Inside the family firms: The impact of family and institutional ownership on executive remuneration

This study empirically examines the impact of ownership structure on executive remuneration of listed family firms in Malaysia. Fixed effects model as the panel analysis of 279-listed family firms from 2010 to 2014 shows that institutional investors could not represent the minority shareholders’ interest in curbing the expropriation via executive remuneration by the controlling family shareholders. When the firm CEOs are non-family directors, both domestic and foreign institutional investors could exert a significant negative impact on executive remuneration. Thus, this study provides a theoretical contribution by affirming that the Type-II agency conflict between controlling shareholders and minority shareholders in family firms is ameliorated when the firm CEOs have no family relationship with the controlling shareholders. In addition, this study also unveils that domestic and foreign institutional investors have a different impact on the executive remuneration, where the governance role of the former prevails over the latter. The findings of this study would be useful for the policy-makers and regulators such as Securities Commission Malaysia and Minority Shareholder Watchdog Group to assess the expropriation issue and corporate governance in family firms. Subjects: Business, Management and Accounting; Accounting; Corporate Governance


PUBLIC INTEREST STATEMENT
Executive remuneration around the world has risen exponentially and grabbed the public attention. This study focuses on family firms as they are the most pervasive form of business organisation in the world. In family firms, the controlling family shareholders have the tendency to misappropriate funds via executive remuneration at the expense of minority shareholders. This research intends to examine whether the institutional investors could represent the interests of minority shareholders to rein in executive remuneration in family firms. Specifically, this study aims to investigate the impact of family and institutional ownership on executive remuneration in family firms. The findings would be useful for the potential investors of family firms to assess the corporate governance of executive remuneration and the possibility of expropriation by the controlling shareholders. Besides, the findings should enable the policy-makers and regulators to appraise the governance measures of remuneration arrangement in family firms.

