CORPORATE GOVERNANCE, OWNERSHIP STRUCTURE AND EARNINGS QUALITY: MALAYSIAN EVIDENCE

Transcription

1 CORPORATE GOVERNANCE, OWNERSHIP STRUCTURE AND EARNINGS QUALITY: MALAYSIAN EVIDENCE Hafiza Aishah Hashim* Doctoral Student, Faculty of Business and Accountancy, Universiti Malaya, Kuala Lumpur, Malaysia / S. Susela Devi, PhD Associate Professor, Faculty of Business and Accountancy, Universiti Malaya, Kuala Lumpur, Malaysia Abstract Contrary to the conflict of interest between outside shareholders and managers in a diffuse ownership such as in UK and US, the agency problems shifts away to conflicts between the controlling owners and minority shareholders in Asia where ownership concentration structures are more prevalent (Claessens and Fan, 2002). This paper examines the relationship between internal governance mechanisms namely the role of board independence and the ownership structure (i.e. managerial ownership, family ownership and institutional ownership) and the financial reported earnings quality. Using data from 280 non-financial companies listed on Bursa Malaysia s Main Board for the year 2004, this study fails to find any significant evidence on the relationship between the traditional functions of board of directors (i.e. proportion of independent non-executive directors) and earnings quality measured by accrual quality model. However, this study finds positive significant associations between proportion of family members and earnings quality which suggest that concentrated shareholdings in family ownership have incentives to reduce agency costs through a better alignment of shareholder and managerial interests. Given the growing role of institutional shareholder in Malaysian capital market, it is interesting to note that this study find positive significant evidence on the relationship between institutional ownership and earnings quality. Concentrated shareholdings by institutional investors provide an incentive for diligent monitoring as they have the resources, expertise and stronger incentives to actively monitor the actions of management and improve financial reported earnings. Keywords: Corporate Governance, Board Independence, Ownership Structure, Earnings Quality, Malaysia *Corresponding author: Hafiza Aishah Hashim, Faculty of Management and Economics, Universiti Malaysia Terengganu (UMT), Kuala Terengganu, Terengganu, Malaysia. hafizaaishah@umt.edu.my or affysya@yahoo.com. Telephone number: Fax number:

2 CORPORATE GOVERNANCE, OWNERSHIP STRUCTURE AND EARNINGS QUALITY: MALAYSIAN EVIDENCE 1. INTRODUCTION The Asian financial crisis in the late 1990s highlights the importance of good corporate governance practices to help restore investors confidence in the East Asian market. The financial crisis together with highly publicized scandals in the United States underscores the critical need for firms in both developed and developing countries to improve corporate governance practices and to gain back investors confidence towards the integrity of accounting numbers. Erosions in the quality of financial reporting raised troubling question about various aspects of corporate governance and most Asian countries 1 have taken proactive actions to improve and strengthen corporate governance systems including Malaysia with the issuance of Malaysian Code on Corporate Governance (2000) that describes the principles and best practices for corporate governance. The Malaysian Code is brought into full effect in January 2001 with the amendments to the Kuala Lumpur Stock Exchange s listing requirement where public listed companies in Malaysia are now required to put in their annual report the statement of corporate governance, a statement of internal control, composition of the board of directors, composition of audit committee, quorum of audit committee and any additional statements by the board of directors. The revamped Listing Requirements is recognized as a major milestone in corporate governance reform and its release bring into effect the Finance Committee Report on Corporate Governance and create an environment that demands higher standards of conduct and higher quality of disclosures from corporate governance participants in Malaysia. However, there are problems associated with Asian developing countries in strengthening corporate governance. Bhattacharyay (2004) highlights seven main problems including (1) excessive government intervention; (2) highly concentrated ownership structure; (3) weak external discipline in the corporate sector; (4) weak legal systems and regulatory framework; (5) lack of quality information; (6) lack of investors protection; and (7) lack of developed capital market that undermine the effectiveness of corporate governance mechanism employed in Asia. Differ to the conflict of interest between outside shareholders and managers in a diffuse ownership such as that commonly found in the UK and US, the agency problems shifts away to conflicts between the controlling owners and minority shareholders in Asia that practice ownership concentration structure (Claessens and Fan, 2002). The concentrated ownership creates agency conflicts between controlling owners and minority shareholders, which are difficult to mitigate through the traditional functions of board of directors (Fan and Wong, 2003). The tightness of ownership allows self-interested behaviours of mangers to go unchallenged internally by the board of directors and externally by takeover markets as the controlling owners who are often also the managers gain effective control of a corporation and have the power to 1 See e.g. Malaysian Code on Corporate Governance 2000, Singapore Code of Corporate Governance 2001, Thailand Code for Best Practice for Directors of Listed Companies 2002, Bangladesh Code of Corporate Governance 2004, Hong Kong Corporate Governance Code 2004 at 1

