Company Board Size and Performance in the UK and USA

This study examined a broad sample of 200 firms for both the UK and USA, ranked by year end market capitalisation 2010; I examine corporate board structure and its association with firm’s performance. Using Return on Invested Capital and Return on Assets as performance measures, the findings of this paper revealed a significant inverse relationship between board size and firm performance. However, it yielded a significantly positive relationship with the number of independent board members and firm performance. Finally, my paper highlighted a positive relationship with CEO duality in the USA; however, this relationship did not prove to be significant. Overall, my results are suggest that small board of directors should be support and the number of independent directors should be sustained and improved upon.

Introduction

Thesis statement:

‘A large board of directors improves corporate financial performance, in the UK and USA’.

The role, structure and efficiency of the board of directors are topics which have been broadly studied due to the importance of the board of director’s role in stipulating and implementing company policies. I therefore revisit the degree by which the size and structure of the boardroom, impact the financial performance of firms.

Board of directors are a body of elected people who oversee the day-to-day activities of the company. They have many responsibilities which include: hiring and firing CEO’s, reporting on the firm’s financial performance to shareholders, overseeing a firms policies are implemented and “helping management make good decisions about firm strategy and actions” Linck et al (2008). It is also the responsibility of the board of directors to “scrutinise management behaviour to guard against harmful behaviour, ranging from shirking to fraud” Linck et al (2008).

In this research paper I will empirically investigate the relationship between board size and performance of companies in both the UK and the USA, using the top 200 companies in each country. To further my research and to extend prior literature, I will examine if the composition of the board; i.e. number of independent directors, has any impact on a firms financial performance. I will also look at the role of CEO duality, which is when the same person holds both the job of Chief Executive Officer (CEO) and Chairman of the board of a firm, to see if any relationships arise between them and firm performance.

Specifically, the research questions which are the basis for hypothesis formation are:

Does board size affect corporate financial performance in the UK & USA?

Does the number of independent directors have any relationship with corporate financial performance in the UK or USA?

Is there any relationship between CEO duality and corporate financial performance in the UK or USA?

There is an extensive body of literature which exists in relation to my research questions. However, there is not one but various conclusions drawn on each question. To date, no one theory exists on board structure. My research involves empirically analysing each research question to determine which view my paper supports.

From these research questions I have formed the following hypothesis to empirically test in my study:

H1: There is a positive relationship between board size and corporate financial performance in the UK and USA.

H2: There is a positive relationship between proportion of independent directors sitting on the board and corporate financial performance, in UK and USA.

H3: There is a positive relationship between CEO duality and corporate performance in the USA.

1.1 Findings of My Paper

In this study, my results show that both my performance measures Return on Assets and Return on Invested Capital showed significant negative correlation with board size in the UK and the USA. Indicating; that there is a strong negative association between large boards and corporate financial performance. My results also suggest that the number of independent directors is positively correlated with firm performance in both the UK and USA; which tells us that the greater the number of independent directors improves firm performance. Finally, my results highlight that CEO duality shows positive but not significant correlation with firm performance in the USA, however, it could not be tested in the UK as none of my UK sample employed CEO duality. Therefore, this study recommends that smaller board size should be encouraged and the composition of outside directors as members of the board should be sustained and improved upon.

The rest of the paper is organised as follows: section two looks at the review of literature; section three is dedicated to data analysis and methodology, section four highlights the results obtained from regressions run; whilst section five draws conclusions to the paper, and looks at possible areas to advance the research. Finally, the reference section outlines papers which I reviewed throughout my study.

2. Literature Review

The effectiveness of the boardroom is an issue which has been highly scrutinised for many years; with a lot of emphasis being placed on the composition of the board. As a legal prerequisite for many organisations, board of directors are seen as a means to ameliorate the agency problem which is a central issue in many firms.

However, there is no recognised theory which exists on board of directors at this point. A majority of the empirical work which has been carried out on boards have been designed to answer the following questions: 1) How is board size and profitability related? 2) Does the number of independent directors affect a firm’s performance? 3) Is board leadership related to a firm’s performance?

