The Audit Committee Characteristics and Firm Performance: Evidence from the UK

*[email protected] / gabriela.zabojnikov[email protected] The Audit Committee Characteristics and Firm Performance: Evidence from the UK Gabriela Zábojníková* Dissertation Master in Finance Supervisor: Prof. Ricardo Miguel Araújo Cardoso Valente Co-Supervisor: Prof. Júlio Manuel dos Santos Martins 2016 i Abstract This study analyses the impact of various audit committee characteristics on firm financial performance using the evidence from non-financial UK companies listed on the London Stock Exchange. After recent accounting scandals, the role of the audit committee has come under continuous scrutiny. However, there are still few studies examining the relationship between audit committee characteristics and firm performance, especially within Europe. Hence, this study aims to fill this gap in the literature by exploring the above mentioned relation and contributing to the body of existing literature. The main findings of this study suggest that the features of audit committees have an impact on UK firm performance. Our findings suggest that there is a significant positive relationship between the audit committee size, frequency of its meetings and its financial experience and firm financial performance. On the contrary, the audit committee independence appeared to be negatively correlated with firm performance. The findings of our study may be used by the shareholders and board of companies to make appropriate choices about audit committee characteristics in order to safeguard the investments of shareholders. Key words: corporate governance, audit committee, firm performance, United Kingdom. JEL classification: G34, M40, M42. ii Biographical Note Gabriela Zábojníková was born in 1990 in Slovakia. She pursued Law Studies at Comenius University in Bratislava with one year exchange programme spent at Ghent University. After completing both bachelor and master in Law with high honours she got accepted to Master in Finance at Porto University after she demonstrated a huge motivation towards Finance and Mathematics. During her studies at U.Porto she took part in a traineeship programme organised by the European Commission. She spent five months in the Internal Audit Service of EC in Brussels dealing mainly with the risk assessment exercise and audits of structural funds, grant management and public procurement. iii Acknowledgements I would like to express my sincere gratitude to my supervisor Prof. Ricardo Valente for his time dedicated to my dissertation, his support, suggestions and precious advice. I would like to thank also to my co-supervisor Prof. Júlio Martins for his help and useful suggestions whenever I needed it and to Prof. Natércia Fortuna for her help with econometrics issues. My sincere thanks also go to Prof. Paulo Pereira, the director of the Master in Finance programme, for providing me this challenging opportunity to be a part of his programme. I would like to express my thanks to my beloved family in Slovakia, without their support I could not study abroad. Finally, my thanks go to my new Portuguese family and my boyfriend for their continuous support during the studies. iv Table of Contents 1 Introduction ................................................................................................................. 1 2 Literature Review ........................................................................................................ 3 2.1 Main Theories ......................................................................................................... 3 2.1.1 Agency Theory ................................................................................................ 3 2.1.2 Stakeholder Theory .......................................................................................... 4 2.1.3 Stewardship Theory ......................................................................................... 4 2.1.4 Resource Dependence Theory ......................................................................... 5 2.2 UK Corporate Governance Framework .................................................................. 5 2.3 Main Definitions ..................................................................................................... 7 2.3.1 Audit Committee .............................................................................................. 7 2.3.2 Firm Financial Performance ............................................................................ 7 2.4 Characteristics of Audit Committees and Previous Studies ................................... 9 2.4.1 Audit Committee Size ...................................................................................... 9 2.4.2 The Frequency of Audit Committee Meetings .............................................. 11 2.4.3 Audit Committee Independence .................................................................... 12 2.4.4 Audit Committee Expertise ........................................................................... 16 3 Hypotheses Development .......................................................................................... 18 4 Methodology and Data .............................................................................................. 20 4.1 Methodology ......................................................................................................... 20 4.1.1 Model ............................................................................................................. 20 4.1.2 Control Variables ........................................................................................... 22 4.1.3 Endogeneity Problem ..................................................................................... 23 4.2 Data ....................................................................................................................... 23 5 Results ........................................................................................................................ 25 5.1 Descriptive Statistics ............................................................................................ 25 5.2 Correlation Analysis ............................................................................................. 26 5.3 Analysis of Regression Results ............................................................................ 26 5.3.1 ROE as a Dependent Variable ....................................................................... 26 5.3.2 Tobin’s Q as a Dependent Variable ............................................................... 29 5.4 Robustness Tests .................................................................................................. 32 v 5.4.1 Log Transformation of Audit Committee Size Variable ............................... 32 5.4.2 Market Capitalisation as a Firm Size Indicator ............................................. 34 6 Conclusion .................................................................................................................. 35 7 References .................................................................................................................. 37 Annexes ......................................................................................................................... 42 Annex 1: List of Companies with Industry Specification .......................................... 42 Annex 2: Audit Committee Data ................................................................................ 45 Annex 3: The Results of the Regressions ................................................................... 51 vi Index of Tables Table 01: Firm performance: dimensions and indicators selected .............................................. 8 Table 02: Studies discovering a negative relationship between AC size and firm perf. ........... 10 Table 03: Studies discovering a positive relationship between AC size and firm perf. ............. 