UNIVERSITY OF GHANA MARKET SHARES AND PROFITABILITY OF UNIVERSAL BANKS IN GHANA BY OPHELIA AMO (10274763) A THESIS SUBMITTED TO THE UNIVERSITY OF GHANA, LEGON IN PARTIAL FULFILMENT OF THE REQUIREMENT FOR THE AWARD OF MASTER OF PHILOSOPHY DEGREE IN ECONOMICS JULY, 2015 University of Ghana http://ugspace.ug.edu.gh i DECLARATION I, Ophelia Amo do hereby declare that except for the references cited, which have been duly acknowledged, this thesis titled “MARKET SHARE AND PROFITABILITY OF BANKS IN GHANA” is the product of my own research work in the Department of Economics, University of Ghana, Legon, from August 2014 to July 2015. This thesis is not published or submitted either in part or in whole anywhere for the award of a degree in any other university. ………………………. ….…………………… Ophelia Amo Date (Student) This thesis has been submitted with our approval as supervisors: …………………………. …………………….. Dr. Augustine Fritz Gockel Date (Supervisor) ………………………… ……………………. Dr. Eric Osei-Assibey Date (Supervisor) University of Ghana http://ugspace.ug.edu.gh ii ABSTRACT The increasing competition among banks in Ghana is largely explained by the liberalisation and implementation of the various financial sector reforms recommended by the International Monetary Fund (IMF) and the World Bank. In today’s competitive world, one of the challenges facing many managers in the banking sector is how to increase business market shares and profits concurrently. Although the correlation between market shares and profitability has been sustained over the years, it remains a generalisation which has been over-extended without its attributes acknowledged. The question frequently asked is whether higher market share leads to higher profits? The study therefore seeks to analyse the factors that influence bank’s market share and also to examine the relationship between market share and profitability of universal banks in Ghana. A panel data of 15 universal banks in Ghana are analysed over a period of 2004 to 2013 using the Generalised Method of Moment (GMM) and random effect regression models. Results from the study indicate that, bank size, operating efficiency, bank age, number of branches, ownership structure, Gross Domestic Product and inflation significantly determine market share. The study also reveals that, there is no significant relationship between market share and profitability. The study recommends that, banks employ more technologies to facilitate their service delivery and ensure the efficient management of bank operations to help alleviate the high operational cost that erodes bank profits. Also, management need to understand the context of their environment before adoption of market share strategy. University of Ghana http://ugspace.ug.edu.gh iii DEDICATION I dedicate this thesis work to God, my parents (Nana Kwabena Amo and Grace Twum) and my siblings (Rita Amo, Justice Amo and Juliet Amo). University of Ghana http://ugspace.ug.edu.gh iv ACKNOWLEDGEMENTS I am grateful to God Almighty for His grace, love, mercies and protection conferred on me throughout my entire study and also seeing me through a successful completion of this course. I also wish to extend my heartfelt gratitude to my supervisors, Dr. Augustine Fritz Gockel and Dr. Eric Osei-Assibey for their insightful and immense contribution to the development of this work in spite of all their busy schedules. For financial support, I want to thank my parents; Nana Kwabena Amo and Grace Twum; my uncle; Charles Kwabena Opoku, Clement Afrifa and Felix Deku for their great support throughout my studies. I am grateful for all your efforts and may the Almighty God replenishes all that you have lost and see you throughout all your endeavours. I would also like to express my profound gratitude to all the lecturers of the Department of Economics, University of Ghana especially Mr. Emmanuel Abbey for his support and guidance for a successful completion of this study. I thank all the Financial Institutions who provided me with the necessary materials and advice. I extend my sincere gratitude to my family whose support is invaluable. Last but not the least, I thank Justice, Samaad, Derick, Bernice, Millicent, Frank, Frederick, Obaayaa so much for their assistance in proof reading and other important roles to bring this work to a success. I also thank my colleagues and all whose names have not been mentioned. University of Ghana http://ugspace.ug.edu.gh v TABLE OF CONTENTS DECLARATION .............................................................................................. i ABSTRACT .................................................................................................... ii DEDICATION ................................................................................................ iii ACKNOWLEDGEMENT .............................................................................. iv TABLE OF CONTENTS ................................................................................ v LIST OF FIGURES ......................................................................................... x LIST OF TABLES .......................................................................................... xi LIST OF ABBREVIATIONS ....................................................................... xii CHAPTER ONE .............................................................................................. 1 INTRODUCTION ........................................................................................................... 1 1.1 Background ....................................................................................................... 1 1.2 Problem statement ............................................................................................ 3 1.3 Objectives of the study ..................................................................................... 6 1.4 Research questions ........................................................................................... 6 1.5 Significance of study ........................................................................................ 7 1.6 Organization of the study ................................................................................. 7 CHAPTER TWO ............................................................................................. 9 LITERATURE REVIEW ................................................................................................ 9 2.1 Introduction ...................................................................................................... 9 2.2 Definition of terms and concepts ...................................................................... 9 2.2.1 Market Share ........................................................................................... 10 2.2.1.1 Market change positioning .................................................................. 10 University of Ghana http://ugspace.ug.edu.gh vi 2.2.1.2 Behavioural differences between high and low market share firms.... 12 2.2.1.3 Limitations of market share concept ................................................... 12 2.2.2 Profitability in the banking system.......................................................... 13 2.2.2.1 Measures of profitability ..................................................................... 14 2.3 Review of theoretical literature ...................................................................... 15 2.3.1 Theory of Economies of scale and Economies of scope ......................... 17 2.3.1.1 Sources of Economies of Scale and Economies of Scope ................... 17 2.3.2 Theory of the market power .................................................................... 19 2.3.3 Efficiency theory ..................................................................................... 20 2.4 Empirical Review ........................................................................................... 21 2.4.1 Relationship between Market shares and profitability ............................ 21 2.4.2 Determinants of Market Shares ............................................................... 26 2.4.2.1 Internal Determinants .......................................................................... 26 2.4.2.2 External Determinants ......................................................................... 35 2.4.2.3 Further Empirical Studies of Determinants of Market Share .............. 37 2.5 Conclusion ...................................................................................................... 40 CHAPTER THREE ....................................................................................... 41 OVERVIEW OF THE BANKING INDUSTRY .......................................................... 41 3.1 Introduction .................................................................................................... 41 3.2 Overview of the Banking industry.................................................................. 41 3.3 Structure of the banking industry ................................................................... 42 3.4 Market share analysis for the banks in Ghana from 2004 and 2013 .............. 