor population, fewer neutral investors results in a lower share price for unacceptable firms and therefore the cost of capital for unethical firms increases to about 9.5% while that of ethical firms would be about 3.9% (Heinkel et al. 2001: 440). Thus, if ethical investing is big enough, 25


the effect would be noticeable in the market through lower share prices and higher cost of capital for unethical firms. However, Schepers and Sethi (2003) conclude that despite exaggerated claims, ethical investment activities have thus far failed to influence the conduct of firms. Michelson et al. (2004), in a balanced view of the effects of ethical investing on firm behaviour, also raise the spectre that firms may not change to become acceptable to ethical investors. Heinkel et al. (2001) make their predictions using an analytical model, but do not perform empirical tests using archival data. Our study provides empirical evidence regarding returns and cost of capital at the firm level comparing “controversial activity” firms versus other firms. This is in contrast with many other studies that consider returns of investment funds (ethical versus other funds), i.e. at the fund level. There are different forms of ethical mutual funds, but many of them are exclusionary, i.e. they screen out socially irresponsible firms (Renneboog et al. 2008). In the following sub-section, we discuss the standing of KLD, whose ratings we use in this study. The motivation for using the KLD controversial business activities screens Corporate social performance is difficult to measure consistently across a large number of firms, because social information is hard to collect and classify. Various ratings for corporate social performance exist, for example, KLD, Calvert, FTSE4Good and the Dow Jones Social Index. Chatterji and Levine (2008: Appendix 1) regard KLD as “one of the oldest and most influential social raters with $8 billion invested in funds based on its index”. This is echoed by Sharfman and Hart (2009) who indicate that KLD data have been more widely used by researchers than any of the other measures of social performance. KLD has been providing performance benchmarks, corporate accountability research and consulting services, analogous to those provided by financial research service firms, since 1988.4 KLD is a leading authority on social research and indices for institutional investors and has one of the largest independent corporate research staff complements in the world. According to Harrison and Freeman (1999) one of the advantages of the KLD ratings is that they are based on the extensive research of independent analysts employed by KLD and KLD clients use these ratings as a basis for investment decisions and advice. KLD data have been used in many research studies, see for example Cho and Patten (2007), Mahoney and Roberts (2007), Entine (2003), Milne and Patten (2002), Agle et al. (1999), Berman et al. 4

Currently, 33 of the top 50 institutional money managers worldwide use KLD‟s research to integrate environmental, social and governance factors into their investment decisions (KLD website). 26

(1999), Griffin and Mahon (1997), Waddock and Graves (1997), Graves and Waddock (1994). Graves and Waddock (1994: 1039) indicate that the assessment scheme used by KLD offers the following benefits over rating mechanisms used in previous studies:  the ratings are applied consistently across all firms and are replicable, because it was done by the same firm using an objective set of screening criteria (multiple attributes and the use of objective measures),  it rates all of the Standard & Poor's 500 firms, because KLD operates a service that supplies information to the investment community at large (large resulting pool of data per year and over time), and  the people doing the ratings consist of knowledgeable individuals not affiliated with any of the rated firms or with academic researchers (independent analysis). KLD‟s process therefore provides unique access to a wide range of consistently rated firms over time and across a number of important social performance attributes. The measurement of corporate social performance used can influence a study‟s results (Griffin and Mahon 1997) and if the metrics are invalid, the study may not find support for the hypothesized benefits of socially responsible investment (Chatterji and Levine 2008) or may not provide credible support. Chatterji and Levine (2008) and Sharfman and Hart (2009) focused on the social responsibility scores given by the various rating companies on a range of topics (for example community relations, corporate governance, diversity, employee relations, environment, human rights, and product quality and safety). They didn‟t comment on the validity of the controversial business activities screens that some of the rating companies use (for example, KLD specifically indicates whether firms take part in tobacco, alcohol, firearms, gambling, military, or nuclear activities). We couldn‟t find any research dealing with the validity of controversial business issue screens. Furthermore, Sharfman and Hart (2009) indicate that in academic research only the social responsibility ratings are used with any regularity (as opposed to the controversial business activities screens). In this study we use the KLD controversial business activities screens to determine if the firms in our sample take part in tobacco, alcohol, firearms, gambling, military, or nuclear activities. KLD provides these screens, because there is a demand from their customers (including ethical funds) for this information. While this provides an additional reason for choosing these six activities as the “controversial activities” in our study, it is also arguably more objective since it is easier to accurately determine if a firm meets the requirements of one of the controversial business activities screens than to


determine its social performance with any measure of accuracy. 5 Using the controversial business activities screens may overcome the concern raised in the literature that corporate social performance ratings are “extremely value-laden” (Sharfman and Hart (2009: 29, emphasis in original), since firms either meet the requirements of the screen (i.e. are involved in tobacco or alcohol) or not. Individual investors and investment funds can use this information and decide how important it is to exclude these firms from their portfolios. We take the six activities included in the controversial business activities screen as the ”controversial activities” and use KLD data to assess individual firms‟ involvement. Using industry classifications would not be satisfactory, because firms in other (non-controversial) activities with subsidiaries or segments involved in “controversial activities” could be incorrectly classified. Previous empirical findings – ethical funds Although our study is at the firm level and not at the fund level, we report previous findings at the fund level briefly to provide some background. Sauer (1997: 137) concludes that a social-responsibility screen “does not necessarily have an adverse effect on investment performance”. Kreander et al. (2005) found no difference in the financial performance of ethical and non-ethical funds in Europe. Similar findings were reported by Bauer et al. (2005) for US, UK and German mutual funds and by Bauer et al. (2007) in a Canadian setting. Bauer et al. (2006) find the same in Australia, although they find evidence of underperformance by ethical funds in earlier years (1992-1996) and attribute it to a catch-up phase. Boasson et al. (2004) examine the issue using faithbased funds as they “have the toughest exclusionary screens” and find no difference in financial performance with unrestricted funds. In a review of several studies of this nature, Fowler and Hope (2007) find that ethical funds either underperform or fail to outperform other funds. Bollen (2007) finds that ethical mutual funds experience lower fund flow volatility than other funds, suggesting that ethical investors are less likely to change their investment behaviour because of changes in financial performance. Furthermore, Schroder (2007) confirms that socially responsible investment indices‟ riskadjusted returns are no different from that of conventional benchmarks. In summary, there is no evidence that the financial performance of ethical funds is different from that of other mutual funds.

5

See Appendix 1 for the determinants of the KLD screens.

Previous empirical findings – ethical and socially-responsible firms Since there is a dearth of studies at the firm level, we review a range of studies related to ethical and socially responsible firms. This includes social performance, reputation, social disclosure and charitable giving. Moore (2001) suggests that good social performance follows prior good financial performance, at least in the UK supermarket industry. Herremans et al. (1993) found that large US manufacturing firms with better social responsibility reputations outperform poor reputation firms with better share market returns and lower risk for the six years from 1982 to 1987. Eberl and Schwaiger (2005) split reputation into an assessment of competence and sympathy. They find a positive correlation between competence and financial performance, but a negative correlation between sympathy and financial performance after controlling for previous financial performance. Roberts and Dowling (2002) also control for previous performance and still find a positive correlation between reputation and profit. Although social disclosure and ethical conduct differ, one could argue that firms predisposed to the one will be likely to be predisposed to the other. Murray and Gray (2006) found no relationship between share returns and social disclosure. Jones et al. (2007) also found very little evidence of a link between sustainability disclosures and abnormal returns. Donations for good causes are arguably one part of good corporate social behaviour. In fact, Gardberg and Fombrun (2006) believe that charitable giving can increase the value of firms and Godfrey (2005) elaborates that perceived genuine giving will be rewarded, but perceived ingratiating giving will be punished by the market. Patten (2008) finds positive 5-day cumulative abnormal returns for US firms which donated tsunami-relief money in 2004. In summary, at the individual firm level, there is no evidence of a correlation between social disclosure and financial performance, but there is evidence of a positive correlation between reputation and financial performance. Development of Hypothesis According to Herremans et al. (1993), ethical investing activities will depress the share prices of “controversial activity” firms (see also, Merton 1987, Bauer et al. 2005, Chatterji et al. 2009, and Renneboog et al. 2008). The reason for a lower share price is that potential shareholders will expect a higher cost of capital (Merton 1987). We examine whether the cost of capital and/or equity values (share price) are affected by ethical investing. A measure of relative share price will capture a decrease in share price. We follow Herremans et al. (1993) and use return on equity (ROE), specifically the accounting 27


return over the market value of equity, as such a measure of relative share price. 6 The accounting profits of “controversial activity” firms will be unaffected by market reactions, but the market value of their equity will decrease, leading to an increase in the ratio of accounting profits measured against market equity. We state our hypotheses as: Hypothesis 1: Equity values of “controversial activity” firms will be adversely affected by ethical investing. Hypothesis 2: The cost of capital of “controversial activity” firms will be adversely affected by ethical investing. Measured by: Hypothesis 1: The return on equity (ROE) of "controversial activity" firms will be higher than the ROE of other firms. Hypothesis 2: The cost of capital of "controversial activity" firms will be higher than the cost of capital of other firms. Control variables Size and risk have been suggested in the literature to be factors that affect firm performance (Fama and French 1993, 1995, Waddock and Graves 1997). We use total sales as a measure for size in line with other studies in the literature (e.g. Waddock and Graves 1997). Since sales may not be normally distributed, we use the log of sales as our control variable for size. Studies of the type we do here, incorporate controls for risk or assess risk separately (e.g., Herremans et al. 1993, Orlitzky and Benjamin 2001, Kreander et al. 2005, Schroder 2007, Chirinko and Elston 2006). We use Beta from the capital asset pricing model (CAPM) and leverage (following Waddock and Graves 1997) to control for risk. In the CAPM, the only firmspecific measure is Beta, therefore we ignore the rest of the model and use firm Betas as a control variable for risk. Where profitability measures (such as ROE) are the dependent variable, other measures of the level and quality of profitability are typically used as control variables (see for example Grullon et al. 2005). We use quality of earnings, measured as cash flow from operations scaled by earnings, and operating margin as control variables.7 According to Richardson

6

The ROE measure that we use (the accounting net profit over the market value of equity) is the inverse of the Price Earnings Ratio, a well regarded market ratio (see for example, Ou and Penman 1989; White et al. 2003). Furthermore, we use a relative measure of share price as actual share prices are not comparable across firms since share prices are influenced by factors such as the number of shares in issue, etc. 7 We controlled for industry as well. This made no difference to our results, except that the adjusted R-square 28

and Welker (2001), disclosure levels can influence the cost of capital, therefore we include a social and environmental disclosure strength measure from KLD in the cost of capital regression to control for this effect. We exclude leverage and Beta from this equation, as these are used to calculate cost of capital. Equations estimated ROE = ƒ(ContrIndicator, Beta, Size (Log of Sales), Quality of Earnings, Operating margin, Leverage) Cost of Capital = ƒ(ContrIndicator, Size (Log of Sales), Quality of Earnings, Operating margin, Disclosure) We hypothesise that “controversial activity” firms will have a higher ROE and Cost of capital and, therefore expect positive correlations. Where: ROE = net income divided by the multiple of share price at balance date and number of shares in issue Cost of capital = weighted average cost of capital calculated by way of the capital asset pricing model ContrIndicator = 1 if the firm has tobacco, alcohol, firearms, gambling,military, or nuclear activities and 0 otherwise Beta = control variable for risk (share price volatility) Size (Log of Sales) = control variable for size Quality of Earnings = cash flow from operations divided by earnings (net income) Operating margin = net income divided by sales Leverage = liabilities divided by total assets Disclosure = 1 if the firm has a social and environmental disclosure strength Method

We use the KLD database to determine if S&P 500 firms are involved in tobacco, alcohol, firearms, gambling, military, or nuclear activities8 (see the background section above for the rationale) and to determine whether firms have a particular strength in social and environmental disclosure. We obtain financial data for the 2004, 2005 and 2006 years from the Compustat database. After discarding firms not of the model increased. We report the results without the industry controls for the sake of simplicity. 8 KLD indicates in their database whether firms are involved in any of six controversial activities - see Appendix 1 for the determinants of the KLD screens. We take the six activities included as the “controversial business activities” for our study and use KLD data to assess individual firms‟ involvement. We individually assessed each S&P500 firm across the six controversial activities using the KLD data to indicate involvement or not.


rated by KLD; firms with missing financial data on Compustat; and two firms with extreme outliers (one in 2004 and one in 2005) with regards to the calculated figure of ROE, our sample includes 1201 firm-year observations. Results Descriptive statistics and Univariate results Table 1 provides the results of our univariate tests that examine the differences of the means of the variables used between the “controversial activity” firms and other firms. The means and standard deviations are shown. Insert Table 1 To emphasise some of the points, we would like to point out that Table 1 shows:  as expected, that the “controversial activity” firms have a higher ROE, however the difference is not significant.  contrary to our expectation, that “controversial activity” firms have a lower cost of capital (significant at the 1% level).  that the “controversial activity” firms have a lower Beta (significant at the 5% level).  that the “controversial activity” firms are bigger in terms of sales revenue (significant at the 1% level).  that the operating margin of “controversial activity” firms is significantly (at the 1% level) lower than that of other firms.  that the leverage of “controversial activity” firms is significantly (at the 1% level) higher than that of other firms. Both our expectations, that the “controversial activity” firms will have a higher ROE as well as have a higher cost of capital, have not been confirmed in the univariate analysis. The observations of significant differences in the other variables between the two groups confirm the need for multivariate analyses where we control for these factors. Multivariate results We consider the correlation between our variables of interest in order to ensure that the multivariate results are not influenced or driven by multi-collinearity. The results of the Pearson correlations are shown in Table 2. From Table 2 it is clear that none of the measures are highly correlated with each other. The highest correlation of measures used in the same equation (neither Beta nor leverage is used in the cost of capital equation), is between our size measure and leverage,

which at 0.337 is far below the 0.70 level of concern noted in the literature (Stevens 1999).9 Insert Table 2 In Table 3 we report our multivariate analyses with ROE as the dependent variable in Panel A and with cost of capital as the dependent variable in Panel B. Overall our multivariate results as reported in Table 3 show fairly low adjusted R-squares (10.4% and 12.3%), but the model is highly significant and individual control variables are highly significant, again confirming the appropriateness of the control variables. Papers with adjusted R-squares as low as 3% and 4% are published in the top finance journals (see for example Atanassov and Kim 2009), as long as the purpose of the model is not to predict outcomes, but to make deductions and conclusions with reference to the significance level of individual variables (Stock and Watson 2007). We are primarily interested in the significance of the independent variable, ContrIndicator. Several robustness tests confirm our results and report higher R-squares of up to 20.9%. Insert Table 3 We hypothesize “controversial activity” firms to have higher ROE than others. However, in Panel A the ContrIndicator variable is not significant, even at the 10% level (see Column 2). We use ROE (calculated using the market value of equity) to find evidence that capital markets notice the involvement of ethical investing activities by depressing the share values of “controversial activity” firms relative to other firms (Herremans et al. 1993: 590). We therefore find no evidence that capital markets depress the share prices of “controversial activity” firms (or increase the share prices of other firms). Given the non-significance of our independent variable (the ContrIndicator dummy variable), we estimate the model without this variable to determine if it contributes to the explanatory value of the model. The results can be seen in Table 3, Panel A, Columns 3 and 4. The adjusted R-square of the model remains the same with and without the independent variable (ContrIndicator), showing that it does not contribute to the explanatory value of the model. In Panel B, we report the results of our comparison of the cost of capital of "controversial activity" firms with others. We expect a positive correlation, i.e. "controversial activity" firms to have higher cost of capital. However, as in the univariate test, the results show that "controversial activity" firms have a lower cost of capital. You may recall that Merton (1987), Bauer et al (2005), and many others 9

We also tested multicollinearity using VIF tests and this confirmed that none of the measures are highly correlated with each other. 29


state that the effect of ethical investing would be to increase the cost of capital of "controversial activity" firms and/or reduce the cost of capital of other firms. We conclude that capital markets do not notice the effect of ethical investing, however, we subject our results to some robustness tests (see Table 4). Insert Table 4 In Table 4 we report the results of our robustness tests with ROE as the dependent variable in Panel A and with Cost of capital as the dependent variable in Panel B. We winsorise the data, by setting the top 5% of variable values to the 95th percentile and the bottom 5% to the 5th percentile (Table 4, Column 1). We also trim the data (removing the top and bottom 5% of all variables) and estimate the equation with trimmed data (Table 4, Column 2). We further estimate separate regressions for each year (Table 4, Columns 3-5). In each case, our primary results are confirmed, i.e. ContrIndicator remains non-significant in Panel A and negative in Panel B. We also estimate the (ROE) regression with industry control variables included and again confirm our main result (untabulated).

1.

Ethical investing strategies are not, at present, detrimental to “controversial activity” firms and these firms may choose to simply ignore the ethical investing community, 2. Investors not screening out “controversial activity” firms are still profiting from their investments and is similarly unaffected by the ethical investing community, 3. Ethical investors should consider new strategies if they aim to influence firms and/or markets, because according to our results their activities currently do not have the results they might expect, and 4. Ethical investors can take heart in the fact that their strategy of avoiding “controversial activity” firms do not , according to our results, result in lower financial returns (other than to reduce diversification options). We limit our investigation to S&P 500 firms. Care should be taken in generalising our findings to other firms and other markets. Using a different definition of “controversial activities” may also lead to different results, although we do use the criteria of the most popular commercial ratings organisations (KLD).

Conclusions and limitations References We expect that if large numbers of ethical investors exclude “controversial activity” firms from their investment portfolios, that the cost of capital and the share prices of these “controversial activity” firms will be adversely impacted (Bauer et al. 2005, Herremans et al. 1993). It has been estimated that if 20% of investment funds are invested in this way, that the effect would become noticeable (Heinkel et al. 2001). If share prices are depressed, accounting returns on the market value of equity will be higher. We find no evidence that capital markets notice the activities of ethical investors, because, according to our tests, accounting returns on the market value of equity are similar for “controversial activity” firms and other firms and furthermore because "controversial activity" firms have lower cost of capital. The reason may be that ethical investing, using the criteria we use, are still below the 20% threshold required before a difference will become noticeable. Furthermore, as Heinkel et al. (2001) point out, in practice different ethical investors apply different ethical screens. If an unacceptable firm is screened out by only a fraction of ethical investors, its cost of capital will not rise as much as it would have if all ethical investors had excluded it. A well developed body of research on the returns from ethical investment funds shows that ethical fund returns are mostly equal to (but sometimes lag) other funds. Our study complements these findings by considering firm level returns and concluding that they are similar for “controversial activity” firms and other firms. The implications for ethical investors are that: 30

1. 2.