3 determine how the company is run and may expropriate the minority shareholders wealth. This study is motivated by several factors. First, while there is extensive literature that discusses the role of ownership structure in corporate governance around the world (Schleifer and Vishny, 1997; La Porta et al. 1999), there is scarce evidence from prior literature that empirically examine the relationship between ownership structure and financial reporting outcomes. Exception of prior studies that examine the relationship between ownership structure and performance (see e.g. Bathala and Rao, 1995; Mitton, 2002; Ng, 2005; Vethanayagam et al., 2006), little is known in terms of the relationship between ownership structure and financial reporting quality especially on the earnings quality issues in less developed economies. Second, given the conflicting theoretical viewpoints with regards to the relationship between family ownership and agency costs, this study try to gather evidence as to whether the existence of family board members on the board provides incentives to reduce or create agency costs in the Malaysian context. Some argues that family controlled firms reduce agency costs as they bring along their family name and reputation into the business while others argue that family controlled firms create agency costs by extracting private benefits at the expense of minority shareholders (Bartholomeusz and Tanewski, 2006). Third, recent literatures acknowledge the role of institutional investors as important force in corporate monitoring to protect minority shareholder s interest. Institutional investors in Malaysia have emerged as a powerful constitution that plays a very significant role in corporate governance. The set up of the Minority Shareholder Watchdog Group (MSWG) that represent 5 largest institutional funds in Malaysia, i.e. Employee Provident Fund (EPF), Lembaga Tabung Angkatan Tentera (LTAT), Lembaga Tabung Haji (LTH), Social Security Organization (SOCSO) and Permodalan Nasional Berhad (PNB) to monitor and deter abuses by company insiders shows Malaysian government commitments to encourage shareholder activism in Malaysia. There is limited evidence from prior study that look into the role of institutional investors around the world and this paper is the first to investigate the role of institutional investors by examining the effect of institutional ownership on earnings quality in Malaysian environment. Fourth, prior study provide evidence regarding the linkage between institutional features and financial reporting quality in East Asian market are cross country study (Fan and Wong, 2002; Leuz et al., 2003; Boonlert-U-Thai, 2006). However, findings of cross country study have been questions for limited sample size, endogeneity problems, noisy variables and severe omitted correlated variables (Gul, 2006). Since different countries have different level of investors protection, legal enforcement and ownership structure, it is important for the researcher to acknowledge these factors when discussing about earnings quality in different countries that based on different socio-economic (Boonlert-U-Thai, 2006; Lo, 2007) to provide a more useful meanings in earnings quality study. The discussion in this paper is organised as follows. Section 2 discusses the relevant literature to develop research hypotheses. Section 3 outlines and explains the sample selection, research method and variable measurement. Section 4 analyses and discusses 2

4 the research result. Section 5 provides the suggestions for future research and finally, the conclusions are considered in Section LITERATURE REVIEW AND HYPOTHESIS DEVELOPMENT Watts and Zimmerman (1978, p.113) state that one function of financial reporting is to constrain management to act in the shareholders interest. Given the increasing complexity of business today, there is a need for the financial reports to include more comprehensive information as investors rely on information they receive from companies in making their investment decisions. In the global institutional investor opinion survey of McKinsey & Company (2002) on corporate governance issues, it was found that the majority of investors agree that corporate governance remains of their great concern with strengthening the quality of accounting disclosure as the top priorities. The majority institutional investors are willing to pay a high premium for companies with good corporate governance, i.e. averaged 12-14% in North America and Western Europe, 20-25% in Asia and Latin America and over 30% in Eastern Europe and Africa. The survey also provides evidence that majority of the respondents (i.e.71 percent) state that accounting disclosure is the most important factor that impacts their investment decisions and 52 percent of the respondents identify that improving financial reporting quality is the governance priorities for policymakers. According to Bushman and Smith (2001), publicly reported accounting information can be used as important input information in various corporate governance mechanisms. A vast body of literature acknowledge the importance of corporate governance mechanisms to improve financial reporting quality and past literature has demonstrated that good governance help reduced the risk of financial reporting problems. As quoted by Hermanson (2003, p.44), Good governance goes in-hand with reduced risk of financial reporting problems and other bad accounting outcomes. Researchers have found evidence on the association between poor governance and poor quality of financial reporting quality including earnings manipulation, financial restatements and frauds (see e.g. Beasley, 1996; Dechow et al., 1996; Peasnell et al., 2000; Klein, 2002; Kao and Chen, 2004; Davidson et al., 2005). This early studies focus mainly on the role of boards as a monitoring tool and the contribution of non-executive directors in enhancing the quality and integrity of accounting information However, Li (1995) argues that the differences in corporate governance across countries emerge as a result of the variations in the ownership structure and understanding the effects of various ownership structure variables is vital to shed light on the corporate governance and control process of firms under difference national types of institutional arrangements. Mitton (2002) finds that firm-level differences in variables related to corporate governance of five East Asian countries, i.e. Indonesia, Korea, Malaysia the Philippines and Thailand had a strong impact on firm performance during the financial crisis. Specifically, he found that firms with higher disclosure quality, higher outside ownership concentration and firms that were focused rather than diversified are 3