45 3.4.1 Share of operating asset........................................................................... 45 3.4.2 Share of industry deposit ......................................................................... 47 University of Ghana http://ugspace.ug.edu.gh vii 3.4.3 Share of industry advances ...................................................................... 48 3.5 Trend of profitability analysis for the banks in Ghana (2004 and 2013). ...... 49 3.5.1 Trend analysis of Return ON Average Asset of the selected banks ........ 50 3.5.2 Trend analysis of Return on Average Equity of the selected banks ........ 51 3.6 Conclusion ...................................................................................................... 52 CHAPTER FOUR ......................................................................................... 53 METHODOLOGY ........................................................................................................ 53 4.1 Introduction .................................................................................................... 53 4.2 The Theoretical Framework ........................................................................... 53 4.3 The model ....................................................................................................... 54 4.3.1 Model specification for the determinants of market shares .................... 55 4.3.2 Model Specification for the relationship between market shares and profitability ............................................................................................................ 57 4.4 The measurement of the variables .................................................................. 60 4.4.1 Market share measures ............................................................................ 60 4.4.2 Profitability measures .............................................................................. 61 4.4.3 Determinants of market share ................................................................. 62 4.4.3.1 Internal determinants of market shares ............................................... 62 4.4.3.2 External determinants .......................................................................... 65 4.5 Diagnostic test ................................................................................................ 67 4.5.1 Stationarity/Unit root test ........................................................................ 68 4.6 Estimation technique ...................................................................................... 69 4.7 Data ................................................................................................................. 70 4.8 Sampling Criteria ............................................................................................ 70 University of Ghana http://ugspace.ug.edu.gh viii 4.9 Conclusion ...................................................................................................... 70 CHAPTER FIVE ........................................................................................... 71 DISCUSSION OF RESULTS ....................................................................................... 71 5.1 Introduction .................................................................................................... 71 5.2 Descriptive analysis ........................................................................................ 71 5.3 Correlation matrix........................................................................................... 73 5.4 Diagnostic results ........................................................................................... 75 5.4.1 Unit root test ............................................................................................ 75 5.5 Regression results ........................................................................................... 76 5.5.1 Industry Operating Asset (IOA) regression results: ................................ 76 5.5.2 Industry Deposits (ID) regression results ................................................ 82 5.5.3 Industry Advances (IA) regression results. ............................................. 86 5.5.4 Composite Index results as a proxy for market share ............................. 91 5.6 Regression for the relationship between market share and profitability ........ 95 5.6.1 Dependent variable: Return On Average Asset (ROA) .......................... 95 5.6.2 Dependent variable: Return On Average Equity (ROE). ........................ 98 5.7 Conclusion .................................................................................................... 101 CHAPTER SIX ............................................................................................ 102 CONCLUSION AND RECOMMENDATION .......................................................... 102 6.0 INTRODUCTION ........................................................................................ 102 6.1 Summary of the main findings ..................................................................... 102 6.1.1 Determinants of Market Share .............................................................. 103 6.1.2 Relationship between market share and profitability ............................ 105 6.2 Conclusion .................................................................................................... 105 University of Ghana http://ugspace.ug.edu.gh ix 6.3 Policy Recommendations ............................................................................. 105 6.4 Limitation of the study ................................................................................. 109 REFERENCES ............................................................................................ 110 APPENDIX .................................................................................................. 122 University of Ghana http://ugspace.ug.edu.gh x LIST OF FIGURES Figure Page 1.1: Changes in market share between 2004 and 2013……………………………………..4 1.2: Changes in profitability between 2004 and 2013…………………………..…………5 2.1: Conceptual framework explaining market share as a predictor of profitability ……16 3.1: Summary of the structure of the banking sector in Ghana………….……………...44 3.2: Trend analysis of industry operating asset…………………………………………46 3.3: Trend analysis of industry deposit……………………………………..…………...47 3.4: Trend analysis of industry advances………………………………………………..49 3.5: Trend analysis of Return On Average Asset (ROA)……………………………….50 3.6: Trend analysis of Return On Average Equity (ROE)……………………………...51 University of Ghana http://ugspace.ug.edu.gh xi LIST OF TABLES Table Page 2.1 Studies on the relationship between market share and profitability………....25 4:1 Description of variables……………………………………………..…........67 5.1 The descriptive statistics of the variables…...................................................72 5.2.1: Correlation matrix of coefficients regress Model for MS………….............74 5.2.2: Correlation matrix of coefficients regression model for profitability……...74 5.3 unit root test……………….……………………………………………….76 5.4 Estimation results for the determination of IOA…………….......................77 5.5 Estimation results for determination of ID………………………………...82 5.6 Estimation results for the determination of IA……………………….…….87 5.7 Estimation results for composite index…………………………………….92 5.8 Estimation results for ROA……………………………………..…………96 5.9: Estimation results for ROE………………………………………….……100 University of Ghana http://ugspace.ug.edu.gh xii LIST OF ABBREVIATIONS ABL Amalgamated Bank Limited ADB Agricultural Development Bank of Ghana AQ Asset Quality ARB Association of Rural and Community Banks BBGL Barclays Bank of Ghana Limited BCG Boston Consulting Group BOA Bank of Africa Ghana Limited BoG Bank of Ghana BSD Banking Supervision Department CONS Constant ECG Ecobank Ghana FABL First Atlantic Bank Limited FBN First Bank of Nigeria GBS Ghana Banking Survey GCB Ghana Commercial Bank GDP Gross Domestic Product GMM Generalised Method of Moment University of Ghana http://ugspace.ug.edu.gh xiii HHI Herfindahl Hirschman Index HFC Home Finance Bank IA Industry Advances ICB International Commercial Bank ICT Information and Communication Technology ID Industry Deposit IMF International Monetary Fund INF Inflation IOA Industry Operating Asset L Liquidity MFIs Micro Finance Institutions NA Not Available NAQ Natural logarithm of Asset Quality NBA Natural logarithm of Bank Age NBFIs Non- Banking Financial Institutions NBS Natural logarithm of Bank Size NGDP Natural logarithm of Gross Domestic Product NNB Natural logarithm of Number of Branches NIB National Investment Bank University of Ghana http://ugspace.ug.edu.gh xiv NINF Natural logarithm of Inflation NL Natural logarithm of Liquidity NOE Natural logarithm of Operating Efficiency NOS Natural logarithm of Ownership Structure NO Number OE Operating Efficiency OFISD Financial Institutions Supervision Department OBS Observations PBL Prudential Bank Limited PG Page PIMS Profit Impart of Market Strategies ROA Return On Average Asset ROE Return On Average Equity RCBs Rural and Community Banks SCB Standard Chartered