There are however, many issues which can complicate the studies carried out on board of directors; with the main issue being endogeneity (Hermalin and Weisbach 2003). Variables such as performance can prove endogenous as they are not accredited to the current board members but also to preceding directors. Therefore, results obtained in many studies on board of directors may be obscured as they overlook this issue when interpreting results.

The findings from empirical studies carried out on boards provide mixed reports. There is literature to prove both a negative and positive relationship exists with corporate performance and board size. Board independence proves the same, as there are both literatures to back the theory that less independent members on boards are related to improved firm performance and more independent directors on boards are associated with improved firm performance. I am now going to look at each section of board composition separately and the literature behind it.

2.1 Board size and corporate financial performance

The issue relating board size and corporate financial performance proves to be highly controversial. Despite the fact that a majority of the literature on board size primarily proposes that smaller boards perform better (Yermack,1996), there is also literature to contradict this (Pearce and Zahra1992).

Larger boards have the advantage that they have a more diverse range of skills than that of a smaller board. They also have the added benefit of having a wider spread of expertise and experience in comparison to smaller boards, which can prove beneficial to a firm. Booth and Del (1996) suggest that large firms are likely to have more external contracting relationships and, therefore require larger boards (Pfeffer, 1972).

However, Jensen (1993) and Lipton and Lorsch (1992) propose that larger boards can be less efficient than smaller board of directors. This is owing to the theory that when board of directors become too large, there is the assumption that as a consequence the agency problem will increase. This view is consistent with Eisenberg et. al. (1998) findings; that, board size and the value of firm’s are negatively correlated. There is also the view of Bhagat and Black (2002), who found no dependable verification on the relationship concerning board size and firm performance. Therefore, I am going to empirically test the thesis statement ‘A large board of directors improves corporate financial performance in the UK and USA’, to find which literature my results are consistent with.

2.2 Independent board member and performance

The evidence to support the relationship between board composition and a firm’s financial performance is assorted. To date, a majority of the literature concludes that there is no positive relationship between board independence and firm performance. For example Bhagat & Black (2001) found a negative relationship between board independence and firm performance.

Link et al (2008) disagrees with this literature, with findings that outside directors, bring valuable expertise and potentially imperative connections to ¬rms. However, (Fama and Jensen, 1983) contradict these findings concluding that inside directors possess more ¬rm speci¬c knowledge. Suggesting that, the composition of boards should consist of more inside directors.

There are numerous reasons to suggest why a majority of literature conclude that no positive significant relationship between board independence and firm performance exists. Board of directors being in place as a means for window dressing the company; incentives to maintain independence; independent directors do not have the same extent of knowledge of a company as inside directors; and finally CEO’s are said to have ‘captured’ the independent board members. Although, as Baysinger and Butler (1985) suggests; it may instead prove more beneficial, to have a mixed board of directors.

.

2.3 CEO Duality

The role of the CEO and the chairman of the board are interrelated. This is why in many firms it is becoming more common to the see that both roles have now been amalgamated and is now known as CEO duality, i.e. one person doing both jobs.

The chief executive officer is monitored by the board of directors; this means that having a chief executive officer who is also the chairman of the board can leave the firm open to the agency problem. Dual roles may result in the CEO acting in their own interest which is what Mallette and Fowler (1992) argued. They argue that in dual roles the chairman of the board will make decisions that influence his personal welfare as CEO, which may lead to a conflict of interest. Also, as chairman the chief executive officer has the power to influence the appointment of new directors of the board and dismissal of existing board members. This power can potentially result in the chairman acting in the interest of himself and therefore increase the agency problem.

For many years now researchers have tried to establish a relationship between CEO duality and firm’s performance. However, as was the case with board size and the number of independent directors; no one conclusion can be agreed on. Brickley, Coles and Jarrell (1997) conclude that firms with a separate chairman and CEO role do not have an improved profit. In contrast to this view, Palmon and Wald (2002) point out that a change

from a single to dual leadership structure leads to positive irregular returns for large firms . In support of this view, Rechner and Dalton (1991) found that firms with a dual leadership outperformed firms with a single leadership structure.