10 Table 04: Studies discovering a positive relationship between AC meetings and firm perf. .... 11 Table 05: Studies discovering a negative relationship between AC meetings and firm perf. ... 12 Table 06: Studies discovering a positive relationship between AC indep. and firm perf. ......... 15 Table 07: Studies discovering a positive relationship between AC indep. and firm perf. ......... 15 Table 08: Studies discovering a positive relationship between AC expertise and firm perf. ..... 17 Table 09: Definition of variables ............................................................................................... 22 Table 10: Descriptive statistics .................................................................................................. 25 Table 11: Correlation matrix ...................................................................................................... 26 Table 12: Regression analysis results for ROE using fixed effects ........................................... 27 Table 13: Regression analysis results for Tobin’s Q using fixed effects ................................... 30 Table 14: Robustness test No. 1: Comparison of regression results for ROE ............................ 32 Table 15: Robustness test No. 1: Comparison of regression results for Tobin’s Q ................... 33 Table 16: Robustness test No. 2: Comparison of regression results for Tobin’s Q ................... 34 Table 17: Overview of the sample of 72 UK companies with the industry specification ......... 42 Table 18: Audit committee data ................................................................................................ 45 Table 19: The estimation output (equation 4.1.1)....................................................................... 51 Table 20: The estimation output (equation 4.1.2)....................................................................... 52 Index of Figures Figure 01: Overview of variables .............................................................................................. 19 Figure 02: Overview of industries .............................................................................................. 44 vii List of Abbreviations AC Audit Committee CFROI Cash Flow Return on Investment DCF Discounted Cash Flow EBITDA Earnings before Interest, Taxes, Depreciation and Amortization EC European Commission EU European Union FEE Federation of European Accountants FRC Financial Reporting Council FTSE Financial Times Stock Exchange IFRS International Financial Reporting Standards IRR Internal Rate of Return NPM Net Profit Margin OLS Ordinary Least Squares ROA Return on Assets ROE Return on Equity ROI Return on Investment ROS Return on Sales SOX Sarbanes Oxley Act UK United Kingdom of Great Britain and Northern Ireland US United States of America 1 1 Introduction The audit committee (hereinafter referred to as “AC“) is regarded as the most important board subcommittee due to its specific role of protecting the interests of shareholders in relation to financial oversight and control (Mallin, 2007). The primary role of the AC is to oversee the firm’s financial reporting process, the review of financial reports, internal accounting controls, the audit process and, more recently, its risk management practices (Klein, 2002). The above stated is true also about audit committees of UK companies which duties have grown after adoption of several Corporate Governance Codes starting by Cadbury’s Report on the Financial Aspects of Corporate Governance. Currently it is the UK Corporate Governance Code adopted in 2010 by Financial Reporting Council (formerly the Combined Code) that sets out the main recommendations regarding audit committees in UK. The role of audit committees and corporate governance as such was particularly strengthened after recent corporate scandals. There are a limited number of previous studies regarding the relationship between different AC attributes, such as its size, frequency of the meetings, financial expertise and qualification of its members and the firm financial performance. The number of studies is limited especially in Europe, therefore the work studies the sample consisting of UK non-financial companies listed on the London Stock Exchange. Moreover, the importance of audit committees in Europe has expanded recently after the European Commission has proposed a reform of the EU statutory audit. According to Federation of European Accountants, this audit reform “brings sweeping changes to the role of the AC. One can claim that it sets this committee on a path towards becoming a key factor within the corporate governance framework of all EU Member States.“ 1 AC enhances the integrity of financial statements and reduces the audit risk thereby enhancing the quality of reported figures (Contessotto and Moroney, 2013). Although companies comply with the regulatory requirements in order to avoid sanctions, not all 1 FEE (2016). The Impact of the Audit Reform on Audit Committees in Europe, Briefing Paper, Federation of European Accountants, Corporate Governance and Company Law. 8 performance in business fundamentals. On the other hand, ROA avoids the potential distortions created by misleading financial strategies. Another ratio used to represent firm financial performance is so called Tobin’s Q ratio. It is calculated as a market value of the company divided by the replacement value of the firm’s assets. In our work, we have examined the relationship between various AC attributes and firm performance represented by ROE 5 , and Tobin’s Q 6 of UK companies listed on the London Stock Exchange. In the table below there are summarized selected performance dimensions and indicators based on Santos and Brito (2012). Table 01: Firm performance: dimensions and indicators selected 7 Dimensions Selected Indicators Profitability Return on Assets, EBTIDA margin, Return on investment, Net income/Revenues, Return on equity, Economic value added Market Value Earnings per share, Stock price improvement, Dividend yield, Stock price volatility, Market value added (market value / equity), Tobin’s Q (market value / replacement value of assets) Growth Market-share growth, Asset growth, Net revenue growth, Net income growth, Number of employees growth Employee Satisfaction Turn-over, Investments in employees development and training, Wages and rewards policies, Career plans, Organizational climate, General employees’ satisfaction Customer Satisfaction Mix of products and services, Number of complaints, Repurchase rate, New customer retention, General customers’ satisfaction, Number of new products/services launched Environmental Performance Number of projects to improve / recover the environment, Level of pollutants emission, Use of recyclable materials, Recycling level and reuse of residuals, Number of environmental lawsuits Social Performance Employment of minorities, Number of social and cultural projects, Number of lawsuits filed by employees, customers and regulatory agencies 5 ROE was measured as a percentage of net income to shareholders’ equity. 6 Tobin’s Q was measured as the total market value of the firm divided by its total asset value. 7 Source: Santos and Brito, 2012. 9 2.4 Characteristics of Audit Committees and Previous Studies It is argued that any differential in performance related to governance is more than likely related to the differences in AC characteristics. The key AC attributes according to the existing literature which will be further examined relate to: (i) size, (ii) meeting frequency, (iii) independence; and (iv) expertise. 