3.

4.

5. 6.

7.

8. 9.

Angel, J.J., and P. Rivoli, 1997, „Does ethical investing impose a cost upon the firm? A theoretical perspective‟, The Journal of Investing 6, 4: 57-61. Agle, B.R., R.K. Mitchell and J.A. Sonnenfeld, 1999, „Who Matters to CEOs? An Investigation of Stakeholder Attributes and Salience, Corporate Performance, and CEO Values‟, The Academy of Management Journal 42, 5: 507-525. Bauer, R., J. Derwall and R. Otten, 2007, „The ethical mutual fund performance debate: New evidence from Canada‟, Journal of Business Ethics 70: 111-124. Bauer, R., K. Koedijk and R. Otten, 2005, „International evidence on ethical mutual fund performance and investment style‟, Journal of Banking & Finance 29: 1751-1767. Bauer, R., R. Otten and A.T. Rad, 2006, „Ethical investing in Australia: Is there a financial penalty?‟, Pacific-Basin Finance Journal 14: 33-48. Berman, S.L., A.C. Wicks, S. Kotha and T.M. Jones, 1999, „Does Stakeholder Orientation Matter? The Relationship between Stakeholder Management Models and Firm Financial Performance‟, The Academy of Management Journal 42, 5: 488-506. Boasson, V., J. Cheng and E. Boasson, 2004, „Are investment managers investing ethically at a disadvantage?‟, The Journal of Applied Management and Entrepreneurship 9, 4: 56-65. Bollen, N.P.B., 2007, „Mutual fund attributes and investor behavior,‟ Journal of Financial and Quantitative Analysis 42, 3: 683-708. Chatterji, A.K., and D.I. Levine, 2008, „Imitate or differentiate? Evaluating the validity of corporate


10.

11.

12.

13.

14.

15. 16. 17. 18.

19.

20. 21. 22.

23.

social responsibility ratings‟, Center for Responsible Business Working Paper Series, No. 37, 1-55. Chatterji, A.K., D.I. Levine and M.W. Toffel, 2009, „How well do social ratings actually measure corporate social responsibility?‟ Journal of Economics and Management Strategy 18, 1: 125-169. Chirinko, R.S., and J.A. Elston, 2006, „Finance, control and profitability: the influence of German banks‟, Journal of Economic Behavior & Organization 59, 1: 69-88. Cho, C.H., and D.M. Patten, 2007, „The role of environmental disclosures as tools of legitimacy: A research note‟, Accounting, Organizations and Society 32, 7-8: 639-647. Eberl, M., and M. Schwaiger, 2005, „Corporate reputation: disentangling the effects on financial performance‟, European Journal of Marketing 39, 7/8: 838-854. Entine, J., 2003, „The myth of social investing: A critique of its practice and consequences for corporate social performance research‟, Organization & Environment 16, 3: 352-368. Fama, E.F., and K.R. French, 1993, „Common risk factors in the returns on stocks and bonds‟, Journal of Financial Economics 33: 3-56. Fama, E.F., and K.R. French, 1995, „Size and Bookto-Market Factors in Earnings and Returns‟, The Journal of Finance 50, 1: 131-155. Fowler, S.J., and C. Hope, 2007, „A Critical Review of Sustainable Business Indices and their Impact‟, Journal of Business Ethics 76, 3: 243-252. Gardberg, N.A., and C.J. Fombrun, 2006, „Corporate citizenship: Creating intangible assets across institutional environments‟, Academy of Management Review 31, 2: 329-346. Godfrey, P.C., 2005, „The relationship between corporate philanthropy and shareholder wealth: A risk management perspective‟, Academy of Management Review 30: 777-798. Graves, S.B., and S.A. Waddock, 1994, „Institutional owners and corporate social performance‟, Academy of Management Journal 37, 4: 1034-1046. Griffin, J.J., and J.F. Mahon, 1997, „The corporate social performance and corporate financial performance debate‟, Business and Society 36, 1: 5-31. Grullon, G., R. Michaely, S. Benartzi and R.H. Thaler, 2005, „Dividend Changes Do Not Signal Changes in Future Profitability‟, Journal of Business 78, 5: 1659-1682. Harrison, J.S., and R.E. Freeman, 1999, „Stakeholders, social responsibility, and performance: Empirical evidence and theoretical perspectives‟, Academy of Management Journal 42, 5: 479-485.

24. Heinkel, R., A. Kraus and J. Zechner, 2001, „The effect of green investment on corporate behavior‟, Journal of Financial and Quantitative Analysis 36, 4: 431449. 25. Herremans, I.M., P. Akathaporn and M. McInnes, 1993, „An investigation of corporate social

26.

27.

28.

29. 30. 31.

32. 33.

34. 35.

36.

37.

38.

39.

40.

responsibility reputation and economic performance‟, Accounting, Organizations and Society 18, 7/8: 587-604. Jones, S., G. Frost, J. Loftus and S. van der Laan, 2007, „An empirical examination of the market returns and financial performance of entities engaged in sustainable reporting‟, Australian Accounting Review 17, 1: 78-87. Kreander, N., R.H. Gray, D.M. Power and C.D. Sinclair, 2005, „Evaluating the performance of ethical and non-ethical funds: a matched pair analysis‟, Journal of Business Finance & Accounting 32, 7&8: 1465-1493. Mahoney, L., and R.W. Roberts, 2007, „Corporate social performance, financial performance and institutional ownership in Canadian firms‟, Accounting Forum 31, 3: 233-253. Merton, R.C., 1987, „A Simple Model of Capital Market Equilibrium with Incomplete Information‟, The Journal of Finance 42, 3: 483-510. Michelson, G., N. Wailes, S. van der Laan and G. Frost, 2004, „Ethical Investment Processes and Outcomes‟, Journal of Business Ethics, 52, 1: 1-10. Milne, M.J., and D.M. Patten, 2002, „Securing organizational legitimacy: An experimental decision case examining the impact of environmental disclosures‟, Accounting, Auditing & Accountability Journal 15, 3: 372-405. Moore, G., 2001, „Corporate social and financial performance: An investigation in the UK supermarket industry‟, Journal of Business Ethics 34: 299-315. Murray, A., and R. Gray, 2006, „Do financial markets care about social and environmental disclosure? Further evidence and exploration from the UK‟, Accounting, Auditing & Accountability Journal 19, 2: 228255. Orlitzky, M., and J.D. Benjamin, 2001, „Corporate social performance and firm risk: A meta-analytic review‟, Business and Society 40, 4: 369-396. Ou, J.A., and S.H. Penman, 1989, „Accounting Measurement, Price-Earnings Ratio, and the Information content of Security Prices‟, Journal of Accounting Research 27, 2: 111-144. Patten, D.M., 2008, „Does the market value corporate philanthropy? Evidence from the response to the 2004 tsunami relief effort‟, Journal of Business Ethics 81, 3: 599-607. Renneboog, L., J. Ter Horst and C. Zhang, 2008, „Socially responsible investments: Institutional aspects, performance, and investor behavior‟, Journal of Banking & Finance 32: 1723-1742. Richardson, A. J., and M. Welker, 2001, „Social disclosure, financial disclosure and the cost of equity capital‟, Accounting, Organizations and Society 26, 7/8: 597-616. Roberts, P.W., and G.R. Dowling, 2002, H „Corporate reputation and sustained superior financial performance‟, Strategic Management Journal 23: 10771093. Sauer, D.A., 1997, „The impact of social-responsible screens on investment performance: evidence from the Domini 400 social index equity mutual fund‟, Review of Financial Economics 6, 2: 137-149. 31

Corporate Ownership & Control / Volume 8, Issue 1, Fall 2010 41. Schepers, D.H., and S.P. Sethi, 2003, „Do socially responsible funds actually deliver what they promise?: Bridging the gap between the promise and performance of socially responsible funds‟, Business and Society Review 108, 1: 11-32. 42. Schroder, M., 2007, „Is there a difference? The performance characteristics of SRI equity indices‟, Journal of Business Finance & Accounting 34, 1&2: 331348. 43. Schueth, S., 2003, „Socially responsible investing in the United States‟, Journal of Business Ethics 43, 3: 189194. 44. Sharfman, M.P., and T.A. Hart, 2009, „Revisiting the Concurrent Validity of the Revised Kinder Lydenberg and Domini Corporate Social Performance Indicators‟, Working Paper, 1-54. 45. Social Investment Forum, 2007, „2007 Report on Socially Responsible Investing Trends in the United States: Executive Summary‟, Social Investment Forum, Washington, D.C., Accessed on 28 April 2009 at http://www.socialinvest.org/pdf/SRI_Trends_Exec Summary_2007.pdf. 46. Stevens, J., 1999, Intermediate statistics: a modern approach – Second edition, Lawrence Erlbaum Associates, Hillsdale, N.J. 47. Waddock, S.A., and S.B. Graves, 1997, „The corporate social performance-financial performance link‟, Strategic Management Journal 18, 4: 303-319. 48. White, G.I., A.C. Sondhi, D. Fried, 2003, The Analysis and Use of Financial Statements – Third edition, Wiley, N.J.

32


Tables Table 1. Descriptive statistics and univariate tests for differences between “controversial activity” firms and other firms for the period from 2004 to 2006 “Controversial activity” firms Other firms Variable Mean (std. dev.) Mean (std. dev.) t-statistic ROE 0.045 (0.0887) 0.040 (0.1911) 0.333 Cost of Capital 0.061 (0.0241) 0.071 (0.0332) 4.178*** Beta 1.035 (0.4869) 1.119 (0.5330) 2.097** Size(Log of Sales) 9.291 (0.9176) 8.952 (1.1941) 3.833*** Quality of earnings 0.724 (0.4627) 0.738 (1.7091) 0.115 Operating margin 0.116 (0.0943) 0.140(0.1229) 2.572*** Leverage 0.676 (0.1467) 0.581 (0.2122) 6.141*** Disclosure 0.141 (0.3486) 0.109 (0.3123) 1.284 N 204 997 “Controversial activity” firms are firms involved in tobacco, alcohol, firearms, gambling, military, and nuclear operations ROE = Return on market value of equity N = Number of firms ***, **, * denotes significance at the 1%, 5%, 10% levels (two-tailed), respectively

Table 2. Pearson correlations

Cost of Capital ROE Beta Size Quality of earnings Operating margin Leverage SE Discl. strength

Contr Indicat -0.120 0.012 -0.062 0.107 -0.003 -0.074 0.175 0.037

Cost of Cap 0.100 0.577 -0.336 -0.042 0.114 -0.763 -0.039

ROE

Beta

Size

-0.154 0.029 -0.072 0.207 -0.106 -0.006

-0.098 -0.023 -0.117 -0.067 -0.011

-0.021 -0.184 0.337 0.249

Qual earn

Oper mar

0.003 -0.021 -0.002

-0.254 -0.032

Lever age

0.009

ContrIndicat = 1 if firm involved in “controversial activities”, namely tobacco, alcohol, firearms, gambling, military, and nuclear operations and 0 otherwise Cost of Cap = Weighted average cost of capital calculated using the capital asset pricing model ROE = Return on market value of equity Size = Size measured by log of sales Qual earn = Quality of earnings measured by cash flow from operations divided by earnings Oper mar = Operating margin measured by net income divided by sales Leverage = measured by liabilities divided by total assets SE Discl. Strength = Social and environmental disclosure strength measure (from the KLD database)

33


Table 3. Multivariate analysis Panel A: Dependent variable – ROE

ContrIndicator Beta Size (Log Sales) Quality of earnings Operating margin Leverage Constant Adjusted R² Significance of the model N

Including ContrIndicator T statistic Significance Column 1 Column 2 1.007 0.314 -7.461 0.000*** 1.951 0.050** -3.240 0.001*** 4.694 0.000*** -4.210 0.000*** 1.683 0.093* 0.104 0.000*** 1201

Panel B: Dependent variable – Cost of Capital T statistic ContrIndicator -3.057 Size (Log Sales) -11.626 Quality of earnings -1.825 Operating margin 1.782 SE Discl. strength 1.735 Constant 20.711 Adjusted R² 0.123 Significance of the model N 1201

Excluding ContrIndicator T statistic Column 3

Significance Column 4

-7.562 2.003 -3.215 4.675 -4.103 1.655 0.104

0.000*** 0.045** 0.001*** 0.000*** 0.000*** 0.098* 0.000***

1201

Significance 0.002*** 0.000*** 0.068* 0.075* 0.083* 0.000*** 0.000***

ROE = Return on market value of equity Cost of Capital = Weighted average cost of capital using the capital asset pricing model ContrIndicator = firms involved in tobacco, alcohol, firearms, gambling, military, and nuclear operations Quality of earnings = Cash flow from operations divided by earnings SE Discl. Strength = Social and environmental disclosure strength measure (from KLD database) N = number of firms ***, **, * denotes significance at the 1%, 5%, 10% levels

34


Table 4. Robustness tests Panel A: Dependent variable: ROE Winsorise T stat 0.477 3.250*** 11.177*** -3.708*** 14.218*** 4.586*** -7.158*** 0.209 0.000*** 1201

Trim T stat 0.733 1.103 5.045*** -3.361*** 3.159*** 7.120*** 0.240 0.124 0.000*** 734

2006 T stat 1.007 1.801 2.314 -0.580 6.193*** -0.495 -1.537 0.095 0.000*** 395

2005 T stat -1.782 0.396 2.938*** 6.756*** 6.650*** 0.484 -3.513*** 0.187 0.000*** 403

2004 T stat 0.985 -5.559*** 0.473 -2.119** 1.864 -2.878*** 2.099** 0.132 0.000*** 403

Panel B: Dependent variable – Cost of Capital Winsorise T stat ContrIndicator -3.566*** Size (Log Sales) -10.373*** Quality of earnings 0.849 Operating margin 2.486** SE Discl. strength 1.548 Constant 17.876*** Adjusted R² 0.117 Significance of the model 0.000*** N 1201

Trim T stat -3.387*** -5.042*** 3.100*** 1.234 0.933 11.012*** 0.062 0.000*** 839

2006 T stat -1.973** -4.702*** 0.747 3.134*** 1.056 8.924*** 0.098 0.000*** 395

2005 T stat -1.996** -5.074*** 1.205 2.640*** 0.544 9.913*** 0.101 0.000*** 403

2004 T stat -1.347 -9.907*** -2.531** -1.998** 1.014 16.098*** 0.208 0.000*** 403

ContrIndicator Beta Size (Log Sales) Quality of earnings Operating margin Leverage Constant Adjusted R² Significance of the model N

ROE = Return on market value of equity Cost of Capital = Weighted average cost of capital using the capital asset pricing model ContrIndicator = firms involved in tobacco, alcohol, firearms, gambling, military, and nuclear operations Quality of earnings = Cash flow from operations divided by earnings SE Discl. Strength = Social and environmental disclosure strength measure (from KLD database) Winsorise – The top and bottom 5% of each variable in the equation was set to the 95th percentile and the 5th percentile respectively Trim – Firm year observations with variables in the top and bottom 5% of observations for the variable was removed N = number of firms ***, **, * denotes significance at the 1%, 5%, 10% levels

35


Appendix 1 Controversial Business Activities – Summary of the KLD Screening Criteria Licensing1 Manufacturers2 Manufacturers of products necessary for the production3 Retailers4

Owners and operators5 Supporting services6 Ownership

Alcohol yes yes

Gambling yes yes

15% or more of revenue 15% or more of revenue

Tobacco Yes Yes

Firearms

Military

Nuclear Power

yes

2% or more or >$50mil 2% or more of revenue or >$50mil

yes

15% or more of revenue 15% or more of revenue yes

yes

15% or more of revenue yes

yes yes For all Issues, company is more than 50% owned by a company with a controversial activity involvement; the company owns more than 20% of another company with a controversial activity involvement

Notes to appendix 1: 1. Yes for Licensing means the company licenses its company or brand name to the product. 2. Yes for Manufacturers means the company are involved in the manufacturing of the product. 3. Manufacturers of products necessary for production relates to companies manufacturing products that are necessary for the production of the controversial product. Yes for this category means any manufacturing. 4. Retailers are companies deriving income from the distribution or the product (wholesale or retail). 5. Yes for Owners and operators means that the company owns or operates the controversial operation. 6. Yes for Supporting services means the company provides supporting services to the controversial operation.

36


CORPORATE GOVERNANCE AND THE USE OF EVA COMPENSATION Ralph DeFeo*, Ehsan Nikbakht**, Andrew C. Spieler***# Abstract The purpose of this paper is to determine if companies that chose to incorporate Economic Value Added (EVA®) as part of their executive compensation package tend to have better corporate governance than similar firms who have not chosen to use EVA®. EVA® is an economic profit metric developed by Stern Stewart & Co., which is calculated by taking the Net Operating Profits after Taxes (NOPAT) and subtracting a capital charge from it. Through the use of binary logistic regression the strength of several key corporate governance measures were tested in order to ascertain whether they have a significant impact on influencing a firm to use EVA®. A major finding is that firms that employed EVA® as part of their compensation package tend to have a weaker corporate governance. Keywords: Corporate Governance, EVA, Executive Compensation

The authors acknowledge the capable research assistance of Xiaohu Deng *Independent consultant **Department of Finance, Frank G. Zarb School of Business, Hofstra University, Hempstead, NY 11549 ***Department of Finance, Frank G. Zarb School of Business, Hofstra University, Hempstead, NY 11549 # Corresponding author

Introduction Economic Value Added (EVA®) is a performance metric developed by Stern Stewart & Co. in the early 1990‟s. Since that time there has been a wealth of empirical studies performed which analyzed the explanatory power of EVA® by looking at the correlation between EVA® and equity returns. There has also been a large body of studies which examined the change in management behavior after the adoption of EVA® as a performance measure. This paper adds to the extant literature by studying the role corporate governance plays with the selection of EVA®. To study the quality of corporate governance within a firm one critically important component will be the composition of the board of directors. The board of directors are elected by the shareholders to oversee management and ultimately approve the compensation packages and the performance metrics by which management will be judged. We have compiled a useable sample of 52 companies that implemented EVA® as a part of their compensation package. Those EVA® companies are matched with 52 companies who do not use EVA®. The results indicate that there is a significant correlation between poorer corporate governance and the selection of EVA® as a performance measure. Part 1 of the paper provides a detailed literature review of EVA® and applications. Part 2 discusses

the data collection, methodology, and analysis. The results and conclusions of the study are provided in Part 3. 1.

