EXECUTIVE COMPENSATION, OWNERSHIP STRUCTURE AND DIVIDEND PAYOUT: EVIDENCE FROM MALAYSIA

This study seeks to present the relationship between executive compensation, dividend payout policy and ownership structure of listed firms in Malaysia. We examine a panel data on a sample of 300 of the largest Malaysian public listed companies (PLCs) on Bursa Malaysia for the years 2008 to 2014. Based on 2,009 firm-year observations, our results show consistent empirical positive evidence on the association between dividend payout and executive compensation across all models. However, the results are inconsistent with Bhattacharyya model of dividend payout. Further, there is evidence that government and family ownerships are positively associated with dividend payout. Our results show that the positive relationship between executive compensation and dividend payout is more evident in politically connected firms (PCON firms) which is consistent with the clientele (catering) theory.


INTRODUCTION
Malaysian business survey 2013 reveals that, in 2012, Malaysia's economy was the third largest in Southeast Asia, and the 29th largest in the world by purchasing power parity, with a gross domestic product (GDP) of USD492.4 billion and per capita income of USD16,922. The survey covered 603 companies and the total payout to directors was RM2.3 billion. For the top 20 paymasters alone, the total directors' payout was RM569.5 million. 1 Further, the business periodical FocusM (9-15 April 2016), revealed that the average remuneration of an executive director grew 14.2% to RM1.53 million (2014: RM1.34 million) while according to the Malaysia-ASEAN Corporate Governance Report 2015, the non-executive directors saw an average fee surge 19.4% to RM117,000. Therefore, Malaysians are concerned about above-average compensation and the growth rate of pay for senior executives. At the same time, Malaysians are concerned about the returns of their investment in the Malaysian stock market and whether these high executive compensations affect the dividend payouts in Malaysia.
Shareholders have had little public success in forcing company boards to change executive pay arrangements, and often their complaints about excessive handouts have fallen on deaf ears (StarBiz, 26 January 2017). Another contentious issue is where, according to a Wall Street Journal, dated 4 October 2016, directors at some companies are paid to lobby for their firms or allied trade groups, while also helping to set the Chief Executive Officer's (CEO) pay. For example, at Louisiana health-care company, LHC Group Inc., the board's compensation committee approved a 90% raise for the chief executive over two years and also gave the director personal use of the company plane. Thus, the inequality in income between the rich and poor is widening in many countries including, the U.S., U.K. and Malaysia, and the compensation gap between bosses and employees is receiving a lot of attention in the mass media.
In Malaysia, for many listed companies, higher executive compensation does not necessarily commensurate with better financial performance, however, despite falling revenues and profits, there are approximately 50 loss-making companies that continue to offer handsome windfalls to their directors (FocusM, 28 May-3 June 2016). Despite the significant payouts to directors and abundant literature on executive pay, there is scant evidence on the relationship between executive compensation and the dividend payout policy of listed firms in emerging capital markets. Bhattacharyya, Mawani and Morrill (2008) suggest that dividends can be used to resolve agency issues in managerial compensation contracts in developed countries. Particularly, in equilibrium, their study suggests that dividend payout ratios should be negatively related to managerial compensation which is a relationship not systematically investigated in developing countries where concentrated ownership and principal-principal agency conflicts (Agency Theory Type II) are prevalent (Bhattacharyya, Elston, & Rondi, 2011).
In an emerging economies context, Young, Peng, Ahlstrom, Bruton and Jiang (2008) opine that instead of traditional principal-agent conflicts espoused in most research dealing with developed economies, principal-principal conflicts have been identified as a major concern of corporate governance in emerging economies. Their study envisages that the principal-principal conflicts between controlling shareholders and minority shareholders result from concentrated ownership, extensive family ownership and control, business group structures, and weak legal protection of minority shareholders. Notably, the focus of their research is on strategy, finance, and economics on principal-principal conflicts with an emphasis on their institutional antecedents and organisational consequences.
Further, prior studies find that pay for performance is weaker at a higher level of managerial ownership, and envision that both the principal-agent (Agency Theory Type I) and managerial power views' are relevant to explaining executive pay (Reddy, Abidin, & You, 2015). In addition to this, Benjamin, Mat Zain and Abdul Wahab (2016) posit that Malaysia's political system affects the severity of agency problem between insider and outsider shareholdings and this could be another unique area of study where the CEO is being politically connected (Kasipillai, Mei Yee, & Mahenthrian, 2017;Minhat & Abdullah, 2014;Tee, Gul, Boon Foo, & Teh, 2017). Given the existence of concentrated managerial ownership and politically connection in Malaysian public listed companies (PLCs), it is very likely that the board of directors want to compensate the insider owners, which may cause the pay for performance relationship to be weak which in-turn influences dividend payouts.
Executive compensation is a controversial subject that has attracted the attention of legislators, the media and the academicians in the U.S., U.K. and Portugal (Bebchuk & Fried, 2003;Conyon & Murphy, 2000;Alves, Couto, & Francisco, 2016). Alves et al. (2016) posits that in Portuguese firms, specific firm factors such as shareholders return, firm characteristics, CEO characteristics, and the profile of board of directors can account for the majority of the variance in total CEO pay. In New Zealand, a study finds that after controlling for size, performance, industry and year effects, the internal features rather than external corporate governance practices influence the CEO's compensation (Reddy et al., 2015). Studies also find that companies that have CEOs on boards have the power to influence board decisions, which causes the boards to become less 50 effective in monitoring CEO compensation (Lee & Isa, 2014;Reddy et al., 2015;Benjamin & Mat Zain, 2015). Further, Conyon and He (2012) document that in China, CEO pay is positively correlated to both accounting and stock market performance, although the link to accounting performance is more robust. Their study also posit that CEO pay dynamics are important as pay in the current year is significantly positively correlated to CEO pay the previous year. In summary, there are three pressing issues for our study of executive compensation in Malaysia. Firstly, executive compensation has increased significantly in the last few decades and there is criticism surrounding the rise in pay, secondly, there is a widely held perception that the executive compensation is insufficiently linked to firm performance. Lastly, the fact executive compensation has outpaced the growth of most other incomes.
Hence, in this study, we aim to examine how the ownership structure impact the relationship between executive compensation and dividend payout and we draw on the agency theory put forth by Jensen andMeckling (1976), Faccio, Lang andYoung (2001) and Setia-Atmaja, Tanewski and Skully (2009) as well as others to develop and test the hypotheses. Overall, there is also a strand of literature which focuses on how executive incentives influence various corporate policies including equity share issuance decisions, pension plans, and cash holdings (Anantharaman & Lee, 2014;Brisker, Colak, & Aitore, 2014), and this study makes an important contribution to extending this stream of research by examining if a firm's dividend payout reflects agency conflicts between concentrated shareholders. Bhattacharyya et al. (2011) advance a theory base on the agency paradigm to expound the dividend puzzle and find that payout ratios are negatively related to managerial compensation. Further, their study also finds that the predicted negative relationship between dividend payout ratio and executive compensation holds in both the U.S. and Canada that are two developed countries with similar legal frameworks. In the context of an emerging market, this study contributes to the growing scholarly work that examines the role of ownership on corporate dividend payout.

LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT
This section reviews the literature on executive compensation and dividend payout and how the firms' ownership structure affects them.

Dividend Payout and Executive Compensation
The question of why companies pay dividends has continued to puzzle researchers, for an extensive review of the literature (Gill & Obradovich, 2012). This literature is based on the assumption that dividend policy is value-relevant for shareholders in emerging capital markets, and is contrary to the prediction of the dividend irrelevance theory put forth by Miller and Modigliani (1961). H1 is developed via consideration of three theories that contribute to explaining why dividend policy may potentially be value relevant using the signalling theory, the life cycle theory, and clientele (catering) theory. These three theories are not mutually exclusive, which helps us to contribute to these dividend theories in finance.
Malaysia has a unique institutional setting; hence, previous evidence from Anglo-American jurisdictions, may not generalise to Malaysia. Harris (2008) and Perel (2003) posit that the business ethics literature raises concerns about unethical and unreasonable compensation policies that deprive shareholders of their fair share of a company's wealth. Smith and Watts (1992) argue that, after controlling for the effects of growth opportunities and firm size, lower dividend yields are associated with higher levels of executive compensation because of the link between a firm's financing and dividend policies. Gaver and Gaver (1993) also show that the results from Smith and Watts (1992) study are consistent at the firm level. On the other hand, Golec's (1994) study provides evidence from the perspective of a real estate industry, which finds that, in typical wage contracts, the total compensation is associated with higher dividend yields than discretionary based compensation. Another study on the banking industry in China, Sakawa, Watanabel and Tanahashi (2017), that focus on the roles of executive incentive compensation packages after the financial deregulation era, find that bonuses tend to be given to executives who achieve higher performance. Their study also find that executive bonus packages at Shinkin banks encourage managers to make greater efforts to ensure their banks achieve higher profit than the regional average. Similarly, White (1996) finds the existence of a dividend incentive in the compensation plan for the oil and gas, the defense/aerospace, and the food processing industries. White expound that management compensation is positively associated with higher dividend payouts and yields, and higher annual changes in dividend levels. White's (1996) study also provide evidence of the association between firm characteristics and the use of compensation contract with a dividend provision, which led him to conclude that his results are consistent with the theory that firms' link compensation incentives to dividend payouts to reduce agency conflicts between shareholders and management.
Another study of 1,650 public listed firms in U.K., Germany, France, Italy, the Netherlands and Spain from 2002 to 2009 by De Cesari and Ozkan (2015) find that director's executive stock option holdings and stock option deltas have a negative impact on the total payout, suggesting that executives do not substitute share repurchase for dividends. Hence, their study suggests that the director's executive share ownership and stock-based pay performance may help mitigate agency conflicts by significantly increasing the level of the total payout. Thus, suggesting that the director's executives do not substitute share repurchases for dividends and that they are used for different purposes. Nevertheless, share repurchases are not common in Malaysia and hence that enables us to focus solely on the dividend payout as the primary means to reduce the vertical agency conflict between shareholders and management.
Theories suggesting that dividend matters to shareholders include the clientele effect (Litzenberger & Ramaswamy, 1982;Miller & Modigliani, 1961); signalling theory (Bhattacharya, 1979); corporate governance and investor protection theory (Mitton, 2004); free cash flow hypothesis (Easterbrook, 1984;Jensen, 1986); and clientele (catering) theory (Baker & Wurgler, 2004). Of these theories, Bhattacharyya et al. (2008) study is particularly relevant as this study finds that dividends payout is negatively associated with the executive compensation, and this result holds when the payout is defined as common dividends or common share repurchases. Further, Bhattacharyya et al. (2008) has advanced a theory base on the agency paradigm that dividends are used to resolve agency issues in managerial compensation contracts and equilibrium, and thus dividend payout ratios are negatively related to managerial compensation (Bhattacharyya et al., 2011). Therefore, we state the below hypothesis: H1: Ceteris paribus, dividend payout is negatively associated with executive compensation in Malaysian PLCs.