2.4.1 Audit Committee Size The first category consists of the size of the AC. On the one hand, the increased number of members is argued to provide more effective monitoring and thus improve firm performance. On the other hand, what is controversial, according to some authors larger audit committees may lead to inefficient governance. Sharma et al. (2009) found evidence that the number of AC meetings is negatively associated with multiple directorships, an independent AC chair and AC independence. Moreover, they found a positive association between the higher risk of financial misreporting and AC size, institutional and managerial ownership, financial expertise and independence of the board. The UK Corporate Governance Code states that “the board should establish an AC of at least three, or in the case of smaller companies, two, independent non-executive directors.” 8 Several authors examined the AC size and firm performance. In the following tables there is an overview of the results of the studies that discovered either negative or positive relationship respectively. Important research regarding the board size and firm performance was done by Hermalin and Weisbach (2003) whose results can be also applied to the case of the AC size and firm performance. In their research they stated that: “Board composition notwithstanding, Jensen (1993) and Lipton and Lorsch (1992) suggest that large boards can be less effective than small boards. The idea is that when boards become too big, agency problems (such as director free-riding) increase within the board and the board becomes more symbolic and less a part of the management process. Yermack (1996) tests this view empirically and finds support for it. He 8 Rule C.3.1. of the UK Corporate Governance Code. 10 examines the relationship between Tobin’s Q and board size on a sample of large U.S. corporations, controlling for other variables that are likely to affect Q. Yermack’s results suggest that there is a significant negative relationship between board size and Q. Confirming the Yermack finding, Eisenberg et al. (1998) document that a similar pattern holds for a sample of small and midsize Finnish firms. The data therefore appear to reveal a fairly clear picture: board size and firm value are negatively correlated (Hermalin and Weisbach, 2003).” Table 02: Overview of the studies that discovered a negative relationship between AC size and firm performance Authors and year Location Sample Methods Dependent Variable Bozec (2005) Canada 500 large firms that were listed on the Canadian Stock Exchange the period was during 1976 to 2000. Multiple regressions ROS, ROA, sales efficiency, net income, efficiency and assets turnover Al-Matari et al. (2012) Saudia Arabia 135 firms which listed on Saudi Stock Market in 2011. Multiple regressions Tobin’s Q MoIlah and Talukdar (2007) Bangladesh 55 firms which were listed on Dhaka Stock Exchange in Bangladesh. The data were obtained from 2002 to 2004. OLS regressions ROA, ROE, log of market capitalization Table 03: Overview of the studies that discovered a positive relationship between AC size and firm performance Authors and year Location Sample Methods Dependent Variable Reddy et al. (2010) New Zealand 50 companies over the period 1999-2007. OLS and 2SLS regression techniques Tobin-Q and ROA Bauer et al. (2009) US 113 observations (firmyears) of real estate investment trusts firms during 2004 and 2006. OLS regression Tobin-Q, ROA, ROE and NPM 11 Al-Matari et al. (2012) De Oliveira Gondrige et al. (2012) Kuwait Brazil 136 non-financial companies. 208 Brazilian companies in 2008. Multiple regression Multiple regression ROA 2.4.2 The Frequency of Audit Committee Meetings The next feature we examined refers to the frequency by which the AC members meet together. It is expected that more active audit committees that meets often will be more effective monitoring bodies. An audit committee that rarely meets (considered inactive) may be less likely to monitor management effectively. The AC meetings frequency in the UK is recommended by the Guide on Audit Committees issued by FRC as not less than three meetings per year. It is for the AC chairman, in consultation with the company secretary, to decide the frequency and timing of its meetings. Although the recommendation is to have at least three meetings per year, most of the chairmen usually call for more frequent meetings. In the following tables there is an overview of previous studies discovering either positive or negative relationship between these two variables. Table 04: Overview of the studies that discovered a positive relationship between AC meetings frequency and firm performance Authors and year Location Sample Methods Dependent Variable Khanchel (2007) US 624 US listed and nonfinancial firms for the period of 1994-2003. Multiple regressions analyses Tobin-Q Kyereboah- Coleman (2007) Africa 103 listed firms drawn from Ghana, South Africa, Nigeria and Kenya covering the five year period 1997-2001. Regressions Tobin-Q 12 Table 05: Overview of the studies that discovered a negative relationship between AC meetings frequency and firm performance Authors and year Location Sample Methods Dependent Variable Hsu and Petchsakulwong (2010) Thailand Public non-life insurance companies in Thailand over the period 2000-2007. Truncated bootstrapped regression DEA 2.4.3 Audit Committee Independence When examining the third category, namely the independence of the AC, we have to at first define what it means. We measured the independence of the AC by the proportion of independent directors over the total number of directors sitting in an AC. The term “independent director” is usually used interchangeably with the term “non-executive director” what is not correct because not all non-executive directors are independent. The approach taken by the UK Cadbury Report was substantially similar in that it refers to independent directors as needing to be only independent of management and free from any business or other relationship which could affect their independent judgment. More recently, the UK Higgs Report 2003 on ‘The Review of the Role and Effectiveness of Non-Executive Directors’ commented on the definition of independence as spelt out in the Cadbury Report. It observed that the definition gives little guidance as to what the test should entail. The Higgs Report further observed that there are over a dozen definitions in the UK, all with different criteria, as promulgated by various shareholder bodies. Finally, the definition of independence according to the rule B.1.1 of UK Corporate Governance Code is as follows: “The board should identify in the annual report each non-executive director it considers to be independent. The board should determine whether the director is independent in character and judgement and whether there are relationships or circumstances which are likely to affect, or could appear to affect, the director’s judgement. The board should state its reasons if it determines that a director is independent notwithstanding 13 the existence of relationships or circumstances which may appear relevant to its determination, including if the director:  has been an employee of the company or group within the last five years;  has, or has had within the last three years, a material business relationship with the company either directly, or as a partner, shareholder, director or senior employee of a body that has such a relationship with the company;  has received or receives additional remuneration from the company apart from a director’s fee, participates in the company’s share option or a performance related pay scheme, or is a member of the company’s pension scheme;  has close family ties with any of the company’s advisers, directors or senior employees;  holds cross-directorships or has significant links with other directors through involvement in other companies or bodies;  represents a significant shareholder; or  has served on the board for more than nine years from the date of their first election.” As to the number of independent directors sitting in audit committees of UK companies, the UK Corporate Governance Code requires at least 3 independent nonexecutive directors. 