Literature Review

Economic Value Added (EVA®) is a metric that was created by Stern Stewart & Co. which is conjectured to provide a better link to value creation then any of the other current metrics in practice including EPS, FCF, RONA, ROE, ROA and other ratios or multiples. Evans and Evans (1998) suggest that “under agency theory, the agent (CEO) is attempting to maximize their utility within the constraints imposed by the principal (owner).” Therefore if the wrong metric is chosen to judge the performance of an executive then he/she will conduct their actions in such a way which will maximize that metric, whether or not it is creating value for the shareholders. This is the reason why it is so important to identify a metric which provides a stronger link to value creation then any other criteria. The effect of this would be a minimization of agency costs, which would create value for shareholders. G. Bennett Stewart (1991) demonstrates that EVA® is the solution. Stewart defines EVA® as Net Operating Profits after Taxes (NOPAT) minus a capital charge. The key difference between EVA and accounting-based performance measures is the capital charge includes the opportunity cost of providing equity capital. Hence, 37


the hurdle rate includes the explicit cost of debt (interest expense) and implicit cost of equity (opportunity cost). Accordingly, NOPAT can be defined as follows: NOPAT = Income available to common equity + Preferred dividends + Minority interest + Interest expense + Increase in equity equivalents Stewart (1996) claims that increasing EVA® will positively correlate with increasing the value of the firm, and thus become the key value driver of any company. EVA® can only be increased in three ways: (1) increase NOPAT, (2) decrease the amount of capital used, or (3) decrease the WACC. In theory, there are no additional information problems between shareholders and managers and so the link to increasing shareholder value must flow through one of the above conditions. Drucker (1995) states the theory behind EVA®. He suggests, “Until a business returns a profit that is greater than its cost of capital, it operates at a loss.” This definition implicitly incorporates the opportunity cost of capital whereas the firm may operate profitably from an accounting point of view, i.e. Net Income is positive. The theory behind EVA® is similar to that behind NPV, in that they both have to exceed their respective costs of capital in order to create value. For both EVA® and NPV if they do not exceed zero the company or project is destroying value for the shareholders. Stewart (1991) examined the relationship between Market Value Added (MVA) and EVA®. His sample consisted of 618 US companies with data gathered from the late 1980‟s. He concluded that the relationship between MVA and EVA® was strong when EVA® was positive. However, the relationship tended not to hold up that well when EVA® became negative. A major reason for this could be the fact that no matter how bad companies do they still have the option of liquidation, which may create a floor for the MVA. Uyemura, Kantor and Pettit (1996) studied 100 bank holding companies again to examine the relationship between MVA and EVA® in the banking sector. The data was collected over a ten year period from 1985 to 1995. They set up a regression with MVA as the dependent variable and EVA®, ROA, ROE, Net income, and EPS as the independent variables. The correlations between the performance measures and MVA were as follows: EVA® (0.40), ROA (0.13), ROE (0.10), Net income (0.80) and EPS (0.60). These results are contrary to the conclusions found in the Economist article, “Valuing Companies – A Star to Sail by?”(1997). In this article the author suggested that EVA® metrics would not work well for financial companies, since they must hold capital on the side for regulatory purposes. O‟Byrne (1996), also of Stern Stewart & Co., reported similar results in his study. He found that over a five year period the 38

changes in EVA® year over year explained 55% of the variation in the year over year changes of MVA. He had even better results once the study was enlarged to look at a ten year period. Over a ten year period the year over year changes in EVA® explained 74% of the variation in the year over year changes of MVA. However, the studies conducted by employees of Stern Stewart & Co. must be qualified since there are obvious conflicts of interest. The following partial list of research studies conducted by independent, i.e. non-Stern Stewart employees, provides additional insight into the relationship. Milunovich and Tsuei (1996) also report a high correlation between EVA® and MVA. In their study they showed that EVA® explained 42% of the variation in MVA, while EPS growth only explained 34% of the variation in MVA, and ROE and EPS only explained 29% of the variation in MVA. Lehn and Makhija (1996) also studied the link between MVA and EVA®. Their data consisted of 241 US companies covering 1987 – 88, and 1992 -93. They concluded that EVA® correlates slightly better than ROA, ROE, or ROS with MVA. More interestingly in their study they found that CEO turnover was significantly related to EVA®. In a follow up study, Lehn and Makhija (1997) concluded that CEO‟s are evaluated more on the basis of EVA® then they are with other accounting metrics. Biddle, Bowen and Wallace (1999) came to an interesting conclusion in their analysis of the annual Fortune 1000 performance over the period of 1988 1997. The authors find that the difference between EVA® and Residual Income is fairly small which would indicate that the majority of the adjustments that Stern Stewart makes in calculating EVA® tend to offset each other. Anderson, Bey and Weaver (2004) also agree with these conclusions. They found very strong correlations between adjusted and unadjusted EVA®, which led them to question the usefulness of calculating and making these set of adjustments. Biddle, Bowen and Wallace (1999) analyzed at a set of 6,174 firms over the period from 1984 -1993. They regressed EVA®, residual income, net income (before extraordinary items), and cash flow from operations (CFO) against annual market adjusted stock returns. The results showed that net income is significantly more associated with market adjusted stock returns than any of the other metrics. The correlation coefficients were as followed, NI (0.13), RI (0.70), EVA® (0.60), and CFO (0.30). The regressions over a cumulative five year window also show net income to generate the highest pair-wise correlation coefficient of (0.31), followed by CFO (0.19), EVA® (0.14), and RI (0.11). These results are contrary to the findings of the previously mentioned studies which seem to indicate a strong relationship between EVA® and share price. Biddle, Bowen and Wallace (1999) also studied the actions of management to see if firms that adopted EVA® changed relative to the period before

Corporate Ownership & Control / Volume 8, Issue 1, Fall 2010 the EVA® adoption. Their results found a statistically significant increase in the amount of asset dispositions and a statistically significant decrease in the amount of asst acquisitions, pos- adoption of EVA®. The authors also report a statistically significant increase in the amount of share repurchases. Broadly speaking, All three of these corporate activities are consistent with the actions expected from a company following an EVA® maximization strategy. They also found that EVA® increased 1300% after the adoption of EVA® into the executive compensation plans, which shows that management will seek to maximize the metric directly tied to their compensation. This finding reiterates why it is so important to find the right metric defined as the one that has the highest correlation with stock returns. Wallace (1997) in his earlier study also found similar results with regard to the decisions managers make when their compensation is linked to EVA® improvements. He observed that when EVA® firms were compared with non-EVA® firms, EVA® firms tended to: 1) make less investments, 2) dispose of more assets, 3) made more share repurchases, and 4) used their assets more intensely. Chen and Dodd (1998) studied a sample of 668 US firms taken from Stern Stewart‟s performance 1000 database over a ten year period. They regressed operating income, residual income, and EVA® against stock returns. Their empirical results found operating income, residual income and EVA® to have a 0.06, 0.05 and 0.02 correlation, respectively. Therefore they concluded that EVA® exhibits less explanatory power than the more traditional accounting measures. It is also to point out that the EVA® metric has limitations as well. This was the focus of Riceman, Cahan, and Lal (2000). Their goal was to how easy it was for corporate executive using EVA® to truly understand what activities would increase it. In their research design, executives at companies employing EVA completed questionnaires. Surprisingly, the authors conclude that the executives had a relatively poor understanding of what activities would increase EVA®. Using a standard grading system (A-F), the average score would have been a D. The analysis also bifurcated the sample into firms that understand EVA and firms that do not. Not surprisingly, the interaction between understanding EVA® and using EVA® revealed interesting insights. In particular, companies that had both understanding and EVA® usage outperformed against the sample companies, while companies that used EVA® but did not fully understand the metric, under performed against the sample companies. Overall the results of EVA® studies seem to be conflicting. Many of the studies conducted by Stern Stewart & Co. and some of the independent studies tend to support a strong relationship between EVA® and price. However, the majority of studies performed, especially the ones with larger sample

sizes, seem to indicate that EVA® is in fact less correlated to market returns than EPS or other commonly used metrics. 2.

Data and Analysis

The universe of sample companies were selected to be used in this study were identified by one of two ways. The first group of firms that use EVA® were identified directly from Stern Steward & Co website. These are firms which selected to use EVA® as part of their compensation system under the guidance of Stern Stewart & Co. The second set of firms was identified by their stated use of EVA® as part of their compensation package in their proxy statement. The second set of firms was hand collected after searching the Lexis-Nexis database with appropriate key word searches. All firms that used EVA® for at least two consecutive years over the period from 1990 to 2000 were included in the sample. The rationale behind this data screen is that if a firm used EVA® for less than two years it was likely not a large part of their compensation package or compensation philosophy. It is likely that these firms would not have the same characteristics of a true EVA® company and would be misclassified. This procedure yielded 87 EVA® companies, 45 of which were being advised by Stern Stewart & Co. and 42 of which were implementing EVA® without the help of Stern Stewart & Co. From this sample only 52 companies (36 Stern Stewart & Co., 16 non Stern Stewart & Co.) could be used in the analysis due to data limitations. The primary data item that was missing was the overnance Index developed by Gompers, Ishii, Metrick (2003). In short, the index compiles 24 anti-takeover and governance variables to rate each firm. The 52 remaining EVA® companies were paired up with a size and industry matched, nonEVA® company. Specifically, the matched sample found the firm with closest total assets within the same two-digit SIC code. This pairwise procedure should control for any systematic size or industry effects. The data used in the study is based upon the first identifiable filing date of the proxy denoting the use of EVA® in the compensation scheme. The paired firms use the closest available proxy relative to the event firm. Return data were compiled from the Center for Research in Securities Prices (CRSP) database. Information relating to SIC codes, total assets, balance sheet ratios and other accounting data were gathered from Compustat. The corporate governance data was extracted directly from firm proxy statements available on EDGAR. The Governance Index (GIndex) data used in this study was constructed by Gompers, Ishii, and Metrick (2003). The GIndex is calculated by adding one point for every corporate provision which reduces shareholder rights, e.g. staggered board. In total, twenty four different provisions are examined so the GIndex for each firm must fall within 0-24 range. A 39


high GIndex would signal that management possesses a relatively entrenched and/or poorly aligned management. The basic empirical methodology is as follows: t-tests are used in the univariate analysis and binary logistic regressions are used in the multivariate analysis. Twelve independent variables have been identified in the extant literature as possible drivers of EVA® selection. A complete description of all variables used in the analysis can be found in Table 2. The corporate governance variables used in this study include the GIndex, percentage if insider directors, percentage of outsider directors, percentage of gray directors, board size, and whether the CEO holds a dual leadership position (also serves as chairman of the board). Besides corporate governance variables, we include return variables in the analysis to examine if there is any systematic influence on the choice of EVA® selection. Two control variables (total assets and long term debt / total capital) appear in all regression models. The reason for including these variables is that firms significantly different in size and leverage may have, on average, drastic differences in board composition and market returns, respectively. 10 By including those two variables in the regression, any systematic influence can be controlled for. The Pearson correlation coefficients between the variables are displayed on Table 3. The only variables which have a correlation which exceeds 0.50 in magnitude are (1) % Inside with % Outside variables with -0.69 correlation and (2) excess (1yr) with excess (3yr), 0.56pair-wisecorrelation. The signs of these correlations fit our intuition: the percent of insiders will move inversely with the percent of outsiders on the board. Similarly, the 3 year excess return will be positively related to the 1 year return. In the empirical models, only one measure is used at a time. The remainder of the other variables are not strongly correlated therefore inclusion should not bias the coefficients or standard errors. Table 4 summarizes the t-test for the difference in means between the EVA® and Non EVA® samples. The results indicate two variables that yielded significant differences. The first variable was the GIndex which is significant at the 1% level. The GIndex difference reported a significantly positive t-statistic which means that EVA® companies tended to have a higher GIndex indicating poorer corporate governance. The dual leadership variable (CEO = CBOD) was significant at the 6% level. The t-statistic was also positive indicating that EVA® companies tended to have the CEO also serve as the chairman of the board more often than non-EVA® companies. This is anecdotally also indicative of poorer corporate governance among EVA® companies. We also note EVA® possess marginally more insiders (p=0.11) also signaling

10

There are many studies that document a firm-size effect.

40

greater inside board representation and weaker corporate governance. Although the univariate analysis reported significance in two of the variables it is necessary to check to see if the influence is retained under the multivariate analysis framework. For the multivariate analysis a binary logistic regression was performed on the data since the dependent variable (whether or not the firm uses EVA®) was constructed as an indicator variable. The results of the regression models are shown on Table 5. Interestingly, the full regression model yields only one significant variable, the GIndex (p=0.021). However, under the multivariate framework the CEO=CBOD variable is no longer significant. We also note that the constant in the equation is not significantly different from zero which indicates the independent variables in the regression model are explaining most of the variation in the dependent variable. The Omnibus tests of model coefficients also supports these findings as the Chisquare test for the model is significant at the 10% level indicating a good fit for the regression model. In a logistic regression the interpretation of the estimated coefficients is subtle. For instance the only significant variable in the regression, the GIndex, has a logit coefficient of 0.211. All of the odds ratios can be found on Table 5 under the heading Exp(B). The interpretation is that for every one unit increase in the independent variable the odds that the dependent variable will be one increases by the odds ratio. So, for every one point increase in the GIndex the odds that the company will use EVA® compensation increases by 1.235. An additional logit regression was performed including the 3-year excess return prior to EVA adoption. The sample size is reduced to 94 based on data availability. The results are displayed in Table 6. From this regression there appears to be no significance in the ability of the prior three years excess returns to predict whether a company will use EVA®. However, the GIndex continues to be significant at the 5% level providing robustness to the previous empirical finding. The final regression that was performed was a regression where the dependent variable denotes if the firm was advised by Stern Stewart Co. The independent variables remained the same as for the first regression model. The sample was necessarily reduced by this limitation and yielded a sample size of 52. The logistic output can be found on Table 7. The results find one variable that distinguishes the firms that use Stern Stewart and those that do not, board size (p=0.011). The results indicate that larger boards tend to use Stern Stewart more than smaller boards, all else equal. The interpretation of this result is not clear. First, there is ample anecdotal evidence and prior research that finds that larger boards are unwieldy and provide less oversight of firm management. Thus, the board may be enacting its fiduciary responsibility to outsource a component of


CEO compensation. Second, larger boards are more likely to be process-oriented and employing an outside consultant is consistent with that interpretation. Smaller boards are more likely to be ad hoc in their decisions and processes. We also validated the model by computing the correct classification. The overall correct classification is 73% as reported in Table 8. In addition, Wald Test was performed to observe how far the estimated parameters differ from zero. Note that the Wald test can be used to test multiple parameters simultaneously. The results are tabulated in Table 9. The interpretation of Wald test for the variables is provided in the forthcoming section of conclusions. 3.

Conclusions

The results of this study provide evidence that EVA® firms tend to have a higher GIndex than non-EVA® firms. Since a higher GIndex number indicates more entrenched / less aligned management, it provides a strong signal of poorer corporate governance. Therefore this study suggests that firms that chose to use EVA® as part of their compensation package tend to have weak corporate governance. This finding is somewhat counterintuitive as firms that use EVA® would a priori be expected to have better corporate governance. This opens a larger question of why do EVA® firms have poorer corporate governance than non-EVA® firms. This question is beyond the scope this research. However, it could be speculated that companies which suffer from poor corporate governance want to window dress the quality of their governance by selecting EVA® as a performance measure. Note that the claim and perception behind EVA® is that it will align shareholder‟s and manager‟s goals more effectively. A second plausible explanation is that firms that exhibit relatively poor corporate governance, could employ the EVA® metric as a conscious choice to reduce agency costs. This study also show finds no significant difference in the returns achieved by companies which use EVA® compared to companies which do not use EVA® after controlling for industry and size. Further, prior performance does not appear to influence the decision to adopt EVA. Since the return data prior to the implementation of EVA® and after the implementation of EVA® shows no significant difference when compared to the non EVA® companies over the same time span, it does not help explain why weak boards tend to select EVA® as a performance measure. It does however show that the choice of using EVA® does not have an observable, significant effect on performance. It is also interesting to note that there was no significance found in the multivariate analysis for any of the other corporate governance variables aside from GIndex. Another question raised by these findings is about the lack of influence on the board structure. Why there was no role for the board

structure if the quality of corporate governance plays such a strong role in determining whether a company will use EVA®? The answer may be that “outsider” directors are sometimes not really outsider directors in substance. The results are also consistent with the robustness of the GIndex subsuming the impact of individual board characteristics for a more holistic measure. The final regression performed showed that the excess returns after the implementation of EVA® were statistically equivalent between firms under the guidance of Stern Stewart & Co. compared with firms that implemented EVA® on their own. It is important to note that this study, as all empirical studies, has limitations. First, this study used a restrictive sample size of only 52 companies employing EVA®. A larger sample size, if available, might yield additional insight into the nuances of corporate governance. Second, firms may differ in their use of EVA® as part of executive compensation. For example, some firms may tie EVA® to salary and others may link EVA® to bonuses or incentive compensation only. Unfortunately, the proxy statements did not provide enough information to further differentiate EVA® beyond a simple binary variable. Nevertheless, the results of this study should bring to light the importance the quality of corporate governance has on the implementation of EVA®, executive compensation and more generally the complex principal-agent relationship between shareholders and manager. References 1.

2.

3. 4. 5.

6.

7.

Anderson, Anne M., Bey, Roger P., Weaver, Samuel C., 2004, “Economic Value Added Adjustments: Much to Do About Nothing?,” Working Paper, Presented at the Midwest Finance Association Meetings. Baysinger, B.D., & H. Butler, 1985, “Corporate Governance and the Board of Directors: Performance Effects of Changes in Board Composition.” Journal of Law, Economics, and Organizations, 1, 101–124. Biddle, Gary C., Bowen, Robert M., and Wallace, James S., 1999, “Evidence on EVA,” Journal of Applied Corporate Finance, Vol. 12, No. 2, 15-20. Brickly, J., Coles, J., and Terry, R., 1994, "Outside directors and the adoption of poison pills," Journal of Financial Economics 34, 371-390. Chen, Shimin and Dodd, James L., 1998, "Usefulness of Accounting Earnings, Residual Income, and EVA: A Value-Relevance Perspective." Working Paper, Drake University, 1998. Dillon, Ray and Owen, James E, 1997, “EVA as a Financial Metric: Attributes, Utilization, and Relationships to NPV,” Financial Practice and Education, 32-40. Drucker, Peter, 1995 “The Information Executives Truly Need,” Harvard Business Review, Jan-Feb 1995.

41

Corporate Ownership & Control / Volume 8, Issue 1, Fall 2010 8. 9. 10. 11. 12.

13.