Introduction
Around the world, executive remuneration has been under the microscope. It is one of the most debated topics in corporate governance literature. The soaring remuneration has sparked off an intense interest among the academia, practitioners, regulators and media (Barontini & Bozzi, 2011;Croci, Gonenc, & Ozkan, 2012). The continuous increase in executive remuneration triggers the curiosity of whether the wealth has been expropriated from the minority shareholders into the pockets of controlling shareholders. The widespread public attention on lavish remuneration packages suggests that the governance measures of executive remuneration deserve a closer scrutiny. One important question is whether the corporate governance structure can play a significant role in determining the level of executive remuneration. According to PWC (2016), executive remuneration has been the central focus of the institutional investors in making their investment decisions. The extant literature suggests that corporate governance mechanisms may help to lessen the agency conflicts between the shareholders and management, and dictate the remuneration policy (Basu, Hwang, Mitsudome, & Weintrop, 2007;Core, Holthausen, & Larcker, 1999).
In a move to enhance corporate governance culture from rules based to market based, Securities Commission Malaysia introduces the Corporate Governance Blueprint 2011. One of its recommendations involves the formulation of an industry-driven code, which empowers the institutional investors to intervene when there are concerns about, among others, inappropriate remuneration packages and failure in internal controls (Securities Commission Malaysia, 2011). Institutional investors are suggested by the classical agency theory (Fama & Jensen, 1983), policy-makers and regulators (Cadbury Committee, 1992;Greenbury Committee, 1995;Hampel Committee, 1998;Securities Commission Malaysia, 2012) and prior literature (Basu et al., 2007;Hartzell & Starks, 2003;Shleifer & Vishny, 1997) as an effective monitoring mechanism to mitigate the conflict between shareholders and management, which is known as the Type-I agency conflict. Voluminous studies are premised on this Type-I agency conflict between the shareholders and management in examining the determinants of remuneration in widely held firms (Conyon & Peck, 1998;Dogan & Smyth, 2002;Gregg, Jewell, & Tonks, 2012;Hartzell & Starks, 2003;Jensen & Murphy, 1990;Kato & Long, 2006;Lin, Kuo, & Wang, 2013;Ozkan, 2011;Rampling, Eddie, & Liu, 2013). Croci et al. (2012) highlight that there are limited studies examining the impact of institutional investors on executive remuneration in the firms with concentrated ownership structure, especially family firms. It is ambiguous whether the institutional investors suggested by the agency theory and prior literature to mitigate the Type-I agency conflict in widely held firms are applicable to ameliorate the Type-II agency conflict in family firms.
In this paper, we focus on the family firms in view of their pervasiveness (Carney & Child, 2013;Claessens, Djankov, & Lang, 2000;Fan, Tan, Guller, Garcia, & Quek, 2011) and significant economic contribution (The Economist, 2015;Fan et al., 2011), yet potential to expropriate wealth at the expense of minority shareholders (Barontini & Bozzi, 2011;De Cesari, 2012;Johnson, La Porta, Lopezde-Silanes, & Shleifer, 2000;Liew, Alfan, & Devi, 2014). In Malaysia, family firms account for about 45% of listed firms (Carney & Child, 2013;Ibrahim, Abdul-Samad, & Amir, 2008) and contribute approximately 67% of nominal GDP (Fan et al., 2011). The aim of this study is to empirically examine the impact of ownership structure on executive remuneration using a sample of 279 Malaysianlisted family firms from 2010 to 2014. Specifically, this study focuses on how the dominant role of family shareholders and the increasing role of institutional investors influence the executive remuneration of Malaysian family firms.
Malaysian firms offer an interesting setting to explore these relationships. A high level of ownership concentration by a family and the significant participation of controlling family shareholders in the management are the main features of Malaysian-listed firms (Abdul-Rahman, 2006;Liew et al., 2014). These features give rise to the conflicts of interests between the controlling family shareholders and minority shareholders. Malaysia has a relatively low level of investor protection and less developed capital markets (Claessens et al., 2000;La Porta, Lopez-De-Silanes, & Shleifer, 1999). The institutional investors are expected to play a significant role in relation to the shareholders' protection in Malaysia especially after the Asian Financial Crisis 1997, which is caused by the weak institutional environment, lack of corporate governance mechanisms and crony capitalism (Abdul-Rahman, 2006;Abdul-Wahab & Abdul-Rahman, 2009;Aswadi Abdul Wahab, Mat Zain, James, & Haron, 2009). According to the Emerging Markets Committee (2012), approximately 31% of Malaysia's market capitalisation is accounted by the foreign institutional investors' investment in 2010. They document that foreign institutional investors, in complying with their own investment criteria, demand a higher standard of corporate governance in investee firms. Malaysia's net foreign direct investment (FDI) inflow has grown by a remarkable 22% from RM32.52 billion in 2012 to RM39.6 billion in 2013(UNCTAD, 2014. In view of the growing presence of foreign institutional investors and their demand for a strong corporate governance framework, their role in influencing and governing executive remuneration calls for a closer scrutiny. Ferreira, Massa, and Matos (2010) point out that domestic and foreign institutional investors are different in the way they monitor the investee firms. On this premise, different from the previous studies on institutional ownership, this study contributes to the literature and fills the gap by separately examining the roles of domestic and foreign institutional investors in alleviating the Type-II agency conflict via executive remuneration in family firms.
In this paper, we examine a panel data of Malaysian family firms over the five-year period from 2010 to 2014, which consists of 1,395 firm-year observations. The findings of this study reveal the entrenchment effect of controlling family shareholders via executive remuneration. Malaysian family firms offer a higher executive remuneration i.e. on average, Malaysian family firms pay a relatively higher remuneration to their executive directors than the total firms sample in the other studies (Abdul- Wahab & Abdul-Rahman, 2009;Jaafar, Abdul-Wahab, & James, 2012;Lim & Yen, 2011); suggesting that controlling family shareholders expropriate minority shareholders via executive remuneration. Our findings show that the managerial ownership held by the family directors has no significant impact on executive remuneration. This suggests that family directors having equity stakes in the firm cannot influence the remuneration design. In this instance, the domestic institutional investors could exert a significant influence in restraining executive remuneration in Malaysian family firms. On the other hand, the foreign institutional investors show an insignificant impact on executive remuneration. These findings fill the literature gap by attesting the different impacts of domestic and foreign institutional investors in monitoring executive remuneration. Furthermore, our findings show a statistically significant and positive association between the total family ownership and executive remuneration. This indicates that controlling family use the total ownership rather than the managerial ownership to influence executive remuneration. The findings concur with Claessens, Djankov, Fan, and Lang (2002) and Shleifer and Vishny (1997) who relate that agency problem persists via the entrenchment effect of controlling family shareholders through excessive remuneration packages. Moreover, our findings unveil that when the CEOs are non-family directors, both domestic and foreign institutional ownership show a significant negative impact on executive remuneration, while the family ownership has an insignificant impact. In essence, this study provides an empirical evidence that the institutional investors take an activist approach to govern and rein in executive remuneration in Malaysian family firms when the CEOs have no family relationship with the controlling shareholders. This paper offers several contributions. The findings theoretically contribute to the agency theory by affirming the Type-II agency conflict in family firms via excessive remuneration. The findings show that the Type-II agency conflict is ameliorated when the CEOs have no kinship with the controlling shareholders. In addition, this paper adds to the literature on institutional ownership by documenting the evidence of prevailing governance role of domestic institutional investors over the foreign institutional investors in Malaysian family firms. When the firm CEOs are recruited from outside and have no family relationship with the controlling shareholders, both domestic and foreign institutional investors could effectively govern and curb the executive remuneration. These findings imply that the institutional investors will only have the incentives to do so in the presence of nonfamily CEOs; that justify their efforts and costs incurred while monitoring. Further, the findings would be useful for the policy-makers and regulators such as Minority Shareholder Watchdog Group and Securities Commission Malaysia to assess the corporate governance and expropriation issues in family firms.
The remaining paper is organised as follows. Section 2 reviews the relevant literature and formulates the hypotheses. Section 3 outlines the research methodology. Section 4 discusses the statistics and empirical results. Section 5 sets out the conclusion, limitations and suggestions for future studies.