5 associated with greater stock performance during the crisis. He suggests that in countries with weak legal protection of minority shareholders, at least some power to protect minority shareholder interest lies at the firm level where firms with higher disclosure quality and transparency, a more favourable ownership structure and less diversified appear to provide protection to the minority shareholder during the crisis. In Malaysia, a one-man or family run companies (Halim, 2001) and significant government equity holdings (Abdullah, 2006a) distinguish the ownership pattern of Malaysian companies that may complicate the corporate governance systems. Large ownership or ownership concentration where controlling shareholder have more than 50% ownership is common in Malaysia that may contribute to deficiencies in corporate governance system employed (Thillainathan, 1999). Study by the World Bank (1999) provide evidence that about 85 percent of companies in Malaysia had owner-managers: the post of CEO, chairman of the board or vice-chairman belonged to a member of the controlling family or a nominee with large shareholders typically owned more than 60 percent of shares. It is argued that extensive occurrence of individual and family run companies tends to discourage professionalism, encourage non-compliance and facilitate false accounting as well as to result in severe conflicts of interests (Halim, 2001). Thus, the following discussion focuses on the board structure and ownership structure literature to develop hypotheses on its relation to financial reporting quality BOARD STRUCTURE Advocates of agency theory believe that board comprising majority of outside directors reduce agency conflicts as they provide effective monitoring tool to the board (Fama and Jensen, 1983). They argue that the inclusion of outside directors 2 increases the boards ability to be more efficient in monitoring the top management and to ensure there is no collusion with top managers to expropriate stockholder wealth as they have incentives to develop their reputations as experts in decision control. Normally, the outside directors are expert managers from other large organizations and with its expertise, independence, objectivity and legal power, outside directors become potentially powerful governance mechanisms to mitigate agency costs and protect shareholders wealth (Li, 1994). The independent directors must be solely outside directors who have no other relationship with the companies except being on the board of directors. For Malaysian companies, independent director is defined in Chapter 1 of Listing Requirements of the Kuala Lumpur Stock Exchange (KLSE) as independent of management and free from any business or other relationship which could be interfere with the exercise of independent judgment or the ability to act in the best interests of an applicant or listed issuer (KLSE, 2001). The KLSE Listing Requirements 2001 also include independence from concentrated and holdings of shares in their definition of director s independence. The Malaysian Code on Corporate Governance views that good corporate governance rests firmly with board of directors and the Code required one third of the board to comprise of 2 Fama and Jensen (1983, p.20) contend that outside directors signal to internal and external markets for decision agents that (1) they are decision experts, (2) they understand the importance of diffuse and separate decision control and (3) they can work with such decision controls system. 4

6 independent non-executive directors in order to bring an independent judgment on the decision process. Efficient monitoring from non-executive directors that free from managerial influence is capable to improve the quality of financial information conveyed to the user of financial statement (Higgs Report, 2003). A number of studies in developed countries have reported a positive role of having higher proportion of independence non-executive directors sit on the board and financial reporting quality (Beasley, 1996; Dechow et. al., 1996; Peasnell et al., 2000; Klein; 2002, Davidson et al., 2005). As outside members do not play a direct role in the management of the company, their existence may provide an effective monitoring tool to the board and thus produce higher quality financial reports (Peasnell et al., 2000). However, evidence from countries with highly concentrated ownership structure is inconclusive. Kao and Chen (2004) and Jaggi et al. (2007) find significant negative evidence between earnings management and the presences of higher proportion of outside directors in Taiwan and Hong Kong sample which suggest that the inclusion of larger proportion of outside members on the board of directors provides better oversight of management to mitigate earnings management activity. Park and Shin (2004) however, fail to find empirical support on the association between earnings management and board independence for their Canadian sample where the ownership structure is highly concentrated and a large block holder controls the public traded firms in Canada. Additionally, study by Abdullah and Mohd Nasir (2004) and Abdul Rahman and Mohamed Ali (2006) also fail to find any significant evidence between independence of boards and earnings management in the Malaysian context. Furthermore, study by Jaggi et al. (2007) provides evidence of insignificant relationship between proportions of non-executive directors and accrual quality in high family-ownership samples of Hong Kong listed companies which suggest that the monitoring effectiveness of independent directors is reduced in family controlled firms. Despite this conflicting result, Malaysian evidence is re-tested and it is hypothesized that H 1 : Board independence is positively associated with earnings quality MANAGERIAL OWNERSHIP While having independent directors appear to be critical to the effectiveness of the boards monitoring function, the extent of management ownership held by management may affect control over the board. Jensen and Meckling (1976) theorize as management ownership increases, their interests will be more closely aligned with owners and the need for intense monitoring by the board should decrease. Bathala and Rao (1995) argue that the role of outside board members is less critical for firms with higher proportion of inside ownership. They find an inverse relationship between the proportion of outside board members and inside ownership of equity of 261 US listed firms which suggest that higher proportions of insider ownership held by inside board members help to closely align the managerial and shareholder interests, thus, reduce the need for intense monitoring from external board members. 5