Ownership Structures
This section reviews the literature on ownership structures (categorised under three main components, i.e., GLCs, family ownership, and institutional ownership) and provides the moderating effect on the relationship between director's compensation and dividend payout.

GLCs
Although the literature on the relationship between government ownership and corporate dividend policy is as yet unexplored in Malaysia, there is growing evidence in the corporate governance literature that suggests a linkage between institutional investors exerting pressure on managers and its impact on managers' compensation (Brickley, Coles, & Jarrell, 1997;Pound, 1988;Bushee, 1998). There is anecdotal evidence that suggests a positive association between government ownership and dividends since firms with government ownership have relatively less difficulty raising funds to finance their investments and pay dividends given that they can borrow from government-owned banks on preferential terms.
Michaely and Roberts (2011) posit that ownership structure and incentives play key roles in shaping dividend policies. In their study, they find that public-listed firms pay relatively higher dividends and that they tend to be more sensitive to changes in investment opportunities than otherwise similar private firms. Further, Jensen and Meckling (1976), argue that owner-managers tend to forego the benefits of diversification and retain large equity stakes in their firms when there are acute agency problems, which can lead to high monitoring and contracting costs. In another study, Al-Najjar and Kilincarslan (2015) find that in Turkey, foreign and state ownerships are associated with less likelihood of paying dividends. On the contrary, Setiawan, Bandi, Phua and Trinugroho (2016) find that in Indonesia, government-controlled firms are positively associated with dividend payout.
In examining ownership concentration of Malaysian PLCs in 1998, Abdul Samad (2002) find the mean ownership of the largest shareholder and the five largest shareholders to be about 30% and 60%, respectively. Malaysian PLCs ownership is less diffused and more concentrated in the hands of families or the government. La Porta, Lopez-De-Silanes, Shleifer and Vishny (1998) note that internationally the use of pyramiding and insider management appointments and cross-ownership arrangements are used to create a divergence between cash flow rights and voting rights. In Malaysia, at the 20% ownership cut-off level, by way of market capitalisation, state ownership becomes much more pronounced at 34.8%, and the presence of widely-held financial institutions and corporations are diminished. Thillainathan (1999) confirmed cross-holdings of share ownership or pyramiding is quite common in Malaysia, which results in the divergence between cash flow rights and voting rights that affect the relationship between insider ownership, firm performance and dividends.
On the aspect of executive compensation in government-linked companies (GLCs), Minhat and Abdullah (2014) find that executive pay is lower in GLCs and that the positive pay-performance relationship is not always evidenced in the GLCs. This suggests that in Malaysian GLCs, the executives are largely guaranteed with a certain level of pay irrespective of the performance of their companies. This finding is consistent with the inefficient pay hypothesis 54 developed by Minhat and Abdullah (2014). On the contrary, (Conyon & He, 2011), a study in the Chinese context, find that consistent with agency theory, the executive compensation is positively correlated to firm performance. Their study also show that executive pay and CEO incentives are lower in state controlled firms and firms with concentrated ownership structures Hence, it is not surprising that the level of equity ownership incentives provides the executives in government-controlled firms with very little incentive to exert effort to improve the firm's performance, and consequently, the dividend payout is largely base on the availability of retained profits and distributable reserves. Therefore, we state the following hypothesis: H2: Ceteris paribus, there is a negative association between GLC ownership and dividend payout in Malaysian PLCs.