3.Data and Methodology

3.1 Data Collection

I have collected data on the top 200 companies within the United Kingdom and the top 200 within the USA, based on their Year End Market Capitalisation for the year 2010. These top companies were chosen because they are the firms most likely to require the board of directors to help ameliorate the agency problem. The measure of financial performance (Return on Assets, Return on Invested Capital) and other data required on these firms have been obtained from the Thompson One Banker Analytics database. The boardroom’s structural configuration information has been retrieved from the companies’ annual reports.

3.1.2 Problems collecting data

Collecting the data proved problematic, as I encountered a few obstacles. I started out with a sample size of 200 companies for each country; however, my final sample size was just 135 for the UK and 128 for the USA, as it was only these companies who met the criteria I required for my study. Ideally, my sample size would have been larger, however, gathering the data proved tedious and time consuming as it involved going into each companies’ annual reports in order to collect the data by hand. There was also the problem that many companies did not specify if the members of their board where independent, therefore these companies had to be removed from the sample to prevent my results from being obscured.

3.2 Performance measures

In order to check the robustness of my model and to verify if my results obtained are consistent with other performance measures, I will run regressions using two different performance measures, namely: Return on Assets and Return on Invested Capital. I will first run my regression using the performance measure Return on Assets which indicates to a firm how profitable they are relative to their total assets. It gives the firm an idea of how efficient management are at using the assets available to them in order to generate profit. I will run a second regression using Return on Invested Capital as my performance measure; this allows a company to assess how efficient it is at allocating the capital under its control to generate returns. Most papers which have researched this topic have produced findings using Return on Assets as their measure of performance. This will able me to make useful comparisons with other studies, whilst at the same time further their research as I have used a second performance measure, Return on Invested Capital.

3.3 Methodology

From the data outlined, I will be able to determine answers to my research questions.

To answer these questions I will use an ordinary least square regression consisting of dummy and non dummy variables. The technique of Ordinary Least Squares (OLS) is attributed to the German Mathematician Carl Friedrich Gauss. It is a technique which estimates the unidentified parameters in a linear regression model by minimising the sum of residuals squared. Its estimators can be shown to have desirable properties that are consistent, unbiased and efficient.

3.3.1 Independent/Dependent variables

The dependent variable for this study is corporate financial performance which is represented by Return on Assets and Return on Invested Capital. Return on assets is given by dividing a company’s annual earnings by its total assets and, return on capital is given by dividing the after-tax operating income by the book value of invested capital.

The independent variables are the board’s size, market capitalisation, total assets, beta, CEO duality and the number of independent directors on the board. Were board size is defined as the total number of directors in the boardroom. Market capitalisation is the total value of the shares of a company. Total assets are the sum of current and long term assets owned by a firm. The beta of a stock is a number describing the relation of its returns with that of the financial market as a whole. CEO duality is when the CEO is also the chairman of the board, as a proxy for CEO duality we use a dummy variable which equals one when the CEO is also the chairman of the board and zero if they are not. The number of independent directors in the boardroom is the number of completely independent directors which are present on the board of directors.

My regression equations are as follows:

ROA= β0 + β1BSIZE + β2BCOMP + β3 CEODUALITY + β4BETA + β5MKTCAP+ β 6TOTALASSETS + e

ROIC= β0 + β1BSIZE + β2BCOMP + β3 CEODUALITY + β4BETA + β5MKTCAP+ β 6TOTALASSETS + e

Each of the above regressions will be run for both the UK and the USA.

Where:

β0 = Intercept coefficient

β1-6 = Coefficient for each of the independent variables

BSIZE = Number of directors on the board

BCOMP = Proportion of independent directors sitting on the board

BCEO = Value 1 if the same person occupies the position of the chairman and the chief executive and 0 for otherwise.