9 An important issue to consider when evaluating the independence of any board or committee is the endogeneity of board/committee composition. Hermalin and Weisbach (1998) suggest that poor performance leads to increases in board independence. In a cross-section, this effect is likely to make firms with independent directors look worse, because this effect leads to more independent directors on firms with historically poor performance. Both Hermalin and Weisbach (1991) and Bhagat and Black (2000) have attempted to correct for this effect using simultaneous-equation methods. In particular, these papers lagged performance as an instrument for current performance. 9 The rule C.3.1 of the UK Corporate Governance Code. 14 The independence of AC has its benefits but also risks. On the one hand, it is argued that having an independent AC within the corporation facilitates more effective monitoring of financial reporting (Beasley, 1996; Carcello and Neal, 2003) and external audits (Abbott et al., 2002; 2004; Carcello and Neal, 2003). On the other hand, being completely separate from management could mean that the independent AC members see less industry issues and are more likely to side with the auditor requiring less negotiations and deliberations and thus fewer meetings. This can have negative impact on the level of monitoring (Sharma et al., 2009). According to some literature sources, the ideal situation arises if the chair of the AC is independent and the most experienced person on the committee due to their pivotal role. However, Sharma et al. (2009) show that some companies appoint an inside director as the AC chair, which consequently leads to less AC independence. Cotter and Silvester (2003) conclude that independent directors on audit committees reduce the monitoring by debtholders when leverage is low. The result is that executives on the AC lead to increased monitoring by debtholders. Additionally, Beasley and Salterio (2001) find that a board chair or CEO on the AC reduces the overall effectiveness of the AC. The independence of the AC may also be influenced by other governance mechanisms. For example, blockholders also form part of the external governance structure but their influence is often exerted internally. Klein (2002) showed a negative association between AC independence and the presence of alternative monitoring mechanisms, such as blockholders, although her results are inconclusive. On the contrary, Morck et al. (1988) and Jensen (1993) claim that the presence of outside blockholders serving on the board enhances governance because these directors have both the financial incentives and the independence to effectively evaluate and monitor management and their policies. Moreover, they have incentives to align their interests with those of management. In summary, the AC independence research suggests the percentage of independent directors, grey-directors, AC chair independence, presence of the CEO and 15 representation of blockholders on the AC may all have an impact on firm performance via the effectiveness of the AC. There are only few studies that examined the relation between AC independence and firm performance. The overview of the studies that found out positive relationship is presented in the table below. Table 06: Overview of the studies that discovered a positive relationship between AC independence and firm performance Authors and year Location Sample Methods Dependent Variable Dey (2008) US 371 firms through 2000 to 2001. Multiple regressions ROA and Tobin- Q Nuryanah and Islam (2011) Indonesia From 315 listed companies, only 46 companies were selected for this study. The sample data was selected from financial sectors over 2002 2004. Multiple regression Tobin-Q Yasser et al. (2011) Pakistan 30 Pakistan listed firms through 2008-2009. Multiple regressions ROE and NPM On the other hand, there are some studies that discovered a negative relationship between AC independence and variables representing firm performance. The summary of such studies is illustrated in the table below. Table 07: Overview of the studies that discovered a negative relationship between AC independence and firm performance Authors and year Location Sample Methods Dependent Variable Dar et al. (2011) Pakistan This study selected 11 oil and gas firms listed on the Karachi stock exchange and this study chooses nonprofitability just over 2004-2010. Multiple regressions ROE 16 2.4.3 Audit Committee Financial Expertise The final category of AC characteristics that might influence the performance relates to the financial expertise which consists of both experience and education. The UK Corporate Governance Code states in regards with the expertise that “the board should satisfy itself that at least one member of the AC has recent and relevant financial experience.” 10 Recent research confirms that accounting expertise within boards that are characterised by strong governance contributes to greater monitoring by the AC and leads to enhanced conservatism (Krishnan and Visvanathan, 2008). It is widely recognized that within each AC, the chair fulfils a key leadership role and therefore should be the most qualified person on the AC. Spira (1999) claims where the AC chair has sufficient auditing background; it is very likely that the chair and the CFO will form a good working relationship. Although it is recognised that the chair of AC should have experience, DeZoort (1998) finds contrary evidence that 76% of AC chairs do not have any auditing experience. Experience alone may not be sufficient to establish financial expertise. Both experience and education are required to become a financial expert (Giacomino et al., 2009). However, the research on this topic is very limited in part due to low incentives to disclose information on backgrounds and careers of directors prior to the post-Enron governance regulatory boom. In the table below, there is a summary of studies proving the positive relationship between AC expertise and firm performance. 10 Rule C.3.1. of the UK Corporate Governance Code. 17 Table 08: Overview of the studies that discovered a positive relationship between AC expertise and firm performance Authors and year Location Sample Methods Dependent Variable Rashidah and Fairuzana (2006) Malaysia 100 companies listed on Malaysia stock exchange. Multiple regression ROE Hamid and Aziz (2012) Malaysia The sample of government linked companies in Malaysia over the period of 2005-2010. Multiple regression ROA 24 financial statements that are necessary for our study and also the majority of the rules apply only to listed companies. The study covers the period of five years from 2011-2015. There are three types of data that were used for our analysis:  Data on audit committees 13 – as we are not aware of any database containing the necessary data regarding the audit committees, we have obtained them from the annual reports of selected companies for 2011-2015, especially from the part “Audit Committee Report“, where the company reports about its AC activity, members, meetings, etc. When obtaining the data about AC size, independent and experienced members, it is important to note that sometimes these numbers differed thorough the year. In such cases, we considered the number in the end of a given year. However, the audit committees usually changed or replaced the members by the end of the year.  Data on firm performance – necessary data for ROE and Tobin’s Q calculation; these data were obtained and calculated from financial statements of selected companies.  Data regarding the control variables – firm’s size was measured by obtaining data about the total assets of the company and firm leverage by the proportion of debt to equity in a company’s capital structure. Most of the time, the total assets value was stated in British pounds but sometimes different currencies such as US dollars or Euro were used. In those cases, we converted the currency using the exchange rates applicable in a given year. 14 Moreover, in one of the robustness test performed, we needed to obtain the data on market capitalisation of the companies included in our sample for the years 2011-2015. This data was also obtained from the financial statements. 