200671838 - Dissertation

Leeds University
Business School
IMPACT OF MANAGERIAL OWNERSHIP ON FIRM PERFORMANCE DURING
FINANCIAL CRISIS – EVIDENCE FROM UK
MOHIT KUMAR
(200671838)
This dissertation is submitted in part fulfilment of the requirements for the degree of
MSc FINANCE & INVESTMENT
THE UNIVERSITY OF LEEDS
LEEDS UNIVERSITY BUSINESS SCHOOL
Dissertation supervisor: Ms HELEN SHORT
Month and year of submission AUGUST 2012
Word count: 10,483

FOR OFFICE USE ONLY
SCRIPT NO
LATE DAYS
Declaration of Academic Integrity
To be attached to any assignment, dissertation or project work submitted for assessment as part of a University assessment.
I have read and understood the Business School and University Regulations on Cheating and Plagiarism.
I state that this piece of work is my/our own and does not contain any unacknowledged work or text sources, which is not referenced.
I agree that this piece of work may be submitted to the plagiarism detection software currently used by the University and
consequently uploaded to the software database.
LUBS/MATH 5 0 5 9 M
MODULE TITLE: MSC FINANCE AND INVESTMENT
Please tick as applicable:
I do not wish this dissertation to be available to students on the same programme in future years.
Or
I have no objections to this dissertation being available to students on the same programme in future years.
SIGNED: DATE: 30/08/2012
NAME (S): [CAPITAL LETTERS]: MOHIT KUMAR

3
ABSTRACT
This paper studies the impact of insider ownership on the economic performance of a firm from sample of 180
firms listed on the FTSE 350 (London Stock Exchange) for the period 2009-2011. A series of OLS regressions
have been used along with 2SLS regressions to tackle the endogeneity that arises from inter-dependence of firm
performance and managerial ownership. Additionally, piece-wise linear regressions have been used to replicate
earlier studies and see if the findings of those papers hold true with the current data. I also look at the role of
institutional ownership as it forms a dominant portion of the shareholdings of the firms in my sample. I find no
evidence of a systematic relation between ownership structure and firm performance. It is also relevant to note
that I find ownership structure to be an endogenous outcome derived from the inter-play of various multi-
dimensional market forces.

4
ACKNOWLEDGEMENTS
I would like to thank my supervisor Ms Helen Short for her guidance, expertise and understanding. Mr
Konstantinos Bozos was a great source of inspiration for me, which helped me complete the project this year. I
sincerely thank him for his invaluable advice. This research would not have been possible with out the support
of my family and friends. Last but not the least, I am grateful to the University of Leeds for bestowing upon me
the opportunity to study MSc Finance and Investment here in the UK.

5
TABLE OF CONTENTS
A) FIGURES/TABLES 6
1) INTRODUCTION 7
2) LITERATURE REVIEW 9
3) RESEARCH METHODS 19
3.1) ECONOMETRIC MODEL 19
3.2) SAMPLE DATA 22
3.3) METHODOLOGY 22
3.4) VARIABLES 23
4) FINDINGS 28
4.1) RESULTS 28
4.2) ROBUSTNESS TESTS 36
5) CONCLUSIONS 37
6) REFERENCES 38
7) APPENDIX 39

6
TABLES
1) Ownership of publicly listed corporations in the UK 17
2) Table describing variables 19
3) Table showing summary statistics of ownership data 24
4) Table showing frequency distribution of ownership data 25
5) Table depicting summary univariate statistics 27
6) Correlation Matrix of all variables 27
7) Piece-wise estimates using 2010 ownership data 28
8) OLS estimates using 2010 ownership data 30
9) OLS estimates using 2011 ownership data 31
10) 2SLS estimates using 2010 ownership data 32
11) 2SLS estimates using 2011 ownership data 34
12) Tools used in STATA 36
13) Fig 1 depicting findings of earlier studies 18

7
CHAPTER 1: INTRODUCTION
There are numerous studies regarding the impact of managerial ownership on firm performance, which have
surfaced in the last three decades. On one hand, Jensen and Meckling (1976), (Morck et al., 1988), Hermalin
and Weisbach (1987), McConnell and Servaes (1990), Kole (1995), McConnell et al. (2008), Short and Keasey
(1999) among others argue that since the managers are taking executive decisions and control the efficiency of
the company, it only makes sense that they be given additional ownership in order to align their interests with
that of the shareholders. On the other hand, Demsetz (1983) Demsetz and Lehn (1985), Cho (1998)
,Himmelberg et al. (1999), Loderer and Martin (1997) , Holderness et al. (1999), Agrawal and Knoeber (1996)
contend that the observed level of share ownership by insiders and firm performance is the outcome of market
forces such that each firm's ownership structure is optimal for that firm. If so, changes in ownership cannot be
used to enhance corporate value. They further argue that any observed cross-sectional empirical relation
between the level of insider share ownership and firm performance must be spurious. Most studies have either
adapted Tobin’s Q or accounting rate of profit as the performance measure and focus on the shares owned by
the Board of Directors, while some also take into account the holdings of officers and CEO. There are various
other studies, which also examine the dynamics of insider ownership and its effect on corporate value.
Nonetheless, the topic is still open to debate and there is no clear agreement on the topic.
In this study, I attempt to understand the impact of insider ownership on firm performance during a financial
crisis. Specifically, I look at the US sub-prime mortgage crisis of 2008 that created ripples in financial
structures around the world the effects of which are visible even today. The advantage of focusing on crises
period is that it allows us to examine unambiguously the effect of corporate governance on firm value. Because
the crisis was, by all accounts, an unexpected event, it presents an interesting opportunity to study the
proximate effect of corporate governance on firm performance during a period of crisis.
My results indicate no proof of a relationship between the ownership structure and performance of firms. These
results are consistent across all three sets of regression utilized in this paper. Also, ownership structure is found
to be an endogenous outcome of different market forces. These results are consistent with results reported by
Demsetz and Villalonga, Cho, Agarwal and Knoeber among others. It signifies that shareholders are aware of
the consequences when they make decisions to dilute the shareholdings and that their goal is value-
maximization because of which such paths are undertaken. The agency costs and disadvantages of these

8
decisions are offset by the advantages of having a larger firm size and capital which implies greater control and
command over the market.
The rest of the paper is structured in the following manner. The second chapter discusses the extant literature
present on the relationship between ownership structure and firm efficiency. The sample data, econometric
model and methodology used are presented in Chapter 3. The relevant findings and conclusions are explained
in chapters 4 & 5 respectively. References are listed in chapter 6 followed by the appendix.