7 While Bathala and Rao (1995) focus on inside members ownership, Beasley (1996) studies the effect of outside directors ownership and financial statement fraud. He provides evidence of significant negative relationship between outside directors ownership and the likelihood of financial statement fraud. His findings suggest that higher level of ownership held by outside directors do help reduce the likelihood of financial statement fraud. Additionally, Peasnell et al. (2005) include managerial share ownership as the intervening variable and posit that the constraining association between earnings management and the proportion of outside directors will be more prominent when the level of managerial share ownership is low. Prior researches by Peasnell et al. (1998 and 2003) find that managerial share ownership is significantly associated with the proportion of outside members. The demand for non-executive directors is lower in companies where the level of managerial ownership is high as shareholders let the management run the companies (Peasnell et al., 1998). Although Peasnell et al. (2005) find only slightly significant association on the three-way interaction of managerial ownership, outside director and income increasing abnormal accrual, this finding shed a light for future research on the interaction of managerial share ownership and the proportion of outside directors in constraining earnings management behaviour. Recognizing the importance of the tightness of ownership of Malaysian companies that influence the governance structure in Malaysia, Abdullah (2006a) investigate the influence of management and non-executives interest on the firm financial distressed of 86-matched sample of distressed and non-distressed companies for a period of Although fail to find evidence on the relationship between board independence and CEO duality on firm value, he found significant effect of management interests on firm value at the lower level and higher level of ownership. Additionally, paper by Abdullah (2006b) that extends Abdullah (2004) research on performance by investigating the extent to which firm s performance, internal governance of board of directors and ownership structure determine directors remuneration of Malaysian public listed companies. Although do not find association between directors remuneration and performance, he finds negatively significant evidence between board independence and the extent of nonexecutive director s interest with directors remuneration levels and suggest that the existence of these two governance mechanisms are effective in constraining the level of directors remuneration in Malaysia. Thus this study also examines the relationship of managerial ownership with accounting earnings quality and it is hypothesized that: H 2 : Managerial ownership is positively associated with earnings quality. Pergola (2005) argues that Beasley (1996) study does not distinguish between the inside and outside stock ownership to see the impact of management s ability to negate the board effectiveness. Therefore, this study also investigates the influence of managerial ownership on earnings quality separately for inside and outside board ownership. The hypotheses are: H 2A : Inside board ownership is positively associated with earnings quality. H 2B : Outside board ownership is positively associated with earnings quality. 6

8 2.3. FAMILY OWNERSHIP Study on the role of family ownership structure is critical to the effectiveness of corporate governance employed by the firms in Asia (Claessens and Fan, 2002). Unlike developed countries such as UK and US with dispersed ownership structure, Asian firms have more concentrated ownership structure where family control is common in both small and established firms (Mak and Kusnadi 2005). La Porta et al. (1999) reviews corporate ownership structure of 27 countries around the world and reveal that except in economies with very good shareholder protection such as United States, families or the state typically controls firms in countries with poor shareholder protection. The controlling shareholders often control the firms through pyramidal structures and have control rights in excess of their cash flow rights. These controlling shareholders even participate significantly in the management process and have the power to expropriate the minority shareholders and raises questions of minority shareholders protections. However, questions of whether family ownership provides incentives to reduce agency costs or create it still remain an open empirical issue. There are two contradictory views with regards to the relationship between family ownership and agency costs. On one hand, several researchers agree that concentrated shareholdings in the hand of family have incentives to reduce agency costs through a better alignment of shareholder and managerial interests. Bartholomeusz and Tanewski (2006) highlight several reasons as noted by prior researchers that favour family firms as agent to reduce agency costs. First, as the benefits and costs of the company are borne by the same person, family firms have more incentives to protect their wealth as it is tied directly to the welfare of the company. Second, family firms have greater expertise concerning firm s operations that make them in a better position to effectively monitor the firm s activities. Third, in order to protect the family s name and reputation, family firms adherence to maximize the long-term wealth of their firms. Fourth, given that the family member s are tied together create special and unique relationship that develops loyalty, efficient communications and effective decision making which in turn reduces the agency costs. However, according to Bartholomeusz and Tanewski (2006), other stream of literatures also draws attention to the possibilities that concentrated ownership by family firms create agency costs. First, family firms might use their concentrated blockholding to expropriate the wealth of outside shareholders through excessive compensation, relatedparty transactions and special dividends. Second, given that their wealth is undiversified, family firms tend to be risk avoidance where they might use their control to invest in less risky projects which are not aligned with other shareholders interest. Third, under the pyramidal control structure (which is common in family business group) family firms may create agency costs if the family members pursue the interest of other members at the expense of outsiders. Study by Bartholomeusz and Tanewski (2006) identify the relation between family control and corporate governance structure of 100 listed companies (i.e. 50 family firms and 50 matched non-family controlled firms) trading on the Australian Stock and found that family firms have lower proportions of independent directors, higher proportion of 7

