Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 Factors Effecting CEO Compensation: Evidence from Listed Banks in Pakistan Ramshah Rashid Lone*, Fatima Hasan* and Meher Afzal* This paper represents the initiative to highlight a developing country’s perspective on justification and thereby identifying the factors affecting CEO compensation. Pakistan’s Banking Sector represents a definite case of very high salaries and compensation packages, especially for the top level staff such as CEO and the directors. This paper tends to explore the reason for such high compensation for CEO’s in the Banking Industry. This paper attempts to use panel data of 22 listed banks in Pakistan for the period 2006-2013 and explores the relationship between CEO compensation and following variables: Firm Performance, Firm Size, CEO from the family, Independence of the board of directors, Shares held by the board, percentage ownership of financial and non-financial institution. Key words: CEO Compensation, Corporate Governance, Banks 1. Introduction Executive compensation has gained importance in public policy debates as well as in academic research. The question that needs to be answered is, if the Chief Executive Officer is at the highest position in an organization how he/she gets compensated and what are the criteria for determining CEO’s compensation. It is actually hard to accurately measure and calculate the level and composition of CEO remuneration in relation to the services he/she provides to the organization. Generally CEO salary comprises of basic salary, bonuses and allowance. However, recently there has been a trend of offering stock options, long term incentive benefits and other settlement plans. Earlier studies have tried to tie CEO compensation with the performance of the firm and proposed the existence of an optimal contracting system in a competitive market. Researchers have also gone to propose the theory of rent extraction by powerful CEO’s. However, agency problem has served as the back ground of all the queries. Komhausar, (2005) have proposed that compensation should be linked with performance. It is also realized that CEO’s have power to influence decisions about their own compensation packages, which brings in the issue of agency problem. Some studies have also focused on family oriented businesses and have looked into whether CEO from the family exercises greater power over setting his/her compensation or not. ________________________________________________________________________________________ Ramshah Rashid Lone * Assistant Professor of the Department of Business Administration at the Lahore School Of Economics, Lahore, Pakistan.ramsha@lahoreschool.edu.pk Fatima Hasan* PhD Student, University of Kansas, US fatima_ijaz@yahoo.com Meher Afzal, *Lecturer Department of Business Administration at The Lahore School of Economics, mehrafzal@gmail.com Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 Researchers have been trying to explain the question of CEO compensation by examining various aspects of performance as well. Daines, Nair & Most of the empirical studies on CEO compensation and corporate governance have been conducted in developed countries. Several economists are of the opinion that conducting such studies in developing countries is not possible. As elaborated by Gosh, (2006) developing economies as managerial markets are not well developed, also corporate laws, governance code, listing agreement and bankruptcy laws are also weak in these economies. In case of Pakistan corporate governance code was implemented in 2002 for listed firms. Implementation of the code of corporate governance in Pakistan primarily falls within the ambit of two entities: the Security Exchange Commission of Pakistan and the State Bank of Pakistan. This paper is an attempt to research in a developing country to find most significant factors affecting CEO compensation in one of the most regulated, stable and growing sector of the country, namely , banking sector. 2. LITERATURE REVIEW According to Murphy (1999), CEO pay packages can be divided into four basic components i) A base salary ii) An annual bonus plan which is tied to some accounting measure of the company e.g. performance iii) Stock options plans and other long term performance based incentive plans, such as restricted stock plans iv) CEO being part of the retirement plans offered by their companies receive special incentives in the form of gratuity/ provident funds. The basic salary is the definite component and other components of CEO compensation may vary depending on company policies Researchers have explored many factors that are likely to impact CEO compensation. Findings about some of the most significant determinants of CEO compensation are discussed below. 