Family ownership
Family firms play a significant role in the global economy and are prevalent around the world. Family businesses generate an estimated 70%-90% of the global GDP annually (Family Firm Institute, 2016). Of all enterprises in North America, 80%-90% are family firms, contributing 64% of the GDP and employing 62% of the U.S. workforce (European Family Business, 2016). Family businesses constitute 65%, 75%, 75% and 80% of the total companies in the U.K., Germany, France and Finland respectively (European Family Business, 2016). Similarly, prior studies indicate that over 35% of S&P 500 firms and about 46% of the S&P 1,500 firms are family firms (Anderson & Reeb, 2003;Chen, Cheung, Stouraitis, & Wong, 2005).
In this study, we focus on whether ownership matter in an emerging market like Malaysia and empirically analyse how ownership impacts dividend payout. We select Malaysia for this study because of its rich institutional settings. Malaysia has been the focus of several studies (Bliss, Gul, & Majid, 2011;Bliss & Gul, 2012;Fung, Gul, & Radhakrishnan, 2015) because of its unique institutional features. The ownership of Malaysian PLCs is predominantly owned by at least a family and based on our dataset of 300 PLCs for a period from 2008 till 2014, 27% of the total firms are in the hands of family ownership and comparatively, the percentage of family-controlled firms in the U.S. is 35% (Anderson & Reeb, 2003), Europe 44.29% (Faccio & Lang, 2002) and East Asia 43.60% (Faccio et al., 2001). (2015) find t hat o ther o wnership variables such as family involvement are insignificant in affecting the probability of paying dividends in Turkey. However, Bhattacharyya et al. (2011) provide evidence that a large number of Italian firms are family-controlled and such control plays a significant role in resolving agency issues and that the increase in family control firms, leads to higher dividend payments. Bhattacharyya, base on the familycontrolled environment, they find that managerial compensations are negatively related to dividend payout ratios. Interestingly, Sakawa and Watanabel (2019) find that family board members do not exploit minority shareholders and rather behave as stewards of the firm. In their study, the authors also find that foreign shareholders interact with family control to increase firm profitability, suggesting that foreign shareholders enhance the role of family board members as stewards.

Al-Najjar and Kilincarslan
In another study base on the Malaysian context, Benjamin et al. (2016) find that family share ownership at the dispersed level of 0% to 5% is negatively associated with dividend payout and positively associated with dividend payout at a dispersed level of 5% to 33%. Further, Setiawan et al. (2016) find that family ownership has a negative effect on dividend payout as they argue that family firms in Indonesia pay lower dividends because they prefer to control the dividend payout and engage in the expropriation of minority shareholders. Chen et al. (2005) find a minimal relationship between family ownership and dividend payouts in Hong Kong where they evidence a negative relationship for family ownership, up to a level of 10% of the equity, indicating possible shareholder expropriation. However, at a higher ownership percentage of between 10% to 35%, the relationship is positive. Taking a cue from the rich institutional setting of Malaysia and the mixed evidence, we are motivated to analyse the role of family ownership in executive compensation and dividend payout. In Malaysia, the family ownership is substantial, yet short of triggering the mandatory general offer at 33.33% of equity shareholdings and also base on Bhattacharyya et al. (2011) and Chen et al. (2005) that evidence no shareholder expropriation at a higher percentage level of family shareholdings, we state the following hypothesis: H3: Ceteris paribus, there is a positive association between dividend payout and family ownership