9 grey directors and greater combination of role between CEO and Chairman compared to the non-family controlled firms which suggest that corporate governance structures adopted by family firms create agency costs as the structures are inconsistent with maximizing the value of the company. Study by Ng (2005) however finds a non-linear cubic relationship between ownership and performance in a family-based environment of listed companies in Hong Kong. She finds that at a relatively low level of family ownership (16.86 percent and below), managers who are normally the professional managers entrench their interests with the companies as their shareholdings are not substantial and tend to seek their self interest. However, at a wide spectrum of medium level of ownership (16.86 percent to percent), the result shows that the family managers align their interests with the companies and the performance is improve with greater ownership concentration. Again when the family managers gain a relatively high level of ownership beyond (63.17 percent), they become so powerful and entrench their interests with the companies to the detriment of minority shareholders interest. In similar environment, Jaggi et al. (2007) provide evidence that outside directors monitoring effectiveness is reduced in family controlled firms which results in lower quality of reported earnings. In the Malaysian context, Haniffa and Cooke (2002) test the relationship between family ownership and voluntary disclosure and find a negative significant coefficient between the proportion of family members on the board and the extent of voluntary disclosures in the annual report of Malaysian companies. Given that many family members sit on the board might be the reason of less demand for voluntary disclosure as they have better excess to inside information. Additionally, Mohd Ghazali and Weetman (2006) that extend prior research by Haniffa and Cooke (2002) also report similar finding between family domination and the extent of voluntary disclosure after the 1997 financial crisis. They argue that family-owned companies remain to be secretive even after the corporate governance reform that suggest they preserve a tradition inherited from the past and resist to change their attitudes towards greater voluntary disclosure at the point of regulatory change. Since there are two different theoretical viewpoints on the role of family ownership and agency costs, the direction of the next hypothesis is indeterminate: H 3 : Family ownership is associated with earnings quality INSTITUTIONAL OWNERSHIP Considering the influence of shareholder activism in governance reforms is important to obtain insight into governance practices (Daily et al., 2003). To date, institutional investors participation has emerged as important force in corporate monitoring to serve as mechanisms to protect minority shareholder s interest. The significant increase in the institutional investors shareholdings has led to the formation of a large and powerful constituency to play a significant role in corporate governance. Example, in US, the California Public Employees Retirement System (CalPERS) has been active in seeking greater director independence and in firms in which they invest their fund, CalPERS request for the firms to have majority independent directors to sit on the board and even identify the lead directors for chairman post and impose age limits on directors (Daily et 8

10 al., 2003). In the UK, institutional investors own between 65 to 75 percent of the United Kingdom stock market which suggest a prominent role that institutional shareholders can play as an agent to the governance systems (Mallin, 2003). In fact, Hermes Investment Management, owned by and is principal fund manager for the British Telecom (BT) Pension schemes manage over 75 billion euro representing equity investments in over 3000 companies worldwide and is the largest institutional investors in the UK (Lee, 2003). To mitigate the problems associated with conflict between controlling owners and minority shareholders in Asia firms, the involvement of institutional investors equity participation may improve corporate governance practices (Claessen and Fan, 2002). Concentrated shareholdings by institutional provide an incentive for diligent monitoring as they have the resources, expertise and stronger incentives to actively monitor the actions of management and prevent managers opportunistic behaviour (Wan Hussin and Ibrahim, 2003). Given they own substantial shareholdings that make it difficult to sell shares immediately at prevailing price, the institutional investors have greater incentives to closely monitor companies with high free cash flow (Chung et al., 2005). Extending prior research that look into the role of internal governance mechanisms and earnings management, Mitra and Cready (2005) provide evidence that active monitoring from the institutional investors also help to prevent managerial opportunistic reporting behaviour and improve the quality of governance in the financial reporting process. They find that institutional shareholders intervene and mitigate the self-serving behaviour of corporate managers in financial reporting based on a sample of 136 companies belong to the S&P 500 group and 237 belong to non- S&P 500 category for eight years period ( ). However, it is depend to the relationship between the institutional investors and the controlling owners as the rent seeking and relationship based transaction may avoid the institutional to monitor the controlling owners and not force them to disclose all information to the public as their value will also be affected (Claessen and Fan, 2002). Chung et al. (2005) find evidence that institutional shareholder moderate the discretionary accrual and surplus free cash flow relationship when the surplus free cash flow is high. The presence of institutional investors with substantial shareholdings restrain managers from engaging in income increasing discretionary accruals when companies have high free cash flow, however, when there is no free cash flow agency problems, the institutional investors do not effectively constrain the management s use of income increasing discretionary accrual. In Malaysia, total institutional shareholdings stood at about 13 percent of the total market capitalization of Bursa Malaysia (for year 2003) that account for higher percentage of institutional shareholdings compared to other countries in the same region (Abdul Wahab et al., 2003). In fact the Employee Provident Fund (EPF) accounts for 86 percent of the total RM173 billion of provident and pension fund (Thillainathan, 1999). Although the institutional shareholdings are growing in the Malaysian capital market, empirical evidence on the effect of institutional shareholding and accounting issues are very limited. Abdullah (1999) is first to examine the relationship between institutional shareholdings and accounting earnings quality (measured by earnings response coefficient) and find 9