9
CHAPTER 2: LITERATURE REVIEW
There has been extensive research on the relationship between ownership structure and corporate performance.
The focus of these studies has been on different aspects of ownership structure like managerial ownership1
,
institutional ownership, influence of the largest shareholder, identity of shareholders etc. As mentioned earlier,
the focus in our study is on the impact of managerial ownership on the performance of firms. The basis for such
a type of study arises from the separation of ownership and control (see Fama and Jensen, 1983), and because
of diffuse ownership in large corporations. The owners of the company, i.e., the shareholders act as principals
and delegate the decision making to managers who act as agents. This agency relationship has a number of
disadvantages wherein the agent can exploit the given authority for private benefits; the agent may not take
risks in pursuance of the principal’s interests; information asymmetry whereby the agent has access to more
information. Thus it can be argued that the interests of the managers need to be aligned with the interests of the
shareholders so that the resources of the firm are not compromised by managers shirking profits. This can be
achieved by providing performance-based incentives like higher pay, managerial commission or by increasing
the manager’s shareholdings. If the managers have a stake in the company, then their goal would be the same as
other shareholders: to improve firm performance, which results in higher dividends and capital gains. The other
alternative is to increase monitoring and including measures to discipline managers so as to ensure that
managers do not abuse their power. The costs related with these measures are called agency costs. (Mallin,
2010)
On the other hand, it can be argued that the diffuseness of ownership is an endogenous outcome of market
forces, which does not affect the performance of firms as the agency costs are offset by the advantages of
having a diffuse ownership. ‘When owners of a privately held company decide to sell shares, and when
shareholders of a publicly held corporation agree to a new secondary distribution, they are, in effect, deciding to
alter the ownership structure of their firms and, with high probability, to make that structure more diffuse.
Subsequent trading of shares will reflect the desire of potential and existing owners to change their ownership
stakes in the firm. In the case of a corporate takeover, those who would be owners have a direct and dominating
influence on the firm’s ownership structure. In these ways, a firm’s ownership structure reflects decisions made
by those who own or who would own shares. The ownership structure that emerges, whether concentrated or
1
The words insider, manager, and director are used interchangeably to refer to the shareholdings of the Board
of Directors in a company. This includes both executive and non-executive directors.

10
diffuse, ought to be influenced by the profit-maximizing interests of shareholders, so that, as a result, there
should be no systematic relation between variations in ownership structure and variations in firm performance.’
(Demsetz and Villalonga 2001) Also, a diffuse ownership is more likely to be representative of a large
corporation where a certain degree of control requires comparatively more capital. Such large corporations have
the benefit large-scale economies of scale, which assist in offsetting agency costs. Moreover, the market for
corporate control and financial institutions providing debt to such firms also support in disciplining managers.
As a result of these contradicting explanations, it is important to look at how companies with different levels of
managerial holdings perform and how this affects companies in general.
The methodology, econometric models and samples used by different researchers to discover the impact of
insider ownership on firm value is varied which has resulted in contradicting conclusions and implications. The
potential problems of the separation of ownership and control were identified in the 18th
century by Smith
(1838) : ‘ the directors of such companies (joint stock companies) however being the managers rather of other
people’s money than of their own, it cannot be well expected that they should watch over it with the same
anxious vigilance (as if it were their own)’. The arguments regarding agency conflict can be traced back to
Berle and Means (1932) who propose the divergence of interest theory. ‘If we are to assume that the desire for
personal profit is the prime force motivating control, we must conclude that the interests of control are different
from and often radically opposed to those of ownership; that the owners most emphatically will not be served
by a profit-seeking controlling group.’
Jensen and Meckling (1976) formalize this relationship between corporate value and insider ownership with
empirical research and conclude that an increase in managerial ownership leads to improvement in the value of
the firm. They argue that at low-levels of ownership, due to information asymmetries between insiders and
outsiders, managers become entrenched and derive private control benefits by consuming perquisites, shirking,
building empires or actually diverting profits and assets at the expense of shareholders. They document that a
firm performs better as the proportion of insider ownership increases. It is also noted that in order to control
management at low insider ownership levels, additional agency costs for monitoring are required which
negatively affect the firm value.

11
Insiders can mask their private control benefits by managing the level and the variability of reported earnings to
reduce the likelihood of outside intervention. Insiders can overstate earnings to avoid detection of
mismanagement or outright diversion by insiders. Within this framework, insiders will become more involved
in the company when they own greater shares of the firm and, consequently, the need for outside monitoring
will be reduced, as long as the interests of insiders and external shareholders converge. Thus, under the
alignment hypothesis, insider ownership can be seen as a mechanism to constrain the opportunistic behavior of
managers.
Following Jensen & Meckling, interest in the relation between corporate value and the allocation of shares
among managers and no managers has continued to evolve on both the theoretical and the empirical front. Stulz
(1988) developed a model focusing on the importance of the takeover market for disciplining corporate
managers in which the market value of the firm first increases, and then decreases, as equity ownership is
concentrated in the hands of insiders.
Morck et al. (1988) also advance an argument concurring with the positive effect of ownership concentration in
the hands of managers, but only up to a certain level. Their study indicates that the relation between value and
inside equity ownership is nonlinear. They estimate a piece-wise linear regression in which the dependent
variable is Tobin’s Q and the primary independent variable is the fraction of shares owned by corporate
insiders. Using a sample of 371 Fortune 500 firms for 1980, the authors find that Q first rises as insider
ownership increases up to 5%, and then falls as ownership increases to 25%, then rises slightly at higher
ownership levels.
Hermalin and Weisbach, (1991) estimate similar regressions for 142 NYSE firms and find a non-monotonic
relation (inverse W-shaped relationship) between Q and the fraction of stock owned by all present and former
CEO’s still on the board of directors. However, their results are different from those of Morck et al. They find
that the relation between Q and CEO stock ownership is positive between 0% and l%, negative between 1% and
5%, positive between 5% and 20%, and negative after that.
McConnell and Servaes (1990) investigate the relation between Tobin’s Q and the structure of equity
ownership for a sample of 1,173 firms for 1976 and 1,093 firms for 1986. They find a significant curvilinear
(U-shaped) relation between Q and the fraction of common stock owned by corporate insiders. At low levels of

12
insider ownership, the relation is strongly positive. The curve slopes upward until insider ownership reaches
approximately 40% to 50% and then slopes slightly downward. In addition, McConnell and Servaes (1995)
replicate their earlier study over a later time period and report similar results. This is consistent with the results
reported by Stulz (1988). The possible explanation for the decrease of firm value at higher/intermediate levels
of insider ownership is managers tend to exert insufficient effort, collect private benefits and entrench
themselves leading to a negative relationship between managerial ownership and performance (entrenchment
effect).
A paper by Kole (1995) suggests that the different findings of Morck et al. and McConnell and Servaes are
attributable to differences in the size of the firms analyzed. Specifically, Morck et al.’s sample contains only
large firms (371 firms from the Fortune 500), while McConnell and Servaes’ sample consists of 1173 firms in
1976 and 1093 firms in 1986. Similarly, McConnell and Servaes (1995), extend their earlier work by adding a
sample of 1943 firms for 1988. Hence, the samples of McConnell and Servaes’ obviously contain firms which
are smaller than those contained in the Morck et al. sample. Kole argues that, ‘on average, the positive
relationship between Tobin’s Q and managerial ownership is sustained at higher levels of ownership for small
firms than it is for large firms.’
McConnell et al. (2008) examine stock price responses to announcements of share purchases by corporate
insiders for a sample of U.S. firms over the interval 1994 through 1999. They control for the endogeneity by
directly examining instances in which changes in share ownership are observed. Their results indicate that stock
price reaction is consistent with the curvilinear relationship between insider ownership and firm value
documented by McConnell and Servaes (1990, 1995) and others. Furthermore, they point out that even though
an optimal ownership structure exists, not all firms are always in that state.
In stark contrast, Demsetz (1983) argues that the ownership structure of the firm that ‘emerges is an
endogenous outcome of competitive selection in which various cost advantages and disadvantages are balanced
to arrive at an equilibrium organization of the firm’. Accordingly, using accounting rate of profit as a measure
of performance, Demsetz concludes that there is no relation between ownership structure and profitability. A
key difference in this paper is that all the measures of ownership structure used by Demsetz and Lehn are based
on the fraction of shares owned by a firm’s most significant shareholders, with most attention being given by
them to the fraction owned by the five largest shareholders. Demsetz and Lehn (1985), present evidence