Ghana SEE South Eastern European SG-SSB Société Générale Social Security Bank UBBL Universal Banking Business Licence US United States University of Ghana http://ugspace.ug.edu.gh xv UMB Universal Merchant Bank WDI World Development Indicator University of Ghana http://ugspace.ug.edu.gh 1 CHAPTER ONE INTRODUCTION 1.1 Background The banking sector plays a critical role in the development of the Ghanaian economy because of its financial intermediation role. In the last ten or more years, there has been a rapid increase in the activity of foreign banks in Ghana due to liberalisation and adoption of the universal banks (Bank of Ghana, 2004) and this has fostered rapid competition among banks in Ghana. Reininger et al. (2002) asserts that, a higher degree of banking sector competition is expected to provide welfare gains in reducing monopoly rents and cost inefficiencies. Since investment is particularly sensitive to a decrease in commercial interest rates, a reduction of monopoly rents should consequently impact positively on investment and economic growth. According to Fries and Taci (2005), these expected gains are particularly a major issue for countries in which bank credit represents the largest source of external finance for companies. In today’s competitive world, one of the challenges facing many management is how to increase business market shares and profits (Mutshinyani, 2009). In order to do this, management need to understand the factors that increase market shares and also, the relationship between market share and profitability. Three major streams of research on market share and profitability relationship are evident. In the first case, researchers have provided empirical results leading to a positive relationship between market share and profitability using the economies of scale and economies of scope theory (Venkatraman et al. 1990; Smirlock, 1985). Studies such as University of Ghana http://ugspace.ug.edu.gh 2 (Hagigie et al. 1999; Kotler and Keller 2006) also report a negative relationship between market share and profitability. Other studies also suggest there is no relationship between these two concepts (Mutshinyani, 2009; Schwalbach, 1991). Prescott and Venkatraman (1986) state that, although the correlation between market shares and profitability has been sustained over the years, yet it remains a generalisation which has been over-extended without its attributes been acknowledged. The general question that has always been asked is whether higher market share leads to higher profits? Scholars who have examined this question using the market and efficiency theory have not succeeded in solving this. While these studies provide insight in different environmental and structural conditions, they generally fail to address the bank specific factors at work in the market share and profitability relationship. Understanding these intervening economic processes would provide a better relationship between market share and profitability. Such information would assist managers’ in carrying out the appropriate policies that would increase market share in diverse environments. Given the relation between the well-being of the banking sector and the growth of the economy (Reininger et al. 2002), knowledge of the underlying factors that influence the financial sector's profitability is therefore essential not only for the management of the banks, but also for numerous stakeholders such as the central banks, bankers associations, governments, and other financial authorities. Knowledge of these factors would be useful in helping the regulatory authorities and bank managers to formulate future policies aimed at improving the market share and profitability of the banking sector in Ghana. University of Ghana http://ugspace.ug.edu.gh 3 Due to the changing banking environment, market share and profitability of banks are some of the most important criteria to measure growth and performance of banks. As the banking sector is increasing due to the liberalisation, about 27 banks in the country are also competing for market shares (Ghana Banking Survey, 2014). For example, in 2004, the first five banks with the highest market share in terms of the industry deposit were Ghana Commercial Bank (GCB) with 20.13%, Barclays Bank Ghana (BBG) with 16.53%, Standard Chartered Bank (SCB) with 16.22%, Ecobank Ghana (EBG) with 8.29% and Merchant Bank of Ghana (MBG) with 5.69%. In 2013, the first five banks were EBG with 14%, GCB with 10%, Stanbic with 9.3%, SCB with 8.3% and BBG with 6.6%. The above analysis shows that, increasing market share is one of the most important objectives of managers. According to the Ghana Banking Survey (2013) the gain in market share of the banks is partly on account of their willingness to offer attractive rates on time. Empirical research is therefore required to know what account for market shares and its relationship with profitability so that relevant empirical information is provided for policy formulation as well as strategy design and implementation by affected parties in the banking sector. 1.2 Problem statement Ghana’s banking sector followed, particularly the liberalisation and the implementation of the various financial sector reforms recommended by the International Monetary Fund (IMF) and the World Bank. This financial sector liberalisation resulted in the increase in the number of banks locally and the entry of foreign banks, which has caused competition University of Ghana http://ugspace.ug.edu.gh 4 among the banks (Akuffo- Duah, 2011 and Awuah, 2011). As the banking sector is increasing due to the liberalisation, there has been competition on bank’s market shares and profitability. Figure 1 and figure 2 show changes in the industry market share and profitability between the period of 2004 and 2013 for the first five banks with the largest share and profitability. Figure 1.1: Changes in market share between 2004 and 2013 Source: Ghana Banking Survey, 2008 and 2014. In figure 1.1, deposit is used as a measure of market share. Competition has led to changes in the market share among the banks between the period of 2004 and 2013 as shown above. 0.00% 5.00% 10.00% 15.00% 20.00% 25.00% EBG GCB STANBIC SCB BBGL m ar k et s h ar e fo r th e fi rs t fi v e b an k s in p er ce n ta g es Banks Changes in market share for the first five banks 2004 2013 University of Ghana http://ugspace.ug.edu.gh 5 Figure 1.2: shows changes in profitability (ROA) between 2004 and 2013 Source: Annual Reports and Accounts of individual banks, 2004 and 2013. In Figure 1.2, Return on Average Asset (ROA) is used as a measure of profitability. From Figure 1.1 and figure 1.2, Ecobank had the largest market share in 2013 but Barclays bank had the highest profitability in 2013. This shows that the largest banks are not necessarily the most profitable. It is clear that the banking sector environment has become very complex, more competitive and more challenging to the management. It is therefore paramount in considering the fact that competition is a part of the banking sector; until a bank develop strategies to compete successfully, it has practically no chance of growth and would remain a firm performing below its potential. In the context of rapid domestic economic and financial sector transformations, an efficient management of banking operations aimed at ensuring growth in market share and profits requires up-to-date knowledge of all those factors that influence the banks. 0 0.2 0.4 0.6 0.8 1 GCB BBG EBG SCB STANBIC p ro fi ta b il it y (R O A ) in p er ce n ta g es Banks changes in profitability for the first five banks 2004 2013 University of Ghana http://ugspace.ug.edu.gh 6 However, most of the research done on the study of banks in Ghana has either focused on bank’s profitability or the competition in the banking industry. Most of the literature focused on how to increase the profits of banks without investigating what accounts for market shares in the Ghanaian banking industry and the relationship between market shares and profitability (Hinson et al., 2006; Kakra and Ameyaw, 2010; Akuffo-Duah, 2010; Kutsienyo, 2011; Awuah, 2011 and Domerher et. al., 2015). Secondly, most of the works on the relationship between market shares and profitability have been focused on developed countries and Ghana in particular, since the reforms, there has not been any specific work on the determinants of market share and how it affects bank’s profitability. In this regard, the study is set out to examine the determinants of market share and the relationship between market share and bank’s profitability. 1.3 Objectives of the study The aim of this study is to determine the factors that affect market share and the relationship between market share and profitability of banks in Ghana. More specifically, the study seeks;  To examine the determinants of market share in the banking industry.  To assess the relationship between market share and profitability in the banking industry. 1.4 Research questions  What are the determinants of market shares for banks? University of Ghana http://ugspace.ug.edu.gh 7  What is the relationship between market share and profitability? 