14. 15.

Economist “Valuing Companies – A Star to Sail By?,” 1997, The Economist, August 2nd 1997, 53-55. Fama, E.F., & M.C. Jensen, 1983, “Separation of Ownership and Control.” Journal of Law and Economics, 26, 301–325. Gompers, P., Ishii, J., and Metrick, A., 2003, “Corporate Governance and Equity Prices,” Quarterly Journal of Economics, 118, 107-155. Jenson, M.C. & Murphy, K.J., 1990, “Performance Pay & Top Management Incentives,” Journal of Political Economy, Vol. 98, No. 2, 225-264. Lehn, Kenneth and Makhija, Anil K., 1996, “EVA & MVA as Performance Measures and Signals for Strategic Change, “Strategy and Leadership, Vol. 24, No. 3, 34. Lehn, Kenneth and Makhija, Anil K., 1997, “EVA, Accounting Profits, and CEO Turnover: An Empirical Examination, 1985-1994,” Journal of Applied Corporate Finance, Vol. 10, No. 2, 90-97. Milunovich, Steven and Tsuei, Albert, 1996, “EVA in the Computer Industry,” Journal of Applied Corporate Finance, Vol. 9, No. 1, 104-115. O‟Byrne, Stephen F., 1996, “EVA and Market Value,” Journal of Applied Corporate Finance, Vol. 9, No. 1, 116125.

16. Pareja, Ignacio Velez, and Grancolombiano, Politecnico, 2001, “Value Creation and Its Measurement: A Critical Look at EVA,” Working Paper, School of Industrial Engineering Bogota, Colombia. 17. Riceman, Stephen S., Cahan, Steven F., Lal, Mohan, 2000, “Do Managers Perform Better Under EVA Bonus Schemes?,” Working Paper, Massey University. 18. Stewart, G. Bennett III, The Quest for Value-A guide for Senior Managers. Harper Business, 1991. 19. Stern Stewart & Co, 2000 http://www.eva.com/content/evaabout/info/evaw orkstable.pdf 20. Uyemura, Dennis G., Kantor, Charles C., and Pettit, Justin M., 1996, “EVA for Banks: Value Creation, Risk Management, and Profitability Measurement,” Journal of Applied Corporate Finance, Vol. 9, No. 2, 94109. 21. Wallace, J, 1997, “Adopting residual income based compensation plans: Do you know what you pay for?,” Journal of Accounting & Finance, December 24th 1997, 275-300. 22. Yermack, David, 1996, "Higher Market Valuation of Companies with a Small Board of Directors," Journal of Financial Economics 40, 185-211.

Appendices Table 1. Paired sample firms using EVA® vs. not using EVA Firms uses EVA® (1) or Not (0)

Name of Firm

Ticker

ARMSTRONG WORLD INDUSTRIES INC

ACK

1

AVERY DENNISON CORPORATION

AVY

1

COOPER CAMERON CORP

CAM

1

CILCORP INC

CER

1

FLEMING COMPANIES INC /OK/

FLM

1

FLOWSERVE CORP

FLS

1

GENESCO INC

GCO

1

GREAT LAKES CHEMICAL CORP

GLK

1

IMMUNOMEDICS INC

IMMU

1

MILACRON INC

MZ

1

PERFORMANCE FOOD GROUP CO

PFGC

1

CHILDRENS PLACE RETAIL STORES INC

PLCE

1

TENNECO AUTOMOTIVE INC

TEN

1

TARGET CORP

TGT

1

TECHNITROL INC

TNL

1

WHOLE FOODS MARKET INC

WFMI

1

Acxiom

ACXM

1

ADC Telecommunication

ADCT

1

Best Buy

BBY

1

Biose Cascade

BCC

1

Becton Dickinson

BDX

1

Briggs & Stratton

BGG

1

42


Table 1 continued Bausch & Lomb

BOL

1

Bowater

BOW

1

Centura Banks

CBC

1

CDI Corp

CDI

1

Crane

CR

1

RR Donnelly & Sons

DNY (RRD)

1

Equifax

EFX

1

Sprint

FON

1

Guidant

GDT

1

Georgia Pacific

GP

1

Hershey Foods

HSY

1

Interntaional Multifood

IMC

1

JC Penny

JCP

1

Coca Cola

KO

1

ELI Lilly

LLY

1

Millenium Chemical

MCH

1

Herman Miller

MLHR

1

Material Sciences Corp

MSC

1

Manitowoc Company

MTW

1

Noble Drilling

NE

1

Olin

OLN

1

Perkinelmer

PKI

1

Ryder Systems

R

1

Silicon Valley Bank

SIVB

1

Standard Motor Products

SMP

1

SPX

SPW

1

Tenet Healthcare

THC

1

Toys R US

TOY

1

Tupperware

TUP

1

Vulcan Materials

VMC

1

Cabot Corp

CBT

0

Central Louisiana Electric

CNL

0

Costco

COST

0

Donaldson Company

DCI

0

Deb Shops

DEBS

0

Goodyear Tire

GT

0

Millennium Phamaceuticals

MLNM

0

Newell Rubermade

NWL

0

Patterson Companies

PDCO

0

Pall Corp

PLL

0

Rex Stores

RSC

0

Sonoco Products

SON

0

Sysco

SYY

0

Trimble Navigations

TRMB

0

Varco International

VRC

0

43


Table 1 continued

44

Weis Markets

WMK

0

Anchor Bankcorp

ABCW

0

Aptar Group Inc

ATR

0

Brunswick

BC

0

BJ Services

BJS

0

BMC Industries

BMC

0

Bristol Myers Squibb

BMY

0

Circuit City

CC

0

Cerner Corp

CERN

0

Chiron Corp

CHIR

0

Dun & Bradstreet

DNB

0

Downey Financial

DSL

0

Gannett Inc

GCI

0

General Mills

GIS

0

W.R. Grace

GRA

0

HCA Inc.

HCA

0

Hertz

HRZ

0

Kimball International

KBALB

0

Kimberly Clark

KMB

0

Lamson & Sessions

LMS

0

Magnetek Inc.

MAG

0

May Department Stores

MAY

0

Martin Marietta Materials

MLM

0

Office Depot

ODP

0

Pepsi

PEP

0

Pentair, Inc.

PNR

0

Qwest Communication

Q

0

Robert Half Intl

RHI

0

Scientific - Atlanta Inc.

SFA

0

Smurfit Stone Container

SSCC

0

St. Jude Medical

STJ

0

Terex Corp

TEX

0

Temple - Inland

TIN

0

Tootsie Roll Ind

TR

0

Varian Medical Systems

VAR

0

WHX Corp

WHX

0

York International

YRK

0


Table 2. Definitions of Variables Definition of variables used in the study. Returns data are from CRSP and accounting data are form Compustat. The GIndex is from Gompers, Ishiii and Metrick (2003). Variable Definition EVA® Companies Dummy variable which represents whether a firm uses EVA® or not Stern Companies Dummy variable which represents whether a firm is advised by Stern Stewart & Co. or not GIndex Represents the GIndex number for the respective firm NBOD Represents the number of members on the board of directors for the respective firm % Outside Represents the % of outside directors on the board % Inside Represents the % of inside directors on the board % Gray Represents the % of gray directors on the board CEO=CBOD Dummy variable which represents whether the CEO is also the chairman of the board of directors or not Excess (1YR) One year return of the company (starting on the date of the proxy release which first used EVA® as a performance measure) minus the one year return of the CRSP value weighted index over the same period Excess (3YR) Three year return for the company (starting on the date of the proxy release which first used EVA® as a performance measure) minus the three year return of the CRSP value weighted index over the same period Excess (-3YR) Three year return for the company (starting three years before the date of the proxy release which first used EVA® as a performance measure) minus the three year return of the CRSP value weighted index over the same period Total Assets Dollar value of the total assets within a firm (in millions) LTD / Capital Long term debt of the company divided by the capital of the company

45


Table 3. Pearson Correlation Statistics Pearson correlation coefficients for the independent variables used in this study. The sample size is 52 EVA® firms and a matched sample of 52 non-EVA® firms. Stock return data are from CRSP and financial accounting data are from Compustat. GIndex is from Gompers, Ishii and Metrick (2003). Board size (NBOD), % insiders, % outsiders and % gray directors are from proxy statements using standard classification. GINDEX GINDEX

Pearson p - value NBOD Pearson p - value % Outside Pearson p - value % Inside Pearson p - value % Grey Pearson p - value CEO=CBOD Pearson p - value Excess (1YR) Pearson p - value Excess (3YR) Pearson p - value LTD / Capital Pearson p - value Total Assets Pearson p- value

NBOD % Outside % Inside % Grey CEO=CBOD Excess (1YR) Excess (3YR) LTD / Capital Total Assets 0.268*** 0.385 -0.168 -0.288*** .193** -0.137 -.195** -0.062 -0.088 0.006 0.156 0.108 0.003 0.049 0.164 0.047 0.531 0.374 0.32*** -0.168* -0.281*** 0.015 -.213** -0.171* .181* .445*** 0.001 0.088 0.004 0.879 0.03 0.082 0.066 0.001 -.691*** -.702*** 0.098 -0.105 -.189* 0.131 0.04 0.001 0.001 0.323 0.29 0.055 0.185 0.69 -0.022 -.259*** 0.107 .177* -0.073 -0.104 0.826 0.008 0.279 0.073 0.46 0.292 0.137 0.047 0.09 -0.111 0.049 0.165 0.637 0.362 0.262 0.62 -0.084 -0.081 -0.011 0.028 0.396 0.413 0.912 0.781 .556*** -0.133 -0.044 0.001 0.177 0.659 -0.076 -0.04 0.441 0.687 .168* 0.088

* Denotes significance at the .10 level ** Denotes significance at the .05 level *** Denotes significance at the .01 level Table 4. Differences between EVA® and non-EVA® sample The sample size is 52 EVA® firms and a matched sample of 52 non-EVA® firms. Stock return data are from CRSP and financial accounting data are from Compustat. GIndex is from Gompers, Ishii and Metrick (2003). Board size (NBOD), % insiders, % outsiders and % gray directors are from proxy statements using standard classification.

GINDEX NBOD % Outside % Inside % Grey CEO=CBOD Excess (1YR) Excess (3YR) LTD / Capital Total Assets

N EVA Mean Non EVA Mean Mean Diff. t-stat p-value 52 10.65 9.05 1.60 3.383 0.001*** 52 10.15 9.75 0.40 0.986 0.329 52 52.55% 49.38% 3.16 0.972 0.336 52 25.34% 29.29% -3.95 -1.641 0.107 52 22.32% 21.11% 1.21 0.581 0.564 52 0.846 0.673 0.17 1.925 0.06* 52 8.38% 7.08% 1.29 0.131 0.897 52 -2.97% 5.85% -8.81 -0.472 0.639 52 37.39% 38.53% -1.15 -0.219 0.827 52 3,518 3,915 -397.06 -0.787 0.435

* Denotes significance at the .10 level ** Denotes significance at the .05 level *** Denotes significance at the .01 level

46


Table 5. Binary Logistic Regression (N=104) The sample size is 52 EVA® firms and a matched sample of 52 non-EVA® firms. Stock return data are from CRSP and financial accounting data are from Compustat. GIndex is from Gompers, Ishii and Metrick (2003). Board size (NBOD), % insiders, % outsiders and % gray directors are from proxy statements using standard classification.

B GINDEX NBOD % Outside % Inside CEO=CBOD Excess (1YR) Excess (3YR) LTD / Capital Total Assets Constant

0.211 0.051 -0.020 -0.031 0.681 0.354 -0.026 0.001 0.001 -1.217

p-value Exp(B) 0.021** 1.235 0.607 1.052 0.291 0.980 0.219 0.969 0.199 1.976 0.479 1.425 0.910 0.975 0.889 1.001 0.603 1.001 0.479 0.296


Table 6. Binary Logistic Regression (N=94) The sample size is 47 EVA® firms and a matched sample of 47 non-EVA® firms are data screens. Stock return data are from CRSP and financial accounting data are from Compustat. GIndex is from Gompers, Ishii and Metrick (2003). Board size (NBOD), % insiders, % outsiders and % gray directors are from proxy statements using standard classification.

B GINDEX NBOD % Outside % Inside CEO=CBOD Excess (-3YR) Excess (1YR) Excess (3YR) LTD / Capital Total Assets Constant

0.204 0.072 -0.024 -0.040 0.239 0.002 0.605 -0.066 0.003 0.001 -0.569

p-value Exp(B) 0.036** 1.226 0.487 1.075 0.221 0.976 0.133 0.960 0.676 1.270 0.291 1.002 0.254 1.832 0.779 0.936 0.720 1.003 0.575 1.001 0.751 0.566


47


Table 7. Binary Logistic Regression – Stern vs. Non Stern (N=52) The sample size is 52 EVA® firms denoted 36 that use Stern Stewart and 16 that did not. Stock return data are from CRSP and financial accounting data are from Compustat. GIndex is from Gompers, Ishii and Metrick (2003). Board size (NBOD), % insiders, % outsiders and % gray directors are from proxy statements using standard classification. B GINDEX NBOD % Outside % Inside CEO=CBOD Excess (1YR) Excess (3YR) LTD / Capital Total Assets Constant

-0.155 0.706 -0.006 0.035 -0.084 0.172 0.230 -0.013 0.001 -4.099

p-value Exp(B) 0.405 0.856 0.011** 2.026 0.873 0.994 0.480 1.036 0.946 0.919 0.811 1.187 0.690 1.258 0.413 0.987 0.613 1.001 0.220 0.017


Table 8. Classification Summary The sample size is 52 EVA® firms and a matched sample of 52 non-EVA® firms. Model output and firm observation are used to construct table.

Predicted by the Model Observed EVA® Non EVA® Overall Classification

Percentage Correct

EVA®

Non_EVA®

42 17

10 35

82 68 73

Table 9. Summary Statistics The sample size is 52 EVA® firms and a matched sample of 52 non-EVA® firms. Stock return data are from CRSP and financial accounting data are from Compustat. GIndex is from Gompers, Ishii and Metrick (2003). Board size (NBOD), % insiders, % outsiders and % gray directors are from proxy statements using standard classification. Output is based on logistic regression and Wald statistic. Variables Used Gindex NBOD %Outside %Inside CEO=CBOD Excess -3Yr Excess 1 yr Excess 3 yr LTD/Capital Total Assets Constant

*** Significant at 1%

48

B .03 .06 -.03 -.05 .25 .003 .58 -.05 .002 .002 -.60

S.E. .025 .14 .16 .29 .15 .8 .73 .42 .32 .40 .65

Wald 16.1 18.2 14.3 2.1 3 13.5 17.1 18.3 17.2 12.4 11.5

Sig .006 *** .092 .121 .152 .183 .232 .182 .132 .121 .139 .143

Ranking 1 N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A


THE VOLUNTARY CSR DISCLOSURE IN CORPORATE ANNUAL REPORTS: EVIDENCE FROM AUSTRALIA Nicholas Andrew*, Mark Wickham** Abstract The relationship between credible Corporate Social Responsibility (CSR) performance and desirable firm outcomes is well established in corporate governance literature. Over the past two decades in particular, there has been an increased recognition of this relationship in the business community and a concomitant increase in the quantity and detail of CSR activities being voluntarily reported by corporations has been observed. The rationale for the increasing levels of voluntary CSR reporting has been attributed to two main corporate strategies: to conform to the expectations of the society and to socially legitimise their operations to their salient stakeholder groups. Whilst there has been extensive academic interest in the concept of CSR, it has focused almost exclusively on normative definitions of the concept, and/or the presentation of empirical evidence that details „why corporations should report their CSR activities‟ and „what CSR activities they should report‟. What is lacking the literature, however, is a focus on the question as to „how do corporations strategically report their CSR activities?‟ We find that there is evidence to support a „Core/Periphery Model‟ of strategic CSR disclosure, which we feel provides a framework for predicting how corporations will voluntarily disclose their CSR performance given the issues, events and/or crises that affect their industry environments. Keywords: Corporate Social Responsibility, Annual Report, Corporate Communication *School of Management University of Tasmania, Hobart, Australia **Dr., School of Management University of Tasmania, Hobart, Australia

1. Introduction The relationship between credible Corporate Social Responsibility (CSR) performance and desirable firm outcomes (such as improved reputation, customer loyalty and long-term profitability) is well established in the corporate governance literature (D‟Orio and Lombardo, 2007; Robins, 2008; Stratling, 2007). Over the past two decades in particular, there has been an increased recognition of this relationship in the business community and a concomitant increase in the quantity and detail of „voluntary CSR disclosure‟ in corporate annual reports has been observed (Boasson, 2009; Matten and Moon 2008). The rationale for the increasing levels of voluntary CSR reporting in annual report documents has been attributed to two main corporate governance strategies: to conform to the expectations of the society within which the corporation operates and to socially legitimise the corporation‟s activities to their salient stakeholder groups (Kurihama, 2007; Samy, Odemilin and Bampton, 2010; Shahin and Zairi, 2007; Thomson and Jain, 2010). Whilst there have been similar levels of academic interest in the concept of CSR over the same period, published research has tended on normative definitions of the concept, and the presentation of empirical evidence that details „why corporations should disclose their CSR activities‟ and

„what CSR activities they should disclose‟ (Garriga and Mele´, 2004; Nelling and Webb, 2009; Schwartz and Carroll, 2008; Syriopoulos, Merikas and Vandzikis, 2007). According to Castello and Lozano (2009), however, there is a real lack of theoretical knowledge about the relationship between „what CSR activities are being voluntarily disclosed‟ and the „how CSR activities are being voluntarily disclosed. This paper, therefore, seeks to contribute to the CSR literature by going beyond the „why‟ and „what‟ questions of voluntary CSR disclosure, to explore the question of „how‟ corporations go about strategically disclosing their CSR performance in their annual report documentation. 2. Literature Review It is now well accepted that a corporation‟s long-term viability depends largely on how it is perceived by its key stakeholders and members of the community in which it resides (Cornelissen, 2004; Oeterli, 2008). In order to link the benefits of CSR performance to the financial bottom line, academic research has undertaken an extensive examination of the strength and causality of the relationship to determine whether „doing good‟ leads to „doing well financially‟ (Dentchev, 2005; Orlitzky, Schmidt and Rynes, 2003). The results of empirical studies of the direct relationship between CSR performance and 49


profitability have been inconclusive, reporting positive, negative, and neutral results (McWilliams and Siegel, 2000). Other research, however, suggest that the benefits of acceptable CSR performance are rather more indirect, and better conceptualised as creating a „virtuous circle‟ for the corporation that creates positive stakeholder relationships that reduce the likelihood of difficulty when dealing with salient stakeholder groups (Castello´ and Lozano, 2009; Waddock and Graves, 1997). Supporting this concept is research by Cheng, Collins and Huang (2006) found that corporations with explicit shareholder rights policies tended to enjoy lower „cost of equity capital‟ than competing firms that did not. Similarly, Ferreira, Sinha and Varble (2008) found that corporations benefit in the form of positive long-run stock performance following certification of quality management. In the case of large companies, therefore, effective CSR performance does tend improve the bottom line in the medium to long term. Given the evidence supporting the „virtuous circle‟, corporations have a vested interest in building and maintaining functional relationships with its key stakeholder groups, and finding the most effective way in which to communicate this effort to salient stakeholder groups. Deegan (2002: 292) states that effective CSR disclosure plays an important part of this process as it provides “information designed to discharge social accountability…corporate disclosure is seen as a method by which management can interact with broader society to influence external perceptions about their organisation”. Some corporate CSR disclosure is of course mandated by government legislation and/or the specific listing rules associated with stock exchanges around the globe (Anderson,

1998; Kercher, 2001; Olgiati, 2003), and requires listed corporations to include information such as „notes to explain the financial reports‟, „explanation of adopted audit processes‟ and „levels of executive remuneration‟ in their annual reports. Evidence abounds, however, of corporations using their annual report document as a marketing communication tool for the voluntary disclosure of CSR performance that is over and above that required by legislation and listing rules (Clarke, 1997; Stanton and Stanton 2002; Waller and Lanis, 2009). Used as a „CSR communications tool‟ in this way, annual reports now include information that helps the corporation market it operations as sustainable and respectful of the needs of various stakeholder groups and the natural environment. Corporations have adopted various methods for communicating their CSR performance, including „triple-bottom-line‟ reporting techniques and the use of dedicated „sustainability reports‟. One of the key benefits of using the annual report in this way is the ability of the corporation to fully control the content and framing of their CSR performance to the needs of their salient stakeholder audiences. Academic research into the voluntary reporting of CSR performance has identified the benefits of such disclosure (the „why?‟ question) (see D‟Orio and Lombardo, 2007; Robins, 2008; Stratling, 2007) as well as defined the broad CSR categories represented observed in disclosure efforts to date (the „what?‟ question). Table 1 provides a summary of the broad CSR categories most often voluntarily disclosed issues by corporations in their annual report documents (Global Reporting Initiative, 2002).