Institutional ownership
Numerous studies have reported drastic increases in global institutional ownership in the corporate sector (Brandes, Goranova, & Hall, 2008;Gillan & Starks, 2003;Tee et al. 2017). According to Khazanah Nasional (2015), institutional investors contribute 76.2% to the total value of equities traded, while retail investors 56 account for only 23.8%. Hence, from the total figure of 76.2%, 43.9% and 32.2% are contributed by local institutional investors and foreign institutional investors respectively. Under the institutional theory, those normative aspects of control are carried out by institutional investors who have the potential to influence managerial decisions and firm performance directly through their share ownership and indirectly through their ability to trade the firm's shares in the open market (Gillan & Starks, 2003).
Sakawa and Watanabel (2020) based on TOPIX 500 in Japan and using 2,924 firm-year observations, is a study of the effect of institutional investors on firm performance. Their study is analysed to test the role of institutional investors in stakeholder-oriented corporate governance. The outcome of their results find that monitoring role of institutional shareholders, or foreign shareholders, functions efectively in Japanese corporations. Their study also show that the monitoring roles of institutions are expected to strengthen firms through higher growth opportunities and hence implies that institutional shareholders contribute to enhancing sustainable firm performance and constructing sustainable corporate governance mechanisms in a stakeholder-oriented system.
Mazna Ramli (2010) posits that Malaysian companies make higher dividend payout as the shareholding of the largest shareholding increase and the magnitude of dividend payout is also larger when there is a presence of the substantial second-largest shareholder in the company. A study in Sri Lanka posits that there is a significant positive association between concentrated ownership structure and dividend policy (Kulathunga & Azeez, 2017). Another study by Benjamin et al. (2016) finds that institutional ownership is associated with higher dividend payout. However, in the context of Turkey (Al-Najjar & Kilincarslan, 2015;Jacob & Lukose, 2018), the studies find that other ownership variables such as domestic financial institutions; total institutional ownership and minority shareholders are insignificant in affecting the probability of paying dividends. To date, prior results are mixed. Amidu and Abor (2006) in their studies find a negative association between dividend payout and institutional holding. Similarly, Grinstein and Michaely (2005) find that higher institutional holdings or a concentration of holdings do not cause firms to increase their dividends.
In this study, we examine whether institutional investors play a monitoring role in Malaysia PLCs that in-turn affects executive compensation. Our study builds on the strand of research on institutional investor monitoring to focus on the effects of institutional investors' monitoring on executive compensation (Shin & Seo, 2011;Conyon, 2013). Accordingly, we state the following hypotheses: H4: Ceteris paribus, there is positive association between institutional ownership and dividend payout.

Sample Selection
The sample consists of 300 of the largest companies listed on Bursa Malaysia for the seven years, 2008 to 2014. The data is collected from the Malaysian bourse, Bursa Malaysia, OSIRIS, DATASTREAM, and the BANKSCOPE databases as well as from the Malaysian Stock Performance Guide. Further, we obtained the ownership data and control variables from the OSIRIS and Bloomberg databases, as well as from company annual reports available on the Bursa Malaysia website.
The detail descriptions of the variables-labels, definitions, and measurements are provided in the variable measurement template (refer to Appendix A). After eliminating the missing data, the sample size is reduced to 287 firms and based on 2,009 firm-years (refer to Table 1). The industry distribution depicts that the majority of observations are from the properties/hotel sector (29.27%), closely followed by the trading/services sector (23.69%) and industrial products sector (22.66%), respectively from the sample size.

Dependent Variable
Gaver and Gaver (1993) use the dividend payout ratio and the dividend yield as two measures of dividend policy. The dividend payout ratio is the dividend per share divided by primary earnings per share before extraordinary items, and the dividend yield is the dividend per share divided by the closing price per share. 58 The dividend yield is sensitive to share prices whereas the dividend payout is not. For this reason, the dividend payout ratio is the primary measure of the firm's dividend payout in this study. Similarly, other studies (Smith & Watts, 1992;Gaver & Gaver, 1993;Gul, 1999;Adam & Goyal, 2008;De Cesari & Ozkan, 2015;Benjamin et al., 2016) have also used this measure. In this study, the dividend payout is measured as the dividend per share scaled by earnings per share before extraordinary items.