13
consistent with Demsetz’ conclusions. They regress the accounting profit rate of 511 U.S. companies in 1980
on different measures of ownership concentration and find no significant correlation.
Moreover, Cho (1998) & Himmelberg et al. (1999) have also shed doubt upon the earlier findings of Morck et
al. (1988) and McConnell and Servaes (1990) by controlling for the effects of endogeneity and unobservable
firm characteristics in their analysis. After controlling for the effects of endogeneity in the corporate value–
managerial holdings relationship, they showed that managerial ownership had little or no effect on corporate
value and investment. Cho (1998), using cross-sectional data and ownership information from value line, first
replicates Morck et al.’s study and finds a similar non-monotonic relation between Q and management share
holdings. However, he then estimates a system of three equations in which insider ownership depends on Q,
investment, and a set of control variables, Q depends on insider ownership, investment and a set of control
variables, and investment depends on insider ownership, Q, and a set of control variables. His estimates for this
system of equations indicates that Q affects ownership structure but not vice-versa.
Himmelberg et al. 1999, extend the Demsetz and Lehn study by adding new variables to explain the variation in
ownership structure. They fit both the quadratic and linear piecewise forms that had been adopted in previous
studies for the performance equation and also control for various possible unobserved heterogeneities. They
find that insider ownership is negatively related to the capital-to-sales and R&D-to-sales ratios, but positively
related to the advertising-to-sales and operating income to sales ratios. Controlling for these variables and fixed
firm effects, they find that changes in ownership holdings have no significant impact on performance. When
they control for endogeneity of ownership by using instrumental variables, they find a quadratic form of the
effect of ownership on performance. Zhou (2001) shows, however, that the power of their approach is
questionable because most changes in ownership are small and large changes are infrequent in the relatively
homogeneous set of firms they study.
Loderer and Martin (1997) use acquisition data to estimate a simultaneous equation model in which Q and
insider holdings are endogenous. Q, log of sales, daily standard deviation of the firm’ stock returns, and daily
variance of the firm’s stock returns are used to explain insider holdings. Insider holdings, log of sales, and a
dummy for whether the acquisition is financed with stock are used to explain Q. Insider ownership fails to
predict Q, but Q is a negative predictor of insider ownership.

14
Holderness et al. (1999) replicate for 1935 and 1995 central aspects of the Morck et al. study and the Demsetz
and Lehn study. As in Morck et al., they find a significant positive relation between firm performance and
managerial ownership with the 0–5% range of managerial shareholdings but, unlike Morck et al. they do not
find a statistically significant relation beyond 5% managerial shareholdings. They also confirm the endogeneity
of managerial shareholdings, which they find depends negatively on firm size, performance volatility, volatility
squared, regulation, and leverage.
The relation between the ownership structure and the performance of corporations if ownership is made multi-
dimensional and treated as an endogenous variable is investigated by Demsetz and Villalonga (2001) and they
find no significant relationship either. It is reasoned that diffuse ownership, while it may exacerbate some
agency problems, also yields compensating advantages that generally offset such problems. Consequently, for
data that reflect market-mediated ownership structures, no systematic relation between ownership structure and
firm performance is to be expected.
Short and Keasey (1999) extend the study to UK where the corporate governance system is very different from
that of the US and test for a cubic form of the relationship between the performance of firms and managerial
ownership and with different measures of the performance of firms. They use a slightly modified model to the
one used by Morck et al., and find results consistent with a non-linear relation between ownership and value.
However, their study implicates that management gets entrenched at higher levels of ownership than their US
counter-parts. This is due to greater institutional monitoring and a lesser ability to mount take-over defenses
within the UK. In addition, the non-linear relation is shown to exist for the accounting rate of profit as well
apart from Tobin’s Q as the primary measure of performance.
Several other studies adopt even more complicated functional forms to describe this relationship. For example,
Davies et al. (2005) use a quintic structure that leads to a two-hump managerial ownership–performance curve.
Florackis et al. (2009) use a semi-parametric estimation approach and their empirical results support the
existence of the initial alignment effect of managerial ownership at levels lower than 15%, while they do not
lead to a strong inference on the relationship between managerial ownership and corporate performance for
intermediate and high levels of managerial ownership Their findings contrast previous findings in the literature,
which indicate a specific complex relationship between managerial ownership and performance at intermediate
and high levels of managerial ownership. Their results also support the view that the nature of the ownership–

15
performance relationship is likely to vary across firms with different characteristics (e.g. size). In addition, they
criticize the parametric approaches used in earlier studies and argue that the semi-parametric approach helps
sidestep concerns associated with the potential misspecification of parametric models (e.g., the arbitrary choice
of a fixed number and/or location of turning points) and enables the consideration of a wider range of non-
linear behaviors.
A critique presented by Cheung (2010) on the paper by Florackis et al. (2009) discusses that the
misspecification problem and the endogeneity problem exists in the model used apart from doubts over the
exact covariance matrix used. But he concludes that a semi-parametric approach represents a promising
approach of investigating the relationship between different variables.
Fahlenbrach and Stulz (2009) investigate the dynamics of managerial ownership and their implications for firm
value. They find that managerial ownership is more likely to fall when the firm’s stock performs or has
performed well. However, managerial ownership is not more likely to increase when the stock is performing
poorly and there is only a weak relation between past poor performance and managerial ownership increases.
However, managerial ownership increases for firms that have become financially constrained.
Coles et al. (2012) present a formal model in which a non-monotonic relation between q and managerial
ownership emerges in equilibrium. They demonstrate through calibrated simulations that their model can
replicate a concave cross-sectional relation between managerial ownership and q. However, they conclude that
the different variables and models used do not sufficiently address the problem of endogeneity and that a more
comprehensive method needs to be developed.
Thus, in summary, using Tobin’s Q as the main measure of the performance of firms, most of the the US and
UK studies2
have found the relationship between the performance of firms and managerial ownership to be,
generally, non-linear–with a movement from alignment to entrenchment and then, possibly, to alignment as
management ownership increases. The precise functional form of the relationship is, however, open to debate.
2
For a more comprehensive list of studies on the topic, visit:
http://e.viaminvest.com/A5OwnershipStructures.asp

16
Institutional Ownership
Institutional investors form a relatively large portion of the shareholdings in UK and US corporations.
Institutions are therefore expected to provide checks and balances to curb shirking by managers. In his seminal
work, Hirschman (1970) identified the exercise of institutional power within an ‘exit and voice’ framework.
This signifies that institutions should either voice their dissatisfaction to the management or sell their
shareholdings. Reports like Cadbury (1991), Greenbury (1995), and Hampel (1998) also emphasize that
institutions need to exercise their power and influence to ensure that the companies in which they have invested
comply with the Code (cited in Mallin 2010). Despite this, institutions do not take a more active role in the
management of companies. They look at the firms as merely tools of investment with no major intrinsic value.
Institutions also face free-rider problems wherein one institution might take the onus to look into the matters of
the company and bear the monitoring costs but the benefits of such actions are borne by other institutions that
also have a stake in the same company. Thus the incentives for institutions to intervene in the functioning of the
company are too few and far in-between. However, there are a few cases where institutions have intervened but
majority of the intervention is carried out in private rather than in public and that too as a last resort during time
of crisis.3
(Short and Keasey 1997)
Institutional ownership for our sample is in the range of 32-33%, which is comparatively lower to the 2010
average shown in Table 1 below. Although institutional ownership has significantly decreased from 60.8% in
1991 to 41% in 2010, this form of ownership still forms a major part of firms and thus it is necessary to look at
its impact on the performance of firms.
3
For the cases of institutional intervention in company matters and circumstances surrounding such
interventions, look at Black and Coffee (2004, cited in Short et al. 1997)