Table 1. Common CSR Issues Disclosed in Annual Report Documentation Corporate Governance

Environmental Protection

Employee Development

Ethics/Ethical Conduct

Emissions (Pollutants)

Diversity

Code of Conduct

Greenhouse Gas Emissions

Training

Whistle blowing

Energy Efficiency

Development

Human Rights

Recycling

Equal Employment Opportunity

Community Support

Health & Safety

Community Initiatives

Health

Fundraising

Safety

Partnerships

Injury

Sponsorships

Employee (health)

Donations/Contributions

Community (health)

Source: The Global Reporting Initiative (2002)

50


3. Research Opportunity Despite the wealth of academic publications concerning the „why?‟ and „what?‟ questions associated with voluntary CSR reporting, there have been calls in the literature for a deeper analysis of the strategy of CSR-related reporting activities that aim to strengthen corporate reputation and bottom-line performance (Halabi, Kazi, Dang and Samy, 2006; Tengblad and Ohlsson, 2010). The study of the methods driving voluntary CSR disclosure will help academics and practitioners alike to reflect on the increasing strategic importance of effective CSR reporting practices. In the current climate of heightened scrutiny of corporate behaviour (see Basu and Palazzo, 2008; Waddock, 2000) and increasing demand for CSR programs by stakeholder groups, there is need for conceptual robustness in order to move CSR research beyond the purely normative perspective towards a more strategic understanding of social and environmental issue management (Castello and Lozano, 2009). As such, the specific research question to be addressed in this paper is: How do Australia‟s largest corporations systematically disclose their voluntary CSR performance in their annual report documentation? The rationale for this centres on a growing agreement that sustainable business success and shareholder value creation cannot be achieved exclusively through maximising short-term profits alone, but rather through the effective communication of market-oriented and socially responsible behaviour (Kotler and Lee, 2005; Samy, Odemilin and Bampton, 2010). We feel that the answer to this question offers an important advancement in the CSR literature, as it will help develop a predictive framework for the how corporations will likely disclose their CSR performance given the issues, events and/or crises that that arise in the future. 4. Method In order to explore this research question, this study undertook a content analysis of the annual reports of the three largest companies (by market capitalisation) in the three largest Australian industries for the years 2005/6 to 2009/10. This timeframe is considered important as it encompasses the period immediately before the Global Financial Crisis as well as its aftermath. The selection of companies and time period offers two other important research opportunities: firstly it provided a longitudinal account of the voluntary CSR activities to be reported by the leading Australian companies across the fiveyear timeframe; and secondly, it provided an opportunity to study how a common corporate governance crisis impacted the voluntary communication of CSR activities across companies in the three largest industries in Australia. In total, 45 annual reports were collected for scrutiny (NB: please

see the results section for the details of the corporation‟s identities). Each of the 45 annual reports downloaded from the respective corporation‟s official websites were subject to a rigorous content analysis process that followed the five-stage protocol identified by Finn, White and Walton (2000), Hodson (1999) and Neumann (2003). In the first stage, the aims and objectives of the research were identified, and the first round coding rules were developed. Coding refers to the process of converting information into contextual values for the purposes of data storage, management and analysis allowing theme identification (Ticehurst & Veal, 2000). Using the literature review as a guide, we decided to initially organise the data by the variables listed in The Global Reporting Initiative (2002) (see Table 1 above). In the second stage of the content analysis, all of the data in the official annual reports were converted into MS World® format, and entered into the codified database. At regular intervals, inter-coder reliability checks were taken to ensure that the data were coded consistently with the rules set in Stage One. In the third stage of the content analysis, the coded data were further interrogated to detect any significant themes that emerged in the voluntary reporting of CSR activities over time. The trends and emergent themes detected in the analysis formed the basis for establishing the second round of data categories. As was the case in Stage One, the second round of coding rules were developed prior to the coding of the data itself (to maintain a consistent approach between researchers), and to provide a protocol for others to follow should they wish to replicate the analysis. In the fourth stage of the content analysis, the second round coding categories were populated with data according to the new coding rules. The interpretation of the data during the second round of coding, and the verification of the conclusions, was facilitated by the use of the NVIVO software package. In the method literature, it has been emphasised that computer software programs such as NVIVO, are of significant value in qualitative analysis and any subsequent theory building (Kelle, 1995; Richards & Richards, 1995; Weitzman & Miles, 1995). Where it was appropriate, data were allocated to more than one node for analysis. Again using the NVIVO software, the contents of each of the initial index nodes were then reviewed to identify common themes that arose in the data. In the final stage of the content analysis, the results of the second round coding were refined and the research findings finalised. In order to facilitate the theory building process, memos were maintained about the data, their categories, and the relationships between them as they emerged. NVIVO has a facility for the creation and retention of such memos for later consideration and analysis. Utilising the memo capability within the NVIVO package, memo reports were generated by the software after „Stage Two‟ 51


coding. From these reports, the trends and emergent themes became clearer. The themes emanating from the „second round‟ of coding form the basis of the discussion section that follows. 5. Discussion

As noted, this study assessed the annual reports of the three largest corporations in Australia‟s three largest industries (according to the Australian Stock Exchange) for the period 2004/5 to 2008/9. Table 2 below presents a summary of the industries and corporations represented in this research.

Table 2. Industry Sectors and Corporations Represented in this Research Mining Industry

Banking Industry

Retail Industry

BHP Billiton

Commonwealth Bank

News Corporation

New Crest Mining

National Australia Bank

Wesfarmers

Rio Tinto Westpac Banking Corporation Source: Australian Stock Exchange (ASX) 2010

5.1 ‘What’ CSR information was voluntarily reported in the annual reports? The content analysis of the annual report documents in this study revealed findings that were consistent with previous studies conducted by Nielson and Thomsen (2007) and Mirfazli (2008) in two main ways. Firstly, that the individual corporations did not voluntarily disclose information regarding all of the possible CSR issues identified in the literature, and secondly that membership of an industry correlated with the array of CSR issues addressed by the corporation. Despite the theoretical equality amongst the array of CSR issues, each of the corporations was found to focus on a small number of CSR issues that were strategically pertinent to their specific stakeholder groups. For example, whilst all three of the banking corporations voluntarily reported their CSR performance in terms of „Corporate Governance‟, the context varied according to their different reputations amongst current and potential clientele. The Commonwealth Bank consistently reported their Corporate Governance CSR performance in terms of „maintaining an effective Code of Conduct‟, where as the National Australia Bank dealt with the issues of „Whistle Blowing policy‟ and „Ethical Conduct‟ in their voluntary CSR disclosure. The difference between these voluntary disclosures may be attributed to the recent history of the two corporations: the Commonwealth Bank has managed to operate free of any major financial fraud issues in its recent history, whereas the National Australia Bank was forced to contend with three major crises of consumer confidence in the first decade of the 21st century- a $360m loss from its foreign exchange operations (which involved rouge trading, documented coverups), $80m dollars in overcharging of customers, and a $50m dollar loss from embezzlement in South Korea. In 2007/9 and 2008/9, all of the banking corporations provided voluntary disclosures regarding the impact (and their exposure to) the Global 52

Woolworth‟s

Financial Crisis. This event was treated in a similar fashion by all three banks, who increased their „Corporate Governance‟ disclosure to include coverage of their „Ethical Codes of Conduct‟ and „Due Diligence‟ when securing loans to clientele. Similar evidence was found in the corporate reports for the three mining corporations, with Rio Tinto emphasising their „Corporate Governance‟ reporting on „Ethical Conduct‟ and the establishment of „Codes of Conduct‟ (arguably in response to the Chinese government‟s prosecution of their executives for corporate espionage). BHP Billiton and New Crest Mining, on the other hand, dedicated their voluntary CSR disclosure on „Environmental Protection‟ – New Crest Mining emphasising their „Energy Efficiency‟ and „Climate Change Initiatives, whereas BHP Billiton emphasised „Ethical Conduct‟ (in light of their ongoing issues with the Ok Tedi environmental disaster). The retailing corporations also selected a specific array of CSR issues to voluntarily disclose; Wesfarmers highlighted their „Corporate Governance‟ performance (specifically in terms of their acquisition strategies), Woolworth‟s emphasised their „Community Support‟ and „Environmental Protection‟ regimes, and News Corporation disclosed their CSR performance on each of the categories listed in the Global Reporting Initiative (2002) categories, although sporadically over the five year sample period. The impact of the Global Financial Crisis was reflected in an increase in the voluntary CSR reportage of „Corporate Governance‟ of all corporations. In the sample cases selected here, there is evidence that Australian listed companies select the bases for the voluntary CSR disclosure dependent on the reputational concerns held primarily by financial stakeholder groups – be they existing and potential investors, and/or existing and potential customer segments. This is perhaps not a revelation, given that the literature recognises the „marketing communications‟ function that the annual report has developed to serve, however, provide a basis for predicting „what‟ voluntary CSR disclosures a


corporation is likely present in their annual will report. We suggest that by defining the reputation of a given corporation (in terms of what their salient financial stakeholder groups perceive to be the areas of ethical concern), a researcher can predict the CSR issues that will be included in that corporation‟s annual report. The voluntary disclosure within the CSR category will likely be concerned with the recent past of the corporation, especially if it has had to deal with a perceived ethical breach of its own making. We also find evidence that corporations will include coverage of a specific CSR issue even if that issue is not specifically applicable to their own organisation or industry. The Global Financial Crisis, for instance, impacted the Mining and Finance industries most prominently; however, the Retail corporations (whose sales and profitability in Australia were largely unaffected) also were at pains to allay investor concerns about the impact of the crisis on their ability to increase shareholder wealth. 5.2 ‘How’ was CSR information voluntary reported? Each of the sample corporations‟ annual report data were arranged in chronological order so that the CSR issues voluntarily disclosed over the five year period (and the context within which they were reported) could be compared. This quasi-longitudinal analysis of the CSR data disclosed by each corporation over time detected what this study will call a „Core/Periphery‟ Model of voluntary CSR disclosure. The quasi-longitudinal analysis indicated that certain voluntary CSR disclosure remained consistent across each of the years, with the only variance detected relating to word-count figures. These „core CSR disclosures‟ remained contextually stable over the entire five year sample period, and we can hypothesise that the framing of the corporation‟s CSR disclosure on these core issues reflected management‟s understanding of their key stakeholders‟ expectations. For example, in the mining sector, the CSR issues surrounding „Health & Safety‟ consistently highlighted the corporations‟ concern for employee and contractor well-being, and the safety investments made over the previous reporting period. In the finance sector, all three corporations consistently reported their commitment to „Codes of Conduct‟ and the manner in which safeguards had been established and enforced to protect shareholder wealth. It is expected that under stable industry and corporate circumstances, the reporting of these „core CSR issues‟ will remain consistent over time, and that using annual reports of the recent past will predict the framing of these core CSR issues in the forthcoming annual report document. The quasi-longitudinal analysis indicated that certain voluntary CSR disclosure was relatively sporadic in nature (i.e. voluntarily reported three

times or less over the five year period). These „peripheral CSR disclosures‟ detailed the corporations‟ response to ad hoc ethical issues or crises that occurred within the previous reporting period. For example, while all of the retail corporations voluntarily report on the „company donations and contributions‟ over the five year period, the individual corporations do not report on it every year (i.e. Woolworths only reported against this topic between 2006/7 to 2008/9, News Corp meanwhile reported on the topic just once in 2005). There are two themes common to the voluntary reporting of the „peripheral CSR issues‟ – firstly, that the CSR reporting relates to a positive outcome from the corporation‟s activity in the area; or secondly, that the corporation has had to deal with a relatively minor ethical indiscretion on the part of itself, or another member of their industry group. For example, in 2005/6, Westpac and the Commonwealth Bank corporations increased their voluntary disclosure of their internal corporate governance controls in the wake of National Australia Bank‟s foreign exchange and overcharging scandals of 2003 and 2004. Whilst the National Australia Bank‟s annual reports now includes such voluntary reportage as „core‟ to their annual reports, the two other banks have refocused their „corporate governance‟ reporting to that required by legislation. Interestingly, the significant effects of the Global Financial Crisis (and the related voluntary CSR disclosures associated with it) have only emerged in the annual reports for the past two years, and can only be viewed as „peripheral‟ according to the definition set forth in this paper. We expect, of course, that the effects of the Global Financial Crisis will endure past 2009/10, and this poses an interesting situation whereby a „peripheral CSR issue‟ may indeed become a „core CSR issue‟ over the medium to long-term. Whilst the data collected in this research does not allow insights into this possible phenomenon, it does pose an interesting relationship between how persistent „peripheral CSR issues‟ are reported given the emergence of a nation-wide (and indeed international) event. The question as to „how long do core CSR issues endure?‟ is also unknown at this point in time. 6. Conclusion and Future Research This research explored the question as to „how‟ Australia‟s largest corporations systematically disclose their voluntary CSR performance in their annual report documentation. It found that the sample of Australian corporations voluntarily disclosed those CSR activities that were directly related to protecting or enhancing their reputation amongst key stakeholder groups, and to this end, with an emphasis on financial stakeholders. It also found that the voluntary CSR disclosure over time conformed to a „Core/Periphery‟ model that could be useful in predicting how

53


corporations will voluntarily disclose their CSR performance given the issues, events and/or crises that affect their industry environments. As this study is preliminary in nature, we recommend the following future research be undertaken to solidify the tenets of the Core/Periphery model introduced here. Firstly, research should explore the veracity of Core/Periphery model concept using a larger sample of corporations and capture data over a longer period of time. We feel that whilst there appears to be evidence of a „core‟ and „periphery‟ in how corporations voluntary report their CSR information, the rules associated with how „core‟ and „peripheral‟ CSR issues are dealt with over time requires attention. In addition, there is an opportunity for researchers to explore whether corporations listed on other international stock exchanges conform to the same system of voluntary CSR disclosure, and the extent to which being listed on multiple exchanges impacts the „what‟ and the „how‟ questions addressed here. Secondly, there is an opportunity to correlate corporate crises with the framing of voluntarily disclosed CSR information over the ensuring period to gauge the effect of these crises as time progresses. Lastly, there is an opportunity for researchers to extent the Core/Periphery Model to encompass a predictive model of voluntary CSR disclosure given the issues, events and/or crises that affect industry environments. Such a predictive model will serve to improve our understanding of the role that voluntary CSR disclosure in the annual report has as a corporate governance mechanism. References 1. 2. 3.

4.

5.

6. 7. 8.

54

Anderson, R.H. 1998. Regulating corporate annual reports in Australia. Business and Economic History. 27(2): 522-534. Basu, K., and Palazzo, G. 2008. Corporate social responsibility: A process model of sensemaking. Academy of Management Review. 33: 122-136. Boasson, E.L. 2009. On the management success of regulative failure: standardised CSR instruments and the oil industry‟s climate performance. Corporate Governance. 9(3): pp. 313-325. Castello, I. and Lozano, J. 2009. From risk management to citizenship corporate social responsibility: analysis of strategic drivers of change. Corporate Governance. 9(4): 373-385. Cheng, A., Collins, D. and Huang, H. 2006. Shareholder rights, financial disclosure and the cost of equity capital. Review of Quantitative Finance Accounting. 27:175–204. Clarke, G. 1997. Messages from CEOs: A content analysis approach. Corporate Communications. 2(1): 3139. Cornelissen, J. 2004. Corporate communications: Theory and practice. London: Sage Publications. Deegan, C. 2002. The legitimising effect of social and environmental disclosures: A theoretical foundation.

9. 10.

11.

12.

13. 14. 15. 16. 17. 18. 19. 20. 21.

22. 23.

24.

25.