BMKTCAP = Year end market capitalisation for each company as of yearend 2010.

BTOTALASSETS = Total assets for each company.

The findings of my regression are discussed in the section 4.

3.3.2 Limitations to my model

It is important to acknowledge that there are limitations to using the ordinary least square method. The model proves to be sensitive to the presence of outliers, which can skew the results obtained and can lead to questioning their validity. There is also the problem that removing the outlier’s will mean that results obtained will no longer be a true representation of the sample.

4. Results

4.1 Summary statistics

The main aim of my research paper was to find if board structure i.e. size, composition and CEO duality impacted upon a firm’s financial performance. Table 1 & 2 below shows the summary statistics across the UK and USA respectively, which highlights the variations in each variable. These tables show the size of the sample I used for each country, as well as the mean, standard deviation, minimum and maximum value for each of the variables used.

Table 1: Summary statistics UK

Variable

Table 1 illustrates that board size has a range of 15, with a minimum number of directors of 4 and a maximum number of 19. This explains that the size of the board of directors varies considerably thorough out the UK, with an average board size of 9.78585.

Table 1 also establishes that the number of independent directors making up the boards of companies within the UK varies from 1 independent member to 19 independent members, which shows a range of 18. The summary statistic also highlight to us that no one company has a fully non-independent board, however 3 companies have fully independent boards.

CEO duality is not shown in the summary statistics for the UK as I found that none of the companies within my sample employed this method.

Table 2: Summary Statistics USA

Variable

Table 2 shows that board size in the USA has a range of 29, with a minimum number of directors of 3 and a maximum number of 32. This shows that the size of the board of directors varies considerably thorough out the USA, with an average board size of 15.6723.

This table also shows us that the number of independent directors making up the board of companies within the USA varies from 1 independent member to 32 independent members, which shows a range of 31. The summary statistic also highlight to us that no one company has a fully non-independent board, however 1 company has a fully independent board.

CEO duality is represented by a dummy variable in the sample. As shown by the summary statistics the average CEO duality is 0.231392; meaning that almost a quarter of the sample employ CEO duality within their firm.

4.2 Regression Results UK

Table 3 below illustrates the results obtained from the 2 regressions which I ran for the UK. I regressed Return on Assets with board size, number of independent members, debt, total assets, beta and market capitalisation; using the method of ordinary least squares. My second regression replaced Return on Assets with Return on Invested Capital regressed against the same explanatory variables.

Table 3: Regression results UK

VARIABLES

Return on Assets

Return on Invested Capital

t-statistics in parentheses

*** p<0.01, ** p<0.05, * p<0.1

The regression results in table 3 above shows, that Return on Assets is negatively correlated (-1.327) with board size; this proves to be significant at the 1% level. However, Return on Assets is positively correlated (0.524) with the number of independent directors, which is significant at the 10% level.

My results obtained from my second regression using Return on Invested Capital as a measure of performance are consistent with that achieved when regressing Return on Assets. Return on Invested Capital is also negatively correlated (-1.448) with board size, this is significant at the 1% level. Return on Invested Capital is positively correlated (1.077) with the number of independent directors; this is significant at the 5% level.

The R2 value, which indicates the explanatory power of the independent variables, is 0.241 and 0.142 respectively. This means that 24.1% of the variation in ROA is explained by the variation in the independent variables, while 14.2% of the variation in ROIC is explained by the variation in the independent variables. A higher R2 value would have been preferred in both regressions; as such low R2 values may indicate that the variables used in the regression may have been insignificant, or it may be an indication that there may have been an important variable omitted from the regression.

Overall, I can see that results from both regressions ran shows that board size has a consistent significantly negative relationship with both Return on Assets and Return on Invested Capital. These findings therefore would suggest that I cannot accept hypothesis 1: A large board of directors improves corporate financial performance’. There also proves to be consistency with results obtained for the number of independent directors and corporate financial performance. Both regressions show positive significant relationships between the number of independent directors and corporate financial performance; which means that I can accept hypothesis 2. As previously mentioned, hypothesis 3 could not be tested for the UK as none of the sample I used for the UK employed CEO duality.