17
Ownership of publicly listed corporations in the UK
1991 (%) 1998 (%) 2008 (%) 2010 (%)
Financial Sectors
Pension Funds a
30.4 21.7 12.8 5.1
Insurance Companies 18.4 21.6 13.4 8.6
Banks b
0.9 0.6 3.5 2.5
Investment Companies and others 11.1 6.0 13.7 24.8
Total Financial sector 60.8 49.9 43.4 41.0
Non-Financial sectors
Non-Financial Businesses 3.6 1.4 3.0 2.3
Households 21.3 18.1 11.0 12.4
Government 2.0 0.1 1.1 3.1
Foreign 12.3 30.5 41.5 41.2
Total 100.0 100.0 100.0 100.0
Table 1
a
Public and private.
b
All types, including bank holding companies.
Source for 1991: Short & Keasey (1999, Table 1, pg. 83)
Source for 1998, 2008 & 2010: Office for National Statistics (Ownership of UK quoted shares 2010, pg.3)

18
Fig 1.
Results of empirical studies showing the relation between Tobin’s Q and insider ownership.
(Demsetz and Villalonga, 2001)

19
CHAPTER 3: RESEARCH METHODS
3.1 ECONOMETRIC MODEL
The general hypothesis examined here is that the performance of firms is non-linearly related to the percentage
of equity shares held by management. Following Short et al., we test for a cubic form of the relationship
between managerial ownership and firm performance. The model to be tested is as follows:
Performance = α + β1 DIR + β2 DIR2
+ β3 DIR3
+ γ Control Variables
This model allows the coefficients on the managerial ownership variables to determine their own turning points.
Table describing Variables
VARIABLES DESCRIPTION
Dependent Variables
VAL Market value of equity at the accounting year end, divided by the book value of
equity at the accounting year end
RSE Return on shareholders’ equity equal to profits attributable to shareholders divided
by shareholders’ equity and reserves
Ownership Variable
DIR Percentage of shares held by directors
INST Percentage of shares owned by institutions owning 5% or more
EXTERNAL Percentage of shares held by other external ownership interests owning 5% or more.
Control Variables
SIZE Log of the firm’s sales
GROWTH Average annual growth in sales
DEBT Total debt divided by book value of total assets
RDTA Research and development expenditure divided by total assets
Table 2

20
We also test for Morck et al.’s piece-wise linear regression model where they pre-determine the turning points
of coefficients. A point to note here is that pre-determining the turning points does not provide altogether robust
results as is argued by Short & Keasey and Demsetz and Villalonga since the motivation behind the turning
points is a qualitative one as opposed to being empirical. However, I still perform the regressions to confirm the
uniformity of results. The model to be tested is as follows:
Performance = α + β1 BRD0-5 + β2 BRD5-25 + β3 BRD25 + γ Control Variables
Here, only the ownership variables change compared to the aforementioned model. The ownership variables
for the piece-wise linear regressions are defined as follows:
BRD0-5 = board ownership if board ownership < 0.05
= 0.05 if board ownership >= 0.05
BRD5-25 = 0 if board ownership < 0.05
= board ownership minus 0.05 if 0.05 <= board ownership < 0.25
= 0.20 if board ownership >= 0.25
BRD25 = 0 if board ownership < 0.25
= board ownership minus 0.25 if board ownership >= 0.25
Furthermore, we also perform 2 Stage Least Square (2SLS) regressions to tackle the issue of endogeneity. This
arises because of the fact that firm performance also affects ownership structure, thus necessitating the
inclusion of VAL as a predictor of ownership structure along with other variables. (Demsetz and Villalonga
2001) For example, due to improvement in performance, managers may be awarded bonus shares. This affects
our regression results and disallows us to draw valid conclusions. 2SLS regressions consist off two regression
equations. In the first equation, firm performance is the dependent variable explained by:
1) Percentage of shares held by directors
2) Percentage of shares held by external shareholders owning 5% or more
3) Percentage of share held by institutions owning 5% or more
4) R&D expenses divided by total assets
5) Log of Sales
6) Total debt divided by book value of total assets
7) Average annual growth in sales

21
The second equation consists off directors shareholdings as the dependent variable explained by:
1) Market value of equity at the accounting year end, divided by the book value of equity at the accounting
year end
2) Total debt divided by book value of total assets
3) Log of Sales
4) INST
5) STDa
– This refers to the standard deviation of annual rates of returns from 2009-2011 4
6) STDm
- This refers to the standard deviation of monthly stock returns for the period 2010-2011 (24
months)5
In the first stage, each endogenous covariate in the equation of interest (VAL, DIR) is regressed on all of the
exogenous variables in the model, including both exogenous covariates in the equation of interest and the
excluded instruments. The predicted values from these regressions are obtained. In the second stage, the
regression of interest is estimated as usual, except that in this stage each endogenous covariate is replaced with
the predicted values from its first stage model from the first stage. (WIKIPEDIA, 2012)
4
The standard deviation of annual rates of return and monthly stock returns had to be manually calculated from
the annual rates of return and monthly stock price data downloaded from FAME.
5
The values for market risk and standard error of estimate from market model were unavailable. Demsetz and
Villalonga (2001) mention that these variables play a important role in determining ownership structure. (Pg.
222)

22
3.2 SAMPLE DATA
The sample consists off 180 companies, which are listed on the FTSE 350 index for the period 2010-2011.
Ownership data had to be collected individually for the companies from their respective annual reports thus
restricting the size of the sample. Due to time constraints the directors, institutional and external shareholdings
were collected for a period of two years. The performance and control variables were calculated through
information gathered from FAME, which is an online database. The following are conditions that eliminated
companies from the sample:
1) Firms that were incorporated after 2007 were not included to ensure that the performance of the firm
was not affected by a new listing.
2) Firms in the financial, oil and gas sectors were avoided due to the different income measuring rules used
by such companies.
3) Privatized firms such as Water, Electricity, Telecommunications, Steel, and Broadcasting were also
sidestepped due to deviations in expectations concerning regulation and control.
3.3 METHODOLOGY
The data analysis comprises Ordinary Least Square regressions utilizing the correction technique for the
unknown heteroskedasticty of White (1980). Morck et al.’s piece-wise linear regression model is also
replicated. The data was diagnosed and treated for outliers, missing values, collinearity, normality of residuals
and simultaneity. Moreover, 2SLS regressions are also performed on the data to treat for the issue of
endogeneity. This problem has been discussed in many articles and it addresses the reverse causality that might
occur from performance to ownership. In other words, the regression results might be compromised because of
the models’ inefficiency to capture the effect of firm performance on managerial ownership. Thus, 2SLS
regression allows us to integrate this effect into the model. In order to resolve the problem of skewness in the
data, logistic transformation of the variable Sales is used. All the regressions and data manipulation was carried
out in STATA.