Ownership Variables
GLCs define as companies whose major ownership and control are held by the main shareholder that is either a government agency (such as Khazanah Nasional, Ministry of Finance, Bank Negara Malaysia, Kumpulan Wang Amanah Pencen [KWAP]) or by a government-related agency in which the government has an interest by virtue of financial or legal exposure. Ownership control is related to the ability to appoint members to the board of directors or to influence the appointment of senior managers who make key decisions e.g., tax planning decisions, strategic restructuring decisions, investment and divestments, and financing decisions (Kasipillai et al., 2017). A dummy variable is used to measure this construct, and companies scored 1 if they are considered GLCs under the above definition and 0 otherwise. Institutional ownership (INST) is measured by the percentage of equity share ownership by institutional investors of the firm. The four main groups of institutional investors operating in Malaysia are funds own by the government, banking groups, insurance companies, and mutual funds. Based on Hartzell and Starks (2003), we only compute the shareholding of the five largest institutional shareholdings in the firm. A firm is categorised as a family firm if 20% or more equity ownership lies with the family, or if a family holds more board seats than any other individual or group on the board. Dichotomous variable coded 1 is applied when the firm is identified as a family own, and 0 otherwise.

Control Variables
The market-to-book value of assets (MBA) ratio is measured at the end of year t. MBA is calculated as the total assets less total common equity plus the market value of equity calculated as the share outstanding multiplied by the closing share price that is scaled by the total assets. Return on assets (ROA) is the natural logarithm of the return on assets, computed as the ratio of its net income in a given period to the total value of its assets.
Political connection (PCON) is measured as 1 if the firm is politically connected and 0 otherwise. A firm in Malaysia is identified as a politically connected firm primarily based on information contained in that firm's annual report for each year from 2008 to 2014. Other important sources are studies by Bliss and Gul (2012) and Fung et al. (2015). The former study identifies politically connected Malaysian firms from 2001 to 2004, whereas the latter study identifies politically connected firms from 2005 to 2010. Both studies also rely on various Malaysian business periodicals and newspapers for their sources of information. Besides, the list used in this study is supplemented by and cross-checked against the sample of politically connected Malaysian firms in earlier studies on Malaysia (Gul, 2006;Johnson & Mitton, 2003).
Duality is defined as present if the CEO also serves as the chairman of the board (CEO_DUAL). This construct is a dummy variable, with firms scoring 1 if duality exists and 0 otherwise. The logarithm of total assets (T_ASSETS) is a proxy for firm size. The debt to total assets (LEV) ratio is measured by the natural log of total liabilities to total assets. Board size (B_SIZE) refers to the total number of executive and non-executive directors on the board, which are shown to have an effect on dividend payout (see Ajay, 2007;Lee & Isa, 2014;Benjamin & Zain, 2015). These variables are winsorised at the 1% and 99% percentiles to minimise the effect of outliers. Further, we control for the industry and year effects too.

Model Specification
This model tests hypotheses H1 to H4. The regression model used to test the hypotheses is as follows: 60 Base Model     Table 3 shows the correlation between the variables. The DIV_POUT is positively and significantly correlated with EXE_COM, the performance variable, MBA, return of assets (ROA) at p < 0.05 level. In terms of ownership structure, GLCs ownership is positively and significantly correlated with MBA, T_ASSETS, B_SIZE and PCON however negatively and significantly correlated with EXE_COM and CEO_DUAL at p < 0.05 level. On the other hand, family ownership is negatively and significantly correlated with MBA, B_SIZE, T_ASSETS, LEV, GLC and PCON at p < 0.05 level. Institutional investor ownership (INST) is negatively and significantly correlated with dividend CEO_ DUAL, ROA and FLY_C and but it is positively and significantly correlated with EXE_COM, LEV, GLC, PCON, B_SIZE and T_ASSETS at p < 0.05 level.