1.5 Significance of study Though quite number of works have been done on banks in Ghana, the main focus has been on profitability of banks other than market shares and also the relationship between market share and profitability. The study aids to bridge the knowledge gap on the determinants of market shares and the relationship with profitability. The study provide policies for management to adopt practices that leads to higher market share in the banking sector. Furthermore, the study adds to the stock of knowledge about the relationship between profitability and market shares and provoke further researchers in this area on Ghana and other similar economies. Last but not the least, the conclusions of the study have important implications for both the present and future financial reforms in Ghana. 1.6 Organization of the study The study is organised into six chapters. The first chapter discusses introduction to the study which includes background to the study, problem statement, research question and objectives, methodology, significance of study, scope and limitations and organisation of the study. Chapter two gives an overview of the banking industry in Ghana and trends in market share and profitability of banks in Ghana as a whole. Chapter three also covers the necessary theoretical and empirical literature on the market share, profitability in the banking industry and its measurement. Chapter four discusses the methodology that is used University of Ghana http://ugspace.ug.edu.gh 8 in carrying out the study. It comprises the model, estimation technique and data used in the study. The fifth chapter focuses on the empirical analysis of the study while the sixth chapter presents the summary, conclusions and recommendations. University of Ghana http://ugspace.ug.edu.gh 9 CHAPTER TWO LITERATURE REVIEW 2.1 Introduction The purpose of this chapter is to present the relevant theoretical and empirical literature on the relationship between market share and profitability of banks as well as the determinants of market share. It provides the theoretical basis for the study as well as the empirical benchmark for discussing the findings of the study. The theoretical review also includes definition of terminologies and concepts as well as theories associated with market share and profitability. The chapter has three main sections. The first section reviews theoretical literature on the nexus between market share and profitability with the emphasis on two main contending theoretical arguments, market share as a predictor of profitability and profitability as a predictor of market share. The second section follows with an empirical review on the relationship between market share and profitability as well as the determinants of market share, with particular attention to the literature that attempted to establish the relationship between market share and profitability. The final part of the chapter is devoted to general conclusions drawn from both theoretical and empirical literature reviewed. 2.2 Definition of terms and concepts The terms and concepts include market share definitions, types, limitations, definition of profitability and it measures. University of Ghana http://ugspace.ug.edu.gh 10 2.2.1 Market Share Dunbar (1999) defines market share as the sum of fees which are spread per share multiplied by the number of shares in all offerings where the bank acts as book manager divided by the sum of all fees charged in the year. Depken (1999) explains market share as the number of wins by the team in a season as a proportion of total wins in the competition. Market shares command the attention of business managers as key indices for measuring the performance of a product or brand in the market place (Cooper et al., 2010). Gaining market share can be a means of obtaining profit (Cooper et al., 2010). Generally, market share in the banking sector context is the percentage gain of the total industry portion in terms of the operating asset, deposit and advances made to customers and other banks (Ghana Banking Survey, 2014). 2.2.1.1 Market change positioning Mostly, it is accepted that increased market share can be equated with success, while as decreased market share is a manifestation of unfavourable actions (O’ Regan, 2002). In any event, the position of the market of any firm can be influenced by voluntary or involuntary actions. The three main include increased market share, static market share and decreased market share. The three main market position can be influenced either by internal or external factors as discussed below. University of Ghana http://ugspace.ug.edu.gh 11 Increased market share O’Regan (2002) suggests that market share can be increased by enhancing the perceived value of the products or by reducing market price. Finlay (2000) proposes that increased market share occurs when:  Current markets are not saturated for the types of offer the firm is making;  Present customers are induced to buy more; Increased economies of scale provide significant competitive advantages; and the firm has spare production or distribution capacity. Decreased market share Decreased market share may be as a result of increased competition or as a deliberate action not to invest in new or improved products or services (O’Regan, 2002). Finlay (2000) proposes that decreased market share occurs when:  There is no money available for the enhancement needed to retain market share;  The firms market are being hit by cheap imports; and  The firm’s reputation has suffered and cannot be reclaimed. Static market share This type of market occurs as a result of a decision not to increase the current market share but to consolidate instead (O’Regan, 2002). On the other hand it may be that despite efforts University of Ghana http://ugspace.ug.edu.gh 12 to increase market share, the firm has only managed to retain its current share. Finlay (2000) proposes that static market share occurs when:  An owner-manager wishes their firm to remain the same size as it is now;  Market share is not important driver of profit; and  No funding is available to support market penetration or extensions. 2.2.1.2 Behavioural differences between high and low market share companies A conclusion made by O’Regan (2002) stated that possibly companies with increased market share behave very different from companies with a perceived decrease in market share. Specifically these behaviours were noted.  As market share increases, there is some tendency for market costs, as a percentage of sales, to decline.  Market leaders develop unique competitive strategies and have higher prices for their higher quality products than to smaller share businesses.  Market leaders spend more on research and development than the others.  As market share rises, turnover on investment rises but profits margin on sale increase sharply. 2.2.1.3 Limitations of market share concept In market share analysis, Kotler and Keller (2006) discusses the limitation of market share concept: University of Ghana http://ugspace.ug.edu.gh 13  The assumption that company’s performance should be judge based on the average performance of all the companies is not valid.  The proposition that higher market share leads to profitability is not always the case. Sometimes a decline in market share is a deliberate attempt to increase performance. For example dropping off unprofitable customers.  If new firms should enter the market, the share of the existing firms should fall is not always true.  The assumption that external forces affect all the companies is not true. 2.2.2 Profitability in the banking system Efficient and effective utilization of resources are key objectives of every banker (Berger et al., 1993). Increasing competition for financial services, technological innovation, and banking consolidation, for example, are all focusing more attention on controlling costs in banking and providing services and products efficiently (Berger et al., 1993). According to Howard and Upton (1991), profitability is an index of efficiency; and is regarded as a measure of efficiency and management guide to greater efficiency. Though, profitability is an important benchmark for measuring the productivity, the degree of profitability cannot be taken as an ultimate evidence of productivity. Profitability means ability to make return from all the business activities of an organization, company, firm, or an enterprise. It shows how efficiently the management can make profit by using all the resources available in the market (Burja, 2011). As stated by Hughes et al., University of Ghana http://ugspace.ug.edu.gh 14 (2008), banks’ ability to operate efficiently to obtain accurate information concerning its customers’ financial prospects and to write effective contracts and to enforce them depends in part on the property rights, legal, regulatory, and contracting environments in which they operate. Such environment includes accounting practices, chartering rules, government regulations, and the market conditions (e.g., market power) under which banks operate. Changes in these structures across governmental influences can lead to alterations in the profitability of banks across jurisdictions. The operating location can also influence the external and internal mechanisms that discipline bank managers (Howard and Upton, 1991). Internal discipline might be induced or reduced by organizational form, ownership and capital structure, governing boards, and managerial compensation. External discipline might be induced or reduced by government regulation and the safety net, capital market discipline (takeovers, cost of funds, stakeholders’ ability to sell stock (stock price)), managerial labor market competition, outside block holders (equity and debt), and product market competition (Howard and Upton, 1991). 