Literature review and hypotheses development
This section discusses the agency theory, which forms the theoretical framework of this study, and reviews the relevant literature, which leads to the hypotheses development.

Agency theory
The cornerstone of the classical agency theory is the misalignment of interests between the principals (shareholders or owners) and agents (managers). The shareholders are interested in maximising the firm value, but the managers tend to enhance personal wealth, job security and prestige (Jensen & Meckling, 1976). This principal-agent problem, or known as Type-I agency conflict, is typical in the corporate setting. Prior literature attests that institutional investors could serve as an external monitoring mechanism to mitigate the agency problem between the shareholders and managers (Hartzell & Starks, 2003;Shleifer & Vishny, 1997). Most scholars agree that the separation of ownership and management creates the agency costs that may not exist if the ownership and management are combined (Fama & Jensen, 1983;Jensen & Meckling, 1976).
The theoretical analysis of the impact of family ownership on classical principal-agent problem is that, family firms are less exposed to the agency problem due to a limited degree of separation between the ownership and management (Carrasco-Hernandez & Sánchez-Marín, 2007;McConaughy, 2000). Nonetheless, a different dimension of agency conflict arises when the controlling shareholders or their family members are directly involved in the management. There is a probability of expropriation of minority shareholders by the controlling shareholders. This gives rise to another form of problem between the principal and principal, or known as Type-II agency conflict (Shleifer & Vishny, 1997;Villalonga & Amit, 2006). Controlling shareholders who often manage and control the firms can expropriate minority shareholders in several ways (Johnson, La Porta, Lopez-de-Silanes, & Shleifer, 2000). They could seek private benefits through managerial entrenchment (Gomez-Mejia, Nunez-Nickel, & Gutierrez, 2001;Shleifer & Vishny, 1997), related party transactions (Liew et al., 2014), excessive salaries or perquisites for family members or insiders (Bhaumik & Gregoriou, 2010;Lim &Yen, 2011). Filatotchev, Zhang, andPiesse (2011) maintain that family shareholders have the motivations of using private information to extract private benefits at the expense of minority shareholders for their own financial gain within a less transparent corporate structure.