13 The data on audit committees can be observed fom the Annex 2 to this work. 14 This was done using official exchange rates obtained from the database of Bank of England, http://www.bankofengland.co.uk/boeapps/iadb/Rates.asp. 25 5 Results 5.1 Descriptive Statistics The results of descriptive statistics are given in the Table 10. On average, there are 4 members of audit committees in the British companies. The minimum number of the AC members is 3 as it is the legal requirement and the maximum is 8. They meet 5 times in a year on average. However, it is interesting to note the differences between the meetings frequency. While some of the AC meet only once per year, others meet on a monthly basis. Nearly 42% of AC members are considered as having recent and relevant financial experience and around 98% of the members are considered to be independent pursuant to the UK Corporate Governance Code. The average return on equity was found to be 18.11% during the examined period and the average Tobin’s Q ratio was 1.82. Table 10: Descriptive statistics Variables15 Mean Median Maximum Minimum Std. Dev. Skewness Kurtosis ACSIZE 4.334302 4.000000 8.000000 3.000000 1.193465 1.065175 4.032258 ACMEET 4.985465 5.000000 13.00000 1.000000 1.773573 1.516689 6.077521 ACINDEP 0.984302 1.000000 1.000000 0.200000 0.095148 -7.080670 54.56633 ACFINEXP 0.416739 0.333333 1.000000 0.125000 0.258645 1.173500 3.136462 FSIZE 0.927433 0.875709 2.360978 -2.208310 0.604473 -0.204640 4.956857 FLEV 0.873796 0.584650 6.540000 -15.67000 1.353169 -4.275370 68.66778 ROE 18.10948 15.70500 179.6300 -66.01000 18.54916 1.981282 21.55090 Tobin’s Q 1.892222 1.185000 24.83000 -0.164600 2.791216 5.984757 43.77849 As we can see from the table above, there are some “outliers“ among our data, especially in ROE sample, with standard deviation of 18.54916. Therefore, we have decided to apply to following rule in order to decide if keeping the respective value of ROE or dropping it from our regression. We have kept the values that belong to this interval: 16 [𝑚𝑒𝑎𝑛 − 3 × 𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑 𝑑𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛, 𝑚𝑒𝑎𝑛 + 3 × 𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑 𝑑𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛] (5.1) 15 Sample size (n) = 72 firms, Time periods (T) = 5 years. 16 It is important to note that we have performed the regression both with and without outliers values of ROE and the results obtained were not differing substantially. However dropping the “outliers” boosted our results. 26 5.2 Correlation Analysis Furthermore, we have performed the correlation analysis of independent variables in order to discover possible correlation among them. This was done because to obtain the unbiased results of the regression, it is necessary that the variables do not correlate with each other. From the table below it is obvious that none of the variables are highly correlated. Table 11: Correlation matrix ACSIZE ACMEET ACINDEP ACFINEXP FSIZE FLEV ACSIZE 1 ACMEET 0.032604 1 ACINDEP -0.066620 -0.104730 1 ACFINEXP -0.324510 -0.078910 0.069416 1 FSIZE 0.155497 0.377897 -0.167610 -0.051310 1 FLEV 0.046181 -0.029820 -0.156090 0.062099 -0.022730 1 5.3 Analysis of Regression Results Finally, empirical analysis was done using fixed effect panel data regression. Two dependent variables (ROE and Tobin‘s Q) were considered in separate models to observe the effect of the corporate governance on each performance measure separately. Results were carried at 10%, 5%, and 1% significance level. Below the results of regressions are presented for each performance measure separately. 5.3.1 ROE as a Dependent Variable Firstly, AC characteristics represented by independent variables were regressed against the dependent variable – ROE and thus their impact was analysed. In this regression we used the panel regression specification since it boosts the power of statistical analysis and we applied the fixed effects model. Firstly, we have used the random effects model but after performing the Hausman test, it suggested to reject null hypothesis and thus random effects model appeared as not suitable for this regression. Consequently, we run regression in fixed effects specifications using a consistent estimator for our covariance matrix and performed likelihood ratio test confirming the use of cross-sectional fixed effects. The results of the regression are presented in the table below. 27 Table 12: Regression analysis results for ROE using fixed effects Note: The table presents the estimates of the equation (4.1.1). The standard errors are presented between parentheses under each estimated coefficient. Statistical significance is represented by * at 10%, ** at 5% and *** at 1%. The variables included in the regression are: ROE (measured as a percentage of net income to shareholders’ equity), ACSIZE (the number of AC members), ACFINEXP (the proportion of members with the recent and relevant financial experience to the overall number of AC members), ACMEET (the number of AC meetings held in respective year), ACINDEP (the proportion of independent members to the overall number of AC members), FSIZE (measured as a natural logarithm of the total assets) and FLEV (measured as a percentage of total debt to total assets). It is important to note that the R-squared value is around 10.94% indicating that only 10.94% of ROE variations are determined by the AC characteristics used in the regression, namely the AC size, the frequency of AC meetings, the independence of AC members and the financial experience of AC members. Whereas the remaining 89.06% of variations is attributed to other variables. However, R-squared has also some limitations, for example it cannot determine whether the coefficients predictions and estimates are biased. Moreover, it does not necessarily indicate if a model is adequate. Therefore, even if the R-squared value is low but the predictors are statistically significant, as we can see from the table below, it is still possible to draw important conclusions about how changes in the predictive value are associated in the response value. Regardless of the value of R-squared, the coefficients that are significant still Independent variables ROE Intercept 4.968897 (4.866478) ACSIZE ACFINEXP 1.413908*** (0.321822) 1.953980*** (0.839501) ACMEET 0.184835*** (0.021856) ACINDEP -1.080597* (0.668504) FSIZE FLEV -3.676244* (2.227430) 1.055618 (0.826902) Observations R-squared Adjusted R-squared F-statistic Prob (F-statistic) 340 0.109373 0.082302 4.040263 0.000030 28 represent the mean change in the response for one unit of change in the predictor while keeping other predictors in the model constant. The model is considered to be overall statistically significant, giving the prob F- statistics value of nearly 0.000 and therefore rejecting the null hypothesis of insignificance. It means that the variables we use in the regression specification can jointly predict the firm performance in our sample of the UK companies. Our first hypothesis (H11) states that there is a potentially positive relationship between the AC size and firm performance measured by ROE. The results of the regression are consistent with this hypothesis. This implies that the AC size can potentially positively influence the firm performance and it is supporting the finding of Bauer et al. (2009) who found out also positive significant relationship between the AC size and the firm performance measured by ROE of the US companies. On the other hand, our result is inconsistent with the finding of MoIlah and Talukdar (2007), who discovered a negative significant relationship between