2.2.2.1 Measures of profitability The Return on Asset (ROA) and the Return on Equity (ROE) have been broadly used as measures of profitability. ROA indicates how effectively a bank is managing its assets to generate income. ROA is the income earned on each unit of asset usually expressed as percentage (Kutsienyo, 2011). As stated by Kutsienyo (2011), the problem with ROA is that, it excludes from the total assets off-balance sheet items (for instance, assets acquired through a lease) thereby understating the value of assets. This can eventually create a University of Ghana http://ugspace.ug.edu.gh 15 positive bias where ROA is overstated in the evaluation of bank performance (Kutsienyo, 2011). Another measure of profitability is the Return on Equity (ROE) is computed by dividing net income by equity (Kutsienyo, 2011). It measures the income earned on each unit of shareholders capital. The shortfall of this measure is that banks with high financial leverage tend to generate a higher ratio. Banks with high financial leverage may be associated with a higher degree of risk although these banks may register high ROE (Mutshinyani, 2009). Thus ROE may sometimes fall short in exposing the true financial health of banks. Another challenge with using ROE is that it is affected by regulation. However, ROE is commonly used in conjunction with ROA (Mutshinyani, 2009). 2.3 Review of theoretical literature Theoretically, this study is based on the concept that market share determines profitability which is supported by various studies (Newton, 1983; Prescott et al., 1986; Jacobson, 1988; Gale and Buzzell, 1993; David, 2001 and Mutshinyani, 2009). Studies investigating the causality between market share and profitability has generally found out market share is a valid predictor of profitability and not the reverse (Prescott et al., 1986; O’ Regan, 2002 and Mutshinyani, 2009). Gale and Buzzell (1993) state that, market share is an important determinant of profitability since large market share is a reward for providing better value. Under most circumstances, companies that have achieved larger market share are considered more profitable than the smaller enterprises. The importance of market share is University of Ghana http://ugspace.ug.edu.gh 16 also acknowledged in the Boston Consulting Group Matrix as a key indicator of industry profitability (David, 2001). The scholars support the evidence with three feasible explanations of market share as a predictor of profitability. The scholars state that, firms with high market share tend to derive profitability from economies of scale and scope, efficiency and market power. This implies that, firms with increased market share are likely to have higher performance which helps to achieve enhanced financial performance. Figure 2.1: Conceptual framework explaining market share as a predictor of profitability Source: Author’s construct University of Ghana http://ugspace.ug.edu.gh 17 2.3.1 Theory of Economies of scale and Economies of scope Armstrong and collopy (1996) define economies of scale as the cost advantages that enterprises obtain due to size, output or scale of production, with cost per unit of output generally decreasing with increasing scale as fixed costs are spread out over more units of output. Economies of scope occur when it is more economical to produce two or more products jointly in a single unit than to produce the products separately (Armstrong and collopy, 1996). They state that economies of scope can arise from two sources;  Spreading fixed cost over an expanded product mix which occurs when the fixed capital of a bank branch is fully utilized by issuing many types of deposit to local residents than building separate offices to fulfil separate demands for transaction accounts, savings accounts, consumer loans, business loans and trust services.  Cost complementarities in producing different products. For example, cost complementarities occurs between deposits and loans when the payment flow information developed in producing deposit services is used to reduce the cost of acquiring credit information about and monitoring loans to the same customers. 2.3.1.1 Sources of Economies of Scale and Economies of Scope Boot (2003) and Walter (2003) discuss possible sources of economies of scale and scope in the banking industry, which can be broadly divided into four groups: University of Ghana http://ugspace.ug.edu.gh 18  Economies of scale and scope that are related to Information and Communication Technology (ICT). The increase of scale economies might be mostly influenced by improvements in ICT technology, which improve information processing within a bank. Boot (2003) mentions that, information technology presents the greatest possible source of scale economies by better use of databases over a wider range of services and customers. These generally have to do with spreading fixed overhead costs of ICT over a larger amount of operations, for instance through the processing of transactions or offering different products through the same distribution channel. Banks with larger size tend to derive profitability than the smaller banks.  Economies of scale and economies of scope that arise through reputation and branding. These arise because a wider range of products can profit from the same (generally fixed cost) reputation and brand. Banks with leading market share tend to derive profitability than the smaller banks and also existing banks perform better than the new banks.  Innovation related economies of scale and scope. This means that larger scale and scope will also help in making a profit from investment in Research and Development, which generally is a fixed cost.  Diversification of risk. These sources of scale and scope is more controversial. Traditional finance theory states that diversification will not bring any advantages to the firm as long as investors can costlessly diversify their own portfolios. In that case, investors University of Ghana http://ugspace.ug.edu.gh 19 will not pay a premium for diversified firms. Armstrong and Collopy (1996) show that, expansions into new markets can be thought of as real options. Entering a new market might open up the possibility for first-mover advantages and as such entering a market has a positive option value for a firm. Armstrong and collopy (1996) state that, larger firms tend to derive profitability from their economies of scale and economies of scope capability as well as their established branding which leads to higher market share. The macro environment might also play a role in the possibility of creating economies of scale. Thus, banks that operate in larger financial markets generally enjoy more economies of scale than those that operate in smaller markets. Bossone and Lee (2004) show that, economies of scale might not only lead to higher rewards to shareholders in the form of higher profits, but may also lead to higher rewards to bank management in the form of additional wages. Therefore, wages may be considered either as an input or output in the production process. 2.3.2 Theory of the market power Market share and profitability can be explained as advantages of market power. Market power is the ability of a firm to raise its price because its rivals are not able to provide consumers a reasonable alternative (Newton, 1983). Market power in banking is generally proxied by taking the Herfindahl-Hirschman Index (HHI), which takes the sum of quadratic asset shares of banks within a country or region, or concentration ratios, which measure the share of total assets within a country held by the big four, five or six banks. According to this theory, market power would enable firms to gain higher profits as they are able to University of Ghana http://ugspace.ug.edu.gh 20 charge a higher premium for their product. Also, firms with market leader status tend to derive profitability from economies of scale capability and that firms with increased market share are likely to have higher performance which helps to achieve enhanced financial performance, greater customer retention and customer satisfaction (David, 2001). Oster (1994) contends that, the size of the market share held by the firm as well the size of the major firms in the market are important considerations for increased profitability. 