11 evidence that the presence of institutional investors lead to lower earnings quality as predicted in their hypothesis. He argues that Malaysian institutional investors prefer short-term investment rather than long-term achievement that that make their decision to dispose their substantial shareholdings inevitably depress the market share price dramatically that support myopic investor hypothesis. However finding by Abdullah (1999) may be arguable for recent capital market development that shows greater institutional investors participation as corporate monitoring. Institutional investors in Malaysia nowadays have become a very large and powerful constitution that plays a very significant role in corporate governance to protect minority shareholder s interest. Recent study by Abdul Wahab et al. (2004) provide evidence of a negative and significant monodirectional causality that run from institutional ownership to performance which suggest that institutional shareholding is a determinant of poor performance but poor performance is not a determinant of institutional ownership. Additionally, they find that institutional investors use corporate governance practice as a measuring tool for their investment decisions which suggest that firms with better corporate governance practices attract higher institutional ownership from institutional investors. For this study it is hypothesized that: H 4 : Institutional ownership is positively associated with earnings quality. 3. METHODOLOGY 3.1. SAMPLE SELECTION As at the end of year 2004, there are 622 financial and non-financial companies listed on Bursa Malaysia s Main Board. Due to different statutory requirements, all banks, insurance and unit trusts companies were excluded from the population of interest. In addition, utilities companies were also excluded from the population of interest because they possess different incentives and opportunities to manage earnings (Peasnell et al, 2000; Abdul Rahman and Mohamed Ali, 2006). After eliminating 55 financial companies and 2 utilities companies, the sample size is reduced to 592 non-financial companies. As there is a strict data requirements for the accrual quality estimation (to calculate earnings quality measure for year 2004, seven years complete accounting data, t = ), the sample is further reduced to 426 non-financial companies. 8 companies belong to industries with less than 10 observations were also eliminated from the analysis. Earnings quality accounting data is obtained from Perfect Analysis Database and any missing financial data were collected manually from respective annual report. Information pertaining to boards of directors characteristics and ownership structure were obtained by examining the disclosures made in annual reports. The final sample of this study comprised of 280 non-financial companies with complete data for earnings quality and corporate governance variable REGRESSION MODEL 10

12 This study used a linear multiple regression analysis to test the association between the dependent variable of earnings quality and the independent variable of board independence, managerial ownership, family ownership and institutional ownership. MODEL 1 EQ = β 0 + β 1 BIND + β 2 MGRLOWNS + β 3 FAMILY + β 4 INSTITUTIONAL + β 5 SIZE + β 6 LEV + β 7 ROA + ε Where: EQ BIND MGRLOWNS FAMILY INSTITUTIONAL SIZE LEV ROA ε = measured by accrual quality based on modified Dechow and Dichev (2002) model 3 = no. of independent NEDs / total no. of boards members = percentage of shares held by independent non-executive directors, executive directors and non-independent non-executive directors = proportion of family members to total number directors on the board = proportion of shares owned by institutional investors to total number of shares issued = log of total asset = the ratio of total liabilities to total assets = the ratio of net income to total assets = error term Model 1 is used to test hypotheses H 1, H 2, H 3 and H 4. In order to test H 2A and H 2B, the following model is estimated: MODEL 2 EQ = β 0 + β 1 BIND + β 2 INSOWNS + β 3 OUTOWNS + β 4 FAMILY + β 5 INSTITUTIONAL + β 6 SIZE + β 7 LEV + β 8 ROA + ε Where: All variables are measured as in MODEL 1 except: INSOWNS OUTOWNS = Percentage of shares held by the inside board members including executive directors and non-independent non-executive directors = Percentage of shares held by independent non-executive directors As consistent with prior studies, this study include firm size, leverage and return on assets as control variables in the regression model. The log of total assets was used to measure firm size and to control for the firm size effect (Jaggi et al., 2007). Leverage was used to 3 Modified Dechow and Dichev (2002) model is explained in section