RESULTS
H1 states that dividend payout is negatively associated with executive compensation. Table 4 regression results show that dividend payout is positively and significantly associated with directors' executive compensation for all the Models (1-5) in Malaysia PLCs. As shown in Table 4 (Model 1), higher compensation leads to significantly higher dividend payout at p < 0.05 significance level, and the basic model explains at 62.6% of the determinants of dividend payout. This positive relationship consistently holds in Models (2-5) that controls for ownership and corporate governance. Thus, our results are not consistent with the findings of Bhattacharyya et al. (2008), as executive compensation in Malaysian PLCs is positively associated with dividend payout and hence demonstrating a contrasting view to the advanced theory of the agency paradigm between an emerging and a developed market. Hence, our results do not support H1, evidencing that emerging Asian markets, in this instance in the context of Malaysia and some other ASEAN countries are different (inconsistent findings) from developed countries (see Bhattacharyya et al., 2008) in influencing the relationship between executive compensation and dividend payout.
H2 states that there is a negative association between GLCs ownership and dividend payouts in Malaysian PLCs. In Table 4 (Models 3 and 5), the results show that GLCs pay higher dividends across both models. This positive association is significant at p < 0.01 level. Thus, our results do not provide support for H2. The rationale for this result may be threefold. Firstly, it is because directors in GLCs are appointed or seconded from civil service, and civil servants assume positions in the private sector GLCs. Lastly, the directors' pay is guided by the government pay scale and practices. Furthermore, as GLCs main shareholders are institutional investors like Employees Provident Fund (EPF) and government-linked trust funds, such as Permodalan Nasional Berhad (PNB), they will only invest in companies that pay dividends. There is also anecdotal evidence that GLCs that belong to the ruling government will adopt a dividend policy that will consistently pay dividends irrespective of the firms' growth performance. Executive compensation of GLCs may be compensated lower than non-GLCs but their variable compensation such as performance bonus may be very significant as this variable compensation is not based on civil service guidelines. H3 states that there is a positive association between family ownership (FLY_C) and dividend payout. Table 4 (Model 2) supports H3 at p < 0.10 level and the findings obtain are consistent with Chen et al. (2005). The results obtain implies that agency conflict Type II is not evidence in Malaysian family-owned PLCs. One plausible reason is that family-owned PLCs do not have the incentive to expropriate the minority shareholders, unlike Setiawan et al. (2016), by shortchanging dividend payout as they will enjoy the biggest portion of dividend payout if declared and paid as they are the largest shareholders. Our findings are also in contrast with the findings of Benjamin et al. (2016), based on Malaysian samples too, as their dataset is dated to the period of World Financial Crisis and encompassing prior periods from 2004 to 2009. H4 states that there is a positive association between institutional ownership and dividend payout in Malaysian PLCs. Table 4 (Model 4) results, has no support for H4. There is some logical rationale as to why the result is not supported and this is attributable to the main characteristics of a GLC and institutional holding. GLCs tend to be the largest companies traded on Bursa, and they are particularly attuned to the wishes of the government to promote sustainable business practices and most GLCs are PLCs that are partially owned by Khazanah. Khazanah constitutes a vital part of the Malaysian economy and make up 49% of the market capitalisation of Bursa Malaysia in 2010 (Subramaniam, Dhoraisingham, & Mahenthiran, 2016). In contrast, institutional shareholdings came about with the implementations of the institutional reforms that are in line with the country's policy on economic development strategies that have been in effect since 1971. These strategies have the intended goal of transforming Malaysia from an agricultural to an export-led economy comprise of among other investments by mutual funds, banks, securities firms and insurance companies that include foreign investors (Subramaniam et al., 2016).

Robustness Test
Several robustness tests of the model are performed. Firstly, as the final dividend is declared only after the board approves of the directors' compensation, which typically only happens after the financial year-end, hence we run the regression base on one-year-ahead dividend payout against existing independent and control variables using the following regression model, shown below. Here, we use lag one-year and lag two-year dividend payout against existing independent and control variables. Thus, we find that both prior year and two-year lag dividend payout regression results (not shown here for brevity reasons) to have no significant results demonstrating that the executive compensation to have no effects on the dividend payouts of Malaysian PLCs. Further, the White test is conducted to establish that the variance of the errors in the regression model is constant. We performed an additional test to examine if PCON firms influence the reported results. This is because Malaysia has a large and well-documented politically connected (PCON) firms (Gomez & Jomo, 1999). For example, in several cross-country studies, Malaysian PCON firms alone comprise 10% to 15% of the sample countries (Boubakri, Guedhami, Mishra, & Saffar, 2012;Chaney, Faccio, & Parsley, 2011;Faccio, Masulis, & McConnell, 2006). Due to the importance of political connection in the Malaysian corporate sector, we are motivated to analyse the role of political connections in relation to the director's compensation and dividend payout in Malaysia. The Malaysian PLCs are segregated into PCON and non-PCON companies. Table 5 ( Model 2) shows that the results obtained in connection to H1, on the relationship between dividend payout and EXE_COM, is positively significant to Malaysian PLCs that are PCON rather than non-PCON firms. Thus, the PCON firms' board's decision to pay higher dividends commensurate with higher EXE_COM, which is consistent with the clientele (catering) theory, when these executive directors are compensated with a higher remuneration package, they endorse a higher dividend payout rate.