the above mentioned variables bringing the evidence from Bangladesh. Furthermore, our finding is also inconsistent with the results of Mak and Kusnadi (2005) who could not provide any relationship between the size of AC and firm performance in Malaysia and Singapore. Moreover, it is also contradictory to the stating of Hermalin and Weisbach (2003) who found a negative significant relationship between the board size in general and the firm performance. The second hypothesis (H21) predicts that the financial expertise of AC members is positively associated with the firm performance measured by ROE. The results of our regression analysis confirm this statement and found positive significant relationship between these two variables. This suggests that the more members with recent and relevant financial experience sitting in audit committees can bring better financial performance of British companies. Such result is consistent with findings of Rashidah and Fairuzana (2006) who examined 100 Malaysian companies and also discovered that as the AC financial experience increases, the firm financial performance increases too. The third hypothesis (H31) predicting that higher frequency of AC meetings is positively associated with the firm performance was also confirmed by the regression. 29 The results showed the positive relationship significant at 1%. The result obtained is consistent with the findings of Carcello (2002). The last hypothesis we tested (H41) predicted that the greater independence of the AC is associated with higher firm performance. However, we have discovered a negative significant relationship between them. Such a result is contradictory to the studies finding a positive association between independence and ROE (Yasser et al., 2011) but on the other hand consistent with Dar et al. (2011) discovering a negative relationship. This can be explained by the fact that independent directors usually suffer from having inadequate knowledge of the business that can lead to wrong advice to the board of directors and consequently to poorer financial performance. As for the control variables, firm size, shows a negative significant relationship, while firm leverage suggests a non-significant relationship with ROE. According to this result, it seems that the benefits of leverage are cancelled by its costs. The negative relationship between ROE and firm size can be explained by the so called “small firm effect“. This theory states that smaller firms, or those companies with a small market capitalization, outperform larger companies. This market anomaly is a factor used to explain superior returns in the Three Factor Model, created by Gene Fama and Kenneth French - the three factors being the market return, companies with high book-to-market values, and small stock capitalisation. According to the theory this effect exists because the small firms have bigger amount of growth opportunities than large companies. Moreover, small companies also tend to operate in a more volatile business environment. As mentioned in a previous part of this work, the results of the regression are presented without using “outliers” values of ROE. 5.3.2 Tobin‘s Q as a Dependent Variable Secondly, AC characteristics represented by independent variables were regressed against another dependent variable measuring the firm performance – Tobin’s Q. Similarly as when testing ROE, we used the panel regression specifications and fixed effects model. We also tried to apply random effects model, but after running the Hausman test we rejected the null hypothesis and considered using fixed effects model 30 as more suitable. Additionally, the likelihood ratio test shown that fixed effects model is suitable, too. The results of the regression are presented in the table below. Table 13: Regression analysis results for Tobin’s Q using fixed effects Independent variables Tobin’s Q Intercept 4.244695*** (0.517840) ACSIZE ACFINEXP 0.093351** (0.041975) -0.611123 (0.354526) ACMEET 0.169638*** (0.017177) ACINDEP -1.092487*** (0.066810) FSIZE FLEV -2.494415*** (0.156438) 0.027589 (0.033699) Observations R-squared Adjusted R-squared F-statistic Prob (F-statistic) 348 0.257308 0.235269 11.67546 0.000000 Note: The table presents the estimates of the equation (4.1.2). The standard errors are presented between parentheses under each estimated coefficient. Statistical significance is represented by * at 10%, ** at 5% and *** at 1%. The variables included in the regression are: Tobin’s Q (measured as a total market value of a firm divided by its total asset value), ACSIZE (the number of AC members), ACFINEXP (the proportion of members with the recent and relevant financial experience to the overall number of AC members), ACMEET (the number of AC meetings held in respective year), ACINDEP (the proportion of independent members to the overall number of AC members), FSIZE (measured as a natural logarithm of the total assets) and FLEV (measured as a percentage of total debt to total assets). The value of R-squared in this case was much higher than in the first case, namely it reached 25.73%. It indicates that 25.73% of Tobin’s Q variations are determined by the AC characteristics that we used in the regression, namely the AC size, the frequency of AC meetings, the independence of AC members and the financial experience of AC members while the remaining 74.27% of variations is attributed to other variables. Although the value of R-squared is higher than in the first model, it is still quite low. However, as we mentioned above, even if the R-squared value is low but the predictors are statistically significant, it is still possible to draw important conclusions about how changes in the predictive value are associated in the response value. Regardless of the 31 value of R-squared, the coefficients that are significant still represent the mean change in the response for one unit of change in the predictor while keeping other predictors in the model constant. The model is considered to be overall statistically significant, giving the prob F- statistics value equals to 0.000. Our first hypothesis (H11) states that there is a positive relationship between the AC size and firm performance measured by Tobin’s Q. The results of regression are consistent with this hypothesis and are statistically significant. Such a result is similar to the first model result regarding ROE. This suggests that the AC size can influence the firm performance also in terms of Tobin’s Q and it is supporting the finding of Bauer et al. (2009) who found out also positive significant relationship between the AC size and firm performance measured by Tobin’s Q of the US companies as well as finding of Reddy et al. (2010) who discovered this relationship in New Zealand. On the other hand, our result is inconsistent with the finding of Al-Matari et al. (2012), who discovered a negative significant relationship between the above mentioned variables examining the companies from Saudi Arabia. The second hypothesis (H21) predicts that the financial expertise of AC members is positively associated with the firm performance measured by Tobin’s Q. However, the coefficient is not significant which implies that the financial experience of the AC members cannot influence the firm performance measured by Tobin’s Q neither positively, nor negatively. Furthermore, the study finds that AC meetings frequency is positively and significantly associated with Tobin’s Q what confirms the third hypothesis (H3). It implies that the AC meetings positively influence the firm performance. This result is consistent with the first model using ROE as a firm performance measure. Moreover, it is supported by the finding of Khanchel (2007) who examined the US companies as well as Kyereboah- Coleman (2007) analysing the African companies. The last hypothesis we tested (H41) predicted that the greater independence of the AC is associated with higher firm