Institutional ownership and executive remuneration
The increasing equity ownership of institutional investors has drawn the attention of the regulators and researchers on their role in monitoring, disciplining and influencing the investee firms (Cornett, Marcus, Saunders, & Tehranian, 2007). Institutional investors are better informed than the individual shareholders because of their large-scale development and timely analysis (Wahal & McConnell, 2000). They could monitor the firms more effectively due to their relatively lower coordination costs (Cornett et al., 2007;Guercio & Hawkins, 1999). Institutional investors are expected to play a significant role in relation to the shareholders' protection in Malaysia, especially after the Asian Financial Crisis 1997, which is caused by the weak institutional environment, lack of corporate governance mechanisms and crony capitalism (Abdul-Rahman, 2006;Abdul-Wahab & Abdul-Rahman, 2009;Aswadi Abdul Wahab et al., 2009). Hoskisson, Johnson, and Moesel (1994) document that the existence of other block-holders reduces the likelihood of family firms to pursue value-destroyed projects and family shareholders to extract private benefits. Besides, prior studies show that the degree of expropriation and the tendency to pursue private benefits are less significant in the family firms with the presence of other block-holders (Jara-Bertin, López-Iturriaga, & López-de-Foronda, 2008;Maury & Pajuste, 2005).
Prior studies report that institutional investors are an effective external mechanism via the adoption of corporate governance (Aggarwal, Erel, Ferreira, & Matos, 2011;Wahab, How, & Verhoeven, 2007), firm performance (Cornett et al., 2007;Filatotchev, Lien, & Piesse, 2005) and earnings management (Chung, Firth, & Kim, 2002;Koh, 2007). The empirical evidences pertaining to the effectiveness of institutional investors in governing executive remuneration are mixed. Cosh and Hughes (1997) find out that the presence or absence of institutional investors makes no appreciable difference to the pay-performance sensitivity and remuneration level in the United Kingdom. Similarly, Dong and Ozkan (2008) report that the institutional ownership, as a whole, does not have any impact on the directors' remuneration and pay-performance link in the United Kingdom. On the other hand, Hartzell and Starks (2003) and Almazan, Hartzell, and Starks (2005), using the sample firms from Standard & Poor's ExecuComp database, report that institutional ownership is positively associated with the performance sensitivity of managerial remuneration and negatively associated with the managerial remuneration level. Using 414 large UK firms for the years 2003and 2004, Ozkan (2007 finds out that institutional ownership has a significant negative impact on CEO remuneration, suggesting an active monitoring role played by the institutional investors. Taken together, the empirical findings suggest that institutional investors are an effective external mechanism in governing remuneration arrangement. Nonetheless, past studies did not differentiate the nationality of the institutional investors. Several past studies point out the differences between the domestic and foreign institutional investors in monitoring and influencing the investee firms (Aggarwal et al., 2011;Ahmadjian & Robbins, 2005;Croci et al., 2012;Ferreira & Matos, 2008;Gillan & Starks, 2003). Aggarwal et al. (2011), using the data from 23 countries, report that the firm-level governance is positively associated with foreign institutional ownership. They reveal that foreign, but not domestic, institutional investors make the boards more likely to have a majority of independent directors and less likely to adopt a staggered board provision. They document that domestic institutional investors play a role in improving the governance of firms located in the countries with strong shareholders' protection; however, in the countries with weak shareholder's protection, the main role of improving governance is played by the foreign institutional investors. Moreover, Aguilera and Cuervo-Cazurra (2004), by collecting the data from 49 countries, note that the presence of foreign institutional investors is positively related to the number of corporate governance codes adopted. Besides, Ferreira and Matos (2008), using the data from 27 countries, find out that the firms with a higher level of foreign institutional ownership have a higher firm valuation, better operating performance, and a lower capital expenditure, suggesting that foreign institutional investors involve in monitoring the corporations worldwide. Overall, these past studies suggest that foreign institutional investors play an influential role in promoting and enhancing the corporate governance systems around the world. With regards to the remuneration payout, Croci et al. (2012) report that foreign institutional investors possess a significant positive impact on CEO remuneration in the context of Continental Europe, while domestic institutional investors do not have any significant impact. They suggest that the internationalised firms offer a higher CEO remuneration. In light of these past findings, this study expects that the domestic and foreign institutional investors would have a different impact on executive remuneration in Malaysian family firms.
In 2009, Malaysian Government disbanded the Foreign Investment Committee (an agency associated with the foreign equity restrictions) and lifted all the foreign equity ownership restrictions in 27 service subsectors. These initiatives are to stimulate the growth of foreign investments (Hill, Tham, & Zin, 2012). According to Emerging Markets Committee (2012), approximately 31% of Malaysia's market capitalisation is accounted by the foreign institutional investors' investment in 2010. Malaysia is the fourth largest recipient of foreign direct investment (FDI) in ASEAN, behind Singapore, Indonesia and Thailand. In Asia, as a whole, Malaysia ranks the seventh top of FDI recipient (Kok, 2014). In 2016, the foreign investments account for about 47% of the total investments approved for the year (Malaysian Investment Development Authority, 2016). The growing presence of foreign institutional investors in Malaysia's equity market and their demand for strong corporate governance practices raise an interesting question relating to their role in influencing and monitoring the investee firms. This study, which covers the study period from 2010 to 2014, makes timely contribution by examining the role of foreign institutional investors in governing executive remuneration immediate after the government's liberalisation of foreign investment in 2009.
To a large extent, prior studies relating to the influence of institutional ownership on executive remuneration are in the context of developed countries, particularly the United Kingdom and the United States (Almazan et al., 2005;Dong & Ozkan, 2008;Hartzell & Starks, 2003;Ozkan, 2007). There are limited empirical studies pertinent to the monitoring role of institutional investors in the developing countries (Lim & Yen, 2011). Hence, this study contributes and fills the gap by examining the monitoring role of institutional investors in a developing country, Malaysia. In addition, this study also contributes to the literature on institutional ownership by segregating it into domestic and foreign. Accordingly, the following hypotheses are generated: H 1 : Domestic institutional ownership has a negative association with executive remuneration.
H 2 : Foreign institutional ownership has a negative association with executive remuneration.