CONCLUSION
In developed countries, over the decades, regulatory reforms and stock exchange requirements have attempted to regulate executive pay to be consistent with firm performance. For example, by stipulating that all or most of the directors on a company's board who set compensation must be independent, and each year firms must reveal the size and structure of their top executives' compensation and the reasoning behind it. As a result, executive directors and their boards know exactly what their peers are making, however, critics say boards are using that information in a dysfunctional manner to ratchet up overall pay. This article empirically analyses the relationship between ownership structure and the executive director's compensation among public listed companies in Malaysia using a panel data from 2008 to 2014. The results of our findings are inconsistent with Bhattacharyya et al. (2011) and that executive compensation is positively associated with dividend payout and that it is more so for politically connected firms. In emerging countries like Malaysia where employees are rewarded for their loyalty to their company, we find that corporations have dividend payout policies where dividends are paid to shareholders despite the high compensation paid to executive directors. Hence, consistent with clientele (catering) theory, when these executive directors are compensated with a higher remuneration package, they endorse a higher dividend payout rate. Executive directors may justify their compensations by ensuring there is some returns or reward to shareholders and avoid being label as a risky firm that does not protect the interests of the minority shareholders (Smith & Watts, 1992).
On the association between GLCs and dividend payout, our results show a positive association between GLCs and dividend payout, which suggests that directors in GLCs appointed or seconded from civil service to assume positions in the private sector, the government pay schemes does have an impact between CEO pay and firm performance. Thus, when the directors' pay is guided by government pay scale and practices, the results suggest that GLCs that belong to the ruling government will adopt a dividend payout policy that will consistently reward higher dividends to shareholders in line with company performance irrespective of the level of directors' compensation.
On the family-owned businesses, the results obtained are consistent with the hypothesis and it is evidenced that family firms pay higher dividends and are based on the fact that they prefer to control it themselves and benefit from those resources. Family firms that are large and in control play a significant role in resolving agency issues and that the higher the increase in family control firm, leads to higher dividend payments. However, on institutional ownership, the findings are not supported.
As for this study's limitations, this study is based on the top 300 highest capitalised Malaysian public listed companies, meaning that the study's conclusions might only be valid and applicable to large companies listed in Malaysia that are influenced greatly by its Islamic culture and associated proclivities. Further, this study is situated in the positivist paradigm and that it should not be construed as a comment about a particular religion or race and relied mainly on a quantitative research approach. An important area for future research might be to consider how shareholder's returns and executive compensations affect dividend payout in other emerging capital markets with different constitutional backgrounds such as Chile that has a civil law jurisdiction.

APPENDIX A Variable Measurement
Variable Descriptions Dependent variable DIV_POUT The dividend payout ratio is measured as the dividend per share scaled by earnings per share before extraordinary shares. Moreover, we use the natural logarithm of dividend payout for the analyses.

Independent variable EXE_COM
The aggregated pay of all executive directors on each firm's board, defined as the sum of salary, bonus, and other cash payments.

Ownership variables GLC
GLCs are defined as companies whose major ownership and control are held by the main shareholder that is either a government agency (such as Khazanah, Ministry of Finance, Bank Negara Malaysia, Kumpulan Wang Amanah Pencen [KWAP]) or by a government-related agency in which the government has an interest by virtue of a financial or legal exposure. A dummy variable coded as 1 when the firm is identified as a government linked company (GLC), 0 otherwise.

INST_I
The percentage of share held by all other institutional investors (excluding EPF, LTAT, PNB, LTH and PERKESO/SOCSO) holding at least 5% of outstanding shares.

FLY_C
Dichotomous variable coded 1 when the firm is identified as family own, and 0 otherwise. A firm is categorised as a family firm if 20% or more equity ownership lies with the family, or if a family holds more board seats than any other individual or group on the board.

Control variables PCON
Indicator of 1 if the firm is politically connected and 0 otherwise.

B_SIZE
The total number of directors on the board of the company.

CEO_DUAL
Dichotomous with 1 if the chairman is also the chief executive officer (CEO) of the company.

MBA
Market to book value of assets at the end of year t [(total assets less total common equity add share outstanding multiplied by closing share price)/total assets].
T_ASSETS Natural logarithm of total assets. LEV Natural logarithm of total liabilities to total assets.

ROA
Return on assets (ROA) is computed as the ratio of its net income in a given period to the total value of its assets.

IND_DUMMY
Dummy variable coded 1 for the specific Industry, 0 otherwise.

YR_DUM
Dummy variable equals 1 for the specific year, 0 otherwise.