13 control firms that are currently facing financial difficulties and return on assets was used to control firms with different performance (Abdul Rahman and Mohamed Ali, 2006) CORPORATE GOVERNANCE VARIABLE Board Independence (BIND) is measured by the proportion of independent non-executive directors on the board, expressed as a percentage. In the context of Malaysia, there are three types of directors including independent non-executive directors, non-independent non-executive directors and executive directors. Given that some non-independent nonexecutive directors are independent, whereas others are not (non-independent nonexecutive directors are sometimes a family member), following prior work by Che Haat (2006), this study focuses solely on independent and non-independent directors instead of executive and non-executive directors. Managerial Ownership (MGRLOWNS) is measured using percentage of shares owned by independent non-executive directors, executive directors and non-independent nonexecutive directors. The measure includes both direct and indirect interests in the company (Che Ahmad et al., 2003). To provide a more comprehensive analysis, the variable Managerial Ownership is further segregated into inside and outside ownership. Inside Board Ownership (INSOWNS) is measured using percentage of shares held by the inside board members including executive directors and non-independent non-executive directors while Outside Board Ownership (OUTOWNS) is measured using percentage of shares held only by independent non-executive directors. Family ownership (FAMILY) is measured using ratio of family members on the board to the total number of directors (Haniffa and Cooke, 2002 and Mohd Ghazali and Weetman, 2006). Each listed companies in Malaysia are required to disclose in the director s information in the annual report any family relationship with any directors and/or substantial shareholders of the company. Institutional Ownership (INSTITUTIONAL) is measured using proportion of shares owned by 5 largest institutional investors to total number of shares issued (Abdul Wahab et al., 2007). The 5 largest institutional investors include Employee Provident Fund (EPF), Lembaga Tabung Angkatan Tentera (LTAT), Lembaga Tabung Haji (LTH), Social Security Organization (SOCSO) and Permodalan Nasional Berhad (PNB) 3.4. ACCRUAL QUALITY VARIABLE To measure earnings quality, this study applies modified Dechow and Dichev (2002) accrual quality model by Francis et al. (2005) which has recently been considered as a better proxy for earnings quality (Jaggi et al., 2007). This measure is based on the observation that accruals map into cash flow realizations and regardless of managerial intent, the accrual quality is affected by the measurement error in accruals. In Dechow and Dichev (2002) approach, the estimated residuals from firm specific regressions of working capital accruals on past, present, and future cash flow from operation capture total accruals estimation error by management and are viewed as an inverse measure of 12

14 earnings quality. Francis et al. (2005) extend the Dechow and Dichev original accrual quality model by adding two additional variables, i.e. change in revenue and property, plant and equipment (PPE) for more complete characterization of the relation between accruals and cash flow. McNichols (2002) in her discussion paper of the quality of accruals and earnings shows that combining original Dechow and Dichev model and Jones model variables significantly increases the explanatory power of accrual quality model. ΔTCA j,t = ϕ 0,j + ϕ 1,j CFO j, t-1 + ϕ 2,j CFO j, t + ϕ 3,j CFO j,t+1 + ϕ 2,j ΔREV j, t + ϕ 2,j PPE j, t + ν j,t Assets j,t Assets j,t Assets j,t Assets j,t Assets j,t Assets j,t Where: ΔTCA j,t = Firm j s total current accruals in year t, = (ΔCA j,t - ΔCL j,t - ΔCash j,t + ΔSTDEBT j,t ); ΔCA j,t = Firm j s change in current assets between year t-1 and year t; ΔCL j,t = Firm j s change in current liabilities between year t-1 and year t; ΔCash j,t = Firm j s change in cash between year t-1 and year t; ΔSTDEBT j,t = Firm j s change in debt in current liabilities between year t-1 and year t; Assets j,t = Firm j s average total assets in year t and t-1; and CFO j,t = Firm j s net cash flow from operation in year t. ΔREV j,t = Firm j s change in revenues in year t-1 and t; and PPE j,t = Firm j s gross value of PPE in year t. For each firm-year, both equation is estimated cross-sectionally for all firms (minimum 10 firms within each industry groups) using rolling 7-year windows. These estimations yield 5 firm- and year-specific residuals, ν j,t, t = t-5, t, which form the basis for accrual metric. Accrual Quality j,t = σ (ν j,t ), equal to the standard deviation of firm j s estimated residuals. Larger standard deviations of residuals correspond to poorer accrual quality and vice versa. Following DeFond et al. (2007) the standard deviation score is multiplied by -1 so that higher score indicate higher earnings quality (EQ). 4. RESULTS AND DISCUSSIONS 4.1. DESCRIPTIVE STATISTICS Table 1 Descriptive Statistics for Dependent and Independent Variables Mean Median Minimum Maximum AQ BIND MGRLOWNS INSOWNS