2.1. FIRM SIZE AND PERFORMANCE Historically firm performance has been considered as the most important variable affecting CEO compensation (Ozkan, 2007). Performance refers to the ability of a firm to earn and maintain profits/returns. Returns can be accounting returns or stock returns. Accounting returns are measured Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 through ROA and ROE (Uneg, 2000), whereas stock returns refer to the stock price performance and is measured by looking at the market price of the firm’s share (Hill, 1991). In literature most of the studies have used one form of performance to find its impact on CEO compensation. Tai (2004) assessed the rapport between CEO’s pay and their companies’ stock performance. He reported a positive relationship between CEO pay and firms performance. However, Boschen et al. (2003) examined the long run impact of rising accounting performance on CEO compensation and investigated whether accounting performance had a greater impact on CEO compensation than stock price. The reported that the two measures of performance had dissimilar long run effects on CEO compensation. They also found significant multi era dynamics i.e. CEO received no compensation for an unanticipated rise in accounting performance while in contrast they did receive a higher long run compensation for an unanticipated rise in stock price performance. The impact of performance on CEO compensation has rarely been studied in isolation. The impact of performance on CEO compensation has been combined with corporate governance variable or firm specific variable. The most common firm specific variable used by researchers is firm size (Tai, 2004). Firm size is basically defined as how small or big an organization is in terms of its sales revenue and/or in terms of book value of the firm’s asset. Ueng (2000) investigated CEO compensation for both small and larger size companies. The study defined small organizations as those that had assets less than $250 million and organizations that appeared on the Fortune 500 companies were taken as large. CEO compensation was taken as the dependent variable while CEO influence, firm size, performance and growth were taken as independent variables. It was reported that CEO pay was significantly related to firm size, CEO influence was not an important determinant of CEO compensation in small organizations while it was Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 found to be significant determinant of CEO compensation in larger organization. Performance as measured by ROA was found in large companies to have an impact on CEO compensation. No evidence was found for the impact of growth rate on CEO compensation The results suggests that there were differences amongst the determinants of CEO compensation for small and larger firms. Ozkan (2007) estimated the relationship between CEO compensation and company’s performance while controlling for corporate governance variables. Ozkan believed that corporate governance factors reduce the agency conflict influencing compensation plans. A regression model was developed with compensation as the dependent variable- measured as the cash compensation i.e. an aggregate of salary, bonus, value of stock options and long-term compensation plans. Firm performance, institutional investors, block holders ownership, director ownership (executive to nonexecutive directors), board size and firm size were taken as independent variables. The results showed a positive and considerable alliance between CEO compensation and company’s performance. Larger board size were found not much of use in monitoring, but were more inclined to control CEO authority. The results also highlighted the role that institutional investors played by holding a large percentage of shares and thus was reported to be in a position to have greater impact on CEO compensation decisions. 2.2. OWNERSHIP STRUCTURE Theory and research has proved that ownership structure of an organization is of a crucial importance as it not only affects the performance of the company but also brings in agency problem for an organization. Measuring an effective ownership structure is not only difficult, but also has been done differently in researches done in the past. Ownership structure can be measured in the form of board control i.e. how Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 many shares are held by the members of the Board of Directors. The extent of board’s control can be measured as – first is the CEO duality (when an individual is acting as a CEO and as a chairman to the board), second is the ratio of insiders members on the board (insider members are those directors who are also part of the firms’ management). Boyd (1994) hypothesized a framework for assessing CEO-board control relations. His research findings supported earlier studies describing the board as the key in-house control means for determining CEO compensation, thus aligning shareholder and management interests which reduces agency problems. In order to test and analyze the hypothesis, total compensation was taken as the dependent variable, described as the sum of basic salary and bonus. The results revealed that CEO duality and director compensation had negative impact on CEO compensation whereas, board member’s stock ownership had positively on board control. Apart from board stock ownership, institutional stock ownership also had a positive relationship. Within the ownership structure the importance of institutional investors has increased in the recent years, and their ability to impact corporate governance policies has emerged as a significant research topic. Researchers have also shown that institutional investors have an important impact upon CEO compensation. Institutional investors are defined as companies like banks and investment companies that collect money from general public investors and invest majorly in corporate shares company thus have an impact on management decisions about CEO compensation package. David et al. (1998) categorized the relationship between organization and institutional investors as “pressure resistant” for an organization i.e. institutional investors do not come across at variance with the management preferences, but at the same time they use their powers to offset managerial governance and may influence CEO pay keeping it in line with investor preferences. The other type Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 of institutional investor’s relationship was termed as “pressure sensitive” that constrains institutional investors from opposing manager’s preferences, which could reduce rights to influence firm policies with respect to CEO compensation. Another type of investors termed as the “pressure indeterminate” for organizations were those whose relationship with the firm was unclear. Earlier research had unveiled that pressure resistant’s investors were found to have a positive impact on firm policies. The author developed hypothesis to test the negative/ positive impact of pressure-resistant investors on CEO compensation level. The model incorporated two dependent variables - one the level of CEO compensation (measured as the sum of basic salary, bonuses and long-term compensations) and two the proportion of long-term inducements in accumulated CEO pay (measured as ratio of total long-term incentives to aggregate CEO salary). The independent variables were the different types of institutional investors described above. The model also incorporated control factors that might impact CEO salary, like the block holders (that held more than 5 percent of the ownership), outside directors, CEO duality, firm performance (measured as return on assets), firms size, firms’ systematic risk (defined as difference of stock price to market portfolio), and industry pay. The results were consistent with the earlier researches and “pressure resistant” institutional investors were found to have influenced the CEO compensation most significantly. Ownership Structure concept cannot be complete, unless it is recognized that a company is either privately held or public limited company. Privately held companies have always been subject of debate when it comes to compensation of their top executives. Family control businesses is quite widespread phenomena, whereas, the concept of ownership by large institutional investors is recent. Mejia et. al. (2003) research was based on family controlled corporations. They tried to research Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 various issues like i) whether family CEOs’ receive low salary packages, ii) whether the concentration of family ownership in an organization effected compensation of their company’s CEO negatively, iii) whether a large ratio of institutional investors effect the compensation of family CEO’s negatively. A firm was selected under the family controlled category if it had two or more directors who had a family relation or had at least 5 percent of voting rights. The research included total compensation as dependent variable. CEO’s association with the family was taken as the dummy variable. Their results supported the hypothesis that in the long run family CEO’s received less compensation than outside CEO’s when family ownership increases. Secondly when institutional investors hold a large chunk of shares, family CEO received less compensation. Croci et. al. (2010) contributed by examining the effect of family ownership