Table 4: Regression results USA

Table 4 illustrates findings for the regressions ran for the USA sample. The same regression where ran for the USA as was for the UK, and results seem to be consistent with that found for the UK sample.

VARIABLES

Return on Assets

Return on Invested Capital

t-statistics in parentheses

*** p<0.01, ** p<0.05,*p<0.1

It was found that performance measure Return on Assets is negatively correlated (-0.715) with board size, significant at the 1% level. On the other hand, it is positively correlated (0.439) with the number of independent directors; significant at the 10% level, which is consistent with the UK results. The dummy variable of CEO duality proves to be positively correlated (0.306) with firm performance; however its results are not significant.

The second performance measure Return on Invested Capital yielded results that were also consistent with UK results when Return on Invested Capital was regressed with the same variables. It proved that Return on Invested Capital is negatively correlated (-0.859) with board size; significant at the 5% level. However, positively correlated (0.732) with the number of independent directors, this is significant at the 10% level. As was found with Return on Assets, CEO duality is slightly positively correlated (0.221) with Return on Invested Capital, although this result did not prove significant either.

The R2 values, 0.672 and 0.558 respectively, meaning that 67.2% and 55.8% of the change in financial performance is influenced by the explanatory variables. Unlike the R2 values for the UK regressions, they proved to be a lot higher for the USA which indicates that the explanatory variables proved to be a lot more reliable and of more significance in the USA sample than the UK sample.

Overall it can be said that comparing results obtained for both the UK and USA prove to be consistent. There were no major anomalies between the results for each country. Both sets of results concluded that there is a negative relationship with board size and firm performance, and a positive relationship between the number of independent directors and firm performance. The USA also yielded the result that CEO duality was positively related with firm performance however, this result did not prove significant.

4.3 Board size and firms performance

From tables 3 and 4 above, it is apparent that when I control for risk (beta), firm size (market capitalisation) and total assets, I find that having smaller boards increases firm financial performance. These findings are not consistent with hypothesis one which states that : ‘Larger board of directors improves a firms financial performance’ and thus disproves my thesis statement.

The reasoning behind such findings is that just as Jensen (1993) found, for a firm to be successful in its monitoring, they should have relatively small boards, as larger boards are “less likely to function effectively”. My results prove consistent with Eisenberg et. al. (1998) who also used Return on Assets as his performance measure, that, board size and the value of firm’s are negatively correlated. It can be said that smaller board of directors have the advantages of quicker decision-making, greater focus on key issues and better management from the chairman; all possible reasons to why my results have proved that smaller boards help improve firm performance in comparison to larger boards.

4.4 Board Independence and firm’s performance

However, the board independence measures in both regressions are statistically significant, therefore supporting my hypothesis 2, that ‘Larger number of independent directors improves a firm’s performance’. As expected, it would appear that having a greater amount of independent board members will have a significant, positive impact on firm performance.

My results obtained for board independence are consistent with the empirical study of Link et al (2008) who established that outside directors bring valuable expertise and potentially important connections to ¬rms. In theory, it is likely that I have found that the more independent directors improve firm performance, as independent directors will not be subject to pressure as inside directors are, and therefore are more likely to act in the shareholders interests.

4.5 CEO Duality and firm’s performance

CEO duality showed positive correlation with both measures of firm performance in the USA. This to some degree supports my hypothesis that ‘CEO duality is associated with improved financial performance’. However, as the result did not prove to be significant I cannot completely rely on the validity of the result.

5. Conclusion

This paper examines the determinants of board structure within the UK and USA, using a sample of 200 firms from each country in 2010. Consistent with empirical evidence, I find that corporate boards are more efficient when they are smaller and are highly concentrated with independent directors.

The main findings of this paper suggest that there is a significant, negative relationship between board size and firm performance in