performance. Similar to ROE results, the study found that there is a significant negative relationship between these two variables. This result is 32 inconsistent with the findings of Dey (2008) and Nuryanah and Islam (2011) who found a positive relationship between the AC independence and firm performance measured by Tobin’s Q in the US and Indonesian companies respectively. As for the control variables in case of the model with the Tobin’s Q, the results are consistent with the model examining ROE. The firm size shows a negative significant relationship, while firm leverage shows a non-significant relationship with Tobin’s Q. 5.4 Robustness Tests Further tests were conducted in this study in order to examine if the main results were sensitive to different measurements with the purpose to obtain clearer results and also to confirm the main findings that were made. 5.4.1 Log Transformation of Audit Committee Size Variable Firstly, the study repeated both regression models using a natural logarithm of the AC size instead of a number representing the AC size. This is usually done as to improve the model fit by altering the scale and making the variable more normal distributed. As we can see from the table below, the results remained the same. Table 14: Robustness test No. 1: Comparison of regression analysis results for ROE using fixed effects Original Model ROE New Model ROE Intercept 4.968897 (4.866478) Intercept 1.397546 (5.014920) ACSIZE ACFINEXP 1.413908*** (0.321822) 1.953980*** (0.839501) LNACSIZE ACFINEXP 9.703961*** (1.589770) 1.551146*** (0.892859) ACMEET 0.184835*** (0.021856) ACMEET 1.142266*** (0.327791) ACINDEP -1.080597* (0.668504) ACINDEP -1.719153* (0.809334) FSIZE FLEV -3.676244* (2.227430) 1.055618 (0.826902) FSIZE FLEV -3.557908* (2.217452) 1.102442 (0.852612) Observations R-squared Adjusted R-squared F-statistic 340 0.109373 0.082302 4.040263 Observations R-squared Adjusted R-squared F-statistic 340 0.101484 0.074174 3.715947 33 Note: The tables present the estimates of the equation (4.1.1). The standard errors are presented between parentheses under each estimated coefficient. Statistical significance is represented by * at 10%, ** at 5% and *** at 1%. The variables included in the regression are: ROE (measured as percentage of net income to shareholders’ equity), ACSIZE (the number of AC members), LNACSIZE (natural logarithm of the number of AC members), ACFINEXP (the proportion of members with the recent and relevant financial experience to the overall number of AC members), ACMEET (the number of AC meetings held in respective year), ACINDEP (the proportion of independent members to the overall number of AC members), FSIZE (measured as a natural logarithm of the total assets) and FLEV (measured as a percentage of total debt to total assets). Table 15: Robustness Test No. 1: Comparison of regression analysis results for Tobin’s Q using fixed effects Original Model Tobin’s Q New Model Tobin’s Q Intercept 4.244695*** (0.517840) Intercept 3.831520** (0.628281) ACSIZE ACFINEXP 0.093351** (0.041975) -0.611123 (0.354526) LNACSIZE ACFINEXP 0.541524** (0.196353) -0.569724 (0.358909) ACMEET 0.169638*** (0.017177) ACMEET 0.169165*** (0.017071) ACINDEP -1.092487*** (0.066810) ACINDEP -1.062034** (0.075029) FSIZE FLEV -2.494415*** (0.156438) 0.027589 (0.033699) FSIZE FLEV -2.498467*** (0.156470) 0.028515 (0.033519) Observations R-squared Adjusted R-squared F-statistic Prob (F-statistic) 348 0.257308 0.235269 11.67546 0.000000 Observations R-squared Adjusted R-squared F-statistic Prob (F-statistic) 348 0.258085 0.236070 11.72299 0.000000 Note: The tables present the estimates of the equation (4.1.2). The standard errors are presented between parentheses under each estimated coefficient. Statistical significance is represented by * at 10%, ** at 5% and *** at 1%. The variables included in the regression are: Tobin’s Q (measured as the total market value of the firm divided by its total asset value), ACSIZE (the number of AC members), LNACSIZE (natural logarithm of the number of AC members), ACFINEXP (the proportion of members with the recent and relevant financial experience to the overall number of AC members), ACMEET (the number of AC meetings held in respective year), ACINDEP (the proportion of independent members to the overall number of AC members), FSIZE (measured as a natural logarithm of the total assets) and FLEV (measured as a percentage of total debt to total assets). Prob (F-statistic) 0.000030 Prob (F-statistic) 0.000097 40 Laing, D., & Weir, C. M. (1999). Governance structures, size and corporate performance in UK firms. Management Decision, 37(5), 457-464. MoIlah, A. S., & Talukdar, M. B. U. (2007). Ownership structure, corporate governance, and firm’s performance in emerging markets: Evidence from Bangladesh. The International Journal of Finance, 19(1), 4315-4333. Morck, R., Shleifer, A., & Vishny, R. W. (1988). Management ownership and market valuation. Journal of Financial Economics, 20, 293-315. Nelson, S. P., & Jamil, N. (2011). Audit Committees and Financial Reporting Quality Following the Government Transformation Program: Evidence from Malaysia. AKEPT Annual Young Researchers X³. Ntim, C. G. (2009). Internal corporate governance structures and firm financial performance: evidence from South African listed companies. University of Glasgow, Glasgow. Nuryanah, S., & Islam, S. M. N. (2011). Corporate governance and performance: Evidence from an emerging market. Malaysian Accounting Review, 10(1), 17- 42. Patro, S., Lehn, K., & Zhao, M. (2003). Determinants of the size and structure of corporate boards: 1935-2000. Financial Management, 38, 1-57. Pfeffer, J., & Salancik, G. R. (1978). The External Control of Organizations: A Resource Dependence Perspective. University of Illinois at Urbana- Champaign's Academy for Entrepreneurial Leadership Historical Research Reference in Entrepreneurship. Rashidah, A. R., & Fairuzana, H. M. (2006). Board, audit committee, culture and earnings management: Malaysian evidence. Managerial Auditing Journal, 21(7), 783-804. Reddy, K., Locke, S., & Scrimgeour, F. (2010). The efficacy of principle-based corporate governance practices and firm financial performance: An empirical investigation. International Journal of Managerial Finance, 6(3), 190-219. Reed, W. R., & Ye, H. (2011). Which panel data estimator should I use? Applied Economics, 43(8), 985-1000. Roodman, D. (2006). How to do xtabond2: an introduction to ‘difference’ and ‘system’ GMM in stata, Working Paper No. 103, Center for Global Development. 41 Santos, J. B., & Brito, L. A. L. (2012). Toward a Subjective Measurement Model for Firm Performance. Brazilian Administration Review, 9, 95-117. Sarbanes-Oxley Act (2002). Sharma, V., Naiker, V., & Lee, B. (2009). Determinants of audit committee meeting frequency: Evidence from a voluntary governance system. Accounting Horizons, 23, 245-263. Spira, L. F. (1999). Ceremonies of governance: Perspectives on the role of the audit committee. Journal of Management and Governance, 3, 231-260. UK Corporate Governance Code (edited in September 2014). Vafeas, N., & Thedorou, E. (1998). The Relationship between Board Structure ad Firm Performance in the UK. The British Accounting Review, 30(4), 383-407. Yasser, Q. R., Entebang, H., & Mansor, S. A. (2011). Corporate governance and firm performance in Pakistan: The case of Karachi Stock Exchange (KSE). Journal of Economic and International Finance, 3(8), 482-491. 