2.3.3 Efficiency theory Mutshinyani (2009) states that, higher market share enables firms to utilize economies of scale to reduce costs and to gain market power. Firms with large market share can exploit increasing returns to scale through the areas of procurement, marketing, advertising, research and development (Jacobson, 1988). Again, the study further states that, it is the role of market share that reduces cost and ensure cost efficiency rather than generating market power that generates the association between market share and profitability (Jacobson, 1988). In summary, one can conclude from the theoretical review that market share is a predictor of profitability. Moreover the theoretical discussions do not provide a firm conclusion on the relationship between market share and profitability, although the majority of opinion tends to be in a favour of the assertion that, increased market share leads to increased profitability so there is a need for empirical review on the relationship between these two notions. University of Ghana http://ugspace.ug.edu.gh 21 2.4 Empirical Review The empirical literature presented here are studies conducted in other countries or markets but are however relevant to the study. The section discusses the relationship between market share and profitability, the internal and the external determinants of market shares. The chapter is concluded with other empirical evidence which has been done on the determinants of market shares. 2.4.1 Relationship between Market shares and profitability in the Banking sector Newton (1983) states that, if we ignore the possibility of double coincidence, the correlation between market shares and profitability can be interpreted as the following:  Market shares determines the profitability of a business  Profitability determines market shares  Or some other variables determine both market share and profitability which can be interpreted as a well- managed and successful firm enjoy higher profitability and natural growth. Newton (1983) analyses whether market share leads to higher profitability by using the Profit Impact of Market Strategies (PIMS) in North American industries. The study concludes that, there is positive but weak relationship between market share and profitability since these two are apparent in some industries whiles others are not in North America. The industries include the cement, food and pharmaceutical industries in North America. The pharmaceutical industry exhibits a positive relationship between market share and profitability whiles the cement industry exhibits a static relationship between University of Ghana http://ugspace.ug.edu.gh 22 them. For the food industry, the study concludes that, even the small firms sacrifice profitability to obtain a higher market share in the long run. The study suggests that the models used in the paper are an oversimplication of the situation and need to be enhanced by the introduction of several subjective measures in order to make them workable. Other studies also support a positive relationship between market share and profitability. (Wernerfelt, 1986; Venkatraman et al. 1990; Davis et al. 1993 and O’Regan 2002). However, the methodology employs by these scholars vary from one another. Wernerfelt (1986) analyses the relationship between market share and profitability using differential game theory in the United States (US), Chicago in 1985. The study suggests that, increased market share leads to higher profitability when the firm attack early in the market otherwise there will be a negative relationship between them. Venkatraman et al. (1990) employ the comparative static methodology to determine the relationship between market share and profitability in individual strategic business in the US between the periods of 1976 to 1986. The study concludes that, the relationship between these two variables are meaningless unless related to a particular environment, strategies pursued and macroeconomic condition. Davis, Schul, Babakus and Pedrick (1993) investigate the relationship between market share and profitability by using the linkages between porter’s generic business strategies of low cost and differentiation, and profitability in the United States paper and pulp manufacturing industries between the periods of 1980 to 1990. The study states that, whiles this approach provides some control over the possible confounding effects of variations in environmental or industry factors, it does not limit the generalizability of the study findings to other industries. O’Regan (2002) examines the relationship between University of Ghana http://ugspace.ug.edu.gh 23 market share and profitability over 1000 sample of small and medium manufacturing firms throughout United Kingdom in 2002 using the likert type scale and recommends that any future research should consider a more in-depth approach since they did not attempt to a more in-depth approach. Nevertheless, Yannopoulos (2012) argues that, the positive association of market share and profitability might be an artifact of the prevailing stability in the economic and competitive environment during their time of study. The competitive and environmental conditions prevailing at the time of study led to that conclusion. In contrast to the theory that states that there is a positive relationship between market share and profitability, Yannopoulos (2012) examines the relationship between market share and profitability using the largest firms in the United States and United Kingdom between the periods of 2008 to 2012 and finds out that market share and profitability may not be related directly but any observed relationship may be as a result of a spurious correlation by testing the hypothesized U-shaped proposed by Porter, Sheth and Sisodia among others using cross sectional data across industries. The results obtained in this analysis support a negative relationship between market share and profitability and is an indication of the hypothesis that when unstable competitive and environmental conditions prevail, higher market shares do not necessarily lead to higher profitability or to a U-shaped relationship and recommends that future studies of the market share effect should look at the moderating role of the economic and competitive environment. This study is in agreement with Hagigie et al., (1999). The study suggests that, an increase in market share requires more investment which might reduce profitability in the short run. This means that, for an organisation to University of Ghana http://ugspace.ug.edu.gh 24 pursue higher market shares, it must sacrifice some profit meaning there is an inverse relationship between market share and profitability. Nevertheless, the methodology employs by these scholars vary from one another. Yannopoulos (2012) used the cross sectional data analysis while Hagigie (1999) used the dynamic panel data. The study concludes that for an organisation to pursue higher market shares, it must sacrifice some profit meaning there is an inverse relationship between market share and profitability. These studies that support the negative relationship between market share and profitability ignored some of the dominant factors that explains the relationship between market share and profitability such as management skills, access to resources etc. In contrast to the former findings, Prescott, Kohli and Venkatraman (1986) examine the nature of the relationship between market share and profitability across a taxonomy of homogenous environments in the US using the comparative static methodology and the Profit Impact Marketing Strategy (PIMS) research data base. The major findings show that, the association between market share and profitability is context-specific. Secondly, there is existence of both direct and spurious relationships between them and lastly, the validity of market share as a predictor of profitability is context –specific. Based on the findings, the authors conclude that, there is no relationship between market share and profitability. The study recommends that, the pursuit of market share as a goal must be done cautiously. This assertion is supported by Mutshinyani (2009) who study the relationship between University of Ghana http://ugspace.ug.edu.gh 25 market share and profitability in South Africa eight retail companies from 2004 to 2008 but the sample size used in the study is small and therefore, the study recommends that, a more sample size should be used in future studies. However, the former used the comparative static methodology whiles the latter used a quantitative through the utilization of hypothesis testing. Summary of empirical review is shown below on table 2.1 Table 2.1 showing prior studies on the relationship between market share and profitability. Study Methodology Market share and Profitability relationship Newton (1983) Profit Impact of Market Strategies(PIMS) Positive relationship Wernerfelt (1986) Differential game Positive relationship Prescott et al. (1986) comparative static methodology No relationship Venkatraman et al. (1990) comparative static methodology Positive and significant Davis et al. (1993) Porter’s generic business strategies Strongly positive Hagigie (1999) Dynamic Panel data Negative relationship O’Regan (2002) Likert type scale Positive relationship Mutshinyani (2009) Quantitative method No relationship Yannopoulos (2012) Cross sectional Negative relationship Source: Author’s construct These results depict that there is no agreement on the relationship between market share and profitability according to table 2.1. University of Ghana http://ugspace.ug.edu.gh 26 2.4.2 Determinants of Market Shares The determinants of market shares are grouped into two; namely the internal determinants or banks specific factors and external determinants which are the macro economic variables. 