Family ownership and executive remuneration
According to Shleifer and Vishny (1997), large shareholders who gain nearly full control of the firm have the tendency to generate private benefits that are not shared by the minority shareholders. They may represent their own interests which do not coincide with the interests of other investors, employees or managers in the firms. Lee (2004) argues that the controlling family shareholders are capable of redistributing benefits from the firms through excessive remuneration or special dividends. DeAngelo and DeAngelo (2000) show the evidence of family shareholders extract private benefits through special dividends, excessive incentive scheme and related party transactions.
The empirical studies on the association between family ownership and executive remuneration are limited. Of the few studies, Cheung, Stouraitis, and Wong (2005) show a significant positive association between the family managerial ownership and CEO remuneration in the context of Hong Kong. They suggest that in the presence of information asymmetry between the entrenched CEOs and outside investors, the former may use their ownership rights to extract a higher salary for themselves. Besides, Haid and Yurtoglu (2006) find out that family shareholders pay their managers significantly high level of remuneration in the context of Germany. On the other hand, McConaughy (2000) reports that the family CEOs in the US family firms are paid less than their non-family counterparts, and the CEO ownership is negatively associated with remuneration level. His findings support the incentive alignment hypothesis; family CEOs possess superior incentives and have less need to receive additional incentives through their remuneration. Similarly, Gomez-Mejia, Larraza-Kintana, and Makri (2003), by drawing 253 family-controlled firms from COMPUSTAT database for the period of 1995-1998, report that family CEOs receive a lower total remuneration than the non-family CEOs in family firms. They explain that, due to incumbent family ties, family CEOs are unlikely to leave for high remuneration elsewhere, which makes it unnecessary to pay them with the remuneration packages that are comparable to those of professional CEOs. Their findings are contrary to the notion that family shareholders extract extra wealth from the firm through remuneration arrangement (Gomez-Mejia et al., 2003;McConaughy, 2000).
Overall, past studies mostly examine CEO remuneration in the context of developed countries and the findings are inconclusive. This study fills the gap by examining the executive directors' remuneration in a developing country, Malaysia. The family ownership for this study is proxied by: (i) managerial ownership of family directors and (ii) total ownership of family shareholders. This is to examine the extent of family ownership that would affect the executive remuneration. The managerial ownership held by the family directors may not make them powerful enough to influence the remuneration; however, they could rely on the shareholdings of the other family members to design the remuneration packages. The following hypothesis is proposed: H 3 : Family ownership has a positive association with executive remuneration.

Research methodology
To be included in the sample, the firms must satisfy the following criteria: (i) family has to hold at least 20% of a firm's equity ( The dependent variable is measured by the total remuneration received by the executive directors, which include salaries, fees, bonuses, allowances, benefit in kind and other emoluments. Executive share options are excluded due to the disclosure inconsistency in annual reports of Malaysian-listed firms (Bacha, Zain, Rasid, & Mohamad, 2009). Past studies excluded share options from their remuneration measurement due to the data unavailability (Abdul- Wahab & Abdul-Rahman, 2009;Cheung et al., 2005;Hassan, Christopher, & Evans, 2003;Kato & Kubo, 2006;Ozkan, 2007). In order to reduce the non-normality and heteroscedasticity, total executive remuneration is transformed using natural logarithm (Abdul- Wahab & Abdul-Rahman, 2009;Croci et al., 2012;Yatim, 2013).
The independent variables included in this study are family ownership, domestic institutional ownership and foreign institutional ownership. Family ownership (fo), domestic institutional ownership (dio) and foreign institutional ownership (fio) are respectively, measured by their proportions of shareholdings over the total shares outstanding. Following previous studies, the control variables comprised of firm characteristics are included: board size (bs) (Core et al., 1999;Ghosh & Sirmans, 2005;Kashif & Mustafa, 2012;Ozkan, 2011), firm size measured by natural logarithm of total assets (lnta) (Hassan et al., 2003;Lin et al., 2013), firm leverage measured by the ratio of total liabilities to total assets (lev) (Dong & Ozkan, 2008;Yoshikawa, Rasheed, & Del Brio, 2010), lagged firm performance proxied by return on assets (roa(-1)) and market to book value of equity (mv) (Gomez-Mejia et al., 2003;Hassan et al., 2003;Yoshikawa et al., 2010).
The study period covers five years from 2010 to 2014. Baltagi (2005) documents several advantages of using panel data, which include: (i) able to control individual heterogeneity, (ii) more informative data, more variability, less collinearity among the variables, more degrees of freedom and more efficiency and (iii) able to identify and measure the effects that are simply not detectable in pure cross-section or pure time-series data. The following panel regression model is constructed to examine the influence of the independent variables and control variables on executive remuneration: where lnexrem = natural logarithm of executive remuneration dio = domestic institutional ownership fio = foreign institutional ownership lnexrem it = 0 + 1 dio it + 2 fio it + 3 fo it + 4 bs it + 5 lnta it + 6 lev it−1 + 7 roa(−1) it−1 + 8 mv it + it https://doi.org/10.1080/23322039.2018.1432095 fo = family ownership bs = board size lnta = natural logarithm of total assets (proxy for firm size) lev = ratio of total debts to total assets (proxy for firm leverage) roa(−1) = ratio of lagged net income to total assets (proxy for firm accounting performance) mv = ratio of book value to market value of equity (proxy for growth opportunity and firm market performance) ε = error term The family ownership (fo) is measured in two ways: (1) total family ownership and (2) managerial ownership held by the family directors on board. This is to examine which extent of family ownership influences the executive remuneration. The managerial ownership held by the family directors may not be influential to have a bearing effect on executive remuneration. The family may have to use the concentrated ownership of the family shareholders to intervene in the remuneration design. In addition, the panel regression model is specifically run for the subsample firms that have a nonfamily CEO, so as to investigate whether the governing role of institutional investors would be enhanced when the management is led by a non-family CEO. To sum up, the above panel regression model is run for three times: (1) family ownership (fo) is measured by the total family ownership; (2) family ownership (fo) is measured by the managerial ownership held by the family directors on board; (3) the panel regression model is specifically run for the subsample firms that have a nonfamily CEO.
Breusch-Pagan Lagrange Multiplier test and Hausman-specific test are conducted to choose the most efficient estimator: pooled ordinary least squares (OLS), random effects model (REM) or fixed effects model (FEM).