and institutional investors on chief executive officers’ salary in continental Europe. The dependent variable was the CEO cash and equity compensation, and independent variables were family firms and institutional investors, firm performance (stock returns), firm size, CEO’s tenure, the experience of a CEO and composition of the board in a firm. Authors defined a business as family controlled, if a family member owned at least 10% of the shares. They used family firms and family CEO’s as dummy variables. The reported that for family owned companies, CEO was compensated far less than what a professional CEO would receive. Although institutional investors had a major part to play in impacting CEO compensation, but in continental Europe institutional investors were found not scrutinizing CEO compensation, which is contrary to earlier research. Also the preference of independent directors, CEO tenure and firm performance were found to have a negative impact on CEO compensation. Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 The literature on ownership structure also reveals an important dimension to compensation in which an issue is bought forward that as the CEO holds multiples positions at the organizations, i.e. act as a CEO and also the head of the board, he automatically gets placed at a higher position, gets more authority in managing operations. Such CEO not only influence their pay packages, but tend to engage in merger and acquisition activities. Mergers and acquisitions are an attractive form of restructuring and reformation for a CEO as the enhanced availability of free cash flows in a postmerger firm adds to CEO power. It has been argued that in merger & acquisitions activities dual position CEO’s tend to only benefit themselves by doing merger activities rather than improving organizational performance. Dorata and Petra (2008) surveyed and evaluated the impact of CEO duality as the driving force for a CEO to involve himself/herself in merger and acquisitions activities in order to increase his/her compensation. The paper inspected as to whether dual serving CEO’s were more involved in merger and acquisitions decision making, with which they tend to benefit more than the corresponding CEO of the non-merging organization. The study used CEO compensation as dependent variable; and CEO duality as independent variable, and also incorporated firm size, returns on equity and the firm’s indulgence in a merger or an acquisition. The results of the regression showed that CEO’s who involved themselves in merger activities received compensation not consistent with their performance. Also it was found that CEO duality was not the basic determinant of compensation in mergers rather the organization size was found the most important factor which was seconded by firm performance. Other variables that have been acknowledged in the literature, but have found to be less significant in affecting CEO compensation and cannot be studied in isolation are the role of compensation Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 committees (Adut et al, 2003; Sun, 2004), CEO tenure (Zheng, 2010; Nourayi and Mintz, 2008) and organizational culture 3. BANKING SECTOR IN PAKISTAN: Pakistan’s banking sector reforms which were initiated in the early 1990s, and have transformed the sector into an efficient, sound and strong banking system. According to joint assessment carried by World Bank and IMF, the conclusion drawn was: “Reforms have resulted in a more efficient and competitive financial system in particular, the predominantly state-owned banking system has been transformed into one that is predominantly under the control of the private sector. The legislative framework and the State Bank of Pakistan’s supervisory capacity have been improved substantially. As a result, the financial sector is sounder and exhibits an increased resilience to shocks.” Pakistan’s Banking Sector is one of the highest paying sector of the economy. Despite increasing inflation, compensation for the top executive, in the form of basic salary, bonuses and allowances have increased over the year. On the other hand, banks in Pakistan have followed an expansionary policy, thus the asset base has gone up, but due to lack of business activity and falling interest rates the performance is showing a declining trend 4. BACKGROUND OF THE RESEARCH PROBLEM: In Pakistan corporate governance code was implemented in 2002 for listed firms. Presently, corporate governance in Pakistan primarily falls within the ambit of two entities: the SECP and the State Bank of Pakistan. In Pakistan research on determinants of CEO compensation is a new area. The studies that have been conducted mostly focused on manufacturing companies (Shah et al, 2009). Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 Despite the financial sector being the fast growing, efficient, strongly regulated with best corporate governance practices, a little work has been done to explore the compensation trends in the financial sector in Pakistan. The objective of this research is thus to identify those factors that affect the CEO compensation specifically in the banking sector of Pakistan. In order to achieve this objective following research hypothesis have been formulated H1: There is a significant and positive impact of Firm size on CEO Compensation. H2: There is a significant and positive impact of firm’s accounting performance on CEO compensation. H3: There is a significant and negative impact of board size on CEO compensation. H4: There is a significant and negative impact of family CEO and his compensation. H5: There is a significant and negative relationship between CEO compensation and proportion of independent directors on the board. H7: There is a significant and negative relationship between CEO compensation and shares held by the board of directors. H6: There is a significant relationship between CEO compensation and institutional participation 5. METHODOLOGY AND MODEL: The data has been hand collected from published audited annual reports of the 22 banks listed on the Karachi Stock Exchange, for the period 2006-2013.The panel data analysis has been used for Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 interpretation purposes. The study includes significant firm specific and corporate governance variables, identified through the literature. In this paper the firm specific variables included are performance and size of the company/bank. To gauge the effect of performance on CEO compensation ROE been used. The size of the bank is measured by taking book value of assets. Ownership structure has been measured as size of the board, the percentage of independent directors on the board, percentage of shares held by institutional investors including percentage of shares held by financial and non- financial institutions, family held business (quite common in Pakistan) and therefore CEO belongs to the family. According to Pakistan’s Code of Corporate Governance, a CEO cannot be the chairman of the company, thus concept of CEO duality is not applicable in Pakistan. (Explanation of the variables have been provided in Table1 Appendix) On the basis of the literature reviewed and data availability in the Pakistan, a following model is developed for the factors that can affect CEO compensation Ln Comp= α0+β1 ROE+β2Firm Size+β3Board Size+β4 CEO Family+β5 Indp Dir+β6ShareFI+β7ShareNFI+β8ShareBoard+€ Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 6. EMPIRICAL FINDINGS: Summary of the descriptive statistics of the dependent variable and independent variables gives an insight about the Banking Sector in Pakistan. Total compensation has a means value of 45821.07, this means that the average salary of a CEO in banking sector is Rs 45.82 million per year. However, the basic pay on average is Rs 27.84 million per year. This points out to the amount of fringe benefits CEO’s in this sector are enjoying in the form of bonuses( average Rs 5.13 million per year) and allowances (average Rs 29.87 million in form of provident fund, medical allowance, house rent and utility allowances) It can also be seen from Table 2(attached in Appendix), that average total assets of the banks were Rs 300 million, with minimum of Rs 4 million to maximum of Rs 1715.2 million. However, the performance was very low, an average of ROE was 10%. With the help of Table2, one can also get a broad picture of corporate governance practices being followed. Banks were generally not operated by families (average value of 0.02), 60% of the directors on the board being independent and more than 50% of the shares are held by institutional investors (Shares_nfi+ Shares_nfi). It can also be seen from the table that average tenure of a CEO in the banking sector was 3.2 years, with minimum of 1 year to maximum of 8 years. Table 3 shows the correlation between explanatory variables. All the correlation co-efficient have value less than 0.5, thus there is no multi -collinearity among the explanatory variables. After checking for multi-collinearity, model was estimated by running multiple regression. Multiple Regression using panel data allows for using either fixed effect or random effect models. Both fixed Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 effect and random effect models were estimated. Hausman test was conducted. The Chi Square value of the test was found highly significant(Prob>chi2=0.187), thus fail to accept null hypothesis that there is a difference in co efficient