42 Annexes Annex 1 – List of Companies with Industry Specification Table 17: Overview of the sample of 72 UK companies with the industry specification Number Company Industry 1 ROLLS-ROYCE HOLDINGS PLC Aerospace and defence 2 BAE SYSTEMS PLC Aerospace and defence 3 GKN PLC Aerospace and defence 4 COCA-COLA HBC AG Beverages 5 DIAGEO PLC Beverages 6 SABMILLER PLC Beverages 7 JOHNSON MATTHEY PLC Chemicals 8 CRH PLC Construction and Materials 9 SSE PLC Electricity 10 MORRISON (WM) SUPERMARKETS PLC Food and Drug Retailers 11 SAINSBURY (J) PLC Food and Drug Retailers 12 TESCO PLC Food and Drug Retailers 13 ASSOCIATED BRITISH FOOD PLC Food Producers 14 MONDI PLC Forestry and Paper 15 CENTRICA PLC Gas, Water and Multiutilities 16 NATIONAL GRID PLC Gas, Water and Multiutilities 17 SEVERN TRENT PLC Gas, Water and Multiutilities 18 UNITED UTILITIES GROUP PLC Gas, Water and Multiutilities 19 REXAM PLC General Industrials 20 DIXONS CARPHONE PLC General Retailers 21 KINGFISHER PLC General Retailers 22 MARKS AND SPENCER GROUP PLC General Retailers 23 NEXT PLC General Retailers 24 MEDICLINIC INTERNATIONAL PLC Health Care Equipment and Services 25 SMITH AND NEPHEW PLC Health Care Equipment and Services 26 BARRATT DEVELOPMENTS PLC Household Goods and Home Construction 27 BERKELEY GROUP HOLDINGS (THE) PLC Household Goods and Home Construction 28 PERSIMMON PLC Household Goods and Home Construction 29 RECKITT BENCKISER GROUP PLC Household Goods and Home Construction 30 TAYLOR WIMPEY PLC Household Goods and Home Construction 31 ROYAL MAIL PLC Industrial Transportation 32 INFORMA PLC Media 33 ITV PLC Media 34 PEARSON PLC Media 35 RELX PLC Media 36 SKY PLC Media 37 WPP PLC Media 38 ANGLO AMERICAN PLC Mining 39 ANTOFAGASTA PLC Mining 40 BHP BILLITON PLC Mining 41 FRESNILLO PLC Mining 43 42 GLENCORE PLC Mining 43 RANDGOLD RESOURCES LD Mining 44 RIO TINTO PLC Mining 45 INMARSAT PLC Mobile Telecommunications 46 VODAFONE GROUP PLC Mobile Telecommunications 47 BP PLC Oil and Gas Producers 48 ROYAL DUTCH SHELL PLC Oil and Gas Producers 49 BURBERRY GROUP PLC Personal Goods 50 UNILEVER PLC Personal Goods 51 ASTRAZENECA PLC Pharmaceuticals and Biotechnology 52 GLAXOSMITHKLINE PLC Pharmaceuticals and Biotechnology 53 SHIRE PLC Pharmaceuticals and Biotechnology 54 SAGE GROUP PLC Software and Computer Services 55 ASHTEAD GROUP PLC Support Services 56 BABCOCK INTERNATIONAL GROUP PLC Support Services 57 BUNZL PLC Support Services 58 CAPITA PLC Support Services 59 DCC PLC Support Services 60 EXPERIAN PLC Support Services 61 INTERTEK GROUP PLC Support Services 62 TRAVIS PERKINS PLC Support Services 63 WOLSELEY PLC Support Services 64 ARM HOLDINGS PLC Technology Hardware and Equipment 65 BRITISH AMERICAN TOBACCO PLC Tobacco 66 COMPASS GROUP PLC Travel and Leisure 67 EASYJET PLC Travel and Leisure 68 INTERCONTINENTAL HOTELS GROUP PLC Travel and Leisure 69 MERLIN ENTERTAINMENTS PLC Travel and Leisure 70 PADDY POWER BEDFAIR PLC Travel and Leisure 71 TUI AG Travel and Leisure 72 WHITBREAD PLC Travel and Leisure 44 Figure 02: Overview of the industries 0 1 2 3 4 5 6 7 8 9 10 Industry Number of companies Overview of Industries Support Services Mining Travel and Leisure Media Household Goods and Home Construction Gas, Water and Multiutilities General Retailers Aerospace and defense Beverages Food and Drug Retailers Pharmaceuticals and Biotechnology Health Care Equipment and Services 45 Annex 2 – Audit Committee Data Table 18: Audit committee data Company Year Size of AC Num. of meetings per year Num. of indep. directors Num. of directors with fin. exp. ROLLS-ROYCE HOLDINGS PLC 2015 5 4 5 3 2014 4 5 4 2 2013 4 4 4 3 2012 4 4 4 3 2011 4 4 4 3 BAE SYSTEMS PLC 2015 3 5 3 2 2014 3 7 3 2 2013 3 6 3 2 2012 3 6 3 1 2011 3 6 3 1 GKN PLC 2015 4 6 4 4 2014 4 5 4 4 2013 4 4 4 4 2012 4 4 4 4 2011 5 5 5 5 COCA-COLA HBC AG 2015 4 9 4 2 2014 3 9 3 1 2013 3 8 3 1 2012 3 8 3 1 2011 3 7 3 1 DIAGEO PLC 2015 8 4 8 1 2014 8 4 8 1 2013 8 4 8 1 2012 8 6 8 1 2011 8 6 8 1 SABMILLER PLC 2015 5 4 4 4 2014 5 4 4 3 2013 5 4 4 2 2012 6 4 5 4 2011 6 4 5 4 JOHNSON MATTHEY PLC 2015 5 5 5 2 2014 5 5 5 2 2013 5 5 5 2 2012 6 4 6 2 2011 6 4 6 2 CRH PLC 2015 4 9 4 1 2014 5 10 5 1 2013 5 8 5 1 2012 5 8 5 1 2011 4 9 4 1 SSE PLC 2015 5 3 5 2 2014 3 3 3 1 2013 4 3 4 1 2012 4 3 4 1 2011 4 3 4 1 MORRISON (WM) SUPERMARKETS PLC 2015 4 7 4 1 2014 4 6 4 1 2013 4 9 4 1 2012 4 6 4 1 2011 4 6 4 1 SAINSBURY (J) PLC 2015 3 5 3 1 2014 3 4 3 1 2013 3 4 3 1 2012 4 4 4 1 2011 4 4 4 1 46 TESCO PLC 2015 6 8 6 4 2014 4 5 4 4 2013 4 5 4 2 2012 4 5 4 2 2011 5 5 5 3 ASSOCIATED BRITISH FOOD PLC 2015 5 5 5 1 2014 4 5 4 1 2013 3 4 3 1 2012 3 4 3 1 2011 3 4 3 1 MONDI PLC 2015 3 4 3 1 2014 3 4 3 1 2013 3 4 3 1 2012 3 4 3 1 2011 4 4 4 1 CENTRICA PLC 2015 4 4 4 1 2014 6 4 6 2 2013 6 4 6 2 2012 5 4 5 2 2011 6 4 6 2 NATIONAL GRID PLC 2015 4 8 1 1 2014 5 6 1 1 2013 6 6 2 2 2012 5 6 1 1 2011 4 6 1 1 SEVERN TRENT PLC 2015 4 4 4 4 2014 4 4 4 4 2013 3 5 3 2 2012 3 4 3 2 2011 3 5 3 2 UNITED UTILITIES GROUP PLC 2015 3 4 3 1 2014 4 4 4 1 2013 4 4 4 1 2012 3 5 3 1 2011 3 4 3 1 REXAM PLC 2015 3 5 3 1 2014 3 4 3 1 2013 3 4 3 1 2012 4 4 4 1 2011 3 4 3 1 DIXONS CARPHONE PLC 2015 3 3 3 1 2014 4 3 4 1 2013 3 3 3 1 2012 3 3 3 1 2011 5 4 5 1 KINGFISHER PLC 2015 4 4 4 1 2014 4 4 4 1 2013 4 4 4 1 2012 4 4 4 1 2011 4 4 4 1 MARKS AND SPENCER GROUP PLC 2015 6 5 6 1 2014 6 6 6 2 2013 5 5 5 3 2012 6 6 6 2 2011 6 5 6 2 NEXT PLC 2015 5 5 5 1 2014 5 4 5 1 2013 5 4 5 1 2012 4 4 4 1 2011 4 5 4 1 MEDICLINIC 2015 4 3 4 1 47 INTERNATIONAL PLC 2014 4 4 4 1 2013 4 3 4 1 2012 4 3 4 1 2011 4 3 4 1 SMITH AND NEPHEW PLC 2015 5 7 5 1 2014 4 5 4 1 2013 4 8 4 1 2012 5 8 5 1 2011 5 6 5 1 BARRATT DEVELOPMENTS PLC 2015 4 4 4 1 2014 5 4 5 1 2013 6 5 6 1 2012 3 3 3 1 2011 4 4 4 1 BERKELEY GROUP HOLDINGS (THE) PLC 2015 4 3 4 2 2014 5 3 5 3 2013 4 3 4 1 2012 3 3 3 1 2011 3 3 3 1 PERSIMMON PLC 2015 4 6 4 3 2014 3 4 3 2 2013 3 6 3 3 2012 3 4 3 3 2011 3 4 3 3 RECKITT BENCKISER GROUP PLC 2015 5 4 5 2 2014 5 4 5 4 2013 3 4 3 2 2012 3 4 3 2 2011 3 4 3 2 TAYLOR WIMPEY PLC 2015 4 3 4 2 2014 4 3 4 1 2013 4 3 4 1 2012 4 4 4 1 2011 3 3 3 1 ROYAL MAIL PLC 2015 6 5 6 2 2014 6 9 6 2 2013 6 5 6 2 2012 6 5 6 1 2011 6 5 6 1 INFORMA PLC 2015 5 3 5 2 2014 4 3 4 2 2013 4 3 4 2 2012 3 3 3 2 2011 3 3 3 2 ITV PLC 2015 3 7 3 1 2014 3 5 3 1 2013 4 5 4 1 2012 3 4 3 1 2011 3 8 3 1 PEARSON PLC 2015 5 4 5 1 2014 6 5 6 1 2013 5 5 5 1 2012 6 4 6 1 2011 7 4 7 1 RELX PLC 2015 4 7 4 2 2014 3 5 3 2 2013 4 5 4 2 2012 4 5 4 3 2011 4 5 4 3 SKY PLC 2015 4 6 4 4 2014 5 6 5 5 48 2013 4 6 4 4 2012 5 6 5 5 2011 3 4 3 3 WPP PLC 2015 7 8 7 1 2014 4 9 4 1 2013 5 7 5 1 2012 4 9 4 1 2011 4 7 4 1 ANGLO AMERICAN PLC 2015 5 4 5 1 2014 6 3 6 1 2013 6 4 6 1 2012 5 3 5 1 2011 4 3 4 1 ANTOFAGASTA PLC 2015 3 4 3 2 2014 3 4 3 2 2013 3 6 3 2 2012 3 5 3 2 2011 3 5 3 2 BHP BILLITON PLC 2015 5 8 5 1 2014 4 9 4 1 2013 4 12 4 1 2012 5 11 5 1 2011 4 9 4 1 FRESNILLO PLC 2015 3 5 3 1 2014 3 5 3 1 2013 3 5 3 1 2012 3 5 3 1 2011 3 5 3 1 GLENCORE PLC 2015 3 4 3 3 2014 3 4 3 3 2013 3 4 3 3 2012 3 5 3 2 2011 3 2 3 2 RANDGOLD RESOURCES LD 2015 4 6 4 4 2014 4 6 4 4 2013 3 6 3 3 2012 3 6 3 3 2011 3 6 3 3 RIO TINTO PLC 2015 5 6 5 1 2014 5 6 5 1 2013 4 7 4 1 2012 4 6 4 1 2011 5 6 5 1 INMARSAT PLC 2015 5 4 5 3 2014 4 4 4 3 2013 4 4 4 3 2012 4 5 4 3 2011 4 4 4 3 VODAFONE GROUP PLC 2015 4 4 4 1 2014 4 4 4 1 2013 4 4 4 1 2012 4 4 4 1 2011 4 4 4 1 BP PLC 2015 5 11 5 1 2014 4 13 4 1 2013 4 12 4 1 2012 4 11 4 1 2011 4 11 4 1 ROYAL DUTCH SHELL PLC 2015 4 6 4 1 2014 4 6 4 1 2013 4 6 4 1 49 2012 4 5 4 1 2011 4 5 4 1 BURBERRY GROUP PLC 2015 8 3 8 1 2014 6 3 6 1 2013 5 3 5 1 2012 5 3 5 1 2011 5 5 5 3 UNILEVER PLC 2015 4 8 4 4 2014 4 8 4 1 2013 4 9 4 4 2012 3 7 3 3 2011 3 5 3 3 ASTRAZENECA PLC 2015 5 5 5 1 2014 5 5 5 1 2013 5 5 5 1 2012 5 7 5 1 2011 5 6 5 1 GLAXOSMITHKLINE PLC 2015 8 6 8 3 2014 8 6 8 3 2013 8 6 8 3 2012 7 6 7 3 2011 7 6 7 3 SHIRE PLC 2015 5 5 5 1 2014 5 6 5 1 2013 3 5 3 1 2012 3 5 3 1 2011 4 5 4 1 SAGE GROUP PLC 2015 5 4 5 1 2014 5 5 5 1 2013 5 4 5 1 2012 4 4 4 1 2011 4 4 4 1 ASHTEAD GROUP PLC 2015 3 5 3 1 2014 4 5 4 2 2013 3 5 3 1 2012 3 4 3 1 2011 3 4 3 1 BABCOCK INTERNATIONAL GROUP PLC 2015 6 4 6 1 2014 5 4 5 1 2013 5 4 5 1 2012 6 4 6 1 2011 5 4 5 1 BUNZL PLC 2015 5 4 5 1 2014 5 5 5 1 2013 5 5 5 1 2012 4 4 4 1 2011 4 3 4 1 CAPITA PLC 2015 5 7 4 1 2014 4 9 3 1 2013 4 6 3 1 2012 4 6 4 1 2011 3 5 3 1 DCC PLC 2015 4 6 4 4 2014 4 4 4 4 2013 4 4 4 1 2012 3 7 3 3 2011 3 5 3 3 EXPERIAN PLC 2015 7 4 7 3 2014 8 5 8 2 2013 7 4 7 2 2012 6 4 6 1