Accounting, Auditing and Accountability Journal, 15 (3): 282-312. Dentchev, N. 2005. Corporate social performance: as business strategy. Journal of Business Ethics. 55(4): 395410. D‟Orio, G. and Lombardo, R. 2007. Corporate governance and corporate social responsibility in Italy: Advantages and disadvantages of a non-explicitly existent system. Corporate Ownership and Control. 4(4): 36-50. Ferreira, E., Sinha, A. and Varble, D. 2008. Long-run performance following quality management certification. Review of Quantitative Finance Accounting. 30: 93-109. Finn, M., White, E.M. & Walton, M. 2000. The analysis of qualitative data: Content analysis and semiological analysis. In Toiurism and Leisure Research Methods: Data Collection, Analysis and Interpretation. Essex: Pearson Education Ltd. Garriga, E. and Mele´, D. 2004. Corporate social responsibility theories: mapping the territory. Journal of Business Ethics. 53(1): 51-71. Global Reporting Initiative. 2010. What is GRI? http://www.globalreporting.org/AboutGRI/WhatIs GRI/ Accessed June 20, 2010. Halabi, A., Kazi, A., Dang, V. and Samy, M. 2006. Corporate social responsibility. Monash Business Review. 12(3): 22-25. Hodson, R. 1999. Analyzing Documentary Accounts: Quantitative Applications in the Social Sciences. Thousand Oaks, CA: Sage. Kelle, U. 1995. Computer-aided Qualitative Data Analysis: Theory, Methods, and Practice. London: Sage Publications. Kercher, B. 2001. The Corporate Image – the Regulation of Annual Reports in Australia. Macquarie Law Journal. 93(4): 93-128 Kotler, P. and Lee, N. 2005. Corporate Social Responsibility: Doing the Most Good for Your Company and Your Cause. Hoboken, NJ: Wiley & Sons. Kurihama, R. 2007. Role for auditing in CSR and corporate governance. Corporate Ownership and Control. 5(1): 109-119. Matten, D. and Moon, J. 2008. „Implicit‟ and „Explicit‟ CSR: A conceptual framework for a comparative understanding of CSR. Academy of Management Review. 33(2): 404–424. McWilliams, A. and Siegel, D. 2001. Corporate social responsibility: A theory of the firm Perspective. Academy of Management Review. 26(1): 117-27. Mirfazli, E. 2008. Corporate social responsibility information disclosure by annual reports of public companies listed at Indonesia Stock Exchange (IDX). International Journal of Islamic and Middle Eastern Finance and Management. 1(4): 275-284. Nelling, E. and Webb, E. 2009. Corporate social responsibility and financial performance: the „virtuous circle revisited. Review of Quantitative Finance and Accounting. 32(2): 197-209. Neuman, W.L. 2003. Social Research Methods (5th ed.). Upper Saddle River: Prentice Hall.


26. Nielsen, A.E. and Christa Thomsen. 2007. Reporting CSR – what and how to say it? Corporate Communications: An International Journal. 12(1): 25-40. 27. Oetterli, G. 2008. CSR Reporting. Keeping Good Companies. ICSA International. 60 (1): 26-31 28. Olgiati, L. 2002. Corporate governance in Switzerland. International Financial Law Review. 29. Orlitzky, M., Schmidt, F.L. and Rynes, S.L. 2003. Corporate social and financial performance: A meta analysis. Organization Studies. 24(3): 403-41. 30. Richards, T. and Richards, L. 1995. Using computers in qualitative research. In Denzin, N.K. and Lincoln, Y.S. (editors) Handbook of Qualitative Research. California: Sage Publishers. 31. Robins, F. 2008. Why corporate social responsibility should be popularised but not imposed. Corporate Governance. 8(3): 330-341. 32. Samy, M., Odemilin, G. and Bampton, R. Corporate social responsibility: A strategy for sustainable business success. An analysis of 20 selected British companies. Corporate Governance. 10(2):203-217. 33. Schwartz, M. and Carroll, A. 2008. Integrating and unifying competing and complementary frameworks. The search for a common core in the business and society field. Business and Society. 47(2): 148-85. 34. Shahin, A. and Zairi, M. 2007. Corporate governance as a critical element for driving excellence in corporate social responsibility. International Journal of Quality & Reliability Management. 24(7): 753-770. 35. Stanton, P. and Stanton, J. 2002.Corporate annual reports: Research perspectives used. Accounting, Auditing & Accountability Journal. 15(4): 478-500.

36. Stratling, R. 2007. The Legitimacy of Corporate Social Responsibility. Corporate Ownership and Control. 4(4): 65-74. 37. Syriopoulos, T., Merikas, A. and George S. Vozikis, G.S. 2007. Corporate social responsibility and shareholder value implications. Corporate Ownership and Control. 5(1): 96-108. 38. Tengblad, S. Ohlsson, C. 2010. The framing of CSR and the globalization of national business systems: A longitudinal case study. Journal of Business Ethics. 93: 653–669. 39. Thomson, D. and Jain, A. 2010. Corporate social responsibility reporting: a business strategy by Australian banks? Corporate Ownership and Control. 7(4): 213-226. 40. Ticehurst, G.W. and Veal, A.J. 2000. Business research methods: A managerial approach. Longman: Pearson Education Pty Limited. 41. Waddock, S. 2000. The multiple bottom lines of corporate citizenship: Social investing, reputation, and responsibility audits. Business and Society Review. 105: 323–345. 42. Waddock, S.A. and Graves, S.B. 1997. The corporate social performance-financial performance link. Strategic Management Journal. 18(4): 303-19. 43. Waller, D.S. and Lanis, R. Corporate social responsibility (CSR) disclosure of advertising agencies. Journal of Advertising. 3891): 109-121. 44. Weitzman E. & Miles, M. 1995. Computer Programs for Qualitative Data Analysis. California: Sage.

55


ENTERPRISE VALUE AND DISCLOSURE LEVEL: EVIDENCES IN THE BRAZILIAN MARKET Fabio Gallo Garcia*, Elmo Tambosi Filho**, Luiz Maurício Franco Moreira*** Abstract There is a strong tendency in global markets towards an enhanced level of corporate transparency regarding the activities of companies and, as a result, information on their performance. The Purpose of this study is to analyze the relationship between greater disclosure levels and shareholder value creation. Increasing levels of disclosure are required from companies‟ management before shareholders and the society in general. Obscure practices that fail to take into consideration the best interests of shareholders increase risks and cause shares to lose liquidity. The São Paulo Stock Exchange‟s “Novo Mercado” (“New Market”) emerged from the intent to improve the Brazilian stock market by adopting best practices in corporate governance, adding transparency to disclosed information, and heightening the respect for the interests of shareholders, whether they may be minority or not. The “Novo Mercado” intends to foster a differentiated environment in which companies committed to corporate governance are recognized and can benefit from better stock prices, resulting in lower placement costs and increased liquidity. Our research will assume that companies with American Depositary Receipts ADRs are committed to a higher level of disclosure as a result of the requirements of the Security Exchange Commission – SEC, and the Financial Accounting Standards Board - FASB; an empiric study about these firms will be performed. We will determine, through a Study Event concerned with cases where ADR have been issued, which consequences of the commitment to higher levels of disclosure as regards shareholder are responsible for value creation, and what are the reflections on the stock price quoted in the Brazilian market. Keywords: American Depositary Receipt (ADR); Differentiated Levels of Corporate Governance; Disclosure; Financial Accounting Standards Boar (FASB); Internacional Accounting Standards Board (IASB); New Market; Security Exchange Commission (SEC) *Dr., Professor of Graduate Management Programme , School of Management Studies Getulio Vargas Foundation E-mail: [email protected] **Dr., Professor of Graduate Management Programme, Methodist University of São Paulo E-mail: [email protected] ***Dr. , Professor of Graduate Management Programme ,School of Management Studies Pontíficia Católica University E-mail: [email protected]

I.

Introduction

must be included with financial statements. The Comissão de Valores Mobiliários (CVM - the Informational asymmetry in the stock market has a Braziliian equivalent of the SEC), in turn, makes marked impact on enterprise valuations and investor recommendations concerning the disclosure of matters interest. Investors wish to reduce the insecurity and it considers relevant for the purpose of better uncertainty that affect their decisions by being aware of understanding the financial statements. risks and mitigating or eliminating them where possible. In Brazil, there is a sense that the information It is therefore very important for companies to increase provided to the market are insufficient or low-quality, the transparency of their financial statements. as compared to other countries such as the United One of the great challenges in the field of States. We admit that our regulation is inadequate. Accounting is establishing the quality and quantity of Which is precisely why the São Paulo Stock Exchange information that meets the needs of the users of is attempting to implement the “Novo Mercado” and the financial statements. Footnotes are information that “Níveis Diferenciados de Governança Corporativa” supplement financial statements and may be descriptive (“Differentiated Corporate Governance Levels”), and or provide analytical charts. why Congress has been trying to establish a new A key issue here is which information companies Corporations Law for several years. should offer to supplement financial statements. In the Brazilian companies seem to resist adopting Brazilian case, the Corporations Law (Law No. higher disclosure levels. Intending to shed light on this 6.404/76) provides general guidance and the notes that discussion, this paper will attempt to investigate value 56


creation at companies that have adopted a more transparent stance. International accounting standards also address the disclosure of accounting and financial information In the United States, the types of information companies must disclose in addition to financial statements are established by the accounting principles set by the Financial Accounting Standards Board (FASB) and the Securities Exchange Commission (SEC). Another positive factor is the voluntary disclosure of additional information, a practice encouraged by the Financial Accounting Standards Board (FASB). This entity is going forward with a broad, important project known as “Business Reporting Research Project”, whose main purpose is to assist companies in improving their reports and financial statement. The basic assumption underlying the project is that increased disclosure makes capital allocation more efficient and reduces the average cost of capital. The reasoning of the FASB is that a company‟s cost of capital is strongly influenced by the transparency level of its accounting and financial information. The cost of capital faced by companies is made up, in part, of a premium to investors that have to do with the potential informational asymmetry between the company and the market. Therefore, adopting the same perspective as LEUZ & VERRECCHIA (2000), we will analyze Brazilian American Depositary Receipts (ADRs) Issuers. This is due to the fact that these companies sport rather high transparency levels as a result of the Securities and Exchange Commission‟s requirements to authorize the issuance of ADR for Brazilian companies in the American market. II. Benefits and Costs of Voluntary Disclosure The main potential benefits relate to the fact that investors benefit from reduced uncertainty, while companies (and their owners) benefit from lower average cost of capital, increased credibility, a better relationship with investors, greater access to liquid markets with stock price fluctuations, and lower risk of lawsuits claiming inadequate disclosure. The benefits for the economy in general are more effective capital allocation, the effects of the lower cost of capital on investments, and more liquid markets. The potential costs of a wider disclosure policy lie in competitive disadvantages for companies, inasmuch as competitors would have more access to information, and disadvantages in negotiations with suppliers, customers and employees. III. Studies on the Benefits of Voluntary Disclosure Earlier studies on this topic generally reinforce the notion that increased transparency in accounting and financial statements and the consequent reduction in

informational asymmetry lead to a lower cost of equity capital by increasing market liquidity and/or reducing uncertainty, thereby enabling companies to create value. Botosan (1997)attempts to relate cost of capital and voluntary disclosure by introducing an index that includes items regarded as helpful to decision-making by investors and analysts. Adopting the index enabled the author to generate a score for each of the 122 sampled industrial companies by assigning marks to each one of the items presented at the footnotes in the 1990 annual report. The results suggest that, for companies to which the market paid less attention, the cost of capital decreased as disclosure levels rose, after adjustment for risk (measured by the market beta) and company size. The results indicate that companies of this description that provided maximum disclosure, equity capital cost dropped by 9% as compared to others that provided little information. On the other hand, for companies that are traditionally in the market‟s focus, increased transparency levels did not bring about lower cost of capital. An article by Sengupta published in 1998 discusses the negative relationship between companies‟ general disclosure policy and the interest rates they must pay when collecting funds from the market. Companies that investors regard as averse to higher levels of disclosure are assigned a greater risk, which ensures higher risk-premium embedded in interest rates. The author measured companies‟ disclosure policy based upon the metrics available at the “Report of the Financial Analysts Federation Corporate Information Committee”, with scores ranging from zero to 100 points. He associated the disclosure measure with two corporate indebtedness metrics: “yield to maturity” in new debt security issues (such as debentures, for example), and the effective interest rate on new debt. Based on a sample of 102 companies, his results show that the two cost metrics maintained a negative relationship with the disclosure measure, after adjustments to interest rate determinants such as company-specific risk, loan features, and market conditions. In another interesting study, Healy, Hutton & Palepu (1999) examined factors relating to improved voluntary disclosure using a 97-company sample in the 1980-1990 period. The results showed that increased disclosure transparency is related to improved stock performance and liquidity, as well as to institutional equity growth. Leuz & Verrecchia (2000) presented a study of the relationship involving informational asymmetry, market liquidity, and cost of capital by means of analyzing the regulations of the German stock market. German accounting standards are characterized by high measurement and low disclosure levels. The authors argue that, as a result, a significant number of German companies traded in global stock markets have adopted strategies under which their accounting statements are prepared based on the International Accounting

57


Standards (IAS) or on the U. S. Generally Accepted Accounting Principles (US GAAP). Therefore, the authors believe that adopting such principles enhances a company‟s commitment to increased disclosure levels. In Brazil, there is a study by Rodrigues (1999) on the effects of the listing of the stock of Brazilian companies in the American market via ADRs, dealing with three hypotheses, as follows:  The visibility, or “investor recognition” hypothesis – prepared by Merton (1987) states that increased visibility reduced total asset risk. The evidence in favor of the hypothesis includes increased market value before the event (listing), with reduced returns, increased stock base, reduced volatility and increased trade volume after the event.  The liquidity hypothesis – dealt with by Amihud & Mendelson(1988) and stipulating that listing in major markets enhances asset liquidity and, among other results, increased disclosure volume; therefore, returns after the event are lower and the pre-event market values is higher.  Segmentation hypothesis – proposed by Alexander, Eun & Janakiranaman (1987) and dealing with asset-pricing models in the context of segmented stock markets, that are subsequently integrated by means of double-listing. It is expected that a doublelisted stock‟s expected return should drop, as long as covariance with the local market portfolio is greater than that with the market where the new listing takes places. Rodrigues concludes that the investor recognition and the return hypotheses are supported by the results collected by the author. ON the other hand, no empirical evidence was found to support the hypothesis of market segmentation/integration. IV. Subject The main purpose of this study is to determine whether companies with greater accounting and financial information disclosure levels perform better than the market. Our study assumes that companies that issue ADRs - as with any other means to raise funds - intend to exploit their investment alternatives with an aim to maximizing shareholder value. However, issuing a security abroad may have an additional meaning as - by meeting American legal requirements and making the stock acceptable to international investors by adopting enhanced disclosure levels - companies may be sending a signal to the market that future prospects are good. Therefore, issuing ADRs is expected to create shareholder value represented by increased stock prices in the Brazilian market after adjustment to the market effect. We will use as proxy the analysis of ADR-issuing companies‟ excess return as compared to market indexes. Our focus will be on the issue of greater information requirements posed by international accounting standards regulating bodies, particularly at

58

the time of the placement of stocks in markets other than the original one. Our hypothesis for the study:  Brazilian companies that have ADRs ib the United States do not perform better than their reference market and, therefore, do not create shareholder value. In this sense our specific objective includes:  Establishing a connection among the “Novo Mercado”, the “Differentiated Corporate Governance Levels”, and international standards as regards disclosure of information. V. Investigative Methodology  The value of ADR-issuing companies will be ascertained by means of an Event Study as per the methodology proposed by CAMPELL, LO & MACKINLAY (1997). Returns measurement and analysis Data on the daily returns of issuing companies were the basic material for the study. The data were operated as neperian logarithm:

 P  Ri  Ln i   Pi 1  Where: Ri  Return of asset i on day 

Pt  Closing price of the share on date  Daily market returns, represented by IBOVESPA, in turn - and similarly to stock returns - are operated in logarithmic form. Therefore:

 P  Rm  Ln m   Pm 1  Where: Rm  Return of market m on day 

Pmt  Closing price of the market (IBOVESPA) on day



Stock and market returns are used to obtain abnormal (excess) returns, ARit , relative to expected figures and according to two models: Market Adjusted Returns Model (MARM), and Market and Risk Adjusted Returns Model (MRARM): MARM  ARit  Rit  Rmt 



MRARM  ARit  Rit  ( i   i Rmt ) Concerning model MRARM, parameters ˆ i and

ˆ i are calculated as square minimum over the data for Rit and Rmt for the period between days 386 and 44,


from day 0 of the shareholders‟ meeting that dealt with the subscription. Rmt is the market return on day  . Observation Periods The tests conducted in this study rely crucially on the definition of time “windows” both to estimate models and to measure performance. These windows are described below. One must bear in mind that they are referenced to day “0”, that is, the date of the shareholders‟ or board meeting that deliberated for subscribing new shares:  MRARM parameters estimation window for the period between 386 and 44 days of excess returns observation:  As regards abnormal returns, we elected periods from 521 to +521 observation days. More specifically, this period is segmented into sub-periods: a) –521 to –261; b) –261 to –20; c) –20 to 0 d); 0 to +20; e) +20 to +261:  As regards the observation of effects in the post-subscription event period, we surveyed abnormal returns in the period +261 to +521. We will, therefore, test the evolution the ADRissuing companies‟ stock prices from the moment they started issuing ADRs to present. Based on the empirical data obtained and the theoretical framework developed, we will perform our analysis of value creation and draw conclusions. Data Treatment We initially surveyed all companies that filed Depositary Receipts with the CVM. At a second stage, we collected daily series of closing prices with Economática (Equity analisys tool). We then eliminated from the sample all companies that did not have a continuous prices series within the observation window (-521 to +521 trade days). This left us with 51 observations for analysis (listed in table 1).

Subsequently, we performed the excess return calculations (considering the MARM and MRARM models) compared to daily temporal series for IBOVESPA (São Paulo Stock Exchange Index), IBA (Brazilian Stock Index), MSCI-World and one index (Ind+10). Ind+10 was assembled from the ten most-liquid stocks on the date of the event (of each ADR-issuing stock in the sample). The liquidity was that provided by the Economática service. VI. Results To analyze our hypothesis we built a table with the compounded excess return index (CAR), whose initial value us 1.00 on day 521, evolving by continuous capitalization for each trade day comprised in the observation window. Ind Car() = 1,0 e   T-student tests were implemented to verify the hypothesis of null average excess returns, against (



AT )

alternate hypotheses AR   0 (double-tail test) and

AR   0 (single tail test). The statistics for these tests were calculated based on the parameters for AR  distribution estimated for the period [-386 ; -44]. The tables (omitted) for both models (MARM and MRARM) do not display significantly negative abnormal returns at significance levels of 2.5% or 5.0%. In addition, significantly positive excess returns took place in no trade days. Our observations lead us to the same conclusions as Furtado (1997) and Garcia (2002) on the market efficiency aspect: no statistically significant evidences were found leading to the rejection of the hypothesis of random distribution around the zero-average of excess returns after the shareholders‟ meeting. The graphs (omitted here) plotting the evolution of the CAR (Compounded Excess Return) index for both models and the several observation windows.