23
3.4 VARIABLES
The key variables of interest are measures of the firm performance, director’s ownership and other ownership
interests. Furthermore, a small number of additional variables are used to control for effects on the performance
of firms not captured by the ownership variables. To smooth fluctuations on an annual basis, the firm
performance and control variables are measured as averages over the period 2010 to 2011.
Performance Variables
The performance measures employed in the study are VAL and Return on Shareholders Equity. RSE is the
accounting based measure as compared to the market-based phenomena of VAL. These two measures are
considered to be backward and forward-looking respectively, which allows us to examine the robustness of our
results and also compare how these variables are affected by managerial ownership. VAL is similar to Tobin’s
Q used in various studies and is calculated as the value of the firm at its accounting year end, divided by the
book value of equity at the accounting year end. The return on shareholders equity is calculated as profits
attributable to shareholders, divided by shareholders’ equity and reserves.
In very basic terms, the market value of the firm measures the discounted present value of its expected future
income stream, while the book value of equity measures the investment by shareholders in the assets utilized to
generate that income stream. VAL, therefore, provides a measure of management’s ability to generate a certain
income stream from an asset base and is, hence, an indication of management performance.
Ownership Variables
The chief ownership variable included in the model is managerial ownership that is measured as the percentage
of equity shares owned by directors’ and their immediate families at the accounting year-end. These were
collected manually from each company’s annual reports. This differs from the ownership measure used in
various other studies that add the officers and CEO’s shareholdings as well. It is however important to note that
ownership data in the UK are only available for individuals who legally hold the position of director of the firm,
and not for other officers/managers of the firm. Also, the ownership data has been calculated at the beginning of

24
the year to reduce any discrepancies regarding the reverse causality wherein firm performance affects
ownership.
Institutional and External shareholding are the other ownership variables in the model. Due to the restrictions
on the level of external ownership reported in the Annual Report, INST measures the percentage of equity
ownership by institutions owning 3% or more of equity at the beginning of 2010. Similarly, EXTERNAL
measures the percentage of equity ownership by other external shareholders, corporations, charities and
individuals, owning 3% or more of equity at the beginning of 2010.
The table below lists summary statistics of ownership data in 2010 and 2011.
2010 2011
DIR% INST% EXT% DIR% INST% EXT%
Mean 0.0516 0.3271 0.0399 0.0529 0.3219 0.0374
Median 0.0043 0.3220 0.0000 0.0040 0.3265 0.0000
Maximum 0.7284 0.7694 0.7059 0.7259 0.7044 0.6934
Minimum 0.0001 0.0000 0.0000 0.0001 0.0000 0.0000
Standard
Deviation
0.1279 0.1640 0.1159 0.1280 0.1645 0.1153
Table 3
As you can see in table 3 above, the mean and median managerial ownership for all the three variables has not
changed much over the two years. However there is quite a difference in the mean, median, and frequency
distribution of the management ownership compared to other studies. The mean and median managerial
ownership in studies like that of Morck et al., Short & Keasey, and Demsetz & Villalonga is in the range of 10-
15% and 3-6% respectively. Whereas, the mean and median managerial ownership for the current data set is 5-
5.3% and 0.4-0.5% respectively, which is comparatively lower. This could due to the size of corporations
included in this sample that consists off 180 of the largest corporations listed on the London Stock Exchange
for which the ownership is quite dispersed.

25
The table below shows the frequency distribution of ownership data for 2010 and 2011.
Director's Ownership 2010 Director's Ownership 2011
Frequency Cumulative Frequency Cumulative
0-5% 151 151 152 152
5-10% 4 155 4 156
10-15% 2 157 2 158
15-20% 7 164 6 164
20-25% 5 169 2 166
25-30% 1 170 2 168
30-35% 2 172 3 171
35-40% 1 173 1 172
40-45% 1 174 1 173
45-50% 1 175 2 175
50-55% 1 176 1 176
55-60% 1 177 2 178
60-65% 0 177 1 179
65-70% 1 178 0 179
70-75% 2 180 1 180
75-80% 0 180 0 180
Table 4
A glance at the frequency distribution table reveals that about 84% of the Companies in this sample have a
managerial ownership of 5% or less and only 12% of the firms exceed 25% of managerial ownership. Even this
is in stark contrast compared to other studies which report that about 45-60% of companies have a managerial
ownership of 5% or less and 13-25% of firms have managerial ownership of 25% or more. It is important to
note here that studies based in the US include officer’s ownership as well apart from ownership of directors.
Unfortunately, the data for officer’s ownership is unavailable in UK. The differences between the distribution
and mean are however interesting and alarming.
Moreover, managerial ownerships of more than 25% usually represent the presence of a family member on the
board who owns a large stake in the firm. Hence, it can be said that not many of these large corporations have
pure management professionals capable of owning a sizeable share of the company. This indicates the inability
of managers to ignore the views of the institutional and external shareholders.

26
Control Variables
Furthermore, additional control variables are included in the sample to account for their conceivable impact on
firm performance. These comprise Size (log of sales), Growth (average annual growth in sales), Debt ( Total
Debt divided by Total Assets) and RDTA ( Average R&D expenditure divided by average Total Assets). The
size of the firms influences the performance through economies of scale, by creating entry barriers for other
competing firms and it also enables optimal use of profitable projects through better financial structures. The
variable Growth is included to control for the potential linkages between firm’s performance, financing
structure and growth.
The variable DEBT is included to control for a number of factors. First, it controls for the possibility that debt
holders exert significant influence over the behavior and operation of the firm and its management. (Stiglitz,
1985) argues that control over management actions is effectively exercised, not by shareholders, but by lenders,
particularly banks. Second, as suggested by (Jensen, 1986), debt may be used by management to signal that
they have bonded themselves to achieving the levels of cash flow necessary to meet the debt repayments. Debt
may, therefore, be used to resolve conflicts between managers and shareholders as it reduces management
discretion to consume excessive privileges and, hence, should increase the value of the firm’s equity (Jensen
and Meckling, 1976). Furthermore, in line with Morck et al. (1988) and McConnell and Servaes (1990) , I
include research and development expenditure as a control variable. Lastly in line with Demstz and Lehn
(1985), I have included standard deviation of annual rates of return for 2009-2011 and standard deviation of
monthly stock returns for 2010-2011 as instrumental variables to predict managerial ownership.
The table 6 below shows the correlations between all the variables used in this sample. Noteworthy correlations
include that of VAL and RSE (0.79), the two performance variables. The correlation of debt with size is 0.45
and 0.37 with RSE which means an increase in debt leads to an increase in sales and RSE. This is
understandable as more capital implies a greater output leading to higher returns. Also, growth is positively
correlated with managerial ownership (0.33). None of the remaining variables are correlated to an extent that
merits noting.

27
The table below depicts the summary univariate statistics.
Mean Median
Standard
Deviation Maximum Minimum
Performance
variables:
VAL 3.10 2.28 3.02 22.78 0.35
RSE 0.21 0.17 0.24 1.84 -0.27
Ownership variables:
DIR 0.05 0.00 0.13 0.73 0.00
INST 0.32 0.33 0.16 0.77 0.00
EXTERNAL 0.04 0.00 0.12 0.71 0.00
Control variables:
SIZE (in £ millions) 6,471.00 1,031.78 24,353.09 243,277.67 3.39
GROWTH 0.26 0.08 1.16 13.34 -0.38
R&D 0.01 0.00 0.03 0.25 0.00
DEBT 0.58 0.59 0.19 0.99 0.07
Table 5
Performance and control variables are calculated as averages over the period 2010–2011.
Ownership variables are measured at the beginning of 2010.
Statistics for the variables VAL, RSE, DIR and SIZE are shown in their unlogged form.
Correlation Matrix
VAL RSE DIR INST EXT SIZE GROWTH DEBT RDTA
VAL 1.0000
RSE 0.7878 1.0000
DIR 0.2640 0.1051 1.0000
INST -0.1126 -0.0536 -0.3318 1.0000
EXT -0.1035 -0.0616 -0.0154 -0.2731 1.0000
SIZE -0.0850 0.0716 -0.2582 -0.3313 -0.0270 1.0000
GROWT
H -0.0263 -0.1121 0.3262 -0.0841 -0.0252 -0.0151 1.0000
DEBT 0.2429 0.3745 -0.0888 -0.1942 0.0392 0.4535 -0.1839 1.0000
RDTA -0.0720 -0.0622 -0.0356 0.0748 -0.0988 -0.0862 -0.0416 -0.1436 1.0000
Table 6