estimates of fixed effect and random effect models are systematic, thus Random Effect model is preferred. This study thus uses Random Effect Regression, where dependent variable is total compensation of all the other variables i.e., company’s performance (ROE) , firm size, board size, CEO and proportion of independent directors, participation of financial and non-financial institutions and CEO family are taken as independent variables. The table below shows significant variables along with T statistics Variable Name Variable Co-efficient Std-Error T-Statistics 9.9e-07 2.96e-07 3.35 -0.617 0.241 -2.55 .0074133 .0028301 2.62 .01808 .01042 1.73 -0.32306 0.1667 -1.94 Symbol Book Value of Assets No of Bv_assets Independent Indp_directors Directors % of share held by share_nfi Financial Institutions % of shares held by share_fi Non-Financial Institutions CEO belonging to the ceo_family family ( dummy variabale) Wald chi2(8) 39.70 Prob>chi2=0.0 000 Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 The regression results points out to the overall goodness of fit for the model. We can interpret that keeping everything else constant, book value of assets, percentage of shares held by independent directors, ceo_family, Shares_nfi and Shares_nfi are highly significant at 5% & 10% level of significance .It can also be interpreted from the table that size of the firm has a positive impact on the CEO compensation as hypothesized (an increase in assets by Rs1 will bring 9.9e-07 percent change in the compensation).Board independence has a statistically significant negative impact on compensation, which shows that as the number of outsiders on the board (independent directors) increases, they affect CEO compensation negatively. However, statistically significant and positive impact of shares held by non-financial and to some extent significant and positive impact of shares held by financial institution shows that role of institutional investor is in favor of CEO for setting their salaries. As far as Pakistan’s banking sector is concerned a family member of the bank holding company being a CEO does not guarantee a high compensation level for the CEO. The most noticeable finding from the regression results is that of performance. Performance as measured by ROE not only has a negative impact on CEO compensation, but is also an insignificant factor. This result is in align with previous compensation studies done in Pakistan. (Shah et al, 2009; Athar &Khan, 2012). 7. CONCLUSION: This study is an attempt to investigate significant factors that affect CEO compensation in the Banking Sector of Pakistan. This study has tried to include both firm specific and corporate governance variable to find out the impact on CEO compensation. The results suggest that: 1. Performance does not play any role in CEO compensation, whereas size of the firm has a significant and positive impact on CEO compensation in the banking sector of Pakistan. Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 2. Board size does not matter, however, the composition of the board is the main factor affecting the CEO compensation. More outsider on the board (independent directors) has a negative effect on CEO compensation. The role of institutional investors on the board is found to be positive and significant. 8. LIMITATION AND FUTURE DIRECTIONS: The data for this study has been collected from audited annual reports of the bank. Unfortunately in case of Pakistan there is no other source, except these annual reports to find about the corporate governance practices. The level of disclosure in these report lacks the information about the demographics of the CEO, especially age, qualification and experience.( 80 % missing values generated). Since all the CEO’s are male, except CEO of First Women Bank, thus data is controlled for gender Although the result of the study are in align with Athar &Khan (2012), a significant value of intercept and low value of R2 =0.204 points out that else than firm specific and corporate governance factors other variables affect the CEO compensation. There is not much work done in literature on analyzing the impact of demographic features on the CEO compensation. Including such variables, and making disclosure in more detail can better capture the true impact of factor that can affect CEO compensation in Pakistan. Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 REFERENCES 1. Adut, D.; Cready, W. H.; & Lopez, T.J. (2003) Restructuring Charges and CEO Cash Compensation, The Accounting Review, 78(1), 169-192 2. Boyd, B.K. (1994): Board Control and CEO Compensation Strategic Management Journal, 15(5), 335-344 3. Boschen, J.F., Duru, A., Gordon, L, A., & Smith, K. J. (2003) Accounting and Stock Price Performance in Dynamic CEO Compensation Arrangements,’ The Accounting Review, 78(1), 143-168 4. Bebchuk, L., Fried, J., & Walker, D. (2002) Managerial power and rent extraction in the design of executive compensation. University of Chicago Law Review, 69, 751–846. 