15 OUTOWNS FAMILY INSTITUTIONAL LGSIZE LEV ROA Table 1 presents the descriptive statistics on the variables used in the regression tests. The mean and median of accrual quality is and respectively. On average, 41.4 percent of directors are independent non-executive directors which suggest a domination of inside directors in the majority of Malaysian listed companies. Although MCCG 2000 required for the companies to have at least one third of the board to comprise of independent non-executive directors, there are 13.2 percent of the companies does not fulfill the requirements of the Code. With respect to the managerial ownership, the percentage is ranged from zero to 82.26% with average value of 28.11%. This average is similar to the study done by Che Ahmad et al. (2003) and Vethanayagam et al. (2006) for Malaysian listed companies. Segregating managerial ownership into inside and outside ownership reveal that significant amount of ownership is held by the inside directors. Independent non-executive directors in Malaysia only hold only small percentage of ownership ranged from zero to 7.76 percent while executive directors and nonindependent non executive directors hold up to percent. In terms of family domination, the proportion is varied from zero to about 71 percent, with average proportion of family members of about percent. There are almost equivalent numbers in terms of family and non-family controlled firms in the sample where 50.4 percent have no family members sit on corporate board while 49.6 percent have at least two or more family members sit on corporate board. The percentage of institutional shareholdings for the sample ranged from zero to percent, with average shareholdings of about 3 percent. Table 2 Correlations among Variables EQ BIND MGRLOWNS INSOWNS OUTOWNS FAMILY INSTITUTIONAL LGSIZE LEV ROA EQ 1 BIND MGRLOWNS * 1 INSOWNS *.999 ** 1 14

16 OUTOWNS **.144 * 1 FAMILY.224 ** **.480 **.477 **.154 ** 1 INSTITUTIONAL * * LGSIZE * ** ** *.252 ** 1 LEV ROA **Correlation is significant at the 0.01 level; * Correlation is significant at the 0.05 level To examine the correlation between the independent variables, a Pearson product moment correlation (r) was computed. The overall correlations among the explanatory variables were relatively low and below 0.5 except the correlation between managerial ownership and inside ownership. As expected, the correlation between managerial and inside ownership is positive and high as managerial ownership is dominated by the inside management. The correlation matrix confirms that multicollinearity is not a problem in this study MULTIVARIATE ANALYSIS Table 3 Regression Results MODEL 1A MODEL 1B MODEL 2A MODEL 2B (Constant) *** *** *** BIND MGRLOWNS INSOWNS OUTOWNS FAMILY 4.047*** 4.044*** 3.929*** 3.933*** INSTITUTIONAL * * LGSIZE * LEV ROA R F-Statistics 3.348*** 5.133*** 3.095*** 4.321*** N ***Significant at 0.01level; **Significant at 0.05 level; *Significant at 0.1 level 15

17 Table 3 presents the results of multivariate regressions used to test the hypotheses stated earlier. This study used two different models in estimating the relationship between earnings quality and the test variables. In the Model 1, the test variable of managerial ownership is considered while in the Model 2, the managerial ownership variable is segregated into inside ownership and outside ownership. Model 1A and Model 2A test the relationship with control variables while Model 1B and Model 2B test the relationship without control variables. All the multiple regression models tested in this study reported a high F-value (all significant at the 0.01 level). In all four models, the coefficient for board independence was not significant. This finding is however consistent with prior study in Malaysia by Che Ahmd et al. (2003), Abdullah (2004), Abdul Rahman and Mohamed Ali (2006) and Vethanayagam et al. (2006). Che Ahmad et al. (2003) suggest that with regards to diversification strategy, the presence independent directors does not seem to influence the decision process either because they do not have contact with daily operation of the firm or they have limited qualifications and merely appointed based on the relationship with the CEO of the firm. Similarly, Abdullah (2004) and Abdul Rahman and Mohamed Ali (2006) argue that the capability of independent directors to fulfil their monitoring role is jeopardised when the management also dominates and control the boards. Due to the dominant role played by CEOs in the director selection process, it is argue that outside directors are incapable to provide independent judgment and raise concern about the quality of independent directors (Abdullah, 2004). Abdul Rahman and Mohamed Ali (2006) and Vethanayagam et al ) also argue that Malaysian independent directors lack expertise, skills and knowledge to understand financial reporting details that explain their insignificant findings on the relationship between board independence and accounting issues they examined. Perhaps, the most important issue addressed by Vethanayagam et al. (2006) is domination of inside directors on the board in Malaysia that bring into questions the quality and accountability of independent directors when some independent directors are not truly independent of management. With regard to the managerial ownership, this study fails to find any significant association between managerial ownership and earnings quality in Model 1A and Model 1B. Segregating managerial ownership into inside and outside ownership also reveal insignificant findings on the relationship between inside and outside ownership with earnings quality. This result is consistent with findings by Che Ahmad et al. (2003) that fail to find any significant evidence on the relationship between managerial ownership and diversification. However, study by Abdullah (2006a) finds that the effect of management ownership on financial distressed status is curvilinear and negatively associated with financial distressed. Similarly, Vethanayagam et al. (2006) also report a non-linear relationship between managerial ownership and firm performance. Results (unreported) to see whether the effects of managerial ownership on earnings quality is curvilinear in this case also reveal insignificant findings thus suggest that increases in managerial ownership does not enhance the financial reported earnings predicted by the agency theory. 16