28
CHAPTER 4: FINDINGS
4.1 RESULTS
First, we study the results for the piece-wise linear regressions given in table 7 below. As you can see, for both
the VAL and RSE regression, the coefficients on the ownership variables BRD05, BRD525 and BRD25 are all
statistically insignificant. These results indicate that there is no relation as such between managerial ownership
Piece-wise estimates using 2010 and 2011 combined ownership data
Variable Dependent Variables
RSE VAL
Coefficient
t-statistic
Coefficient
t-statistic
BRD05 -0.111647 -7.614198
-0.16 -0.71
BRD5-25 0.139226 5.094596
0.44 1.28
BRD25 0.324605 0.797699
1.50 0.33
INST -0.055457 -1.355957
-0.79 -1.45
EXT -0.100612 -2.555180
-1.16 -2.96**
SIZE -0.010779 -0.286027
-1.63* -3.24**
GROWTH -0.022501 -0.156557
-4.97** -2.62**
DEBT 0.278112 2.322490
4.99** 3.07**
RDTA -0.075698 -2.907126
-0.36 -1.22
Intercept 0.129061 3.984768
2.02** 3.92**
Adjusted R2
0.2311 0.1512
F-Statistic 11.13** 4.15**
Table 7
*Significant at 10% confidence level using two-tailed test
**Significant at 5% confidence level using two-tailed test

29
and firm performance. Moreover, the signs of the coefficients of these variables differs from those reported by
Morck et al.
The coefficients for the variables BRD525 and BRD25 are positive while that on the variable BRD05 is
negative whereas Morck et al. report positive coefficients on BRD05 & BRD25 and a negative coefficient on
BRD525. This signifies that management move from entrenchment to alignment and continue aligning for
higher levels of ownership. The other ownership variables of INST and EXT are insignificant as well although
EXT is significant and negative for the VAL regression. The coefficients on the variables SIZE, GROWTH and
DEBT are all statistically significant for both regressions. It can be understood that SIZE and GROWTH have a
negative impact on the performance variables. The positive coefficient on DEBT means that leverage does have
a positive effect on firm performance in the sense that lenders increase their monitoring and reduce the potential
of shirking by managers. Overall, the above model explains about 15-24% of the variations in the dependent
variables.
Second, we look at the results for the model used by Short et al. shown in tables 8 and 9 below. The coefficients
on the ownership variables for DIR, DIR2 and DIR3 are all statistically insignificant for both VAL and RSE
regressions. This holds true for 2010 and 2011 as well, which is inconsistent with the findings of Short et al.
who provide support for a non-linear cubic form of the relationship between managerial ownership and firm
performance. Even the signs on the coefficients of the ownership variables, which suggest that management
moves from entrenchment to alignment and back to entrenchment, are the exact opposite of those reported by
Short et al.
In terms of other ownership variables, there is no support for the hypotheses that institutions affect the
performance of a company. This is consistent with the findings of Short et al. but contrasts with results reported
by McConnel and Servaes. It is interesting to note here that similar to the piece-wise estimates of table 7 above,
the coefficient on the variable EXT is statistically significant and negative for the VAL regression during both
years. This too is inconsistent with findings of Short et al. who report a statistically insignificant relation. Also
similar to the piece-wise estimates; the control variables SIZE, GROWTH and DEBT are all statistically
significant and of the same sign as reported above. This again, is in contrast with the findings of Short et al.
who report a positive effect of SIZE & GROWTH and a negative effect of DEBT. Moreover, even the
coefficients on RDTA are insignificant. The negative effect of SIZE and GROWTH, although negligent, is

30
surprising, as an increase in sales should represent an increase in the profits of a company. This can probably be
attributed to the financial crisis for which period the results were calculated and analysed in the sense that
despite a minor growth in sales, companies still had to contend with financial problems caused by recession.
Overall the above model explains about 15-24% of the variations in the dependent variables.
Regression estimates using 2010 ownership data
RSE VAL
Coefficient
t-statistic
Coefficient
t-statistic
DIR -0.3464764 -2.722784
-0.77 -0.49
DIR2
2.226005 22.50462
1.20 0.89
DIR3
-2.042603 -23.70606
-1.10 -0.90
INST -0.053061 -1.309865
-0.76 -1.41
EXT -0.093750 -2.435938
-1.09 -2.87**
SIZE -0.011085 -0.270968
-1.80* -3.32**
GROWTH -0.020611 -0.136558
-4.12** -2.03**
DEBT 0.2776591 2.416878
4.99** 3.23**
RDTA -0.092725 -3.018275
-0.43 -1.32
Intercept 0.132211 3.765921
2.30** 4.09**
Adjusted R2
0.2358 0.1502
F-Statistic 10.71** 4.47**
Inflection Points1
Table 8
+ β3 DIR3
1
Inflection points have not been calculated because of insignificant coefficients on ownership variables

31
Regression Estimates using 2011 ownership data
RSE VAL
Coefficient
t-statistic
Coefficient
t-statistic
DIR -0.282975 -0.305467
-0.59 -0.05
DIR2
1.331834 2.557803
0.67 0.09
DIR3
-0.829748 0.489455
-0.40 0.02
INST -0.087427 -1.620041
-1.14 -1.60
EXT -0.090751 -2.555187
-1.02 -2.82**
SIZE -0.012595 -0.288473
-1.91* -3.50**
GROWTH -0.023959 -0.181841
-3.25** -1.86*
DEBT 0.2842524 2.500421
5.11** 3.31**
RDTA -0.086438 -2.958823
-0.40 -1.17
Intercept 0.151875 3.960256
2.35** 4.09**
Adjusted R2
0.2314 0.1351
F-Statistic 10.61** 3.60**
Inflection Points1
Table 9
+ β3 DIR3
1
Inflection points have not been calculated because of insignificant coefficients on ownership variables
Until now, the regression estimates of both the piece-wise and cubic specifications give no indication of a
relationship between ownership structure and performance of firms. This is consistent with the findings
reported by Demsetz and Villalonga (2001) and Cho (1998) who find no systematic relation between ownership
structure and firm performance either. It can be said here that ownership is an endogenous outcome influenced
by different market forces and that it bears negligent influence on the economic functioning of the firm.