2.4.2.1 Internal Determinants The internal determinants (bank specific factors) of market shares can be defined as those factors that are influenced by the bank management decisions and policy objectives. The internal determinants may include but not limited to the size and location of branches, number of employees, operational efficiency, marketing competencies, management competencies, motivation, quality of customer service and strategy. Whereas it may be challenging to measure some of these variables, they are indirectly reflected in the operating performance of the banks which can be derived from the balance sheet and income statements of the firms under study. In this study, factors such as bank liquidity, asset quality, bank size, operating efficiency, ownership structure, bank age and number of branches are analysed. Liquidity Liquidity is a major concern for banks and the shortage of liquidity can trigger bank failure (Kosmidou, 2008). Banking regulators also view liquidity as a major concern. This is because banks without sufficient liquidity to meet demands of their depositors risk experiencing bank run. Liquidity measures the ability of banks to meet short-term obligations or commitments when they fall due. Traditionally, banks take deposits from University of Ghana http://ugspace.ug.edu.gh 27 customers and give out loans. For this reason, the ratio of bank’s advances to customer deposits is used as a proxy for liquidity (Kosmidou, 2008). Studies by (Molyneux and Thornton, 1992; Guru et al., 1999 and Tabar et al., 2013) found that banks performance is negatively related to liquidity which also affect market shares. These studies used the panel data analysis. Holding assets in a highly liquid form tends to reduce income as liquid asset is associated with lower rates of return (Molyneux and Thornton, 1992). For instance, cash which is the most liquid of all assets is a non-earning asset. It would therefore be expected that higher liquidity would negatively correlate with banks performance. Conversely, Bourke (1989) examines the performance of banks in twelve European, Northern American and Australian countries. Using the international data for 1972-1981, the study found that both ratios of capital and liquidity have a positive relationship with the market performance using panel data analysis. Kosmidou et al. (2005) also found a significant positive relationship between liquidity and bank’s performance in the domestic United Kingdom’s banking industry between the periods of 1995 to 2002 using a panel data. Bordeleau and Graham (2010) also establish a mixed relationship between these two variables using nonlinear regression model. The dataset used for estimation contains a panel of quarterly observations for 55 United States (US) bank holding companies and 10 Canadian banks, spanning the period from 1997 to 2009. The study concludes that, all University of Ghana http://ugspace.ug.edu.gh 28 things been equal, if a bank is more reliant on short‐ term funding, it may need to hold more liquid assets in order to maximize profits and if a bank is not more reliant on short term funding, it would hold less liquidity. Thus, the conclusion on the impact of liquidity and bank profitability is uncertain and may require further empirical work. Asset quality The loans to total assets ratio is a measure of income source of banks and it is expected to affect performance positively unless bank takes on unacceptable level of risk. This ratio is one of the important measures of asset quality (Alper and Anbar, 2011). According to Heffernan (1996), credit risk is when an asset or a loan becomes irrecoverable in the case of outright default, or the risk of delay in the servicing of the loan. Credit risk can have a rippling effect thus leading to insolvency (Bessis, 2002). The higher the provision for bad debt to advances ratio, the higher the credit risk and the higher the accumulation of unpaid loan and interest. Additionally, present value of the asset declines, thereby undermining the solvency of a bank. The Risk-Return Hypothesis implies that, high risk should be associated with high profitability indicating a positive relationship. However, according to Kosmidou (2008), poor asset quality can have adverse impact on bank profitability, reducing interest income revenue, and by increasing the provisions cost. This shows that, the relationship between asset quality and bank’s performance are inconclusive and therefore, further empirical research need to be done on it. University of Ghana http://ugspace.ug.edu.gh 29 Flamini et al. (2009) using a sample of 389 banks in 41 Sub-Saharan Africa (SSA) countries to study the determinants of bank profitability between the periods of 1998 to 2006 found a positive relationship between asset quality and bank’s performance. The study finds that, apart from credit risk, higher returns on assets are associated with larger bank size, activity diversification, and private ownership which also affect the performance of banks and their market shares using panel data analysis. The study ignores the fact that bad loans may reduce profitability which resulted in the positive relationship between asset quality and bank performance. In contrast, Acaravci and Calim (2013) establish a significant and negative impact of quality asset on profitability for privately-owned bank and a significant and positive impact on profitability for a foreign bank in Turkey. According to study, the researchers expected a positive relationship between asset quality and performance. But, the coefficient of this ratio is also expected to be negative because bad loans are expected to reduce profitability. They used Johansen and Juselius cointegration test approach by collecting data from the three biggest state-owned, privately-owned and foreign banks in Turkey. The sample period spans from 1998 to 2011. Bank size Bank size is associated with diversification which may impact favourably on risk and product portfolio. Economies of scale will reduce the cost of gathering and processing information (Boyd et al., 1993) so that a positive effect of bank size is associated with University of Ghana http://ugspace.ug.edu.gh 30 market share. These studies have the same findings but used different methodological approaches. Bourke (1989), Molyneux et al., (1992) Bikker et al., (2002) and Goddard et al., (2004) have all linked bank size to capital ratios, which they claim to be positively related to size using the panel data analysis. The results imply that, as size increases, market share increases. This is especially true for banks that have large assets since those banks have the highest share in the market. Scholtens (2000) examines the hypotheses on whether size matters for individual banks' profit performance in Canada, France, Germany, Italy, Japan, the UK, and the US during the years of 1981 to 1997.by using the extreme bounds stability analysis and a stability analysis in line with Sala-i-Martin (1997) who also used to test for the reliability of the regression outcomes. It turned out that there is a positive relationship between bank shares and their size. Berger and Bouwman (2013) also found out that there is a positive relationship between market shares and bank size from 1984 to 2009 in Russian banks using the logit regression. Operating Efficiency The expense-to-income ratio is used as a proxy for operating efficiency. The expense-to- income ratio is defined as the operating costs over total generated revenues. A negative correlation is expected between the operating cost and performance implying that higher operating cost means lower profit and vice-versa (Kutsienyo, 2011). University of Ghana http://ugspace.ug.edu.gh 31 Antonio (2012) analyses the determinants of Spain bank’s performance from 1999 to 2009 by applying the system-Generalised Method of Moment (GMM) estimator and found out that there is a negative relationship between operating efficiency and banks’ performance which also affect market shares negatively. This is confirmed by Ayanda et al., (2013) who also found a negative relationship between market performance of banks and operating efficiency in Nigeria. The former used the system-GMM estimator while the latter used the econometric analysis of cointegration and error correction technique. Ownership Structure The structure of the ownership of banks is subject to numerous regulations aiming to protect the rights of depositors, shareholders and to make banks operate more efficiently (Micco, 2004). Since many parties participate in banking activities, it is expected that other ownership characteristics namely institutional ownership, government ownership, block holder ownership and its concentration