Moram_Ibovespa 1,2000

1,0000

0,8000

0,6000

0,4000

0,2000

0,0000 -600

-400

-200

0

200

400

600

Graph A. CAR Index Evolution (base MARM/IBOVESPA) for observation window [-521;+521] 59


Moram_Ibov20 1,1200 1,1000 1,0800 1,0600 1,0400 1,0200 1,0000 0,9800 -25

-20

-15

-10

-5

0

5

10

15

20

25

Gráfico B. CAR Index Evolution (base MARM/IBOVESPA) for observation window [-20;+20]

Gráfico C. CAR Index Evolution (base MRARM/IBOVESPA) for observation window [-521;+521] Momer_Ibov20 1,0600 1,0500 1,0400 1,0300 1,0200 1,0100 1,0000 0,9900 -25

-20

-15

-10

-5

0

5

10

15

20

25

Gráfico D. CAR Index Evolution (base MRARM/IBOVESPA) for observation window [-20;+20]

60


We were able to determine, by observing the charts that the period in the vicinity of the event sports marked evolution of abnormal returns, indicating that there is a positive effect at the time of the placement of ADRs, but that this effect doesn‟t persist in the long run. We also observed, as did Euchério (1999), that after the ADR placement event, average excess return drops, which is consistent with his observations. Statistical tests performed based on the two models (MARM and MRARM) - see the descriptive statistics, attached – showed that average abnormal returns ex-post (based on the different benchmarks) are more acutely negative, confirming the new observations performed above. VII. Conclusions We observed that longer observation periods for compounded excess returns do not allow rejecting the hypothesis that Brazilian companies with American Depositary Receipts - ADRs in the U.S. market perform no better than their reference market. However, considering compounded excess returns in the period close to the event of ADR placement when the market become aware of the company‟s new status, therefore - there is a positive movement of the stock‟s return. We have found evidence that confirms the studies by Euchério (1999), according to which, after the placement of ADRs, average excess return drops, which is consistent with the liquidity hypothesis of Amihud & Mendelson (1988) and the visibility hypothesis formulated by Merton (1987) Finally, empirical evidence indicates that the performance of the stocks of the set of ADR-issuing companies is better than BOVESPA‟s corporate governance index and than IBOVESPA, indicating that increased information disclosure leads to shareholder value creation.

References 1.

BOLSA DE VALORES DO ESTADO DE SÃO PAULO. Níveis Diferenciados de Governança Corporativa. São Paulo, 2001. 11p. 2. ______________________________ Novo Mercado. São Paulo, 2001. 38p. 3. BOTOSAN, Christine A., Disclosure Level and the Cost of Equity Capital, published in the Accountng Review, vol. 72, nº 3, July 1997, pages 323-349 4. CAMPBELL, John Y.; LO, Andrew W.; Mackinlay, A. Craig. The Econometrics of Financial Markets. Princenton: 1997. 5. GARCIA, Fabio G. Verificação da Existência de Assimetria de Informações no Processo de Emissão de Ações no Mercado Brasileiro: Uma forma de medir a Importância da estrutura de Ativos da Empresa. 2002. 246 f. Doctoral Thesis – Escola de Administração de Empresas de São Paulo, Fundação Getúlio Vargas, São Paulo. 6. HEALY, Paul M., Amy P. Hutton e Krishna G. Palepu, Stock Performance and Intermediation Changes Surrounding Sustained Increases in Disclosure, in Contemporany Accounting Research, vol. 16, nº 3, fall 1999, pages 485-520, The Canadian Academic Accounting Association. 7. LEUZ, Christian; VERRECCHIA, Robert E. The Economic Consequences of Increased Disclosure. Working Paper Series: Finance & Accounting, Frankfurt, n.41, p.35, June 2000. 8. SAUDAGARAN, Shahrokh M. International Accounting: a user perspective. 1st ed. Cincinnati,OH: South Western College Publishing, 2001. 228p. 9. SENGPUTA, Partha, Corporate Disclosure Quality and the Cost of Debt, Accounting Review, Vol. 73, nº 4, October 1998, pages 459-474. 10. RODRIGUES, Euchério L. Maior Visibilidade ou Integração do Mercado de Capitais Brasileiro. Revista da CVM, No. 29 August 1999.

61


INSIDE AND OUTSIDE SHAREHOLDERS AND MONITORING: EVIDENCE FROM DEVELOPING COUNTRY Mazlina Mustapha*, Ayoib Che Ahmad** Abstract This paper tests the effect of managerial (inside) and block-holders (outside) ownership in relation to agency theory in Malaysian business environment. This study tests the agency relationship in different culture and social contact and provides evidence whether agency theory in non-western organizations have equal impact in Asian organizations. Consistent with agency theory and the convergence of interest hypothesis, managerial ownership (insiders) in Malaysia indicate a negative relationship with the demand for monitoring. This finding may be due to the fact that as the managers are also the owners, there is less conflict, less information asymmetry and less hierarchical organization structure in the companies, which lead to lower monitoring costs. However, another ownership structure, outside block-holders appear to demand more monitoring. This positive relationship may be explained by their effort to compensate their lack of involvement in the daily transactions and internal decisions of the company, especially in the concentrated business environment in the country. Keywords: agency theory, block-holders, managerial ownership, monitoring costs, ownership structure * Faculty of Economics and Management, Universiti Putra Malaysia, 43400 UPM, Serdang, Selangor, Malaysia. [email protected] Tel: 603-89467636, Fax: 603-89486188 ** College of Business, Universiti Utara Malaysia, 06010 Sintok, Kedah, Malaysia

I.

Introduction

The divorce between ownership and management functions may lead to the possibility of principal– agent conflicts as the managers may not always act in the shareholders best interests and may misuse the corporate assets (Jensen and Meckling, 1976; Shleifer and Vishny, 1986). This divergence of interest between managers and shareholders may lead to “agency problems”, and results in agency costs as described in agency theory (Farrer and Ramsay, 1998). Various factors have been considered to overcome this problem and reduce the costs. Fleming, Heaney and McCosker (2005) claim that ownership structure such as concentrated ownership and managerial structure can mitigate and reduce agency costs of an organization. When managers own the shares of the firm, they have the incentive to increase the value of the firm rather than shrink as they have entrepreneurial gain in the company (Jensen and Meckling, 1976). It is believed that incentive to consume perquisites declines as manager‟s share ownership increases because his share of firm‟s profit increase with ownership while his benefits from perquisite consumptions are constant (Ang, Cole and Wuh Lin, 2000; Fleming et al., 2005), and accordingly, the incentive to pursue personal benefits increases when he own smaller portion of the firm‟s shares (Mat Nor 62

and Sulong, 2007). Furthermore, as the owners are actively engaged in day to day activities of the company (Nimie, 2005), there will be less information asymmetry, less conflicts and less hierarchical organization structure. This less complex organization structure reduces the need for assurance and monitoring thus require less monitoring and agency costs. Besides managerial ownership, another ownership structure suggested by the literature to reduce the agency problem is through concentrated ownership. It is claimed that concentrated ownership by outside shareholders (such as block-holders), have greater incentives to align management and shareholders‟ interests (Li and Simerly, 1998). Blockholders are also said to facilitate behavior-based monitoring from the capital market (Eisenhardt, 1989). Prior studies claim that share ownership by blockholders can help to monitor agency problems (Agrawal and Knoeber, 1996; Fleming et al., 2005; Fosberg, 2004; O‟Sullivan, 2000). This is due to the fact that shareholders of an organization have a residual claim on the earnings and assets of the organization and therefore bear proportional to their share ownership, the economic consequences of actions taken by organization managers and directors. If managers engage in opportunistic behavior, shareholders bear a portion of the costs of such actions (Fama and Jensen, 1983a). Large shareholders


are also claimed to have greater incentive to monitor management and have the necessary power to influence the company‟s policies since they will bear a significant proportion of managers‟ value destroying actions (Haniffa and Hudaib, 2006). However, agency theory is also criticized for its ignorance of the existence of social and authority relationship and assumes social life is a series of contract (Johnson and Droege, 2004). It is unknown whether the agency theory findings in western countries have equal impact in Asian organizations (Ekanayake, 2004; Johnson and Droege, 2004). Previous literature (Conlon and Parks, 1990; HassabElnaby and Mosebach, 2005; Ekanayake, 2004) indicates that there is a possibility that given the cultural differences, the typical nature of agents in agency theory may not be the case with regard to nonwestern countries. Sharp and Salter (1997) argue that the agency effects are lower in Asia. It is also claimed that there is a limited empirical research that directly tests agency theory in different culture context (Ekanayake, 2004). Thus, this study empirically examines the agency relationship in Malaysian organizations, one of the Asian countries. Besides being a developing country with an emerging market, Malaysia is chosen in this study because of its unique concentrated business environment. It is claimed that owner managed firms are common among Malaysian companies (Mat Nor and Sulong, 2007), and large shareholders also exist in these companies. This study defines block-holders as those shareholders who hold at least 5% or more of a voting right in an organization and are not linked to the organization management in either business or family relationship. Specifically this study focuses on the effect of managerial (insider) and block-holders (outsider) ownership on the agency costs of Malaysian public listed companies. This study uses the direct measure of agency costs, which are the cost of monitoring the companies as recommended by Malaysian Code of Corporate Governance (FCCG, 2001). This study aims to provide evidence that support or reject prior research findings in western countries relating to the effect of these ownership structures on the agency relationship which is reflected in its agency costs. The results indicate an inverse relationship between managerial ownership (insider) and monitoring costs as predicted in agency theory. This is supported by the independent t-test which indicates that those companies which have high managerial ownership in their organizations have significantly lower monitoring costs compared to those with low managerial shareholdings. And, another ownership structure, outside block-holders appear to demand more monitoring to compensate for their lack of involvement in internal decisions of the companies. The finding indicates that as the percentage of shareholdings by block-holders increase, the monitoring costs also increase.

The remainder of the paper is organized as follows. The next section discusses the relevant literature on the role played by managerial and blockholders ownership in agency setting and how it affects the agency costs, which lead to the hypotheses development. The methodology employed in this study is outlined in Section III and the results of empirical testing are presented in Section IV. The paper ends with the conclusion of the research. II. Literature review and hypothesis development A. Principal-agent relationship in agency theory Initially, physical assets defined an individual‟s net worth to denominate wealth (Carlson, Valdes and Anson, 2004). Examples of such assets are lands. Kings and members or royalty defined their power based on the land that they owned. Later, as the economic activities changed from agricultural to industrial economy, this basis changed from ownership of land to ownership of legal entities. In their discussion of the origin of the word “share ownership”, Carlson et al. further claim that as a consequence of the industrial revolution, public organizations are established to create goods and services and stocks and bonds are created to support the financing of the new enterprises. These stocks also reflect the ownership of the organizations. If in the past, banks are the custodians of physical assets of their clients (such as coins, jewels, and land deeds), with the full force of industrial revolution, banks begin to “hold shares of ownership” in public organizations, which create the term “shareholders”. With the acceptance of industrial revolutions also, organizations grow bigger, and the owners are no longer the managers of the organizations. It is not practical for the shareholders to make day to day decisions of the organizations and this job is delegated to the managers. This separation between the owner and managers tends to create agency problems as claim in agency theory. Agency theory postulates that the firm consists of a contract between the owners of economic resources (the principal) and management (the agents) who is charged with using and controlling these resources (Jensen and Meckling, 1976). This theory posits an inherent moral hazard problem in these relationships, which in turn give rise to agency costs for the organization. The agency relationship between the principal and the agent give rise to agency costs because the managers may not act in the owners‟ best interest, such as consumption of excessive perquisites and sub-optimal investments (Fleming et al., 2005). Agents normally have more information than principals and this information asymmetry adversely affect the principal‟s ability to monitor whether their interest are being properly served by the agents

63


(Adams, 1994). The principals want to ensure that their resources are being utilized in the best manner possible, which later will flow back to them in the form of dividend. Whereas the agents are also concern as this would be the measurement of their efficiency in managing the company, and may be the source for the determination of their salary/remuneration in the future. In the process of discharging the duties, agency theory assumes that the agents and principal will act rationally and they will use the contracting process to maximize their wealth. According to Kren and Kerr (1993), to ensure the efficiency in the contracting process, both principal and agents will incur contracting cost. For instance, to minimize the risk of shirking by agents, the principal will appoint the board of directors (Fama and Jensen, 1983a) and auditors. The board of directors will ensure that the management acts on behalf of shareholders, i.e. increase the wealth of the corporation (Iskandar and Mohd Salleh, 2004). An effective board of directors will provide a measure of reducing the agency problem, which will then lead to transparency of financial reporting and good governance of the organization. And, the external auditors will examine the financial statements prepared by the management to ensure their compliance to the standards, rules and regulations required and reflect the true and fair view of the organization‟s transactions. Agents on the other hand will incur bonding cost, for example, the cost of internal audit in order to signal to the owner that they are acting responsibly and consistent with their contract of employment (Adams, 1994). B. Agency costs and managerial ownership Prior literature suggests various ways to overcome this agency problem. Among others, it is claimed that managerial shareholdings can reduce and mitigate agency costs (Jensen and Meckling, 1976; Agrawal and Knoeber, 1996; Ang et al., 2000; Chow, 1982; Fleming et al., 2005; O‟Sullivan, 2000). They argued that the agency costs of equity arise from the direct expropriation of funds by the managers, consumption of excessive perquisites, shirking, sub-optimal investment and entrenching activities. Thus, earlier studies suggest that managers are encouraged to own the organizations‟ share to motivate management monitoring (Agrawal and Knoeber, 1996; Fleming et al., 2005). This is because the higher the portion of the shares, the more responsible is the manager to increase the value of the companies. According to the original agency theory by Jensen and Meckling (1976), and Fleming et al. (2005), equity agency cost is zero when there is a 100% owned manager organization, and there is a positive relationship between equity agency costs and the separation of ownership and control. As owner manager equity ownership falls below 100%, the equity ownership

64

becomes relatively dispersed. In this circumstance, the manager has a greater incentive for shrinking or the consumption of excessive perquisites. This is due to the fact that although the firm‟s value falls, the managers only bear a portion of the expense related to their ownership stake (Farrer and Ramsay, 1998). In other words, a lower managerial equity holding is associated with lower incentive and effort exert by the managers in their responsibilities to seek profitable investments. Chow (1982) suggest that when managers own smaller equity stake in their firms they have an increased incentive to falsify financial disclosures, since such disclosures are likely to be utilized by shareholders in setting managers‟ remuneration. This notion is also supported by Ang et al. (2000) and Fleming et al. (2005) who find that there is an inverse relationship between the agency cost and managers‟ ownership share and a direct relationship with the number of non-manager shareholdings. The incentive to consume perquisites declines as his ownership share rises, because his share of the firm‟s profits rises with ownership while his benefits from perquisite consumption are constant. It is also suggested that managerial shareholdings help align the interests of shareholders and managers in its convergence of interest hypothesis (Jensen, 1993). The higher the ownership of the firm by the management, the less the conflicts among the stakeholders, the less the agency problem and cost associated with it (Friend and Lang, 1986; Jensen and Meckling, 1976). This is because the insiders have incentives to protect shareholders interests and need less supervision by the board, since board activity is a costly monitoring alternative (Vafeas, 1999). It is also said that increased agent ownership reduces the need for monitoring as the incentive alignment is enhanced. The convergence of interest model suggested by Jensen and Meckling (1976) claim that an increase in the proportion of firm‟s equity owned by insiders is expected to increase firm value as the interest of inside and external shareholders are realigned, and consequently there is a reduced need for intensive audit. O‟ Sullivan (2000) finds that significant managerial ownership results in a reduced need for intensive auditing which may be due to the merging functions of ownership and management, and consequently minimize the monitoring motivation for audit. The auditors are also said to be less inclined to undertake additional testing when managers are also significant equity holders, since owner managers are less likely to deliberately mislead themselves (O‟Sullivan, 2000). Publicly traded firms in which top management has a larger ownership stake experience corporate crime (proxy for agency cost) less frequently (Alexander and Cohen, 1999). Managers also will have more powerful incentives to make value maximizing decision about capital structure as their stock ownership is high (Berger, Ofek, and Yermack, 1997). Besides increase incentive to


maximize the firm value, holding common stocks also motivate the managers for its underlying voting rights, such as increase their influence on board of directors and hence on the firm‟s general policy (DeAngelo and DeAngelo, 1985). However there are also studies which suggest contradict and mix findings, such as Singh and Davidson (2003) who conclude that managerial ownership does not serve as a significant deterrent to excessive discretionary expenses which is used as a proxy for agency cost in their study. Prior studies also claim that managers who owned excessive shareholdings in their companies and not diversified would be more risk averse than other shareholders. This may motivate the managers to adopt overly conservative approach which suppresses shareholders return, as the financial collapse of the company may cause the financial collapse of the directors (Farrer and Ramsay, 1998; Fama and Jensen, 1983b; Loh and Venkatraman, 1993). In other words, if the managerial shareholdings are too high, their interests may not be aligned with the interests of other shareholders. In terms of this ownership structure‟s association with another monitoring mechanism, that is auditing, it is found that the lower the managerial share ownership in a company, the greater the probability of the company being audited (Tauringana and Clarke, 2000). Another literature claims that agency theory suggests that in the absence of regulation, the propensity of firms to demand independent audit is a function of the extent of the divorce between ownership and control (Chan, Ezzamel, and Gwilliam, 1993). This is supported by Fan and Wong (2005) who claim that external auditors play a monitoring and bonding role in order to mitigate the agency conflict between the controlling owners and the outside investors. Thus, it is claimed that management monitoring can be provided by encouraging the managers to own the shares of the organizations (Fleming et al., 2005; Agrawal and Knoeber, 1996). This is supported by Jensen and Meckling (1976) who claim that agency problems can be mitigated by making managers owners of the organizations, because when managers have entrepreneurial gain, they have the incentive to increase the value of the organization rather than shirk. It is further claimed that the higher the managerial ownership, the lower the demand for monitoring mechanisms as the owners are actively engaged in day to day activity (Niemi, 2005), and therefore will lead to less conflicts and less information asymmetries. Ang et al. (2000) and Fleming et al. (2005) claim that the incentive to consume perquisites declines as managers ownership share arises, because their share of the organization‟s profit rises with ownership while their benefits from perquisite consumption are constant. This is supported by Jensen and Meckling (1976) and Mat Nor and Sulong (2007) who argue that when managers own a

smaller portion of the organization‟s share, they have greater incentive to pursue personal benefits and less incentive to maximize organization value. Thus, one of the ways to reduce the associated increase in agency costs is to increase the shares held by managers. It is also assumed that owner-managers are more efficient in controlling corporate assets than hired managers, which result in less hierarchical organizational structure. This less complex organization structure also reduces the need for assurance and monitoring (Abdel-Khalik, 1993). Jensen (1986) further argues that management risk averse behavior will give an impact on audit effort through the audit risk assessment, which would result in lower audit effort and fees (Niemi, 2005). In summary, manager owned organizations are predicted to have lower perquisite consumption, less conflict, lower information asymmetry, less hierarchical and lower organizational complexity and lower risk, which would result in lower monitoring needed. Furthermore, the greater the shares held by the managers, the greater is their incentive to maximize the organization value and reduce the monitoring costs. Therefore, it is hypothesized that: H1: The greater the ownership control by the managers (insider), the lower is the total monitoring costs of the organization. C. Agency costs and block-holders ownership Another ownership structure, that is block-holder equity stake, also indicates greater incentives and capability to monitor management (Singh and Davidson, 2003; Fleming et al., 2005; Fosberg, 2004). Prior studies claim that share ownership by blockholders can help to monitor agency problems (Agrawal and Knoeber, 1996; Fleming et al., 2005; Fosberg, 2004; O‟Sullivan, 2000). This is due to the fact that shareholders of an organization have a residual claim on the earnings and assets of the organization and therefore bear proportional to their share ownership, the economic consequences of actions taken by organization managers and directors (Haniffa and Hudaib, 2006). If managers engage in opportunistic behavior, shareholders bear a portion of the costs of such actions (Fama and Jensen, 1983a). It is also said that block-holders existence in an organization can resolved the conflict of interests over financing policy arise between managers and shareholders because of the fact that managers preference for lower organization risk due to their under-diversification (Fama, 1980), and managers‟ dislike to being subject to performance pressure that large fixed interest payment entails (Jensen, 1986). Managerial insiders are reluctant to use the optimal amount of debt financing for the organization because of the additional bankruptcy risk associated with higher level of debt engender (Fosberg, 2004).