5. Croci, E., Gonenc, H., & Ozkan, N. (2010) CEO Compensation, Family Control and institutional investors in continental Europe,’ School of Economics, Finance and Management,1-29 6. David, P., Kochar, R., & Levitas, E. (1998) The Effect of institutional Investors on the Level and Mix of CEO Compensation The Academy of Management Journal, Special Research Forum on Managerial Compensation and Firm Performance, 41(2), 200-208 7. Dorata, N. T., & Petra, S. T. (2008) CEO duality and Compensation in the market for corporate control, Managerial Finance, 34(2), p. 342-353 8. Hambrick, D. C., S. Finkelstein. (1995) The effects of ownership structure on conditions at the top: The case of CEO pay raises. Strategic Management J. 16 175-193. 9. Hill, C. W. L. & Phan P. (1991) CEO Tenure as a Determinant of CEO Pay, The Academy of Management Journal, 34(3), 707-717 Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 10. Luis R. Gomez-Mejia, Martin Larraza-Kintana & Makri, M. (2003) the Determinants of Executive Compensation in family- Controlled Public Corporations, The Academy of Management Journal, 46(2), 226-237 11. Nourayi, M. M., & Mintz, S. M. (2008) Tenure, Firm’s performance, and CEO’s compensation,’ Managerial Finance, 34(8), p. 524-536 12. Ozkan, N. (2007) CEO Compensation and Firm Performance: An Empirical Investigation of UK Panel Data School of Economics, Finance and Management, p. 1-27 13. Sun, J., & Cahan, S. (2009) The effect of compensation Committee quality on the association between CEO cash compensation and Accounting performance Corporate Governance: An International review, 17(2), 193-207 14. Shah, S. Z. A., Javed, T., & Abbas, M. (2009) Determinants of CEO Compensation, Empirical Evidence from Pakistani Listed Companies International Research Journal of Finance and Economics, 32, p. 148-159 15. Tai, L. S. (2004) Firm Size, Performance and CEO Pay: Evidence from a Group of American Quality Companies Total Quality Management, 15(1), p. 35-42 16. Tosi, H. L., L. R. Gomez-Mejia. (1989) The decoupling of CEO pay and performance: An agency theory perspective, Admin. Sci. Quart, 34, p. 169-189. 17. Tosi, H. L., S. Werner, J. P. Katz, L. R. Gomez-Mejia. (2000) How much does performance matter? A meta-analysis of CEO Pay studies. J. Management , 26, 301-339. 18. Ueng, C. J. (2000) CEO influence and Executive Compensation: Large firms vs. Small firms 26, 3-12 19. Zheng, Y. (2010) The effect of CEO tenure on CEO compensation: Evidence from inside CEO’s vs. outside CEO’s, Managerial Finance, 36(10), p. 832-859. Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 20. Iqbal.A ; Khan.I ;Ali.S (2012). CEO compensation and Bank Performance, American Journal of Scientific Research, 78, p. 93-100 Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 APPENDIX Table 1. Variable Description Variables Dependent Variable Ln Comp Definition/Description CEO compensation calculated as sum of salary, bonus and allowance (Stock options concept is not applicable in Pakistan Independent Variables ROE Firm Size Ownership Structure Board Size CEO family Profit Before Tax/Share Holder Equity Book Value of Assets CEO Duality Independent Directors(IndpDir) Share of Financial Institutions(ShareFI) Share of Non- Financial Instituion(ShareNFI) Shares held by the board(ShareBoard) No of directors on the board Equals 1 if CEO is the member of the controlled family Equals 1 if CEO is also the Chairman of the Comoany % of independent directors on the board % of shares held by Financial Instituions % of shares held by Non-Financial Instituion % of shares held by the board TABLE 2: DESCRIPTIVE STATISTICS OF THE VARIABLES Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 Table2. Descriptive Statistics of the Variables . summarize totcomp_ceo basicpay_ceo000 ALLOW_CEO BONUS_CEO bv_assets000000 roe boardsize CE > o_tenure indp_directors SHARES_BOARD SHARES_FI share_nfi Variable Obs Mean totcomp_ceo basicpay~000 ALLOW_CEO BONUS_CEO bv_as~000000 175 175 176 175 175 45821.07 27845.22 29878.51 5130.189 300505.3 roe boardsize CEO_FAMILY ceo_tenure indp_direc~s 175 175 175 175 175 SHARES_BOARD SHARES_FI share_nfi 167 175 175 Std. Dev. Min Max 37869.41 32836.27 37605.02 10866.38 316588.2 3848 1108 0 0 4025 277516 264439 277516 63212 1715271 10.24311 8.554286 .2628571 3.217143 .5869714 37.98785 1.416719 .4414486 2.078556 .2228588 -125.25 4 0 1 .1 184.06 13 1 8 .92 6.683883 8.044107 49.2327 11.7094 9.502677 30.91336 0 0 .02 63.41 46.57 98.99 Table 3 Correlation Matrix between