Results and analysis
This section discusses the descriptive statistics for the variables included in this study and the panel regression results. Table 1 presents the descriptive statistics for the variables included in this study.

Descriptive statistics
The mean value of executive remuneration (exrem) is RM 3.413 million, ranging from RM0.005 million to RM126.768 million. This finding is higher than the mean values reported by prior studies (Abdul-Wahab & Abdul-Rahman, 2009;Jaafar et al., 2012;Lim & Yen, 2011). This may be attributed to the inclusion of non-family firms to their sample sets. Abdul-Wahab and Abdul-Rahman (2009), Lim andYen (2011), andJaafar et al. (2012) report the mean values of RM1.568 million, RM2.570 million and RM1.854 million, with the maximum values of RM66.743 million, RM101.000 million and RM69.621 million, respectively. This finding reveals that the executive directors in Malaysian family firms receive a comparatively higher remuneration compared to the overall firms.
Pertaining to the independent variables, the mean values of the domestic (dio) and foreign institutional ownership (fio) are 10.979 and 4.530%, respectively, with the maximum values of 57.330 and 41.570%, respectively. The mean value of dio is more than twice of fio. Abdul-Wahab and Abdul-Rahman (2009) report a maximum institutional ownership of 90.550% with the mean value of 12.650%; while Ghazali (2010) reports a maximum foreign institutional ownership of 80.160% with the mean value of 23.830%. Apparently, the institutional ownership in Malaysian family firms is relatively lower than that of the overall firms. The mean value of family ownership (fo) is 47.425%, with the maximum value of 88.720%. It is notable that the family ownership is comparatively higher than the institutional ownership. The relatively low institutional ownership in Malaysian family firms could be interpreted in two ways. On the one hand, family shareholders are reluctant to disperse the ownership concentration to external institutional investors for the fear that the firm's direction and management would be influenced by the institutional investors and deviated from the interest of the family. On the other hand, the institutional investors are aware of the prominent misappropriation and expropriation issues in family firms, thus have less investment preference. Schultz, Tan, and Walsh (2010) maintain that foreign institutions invest less in the firms that reside in the countries with poor outsider protection and corporate disclosure, and have the ownership structure that is conducive to governance problems. The finding of low fio implies that the foreign institutional investors are less confident about the corporate governance system of Malaysian family firms, which are perceived to have poor minority shareholders' protection and high possibility of expropriation by controlling shareholders.
With regard to the control variables, the mean value of the board size (bs) is 7.599 with the smallest board number of4 directors and the maximum number of 13 directors. Compared to the average board size of 10.420 among the family firms in the Continental Europe (Croci et al., 2012), Malaysian family firms have a smaller board size. The mean value of total assets (ta) is RM1.249 billion. The variation in ta is huge with the minimum of RM0.021 billion and the maximum of RM60.600 billion. The mean value of firm leverage (lev) is 37.412%. This is lower than the past studies which report the mean leverage of more than 40% (Benjamin, Zain, & Wahab, 2016;Lins, 2003). This suggests that Malaysian family firms do not use great extent of debts to finance the business. The mean value of the lagged return on assets (roa(−1)) is 5.731%. This is higher than the past findings which included non-family firms in the sample (Abdul- Wahab & Abdul-Rahman, 2009;Lim & Yen, 2011). This suggests that family firms perform better than the overall firms in Malaysia. This is in accordance with the past findings in the context of developed countries (Ali, Chen, & Radhakrishnan, 2007;Anderson & Reeb, 2003;Andres, 2008;Maury, 2006). Likewise, the mean value of the market to book value of equity ratio (mv), which is 95.52%, is higher than the past findings (Lim & Yen, 2011).