32
Third, to further investigate the effect of endogeneity and how different variables interact with the ownership
variables, I performed the instrumental variable regression or two stage ordinary least square regression as it is
2SLS estimates using 2010 ownership data
Variable Dependent Variable
DIR (OLS) 2SLS VAL (OLS) 2SLS
Coefficient
t-statistic
Coefficient
t-statistic
Coefficient
t-statistic
Coefficient
t-statistic
DIR 1.853294 4.353836
1.71* 1.17
INST -0.385820 -0.340678 -1.348765 -0.410960
-4.06** -4.91** -1.46 -0.25
EXT -2.534708 -2.099777
-3.02** -1.68*
SIZE -0.027125 -0.016664 -0.262347 -0.1894345
-3.74** -1.60 -3.22** -1.36
DEBT -0.019289 -0.110539 2.423657 2.384758
-0.46 -1.26 3.28** 3.49**
GROWTH -0.161992 -0.207845
-3.03** -1.62*
RDTA -2.707025 -2.298698
-1.15 -0.57
VAL 0.008958 0.045828
1.69* 1.60
STDa
-0.001332 0.000216
-1.39 0.11
STDm
1.211632 1.121187
2.70** 2.75**
Intercept 0.269364 0.134442 3.677798 2.735955
3.18** 1.10 3.97** 1.70*
Adjusted R2 0.3212 0.1263 0.1500 0.1208
F-Statistic 3.79** 10.81** 5.18** 4.00**
Table 10
commonly known based on the model of Demsetz and Villalonga (2001). As the name suggests, the analysis
for this method is carried out in two stages where the first stage consists off regressions of the endogenous

33
variables (VAL and DIR) against all the other exogenous variables. In the second stage, the variable of interest,
i.e., VAL is regressed with all the other variables but the values of VAL and DIR used in this analysis is that of
the first stage. The results for 2SLS regressions are given in table 10 and 11 for the years 2010-2011 along with
OLS estimates in order to merit a comparison of the two.
The OLS results indicate that at least two of the ownership variables affect VAL in a significant way which is
consistent with the findings of Demsetz et al. (2001). DIR has a positive influence on VAL and, INST & EXT
have a negative effect which is also consistent with Demsetz et al. As they mention, there is cause to suspect
these results as the signs taken by INST & EXT and DIR are the reverse of what one would expect if greater
ownership concentration by outside investors does lead to greater firm performance and larger shareholdings by
insiders does lead to a more entrenched management. These OLS results concur with the findings of Morck et
al. , Short et al., among others that provide an effect of ownership structure on firm performance.
However, the results shown in these tables for the 2SLS estimates of the coefficients of INST, EXT and DIR
cast considerable doubt on this asserted affect of ownership structure on firm performance. None of the 2SLS
estimates of the three ownership coefficients are statistically significant in the performance equations for both
2010 and 2011. Thus, the 2SLS estimates also reject the hypothesis that ownership structure affects firm
performance. This is consistent with the view that ownership structure is chosen so as maximize firm
performance, and that greater diffuseness in ownership, although it makes the agency problem more severe,
conveys compensating advantages on firms that choose to rely on a diffuse ownership structure. The results
also provide support to the findings of Demsetz and Villalonga. Among other variables, constant with our
earlier findings, SIZE and GROWTH have a negative effect on VAL whereas a significant positive effect of
DEBT can be seen for both OLS and 2SLS estimates. This confirms that debt facilitates a more reliable control
measure for firms to keep managers in check and discourage them from indulging in unscrupulous activities
detrimental to the progress of companies and profitability of ultimate owners. The coefficient for RDTA
remains insignificant for all our estimates.

34
2SLS estimates using 2011 ownership data
Variable Dependent Variable
DIR (OLS) 2SLS VAL (OLS) 2SLS
Coefficient
t-statistic
Coefficient
t-statistic
Coefficient
t-statistic
Coefficient
t-statistic
DIR 1.066365 6.351146
0.91 1.43
INST -0.363865 -0.299287 -1.655858 0.270203
-4.28** -3.90** -1.66* 0.15
EXT -2.608868 -1.484869
-2.95** -1.00
SIZE -0.029953 -0.015713 -0.282651 -0.117742
-3.84** -1.37 -3.43** -0.71
DEBT -0.020044 -0.146458 2.460882 2.430237
-0.44 -1.52 3.31** 3.37**
GROWTH -0.151939 -0.236061
-2.41** -1.74*
RDTA -2.790926 -1.447905
-1.10 -0.33
VAL 0.005493 0.0554872
0.98 1.76*
STDa
-0.001841 0.000414
-1.75* 0.19
STDm
1.009269 0.886883
2.28** 1.96**
Intercept 0.309068 0.127388 3.926992 1.837886
3.41** 0.93 4.03** 0.95
Adjusted R2 0.3003 -0.0517 0.1339 0.0205
F-Statistic 3.89** 7.48** 4.35** 3.61**
Table 11
Among the variables included to explain variations in director’s shareholdings, it can be noticed that managers
tend to hold fewer shares in large firms and in companies with high institutional ownership. One thing, which is
inconsistent with the findings of Demsetz and Villalonga, is the positive significant effect of VAL on
managerial shareholding signifying that managers increase the stakes in companies that are performing well.

35
The variables DEBT and STDa
are insignificant for both the years in explaining fluctuations of director’s
shareholdings. However, it is interesting to notice that the variable STDm
has a significant and strong positive
effect on the holdings of managers meaning that managers hold more shares in companies with higher standard
deviation of stock returns. There is no logical explanation for this as one would assume that managers, just like
any other reasonable person, are risk-averse and would invest less in high firm-specific risk firms than in low
firm-specific risk firm. This is also consistent with findings of Demsetz and Villalonga.
After analyzing the estimates for all the three sets of regressions, the signs of and the coefficients themselves
beg me to ask the question whether the non-monotonic relation reported by earlier studies were accidental
occurences.

36
4.2 ROBUSTNESS TESTS
In order to ensure that the results are robust, three different models were tested – piecewise model, cubic model,
and 2SLS model. Besides, I conducted supplementary analysis to determine the robustness of our results.
Following Demsetz and Lehn (1985) and Short and Keasey (1999), it is possible that the relationship between
the performance of firms and managerial ownership is spurious because the relationship between these
variables is industry-specific and no control has been included in the regressions for this possibility. To
examine whether our regression results reflect industry effects, we repeated all of the above regressions after
incorporating industry five-digit SIC Codes. The results shown in tables 7,8,9 were not affected by the
inclusion of industry dummy variables with only slight variations in the values.
Additionally, to further investigate the robustness I regressed the variables with both 2010 and 2011 ownership
data to ensure that the results are not biased because of ownership being defined at the beginning of 2010.
Furthermore, a series of tests was conducted ensuring that all the variables met the necessary assumptions to
draw valid regression conclusions. The different tools used in STATA to ensure this are listed below:
Problem Diagnosis Solution
Outliers Predict, stem and leaf, avplots, scatterplots Winsorization, Elimination
Missing Values - Mean Substitution
Normality of Residuals Kdensity, pnorm, qnorm, iqr, Shapiro wilk W test Transformation e.g., Log values
Collinearity Vif, collin Elimination of variable
Homoskedasticity Hettest, imtest, white test Robust regression
Model Specification Linktest, ovtest Add variable
Table 12
The consistent findings across all three regression models indicate the general robustness of my results.

37
CHAPTER 5: CONCLUSIONS
In this paper, I investigated the impact of ownership structure on firm performance through piece-wise linear
regression, ordinary least squares regression and 2SLS regression. I could find no evidence of a systematic
relation between ownership structure and firm performance across any of the set of regressions. It is therefore
implied that ownership structure is an endogenous outcome of different market forces and while it may
aggravate agency problems, these are offset by the advantages of such an outcome. This paper also provides
evidence that the coefficients of single equation models of the effect of ownership structure on performance are
biased. Bias is also likely to result from studies that fail to take account of the complexity of interests that are
involved in an ownership structure. It can also be understood that the market responds to changes in the
economic environment and influences firms to arrive at optimum ownership structures.
However, there are studies like that of Coles et al. and Florackis et al. which point out the inefficiency of
different models and variables to treat for the problem of endogeneity. Consequently, the issue of reverse
causality still remains. For example, managers of successful firms are more likely to be rewarded with
additional forms of stock ownership; more successful firms might award directors equity shares. Also, due to
time constraints, ownership data was gathered for only 180 companies, which limits the scope of the study to a
certain extent as there are comparatively a small set of observations. A wider sample would also have yielded a
more natural normal distribution of the variables. In addition, there are certain variables used in other studies
like Standard Error of estimate from market model, beta and ratio of advertising expenditure to assets/sales that
were unavailable. This might affect the 2SLS regression estimates for the sample.
Hence, future studies on this topic can concentrate on new approaches and avenues like semi-parametric models
to study the effect of ownership on performance. One thing is for certain though, the interaction between
ownership and firm performance (if any) is a complex one that involves multi-dimensional factors apart from
those used in this study. The relation transcends beyond quantitative analysis of firm factors and involves
qualitative, economic environment, industrial, regulatory factors as well thereby creating the need for a more
comprehensive and conclusive data analysis plan on the issue.