65


Therefore managers will not issue the optimal amount of debt without pressure from a disciplining force (Jensen, 1986). However, the shareholders want the leverage to be used at its optimal level in order to maximize the organization value. Berger et al., (1997) and Borokhovich, Brunarski, Harman, and Kehr (2005) claim that this conflict can be resolved by having block-holders in the organization as they find that leverage rises in the presence of significant blockholders. Again, this suggests that block-holders have a strong incentive to monitor the opportunistic behavior of organization managers. A local study by Mat Nor and Sulong (2007) postulates that large share ownership provides the incentive of controlling shareholders to use their influence to maximize value, exert control and to protect their interest in the company. They further claim that majority control gives the largest shareholders considerable power and discretion over organization‟s important decisions. This is supported by Haniffa and Hudaib (2006) who claimed that large shareholders have greater incentive to monitor management and have the necessary power to influence the company‟s policies since they will bear a significant proportion of managers‟ value destroying actions. On the other hand, a well-diversified investor is not particularly worried as the bankruptcy risk of any one organization in the portfolio of investments will not have a large impact on their wealth. Consequently, a shareholder‟s incentive to monitor insiders and ensure that the organization is properly managed is directly related to the proportion of the organization‟s shares that the shareholder owns. And it is further expected that block-holders would favor more extensive audit and consequently pay higher audit fees as they have the financial incentives to ensure maximum monitoring is undertaken (O‟Sullivan, 2000). Furthermore, this scenario may be expected in Asian countries such as Malaysia as the businesses are claimed to be very concentrated (Ow-Yong and Guan, 2000; Mat Nor and Sulong, 2007), especially with family businesses (Haniffa and Hudaib, 2006). Thus the outside shareholders have no idea about how the businesses are conducted unless through an external verification such as an audit. This is supported by another literature, Hay et al., (2008), who suggest two possible outcome of having block-holders in the governance structure of an organization. The first outcome is consistent with agency theory. A block-holder that is actively involved in operations and decision making (such as managerial shareholdings) may have such a broad span of control over activities and internal control that the need for other mechanisms such as external auditing may be reduced. On the other hand, a major outside shareholder (such as outside block-holders) may also use this influence to increase external auditing to compensate for a lack of control over other internal decisions.

66

Hence, it is argued that block-holders will assist in the monitoring of the organization as they have the incentive to do so and this monitoring is directly related to the proportion of the organization‟s shares that the shareholders own. And, as the data is collected using Malaysian sample companies which are claimed to be concentrated, especially with family businesses, and the study defined block-holders as outside large shareholders who are not involved in the daily activities of the companies, these large shareholders are expected to demand more monitoring to compensate for their lack of control over the internal decisions and daily transactions happen in the companies. This is also consistent with the earlier studies (Adams, 1994; Kren and Kerr, 1993) which claim that the principals and agents in the agency relationship will incur contracting costs to minimize the risk of shirking by the agents, ensure optimal investment of the organizations and motivates more transparent reporting. Thus, it is hypothesized that: H2: The greater the ownership control by the block-holders (outsider), the greater is the total monitoring costs of the organization. III.

Data and Methodology

A. Data and sample Data for the study was collected using primary and secondary sources. Primary data was collected using cross-sectional surveys which were sent to Malaysian public listed companies. Data collection cannot be done solely by using secondary data, as some of the information needed (such as internal audit costs) for the study is not available from secondary sources (such as annual reports). The population of the study includes all companies listed on the Main and Second Board of Bursa Malaysia. However, the companies classified under finance sector were excluded in this study because of their unique features and business activities, as well as differences in compliance and regulatory requirements (Yatim, Kent and Clarkson, 2006; Mat Nor and Sulong, 2007). Questionnaires were sent to all 867 companies in the population. Once the questionnaires were returned, the annual reports of those companies with completed questionnaires were scrutinized for further information to be used in the study. The secondary data was hand-collected from the companies‟ annual reports which were available at Bursa Malaysia‟s website (http://www.bursamalaysia.com.my). In the annual report, the Directors‟ Report, Statement on Internal Control, Corporate Governance Statement, directors‟ profile, Shareholdings Statistics, Corporate Information, Statement of Directors‟ Shareholdings, the financial statements and notes to the accounts are scrutinized. Information on directors‟ shareholding and directors‟ background can be


gathered from the directors‟ profile, Corporate Governance Statements, Shareholdings Statistics and notes to the accounts. The external audit fees, book value of the assets, total receivables, total inventories, total long term debts and number of subsidiaries can be gathered from the financial statements and notes to the accounts. The information about the existence of the internal audit department is normally included in the Statement of Internal Control; however, it is not mandatory to disclose the internal audit cost. Only 3 companies voluntarily disclose their internal audit costs. Information needed to calculate Tobin‟s Q and return to total assets (ROA) can be gathered from the financial statements. Data from the annual reports were then transferred to the worksheets. The information gathered from the questionnaires was also tabulated in the worksheet and further matched and validated with the information obtained from the annual report. This will then address the reliability concern of our survey data as conducted by Anderson, Francis and Stokes (1993) in their study of Australian companies. Non response bias was also conducted for the data collected from the questionnaires. After considering the incomplete and inconsistence questionnaires, there were 235 usable samples for the study. The data was also inspected for outliers by means of standard regression diagnostics at three standard deviations (as suggested by Hair, Anderson, Tatham and Black, 1998, p. 65). Normality check of the data was also carried out and some of the measures were transformed into logarithm to control for skewed nature of data. As multivariate regression is used to analyze the data in this study, assumptions

of multicollinearity, homoscedasticity and linearity are also tested. B. Variable definition Dependent variable in this study is the monitoring costs of the companies listed in Bursa Malaysia. Earlier studies use indirect measurement such as asset utilization ratio (Singh and Davidson, 2003), ratio of selling and administration expenses to sales (Singh and Davidson, 2003) and ratio of operating expenses to sales (Ang et al., 2000) as proxies for agency costs incurred by the firms in monitoring their firms. But this study uses measurements that are directly related to these firms in monitoring the shareholders wealth of their companies. Directorship and auditing (internal and external) are specified as monitoring mechanisms in the Malaysian Code of Corporate Governance (FCCG, 2001). Thus, the dependent variables in this study involve the costs of these monitoring mechanisms demanded by the organization in Ringgit Malaysia (RM). However, as the executive directors are in-charged of managing the companies, and the non-executive directors are said to monitor and controlling the opportunistic behavior of the management (Jensen and Meckling, 1976; Haniffa and Hudaib, 2006), this study does not include executive directors‟ remuneration as monitoring costs. Hence, total Monitoring (MONITOR) is measured by the sum of organization investment in non-executive directors‟ remunerations (DIRREMNED), internal auditors‟ costs (INTCOST) and external auditors‟ costs (EXTCOST).

Table 1. Operationalization of the research variables Variable

Explanation

Measurement

Dependent variable: MONITOR

Total of external audit costs, internal audit costs and non-executive directors remuneration

Total Monitoring = External audit costs (EXTCOST) + Internal audit costs (INTCOST) + Non- executive Directors remunerations (DIRREMNED)

Managerial ownership(Insiders) Block-holders shareholdings (Outsiders)

Percentage of executive directors‟ shareholdings (%) Percentage of block-holders‟ shareholdings (%)

Size of the organization Complexity of an organization‟s operation Complexity of an organization‟s assets Debt of the organization Risk of an organization Performance of an organization Growth of an organization Listing status of an organization Industry

Natural log of total assets Natural log of number of subsidiaries (including its head-office)

Independent variable: MGROWN BLKOWN Control Variables: SIZE COMPLEX RECINV DEBT RISK ROA GROWTH LISTSTAT INDUSTRY

(Inventories and Receivables)/ Total assets Long term debt / Market value of the firm 1 if company has a loss in current year and 0 otherwise. Profit before interest and tax /Total Assets (ROA) Market value of the firm / total assets (Tobin‟s Q) 1 if company is listed in the main board, and 0 otherwise CONTRASE – for companies in consumer, trading and services sectors; INDPROP - for companies in industrial, construction and property sectors

The independent variables in this study are managerial ownership (MGROWN) who are the insiders and block-holders ownership (BLKOWN),

who are the outsiders. The study defines managerial ownership as the total percentage of executive directors‟ shareholding, while the block-holders is the 67


total percentage of shareholding of block-holders who hold at least 5% or more of a voting right in an organization and are not linked to the organization management in either business or family relationship. The controlled variables include in the study are size, complexity, performance, risk, growth, listing status and industry. The following model is used to analyze the relationship between the monitoring costs and ownership structure: MONITOR = αi - b1MGROWNi + b2 BLKOWN + b3RECINV + b4DEBT b5COMPLEXi + b6SIZEi – b7RISKi – b8ROAi + b9GROWTHi + b10LISTSTATi + b11CONSTRASEi + b12INDPROPi + εi Variable definitions, labels and measurement used are reported in Table 1. IV A.

Results and discussions Descriptive statistics

Panel A and Panel B of Table 2 presents the descriptive statistics for the variables used in the study. Panel A reports those for continuous variables and Panel B presents those for dichotomous variables. Panel A shows that non-executive directors‟ remunerations constitute the largest component of monitoring costs, followed by internal audit costs and external audit costs ranking second and third respectively. The mean percentage of shareholdings by the managers is about 27%, which is approximate

the 34% average of Haniffa and Hudaib (2006) findings. The mean percentages of block-holders shareholdings, which do not include those parties involved in the management or have any family relationship with the managers, is about 15%. Further analysis indicates that about 21.3% of the cases have the cumulative largest shareholders owning more than 25% of the issued shares in the companies, and about 42.5% with accumulative largest shareholders owning between 5% to 25%. This suggests that Malaysian companies are concentrated and less diffused. The ratio of long term debt to the market value ranges from 0% to 93% with the average close to 15%. The descriptive statistics also show that the sample companies cover a wide range of companies, some moderately small and some relatively large, range from those with RM18 millions to RM65,092 millions of total assets. The complexity of the companies in terms of their operations range from simple, where there are companies with only their head office with no subsidiary, to more complex. The complexity of their assets‟ compositions also reflect the same pattern, the ratio of inventories and receivables to total assets range from 0.19% to 80% and the average is about 31%. On average, the respondent companies have the total assets of RM1,564 millions and 20 subsidiaries, while the average Tobin‟s‟ Q is 1.05. Panel B reports that about 75% of the companies are listed in the main board of the Bursa Malaysia, and the balance in the second board. Only 20% of the companies suffer a loss in the current year.

Table 2 . Descriptive statistics of variables Panel A:Continuous variables Variables INTCOST (RM) EXTCOST (RM) NEDREMM (RM) MONITOR (RM) MGROWN BLKOWN DEBTSTRC RECINV COMPLEX SIZE (RM) ROA GROWTH Panel B : Dichotomous variables LISTSTAT RISK CONTRASE INDPROP

Mean

Std dev

Minimum

Maximum

280,896 263,487 302,249 846,632 0.272734 0.151725 0.1468 0.308798 19.74 1,564,597,791 0.010054 1.051495

971,753 732,805 435,358 1,799,424 0.2323824 0.1910793 0.1584435 0.1945093 34.801000 5,679,828,495 0.2258620 0.7091715

0 15,000 0 56,900 0.0000 0 0.0000 0.0019 1.0000 18,261,685 -3.0172 0.3081

10,000,000 9,700,000 4,045,000 21,010,000 0.8637 0.7657 0.9328 0.8046 445.00 65,092,100,000 0.2037 7.9680

Yes 175 46 78 126

% 75 20 33 54

No 60 189 157 109

% 25 80 67 46

Variable definition: INTCOST = Total internal audit cost in RM; EXTCOST = Total external audit costs in RM; NEDREMM = Total NED remunerations in RM; 68

MONITOR = Total monitoring costs in RM; MGROWN = Executive directors‟ shareholdings (%);DEBTSTRC = Long term debt to market value of the firm; RECINV = Ratio of inventories and receivables to total assets; COMPLEX = number of


subsidiaries(including the head office); SIZE = Total assets in RM; ROA = ROA; GROWTH = Tobin‟s Q; RISK = Current year loss(Dummy); LISTSTAT = Board listing (Dummy); CONSTRASE = Companies in consumer, trading and service sectors (Dummy); INDPROP = Companies in industrial, constructions and property sectors (Dummy). The results of standard tests on skewness and kurtosis in Table 3 indicate that there is no problem with normality assumption 11 . A visual check for normality using histogram and normal probability plots is also carried out. All the histograms appear to be reasonably normally distributed and the normal distribution of the probability plot forms a straight line and the values appeared to fall approximately on this normality line. Thus, these variables can reasonably be considered as normally distributed. In summary, the model does not violate the basic OLS assumptions and could be used to test the expected hypotheses. Table 4 presents the correlation matrix for the dependent and independent variables. The result indicates that there is no multicollinearity problem, as the correlations are below the threshold value of 0.8 (Gujarati, 2003, p. 359). Variable definition: MONIITOR = Total monitoring costs(ln); MGROWN = Executive directors‟ shareholdings (%); DEBTSTRC = Long term debt to market value of the firm; SIZE = Total assets(ln); COMPLEX = number of subsidiaries(ln); RECINV = Ratio of inventories and receivables to total assets; ROA = ROA; RISK = Current year loss(Dummy); GROWTH = Tobin‟s Q; LISTSTAT = Board listing (Dummy); CONSTRASE = Companies in consumer, trading and service sectors; INDPROP = Companies in industrial, constructions and property sectors.

11

The data is said to be normal if the standard skewness is within ±1.96 and standard kurtosis is between ±3.0 (Mat Nor and Sulong, 2007; Abdul Rahman and Mohamed Ali, 2006; Haniffa and Hudaib, 2006). 69


Table 3. Normality test statistics of sample companies Variable

MONITOR MGROWN BLKOWN REVINV COMPLEX RISK SIZE LISTSTAT CONSTRASE INDPROP ROA GROWTH

Mean

Minimum

Maximum

Std Dev

Skew ness

Kur tosis

12.9841 0.2727 0.1517 0.3088 2.4998 0.2000 19.744 0.7400 0.3300 0.5400 0.0101 1.0515

10.9491 0.0000 0.0000 0.0019 0.0000 0 16.720 0 0 0 -3.0172 0.3081

16.8605 0.8637 0.7657 0.8046 6.0981 1 24.8991 1 1 1 0.2037 7.9680

1.0005 0.2324 0.1911 0.1945 0.9091 0.3980 1.4171 0.4370 0.4720 0.5000 0.2259 0.7092

0.864 0.210 1.500 0.329 0.232 1.544 0.911 -1.130 0.718 -0.146 -10.814 5.424

0.922 -1.230 1.470 -0.888 1.430 0.386 0.887 -0.731 -1.497 -1.996 140.20 42.856

Note: Figure in the parenthesis is the P value Table 4. Correlation matrix

0.09*

0.07

0.28* **

0.42* **

0.05

0.22* **

-0.04

1.00 0.50 ***

1.00

0.47* **

0.21***

0.18 ***

0.06

1.00

0.02

0.09*

0.07

0.04

0.09*

-0.07

-0.08

-0.08

0.00 0.09 ***

1.00 0.23 ***

1.00

INDPROP

1.00

CONSTRASE

0.16* **

LISTSTAT

0.19 ***

1.00 0.37* **

GROWTH

0.08

ROA

RECINV

-0.01

COMPLEX

DEBTSTR

1.00

SIZE

BLKOWN

1.00 0.45 ***

RISK

MGROWN

MONITOR

Variable

1.00 0.76* **

1.00

MONITOR MGROWN

BLKOWN DEBTSTRC RECINV

RISK

1.00 0.26 *** 0.31 *** 0.24 *** 0.21 *** 0.25 ***

SIZE 0.82 *** COMPLEX 0.61 ***

-0.03 0.21 *** 0.10 *

0.00 0.40 *** 0.14 **

ROA 0.15 ** GROWTH 0.09 * LISTSTAT 0.32 *** CONSTRASE

INDPROP

0.11 * 0.15 **

-0.02 0.07 0.13 ** 0.13 ** 0.11 * 0.10 *

0.05

0.18* **

0.02 0.16* *

0.15* **

0.06

0.00 0.23 ***

0.14* *

-0.02

0.09 *

0.01 0.28 *** 0.10 *

0.13* *

0.01

0.09 *

0.09 *

Notes: *** significant at 1% level ** significant at 5% level * significant at 10% level (See variable definition in Table 3)

70

-0.04 0.43 ***

0.52* **

1.00

0.20* **

-0.05

0.05


B. Results of main model Column two of Table 5 presents the multiple regression analysis used to test the main model. The adjusted R squared for the model is 0.757 and the Fvalue of 61.837 is significant (p