Panel regression results
Breusch-Pagan Lagrange Multiplier and Hausman-specific tests have been conducted in order to choose the most efficient estimator for the five-year panel data analysis. The rejection of null Table 1

. Descriptive statistics
Legend: exrem is the total remuneration received by the executive directors; dio is domestic institutional ownership; fio is the foreign institutional ownership; fo is the family ownership; bs is the board size, ta is the total assets, lev is the ratio of total debts to total assets, roa(−1) is lagged return on assets, mv is the ratio of market value to book value of equity. hypotheses of both tests (p-value <0.05) indicate that fixed effects model (FEM) is more efficient and appropriate for this study. The robust standard errors estimation has been used to address the nonnormality and heteroscedasticity. Table 2 shows the panel regression results using FEM with robust standard errors estimation. The regression model accounts for 34.1% (Panel A), 34.2% (Panel B) and 32.7% (Panel C) of the variability of executive remuneration.
Based on Table 2 Panel A, both domestic (dio) and foreign institutional ownership (fio) do not have any significant impact on executive remuneration (lnexrem) in Malaysian family firms. Thus, H 1 and H 2 are rejected by these results. These findings are inconsistent with the negative association reported by Hartzell and Starks (2003), Dong and Ozkan (2008) and Abdul-Wahab and Abdul-Rahman (2009) who examine in the context of the United Kingdom, the United States and Malaysia, respectively. This may be attributed to the inclusion of non-family firms in their sample sets. These findings suggest that the institutional investors, both domestic and foreign, do not play an effective role in monitoring executive remuneration in Malaysian family firms. This may be due to their relatively low shareholdings in family firms, hence making their voices less powerful and influential. The institutional investors may simply sell their shares and exit the firms rather than monitoring because the monitoring cost involved may exceed the benefit. Besides, they may face several impediments that hinder their ability to monitor, such as liquidity concerns, free rider problems and internal conflict of interests (Cvijanović, Dasgupta, & Zachariadis, 2016;Ivanova, 2017;McCahery, Sautner, & Starks, 2016). In addition, they may lack knowledge and experience of how to effectively engage with the investee firms. The Malaysian Code for Institutional Investors, which gives the guidance to institutional investors, was introduced in 2014; prior to that, the institutional investors in Malaysian firms generally did not have any explicit guides. Succinctly, these findings suggest that institutional investors could not ameliorate the Type-II agency conflict via executive remuneration in Malaysian family firms. They are not an effective external monitoring mechanism that the minority shareholders could rely on to mitigate the family opportunism via executive remuneration.

Table 2. Panel regression results using FEM with robust standard errors estimation
Legend: Panel regression model is run under three different scenarios: Panel A shows the regression results where the fo is measured by the total family ownership; Panel B shows the regression results where the fo is measured by the managerial ownership held by the family directors on board; and Panel C shows the regression results for the subsample firms that have a non-family CEO. The dependent variable under all of the scenarios is the natural logarithm of executive remuneration (lnexrem). dio is the domestic institutional ownership, fio is the foreign institutional ownership, fo is the family ownership, bs is the board size, lnta is the natural logarithm of the total assets, lev is the ratio of total debts to total assets, roa(−1) is lagged return on assets, mv is the ratio of market value to book value of equity. *0.10 significance level respectively. **0.05 significance level respectively. ***0.01 significance level respectively.

Variable
Panel On the other hand, Table 2 Panel A reports a positive and statistically significant association between the total family ownership (fo) and lnexrem at 0.01 significance level. As such, H 3 is supported. This finding supports Cheung et al. (2005) that controlling family shareholders use their concentrated ownership to influence the remuneration design. This finding indicates the entrenchment effect of family ownership and shows the evidence of Type-II agency conflict in Malaysia family firms via executive remuneration. Further, the insignificant association between the lagged firm performance (roa(−1)) and lnexrem highlights the possible expropriation of controlling family shareholders, whereby they remunerate their family executives on board without linking to the firm performance. Taken together, the controlling family shareholders use their concentrated ownership to influence the executive remuneration; the institutional investors, both domestic and foreign, are not an effective mechanism in governing executive remuneration and could not mitigate the pervasive Type-II agency conflict in Malaysian family firms.
Under Table 2 Panel B, family ownership (fo) is measured by the managerial ownership of family directors on board. It is found out that the family managerial ownership has no significant impact on the lnexrem. Meanwhile, the dio shows a negative impact on the lnexrem at 0.05 significance level. This could be interpreted that the managerial ownership held by family directors on board do not make them powerful enough to influence the remuneration arrangement. In this circumstance, the domestic institutional investors could monitor and play an effective role in reining in executive remuneration. Hence, H 1 is supported and H 3 is rejected by these results. The fio shows an insignificant association with the lnexrem; H 2 is rejected. The different impacts of dio and fio may be attributed to the different stakes of shareholdings in the firms (10.979% vs. 4.530%, as reported in Table 1). The foreign institutional investors with a lower shareholdings may find the costs of monitoring exceed the benefits of monitoring the management. In essence, these findings suggest that the role of domestic institutional investors in governing executive remuneration prevails over that of the foreign institutional investors in Malaysian family firms. Table 2 Panel C reports the panel regression results for the subsample firms that have a nonfamily CEO. It is shown that the total family ownership (fo) has no significant impact on the lnexrem while both domestic (dio) and foreign institutional ownership (fio) show a significant negative impact on the lnexrem. The findings indicate that the influence of controlling family shareholders on the executive remuneration is diluted when the firm CEOs are non-family directors. In this instance, the institutional investors, both domestic and foreign, are able to govern and restrain the executive remuneration. These findings theoretically contribute to the agency theory by asserting that Type-II agency conflict in family firms could be mitigated when the firm CEOs have no family relationship with the controlling shareholders. The institutional investors could have more power to voice on the executive remuneration and represent the minority shareholders when the firm CEOs are recruited from outside rather than having kinship with the controlling shareholders.
All of the control variables, board size (bs), firm size (lnta), growth opportunities (mv), show a significant positive impact on the lnexrem, except the firm leverage (lev) and lagged firm performance (roa(-1)), under all of the three scenarios from Panel A to Panel C. The insignificant association between the lev and lnexrem may be due to the relatively low level of debts used by the family firms to finance the business; thus, the firm leverage has an insignificant impact on the lnexrem. The insignificant association between (roa(−1)) and lnexrem indicates that the executive remuneration in Malaysian family firms is not linked to the firm performance. This raises the concern about the expropriation of profits at the expense of minority shareholders in family firms through executive remuneration.