38
CHAPTER 6: REFERENCES
1. AGRAWAL, A. & KNOEBER, C. R. 1996. Firm Performance and Mechanisms to Control Agency
Problems between Managers and Shareholders. Journal of Financial and Quantitative Analysis, 31,
377-397.
2. BENNINGA, S. 2010. Principles of Finance With Excel, Oxford University Press.
3. BERLE, A. A. & MEANS, G. C. 1932. The Modern Corporation and Private Property, Transaction
Publishers.
4. BLACKMORE, D. 2010. Ownership of UK Quoted Shares [Online]. Office for National Statistics.
Available: http://www.ons.gov.uk/ons/rel/pnfc1/share-ownership---share-register-survey-
report/2010/stb-share-ownership-2010.html [Accessed 18/06/2012.
5. CHEN, X., ENDER, P., MITCHELL, M. AND WELLS, C. . 2003. Regression with Stata [Online].
Available: http://www.ats.ucla.edu/stat/stata/webbooks/reg/default.htm [Accessed 28/07/2012.
6. CHEUNG, A. W.-K. 2010. Managerial ownership and performance: A commentary essay. Journal of
Business Research, 63, 292-293.
7. CHO, M.-H. 1998. Ownership structure, investment, and the corporate value: An empirical analysis.
Journal of Financial Economics, 47, 103-121.
8. COLES, J. L., LEMMON, M. L. & FELIX MESCHKE, J. 2012. Structural models and endogeneity in
corporate finance: The link between managerial ownership and corporate performance. Journal of
Financial Economics, 103, 149-168.
9. DAVIES, J. R., HILLIER, D. & MCCOLGAN, P. 2005. Ownership structure, managerial behavior and
corporate value. Journal of Corporate Finance, 11, 645-660.
10. DEMSETZ, H. 1983. The Structure of Ownership and the Theory of the Firm. Journal of Law and
Economics, 26, 375-390.
11. DEMSETZ, H. & LEHN, K. 1985. The Structure of Corporate Ownership: Causes and Consequences.
Journal of Political Economy, 93, 1155-1177.
12. DEMSETZ, H. & VILLALONGA, B. N. 2001. Ownership structure and corporate performance.
Journal of Corporate Finance, 7, 209-233.
13. FAHLENBRACH, R. D. & STULZ, R. M. 2009. Managerial ownership dynamics and firm value.
14. FAMA, E. F. & JENSEN, M. C. 1983. Separation of Ownership and Control. Journal of Law and
Economics, 26, 301-325.
15. FLORACKIS, C., KOSTAKIS, A. & OZKAN, A. 2009. Managerial ownership and performance.
Journal of Business Research, 62, 1350-1357.
16. HERMALIN, B. E. & WEISBACH, M. S. 1991. The Effects of Board Composition and Direct
Incentives on Firm Performance. Financial Management, 20, 101-112.
17. HIMMELBERG, C. P., HUBBARD, R. G. & PALIA, D. 1999. Understanding the determinants of
managerial ownership and the link between ownership and performance. Journal of Financial
Economics, 53, 353-384.
18. HOLDERNESS, C. G., KROSZNER, R. S. & SHEEHAN, D. P. 1999. Were the Good Old Days That
Good? Changes in Managerial Stock Ownership since the Great Depression. The Journal of Finance,
54, 435-469.
19. JENSEN, M. C. 1986. Agency Costs of Free Cash Flow, Corporate Finance, and Takeovers. The
American Economic Review, 76, 323-329.

39
20. JENSEN, M. C. & MECKLING, W. H. 1976. Theory of the firm: Managerial behavior, agency costs
and ownership structure. Journal of Financial Economics, 3, 305-360.
21. KOLE, S. R. 1995. Measuring managerial equity ownership: a comparison of sources of ownership
data. Journal of Corporate Finance, 1, 413-435.
22. LEWIS, P., SAUNDERS, M. N. K. & THORNHILL, A. 2009. Research Methods for Business
Students, Pearson.
23. LODERER, C. & MARTIN, K. 1997. Executive stock ownership and performance tracking faint traces.
24. MALLIN, C. A. 2010. Corporate governance / Christine A. Mallin, Oxford ; New York :, Oxford
University Press.
25. MATHIESEN, H. 1999-2012. ENCYCOGOV [Online]. ViamInvest. Available:
http://e.viaminvest.com/A5OwnershipStructures.asp [Accessed 15/06/2012.
26. MCCONNELL, J. J. & SERVAES, H. 1990. Additional evidence on equity ownership and corporate
value. Journal of Financial Economics, 27, 595-612.
27. MCCONNELL, J. J. & SERVAES, H. 1995. Equity ownership and the two faces of debt. Journal of
Financial Economics, 39, 131-157.
28. MCCONNELL, J. J., SERVAES, H. & LINS, K. V. 2008. Changes in insider ownership and changes in
the market value of the firm. Journal of Corporate Finance, 14, 92-106.
29. MORCK, R., SHLEIFER, A. & VISHNY, R. W. 1988. Management ownership and market valuation:
An empirical analysis. Journal of Financial Economics, 20, 293-315.
30. OAKSHOTT, L. 2009. Essential Quantitative Methods: For Business, Management and Finance,
Palgrave Macmillan.
31. SHORT, H., KEASEY, K., 1997. Institutional shareholders and corporate governance in the UK. In:
Keasey, K., Thompson, S., Wright, M. (Eds.)., Corporate Governance: Economic and Financial Issues,
Oxford University Press, Oxford, pp. 18–53.
32. SHORT, H. & KEASEY, K. 1999. Managerial ownership and the performance of firms: Evidence from
the UK. Journal of Corporate Finance, 5, 79-101.
33. SMITH, A. 1838. An Inquiry Into the Nature and Causes of the Wealth of Nations, T. Nelson.
34. STIGLITZ, J. E. 1985. Credit Markets and the Control of Capital. Journal of Money, Credit and
Banking, 17, 133-152.
35. STULZ, R. M. 1988. Managerial control of voting rights: Financing policies and the market for
corporate control. Journal of Financial Economics, 20, 25-54.
36. WHITE, H. 1980. A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for
Heteroskedasticity. Econometrica, 48, 817-838.
37. WIKIPEDIA. 2012. Instrumental variable [Online]. Available:
http://en.wikipedia.org/wiki/Instrumental_variable [Accessed 18/08/2012.
38. ZHOU, X. 2001. Understanding the determinants of managerial ownership and the link between
ownership and performance: comment. Journal of Financial Economics, 62, 559-571.
APPENDIX
XXX

200671838 - Dissertation

More Related Content

What's hot (19)

Similar to 200671838 - Dissertation (20)

200671838 - Dissertation