Explanatory Variables Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 . corr roe boardsize ceo_family indp_directors shares_board share_fi share_nfi (obs=167) roe boards~e ceo_fa~y indp_d~s shares~d share_fi shar~nfi roe boardsize ceo_family indp_direc~s shares_board share_fi share_nfi 1.0000 0.1351 -0.0999 0.0148 0.0213 0.0073 -0.1394 1.0000 0.2065 0.0215 0.1014 0.1107 0.0274 1.0000 0.1452 1.0000 0.3163 -0.0684 1.0000 0.4714 0.0508 0.0489 1.0000 0.0670 0.1680 -0.1764 -0.2896 xtset id years panel variable: id (strongly balanced) time variable: years, 2006 to 2013 delta: 1 unit Table 4. Random Effect Model 1.0000 Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 . xtreg lncomp bv_assets000000 roe boardsize ceo_family indp_directors shares_board share_fi share_nfi, > re Random-effects GLS regression Group variable: id Number of obs Number of groups = = 167 21 R-sq: Obs per group: min = avg = max = 7 8.0 8 within = 0.2040 between = 0.1770 overall = 0.1857 corr(u_i, X) Wald chi2(8) Prob > chi2 = 0 (assumed) lncomp Coef. bv_assets000000 roe boardsize ceo_family indp_directors shares_board share_fi share_nfi _cons 9.91e-07 -.0005099 -.0061277 -.3230661 -.6175057 .0037744 .0180854 .0074133 10.13193 2.96e-07 .0015659 .0497286 .1667783 .2416923 .0078569 .0104249 .0028301 .5330822 sigma_u sigma_e rho .58310941 .50252511 .5738211 (fraction of variance due to u_i) . Std. Err. z 3.35 -0.33 -0.12 -1.94 -2.55 0.48 1.73 2.62 19.01 P>|z| = = 0.001 0.745 0.902 0.053 0.011 0.631 0.083 0.009 0.000 39.70 0.0000 [95% Conf. Interval] 4.11e-07 -.0035789 -.1035939 -.6499455 -1.091214 -.0116249 -.002347 .0018665 9.087112 1.57e-06 .0025592 .0913385 .0038132 -.1437975 .0191736 .0385179 .0129601 11.17676 Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 Table 5. Fixed Effect Model . xtreg lncomp bv_assets000000 roe boardsize ceo_family indp_directors shares_board share_fi share_nfi, > fe Fixed-effects (within) regression Group variable: id Number of obs Number of groups = = 167 21 R-sq: Obs per group: min = avg = max = 7 8.0 8 within = 0.2153 between = 0.0545 overall = 0.0955 corr(u_i, Xb) F(8,138) Prob > F = -0.2634 lncomp Coef. bv_assets000000 roe boardsize ceo_family indp_directors shares_board share_fi share_nfi _cons 1.07e-06 .0006142 -.0491509 -.3152887 -.7795553 -.000498 .0275795 .0050561 10.62386 4.21e-07 .0018931 .0553437 .177142 .2814306 .0102383 .0129512 .0033824 .6023598 sigma_u sigma_e rho .67480716 .50252511 .64326481 (fraction of variance due to u_i) F test that all u_i=0: . Std. Err. F(20, 138) = t 2.53 0.32 -0.89 -1.78 -2.77 -0.05 2.13 1.49 17.64 9.41 P>|t| = = 0.013 0.746 0.376 0.077 0.006 0.961 0.035 0.137 0.000 4.73 0.0000 [95% Conf. Interval] 2.33e-07 -.003129 -.1585823 -.6655522 -1.336029 -.0207422 .0019712 -.001632 9.432814 1.90e-06 .0043575 .0602805 .0349749 -.2230815 .0197462 .0531879 .0117442 11.81491 Prob > F = 0.0000 Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 Table 6. Hausman Fixed Random Test . hausman fixed random Note: the rank of the differenced variance matrix (7) does not equal the number of coefficients being tested (8); be sure this is what you expect, or there may be problems computing the test. Examine the output of your estimators for anything unexpected and possibly consider scaling your variables so that the coefficients are on a similar scale. Coefficients (b) (B) fixed random bv_as~000000 roe boardsize ceo_family indp_direc~s shares_board share_fi share_nfi 1.07e-06 .0006142 -.0491509 -.3152887 -.7795553 -.000498 .0275795 .0050561 9.91e-07 -.0005099 -.0061277 -.3230661 -.6175057 .0037744 .0180854 .0074133 (b-B) Difference 7.48e-08 .0011241 -.0430232 .0077775 -.1620496 -.0042724 .0094941 -.0023572 sqrt(diag(V_b-V_B)) S.E. 2.99e-07 .0010639 .0242899 .0597019 .1441806 .0065644 .0076846 .0018525 b = consistent under Ho and Ha; obtained from xtreg B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic chi2(7) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 10.07 Prob>chi2 = 0.1847 . Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 BANKING SECTOR GRAPHS Table 2.1 Trend of average Total Compensation 70000 60000 50000 40000 30000 20000 10000 0 2006 2007 2008 2009 2010 2011 2012 Table 2.2 Trend Average Basic Pay 40000 30000 20000 10000 0 2006 2007 2008 2009 2010 2011 2012 2013 2013 Proceedings of 10th Annual London Business Research Conference 10 - 11 August 2015, Imperial College, London, UK, ISBN: 978-1-922069-81-8 Table 2.3 Average Allowance & Bonus Trend 50000 40000 30000 20000 10000 0 2006 2007 2008 2009 2010 2011 2012 2013 Table2.4. Average Bank Size 500000 400000 300000 200000 100000 0 2006 2007 2008 2009 2010 2011 2012 2013 2012 2013 Table2.5Performance Trend (ROE) 30.00 25.00 20.00 15.00 10.00 5.00 0.00 2006 2007 2008 2009 2010 2011