Conclusion, limitations and suggestions for future studies
This study examines the impact of ownership structure on the executive remuneration of listed family firms in Malaysia. Specifically, this paper investigates whether the institutional investors could represent the interest of minority shareholders to oversee the executive remuneration and mitigate the Type-II agency conflict in family firms, or on the other hand, the family shareholders use their concentrated ownership to influence the remuneration design and exacerbate the Type-II agency conflict.
The findings of this study show that Malaysian family firms pay a higher remuneration to the executive directors, which is, on average, higher than the findings reported by the other studies in the Malaysian context. This raises the concern about fund misappropriation via executive remuneration in Malaysian family firms. Moreover, this study reports that institutional ownership, both domestic and foreign, have no significant impact on executive remuneration. Instead, the total family ownership shows a significant positive impact on executive remuneration, which affirms the evidence of Type-II agency conflict in Malaysian family firms. This study further finds out that when the firm CEOs have no affiliation with the controlling shareholders, both domestic and foreign institutional investors could exert a significant negative impact on executive remuneration. The findings suggest that institutional investors could play an effective role in reining in executive remuneration when the CEOs are non-family directors. These empirical findings provide a theoretical contribution, which is, the institutional investors could ameliorate the Type-II agency conflict via executive remuneration when the CEOs have no family relationship with the controlling shareholders. In addition, this study also reveals that family managerial ownership has an insignificant influence on the executive remuneration. In this instance, the domestic institutional ownership shows a significant negative impact on the executive remuneration; on the other hand, the foreign institutional ownership possesses an insignificant impact. These findings imply that the role of domestic institutional investors in monitoring executive remuneration prevail their foreign counterparts. Besides, this study finds out that the executive remuneration is not linked to the firm performance; which attests the expropriation of profit at the expense of minority shareholders and reaffirms the Type-II agency conflict in Malaysian family firms. Furthermore, the findings show that the firms with larger size in term of total assets, larger board size and higher growth opportunities pay a higher executive remuneration, which can be interpreted as reflecting their demand for executive directors' talent and expertise.
The study identifies several limitations. The main limitation of this study is the exclusion of share option from the measurement of executive remuneration. This is due to the data unavailability; the public-listed firms in Malaysia are not mandated to disclose the share options granted to the directors in their annual reports. Second, this study focuses on Malaysian-listed family firms over the five-year period from 2010 to 2014; the results potentially do not apply to other forms of organisation in other legal frameworks and economic environments, and may not be generalisable to other periods. Third, the five-year study period may be a narrow window to investigate the impact of ownership structure on executive remuneration. Future studies could extend the study period, i.e. 10 years or more, to examine how the changes in ownership structure influence the executive remuneration. Fourth, the measurement of institutional investors disregards the types of institutions, for instance, mutual funds, insurance firms and pension funds. The objectives and monitoring role of different institutional investors may be different. Future studies could segregate the institutional investors according to their type and examine whether they have a different impact on the executive remuneration.
Notwithstanding the limitations, this study provides an empirical evidence on the impact of family and institutional ownership on the executive remuneration with respect to a market that, at the same time, is highly representative of East Asian firms' characteristic of concentrated family ownership and significantly different from the markets of the United States and the United Kingdom. The findings of this study have several practical implications for the policy-makers and regulators. The paper shows the evident of Type-II agency problem in Malaysian family firms. The policy-makers and regulators such as Securities Commission Malaysia and Malaysian Institute of Corporate Governance should play an active role in governing executive remuneration in family firms. Besides, Minority Shareholder Watchdog Group should regularly target the family firms and question their remuneration design during the annual meetings. This study reveals that the executive remuneration in family firms is not associated with the firm performance. The substance of remuneration committee makeup should be enhanced as to ensure the effective governance in designing the remuneration