would also affect bank market performance. Micco et al. (2004) study on the relationship between bank ownership and performance of their market share by building a dataset on bank ownership and bank performance covering approximately 50,000 observations for 119 countries over the period of 1995-2002 using the dynamic panel data analysis. They found that, while ownership is strongly correlated with performance in developing countries, it is not in industrial countries. The paper further suggests that, state owned banks operating in developing countries tend to have a lower market share and higher costs than their private counterparts. This is supported by Setiyono and Tarizi (2014) using data from bank scope for publicly-listed commercial banks in 11 University of Ghana http://ugspace.ug.edu.gh 32 Asian countries over the 2004 to 2010 period. The findings show that, public ownership is negatively associated with banks performance but positive with private or foreign banks using the dynamic panel data analysis. This assertion is criticized since it is not always true that foreign banks perform better local banks but rather banks that have high operational cost will not perform better. However, a study by Cao et al. (2010) examine the relationship between market performance and the state owned banks from 1999 to 2004 using the ordinary least squares and the two stage least squares. The data is collected from the China Stock Market and Accounting Research Databases. The result indicates that, firms with worse market performance borrow more and banks within strategic industries borrow less due to the alternative financial support from the government. Zouari and Taktak (2014) study the ownership structure and financial performance in Islamic banks in a panel data sample of 53 Islamic Banks scattered over 15 countries for a five-year period (2005-2009). Empirical evidence shows that, there is no obvious correlation between ownership concentration and Islamic bank performance. In addition, the results reveal that, family and state ownership affect positively bank’s performance. The study also indicate that, banks with institutional and foreign shareholders do not perform better either, since they either borrow more or the government support the local banks than the foreign banks. This result is due to the economic instability during the time of the study. University of Ghana http://ugspace.ug.edu.gh 33 Bank Age The Age of the bank is the number of years that the bank has been in existence which is equivalent to the year of establishment (Arjan, 2012). Banks are significant entities in the economy and for that matter the number of years of which it came into existence is important to check if the age of the bank have an effect on bank performance and for that matter market share. Göbel and Zwick (2009) study on the relationship between age and productivity by using the linked employer employee data which is an institute of employment research and applied the methods of Generalised Method of Moment (GMM) to estimate the results. The study is carried in German companies from 1997 to 2005. The study found a negative relationship between age of companies and their market performance, but the results are biased since they either did not take into account endogeneity of the age of the company, time dependencies, or essential information correlated with productivity and age shares such as workforce tenure, the technical state of capital, and qualification. This is supported by (Ruis and Scholman, 2012) who also found a negative relationship between them. The former used the linked employer employee data which is an institute of employment research and applied the methods of Generalised Method of Moment (GMM) to estimate the results while the latter used a combination of a descriptive analysis with econometric estimation techniques. Uschi and Veen (2010) assess the impact of age on the Company’s performance by using linked employer-employee data set for Germany based on more than 18,000 companies University of Ghana http://ugspace.ug.edu.gh 34 over a ten year period of young and old companies in Switzerland. The results indicate that, there is a positive relationship between age and the performance of the companies and made a conclusion that the exact shape of the cost and benefit curves depends on what type of skills and on what type of production process or business task an organisation is characterized by and the old companies have skills advantage, experience and larger networks to the new companies. Number of Branches A study done by Shukla and Sinha (2013) show that number of branches are assumed to have a positive relationship with market share. This is because the number of branches show the number of workers and assets in the sector and these workers are being trained to acquire knowledge and skills to affect productivity. These studies below have the same findings, but used different approaches. Samad et al. (2006) examine the association between the number of branches of commercial banks with headquarters in the State of Utah to its market performance between the periods of 2000 and 2004 by applying T-tests and Kruskal-Wallis tests. The paper found no significant difference in performance between small and large banks between the years 2000 and 2004. This is confirmed by Shukla and Sinha (2013) who also found a positive relationship between the number of branches and bank’s performance in the Indian banking industry by using a self-developed questionnaire, measured on a likert scale to collect data from respondents in the year 2013. University of Ghana http://ugspace.ug.edu.gh 35 2.4.2.2 External Determinants These are factors the banks do not have control over them, although banks can anticipate changes in the external environment and position themselves strategically to take advantage of them. The external environment defines the legal, political, economic, technological, and social landscapes in which banks operate (Kutsienyo, 2011). As stated by Kutsienyo (2011), there are two types of external determinants; financial structure determination and macroeconomic determinant. In this study, the macroeconomic determinant would be employed. The macroeconomic-specific determinants reflect the general macroeconomic and market conditions in the country. Gross Domestic Product (GDP) and inflation which are macroeconomic factors are adopted as external factors to be examined as they are widely studied in other countries. Inflation Perry (1992) suggests that, the extent to which inflation affects bank performance depends really on whether inflation expectations are fully anticipated or not fully anticipated. If the bank fully anticipates the inflation rate, then this implies that it can accordingly adjust its interest rates in order to increase their revenues faster than their costs and thus acquire higher economic profits and if banks do not fully anticipate then banks’ performance would be reduced. Athanasoglou et al. (2006) examine the profitability behaviour of bank-specific, industry related and macroeconomic determinants, using an unbalanced panel dataset of South Eastern European (SEE) credit institutions over the period 1998-2002. The estimation University of Ghana http://ugspace.ug.edu.gh 36 results indicate that the macroeconomic environment has a direct impact on the aggregate performance of the industry. With respect to the macroeconomic variables, inflation has a strong effect on performance, especially when it is anticipated. This is supported by Gul et al. (2011) who examine the relationship between bank specific and macroeconomic characteristics over bank profitability by using data from the top fifteen Pakistani commercial banks over the period 2005-2009. The authors investigate the impact of inflation and on major profitability indicators i.e., return on asset, return on equity, return on capital employed and net interest margin separately. The empirical results found strong evidence that inflation has a strong influence on the profitability especially when is anticipated and affects profitability when not anticipated. However, the former used unbalanced panel data while the latter used a balanced panel data. Gross Domestic Product (GDP) The GDP is the measure of total economic activity within the economy and it is commonly used economic indicator. Staikouras and Wood (2004), review the performance of European banking industry for the years 1994-1998. Using ordinary least square method and fixed effects model they conclude that growth of the GDP exerts a significant negative impact on bank’s market performance. This is supported by Sufian (2011) who analyse 11-29 Korean commercial banks during year 1992-2003 by using linear regression model. Kanwal and Nadeem (2013) also support this assertion by using the pooled ordinary least square and fixed effect. University of Ghana http://ugspace.ug.edu.gh 37 Gul et al. (2011) examine the relationship between bank macroeconomic characteristics over bank’s performance by using data from top fifteen Pakistani commercial banks over the period 2005-2009. They investigate the impact of economic growth on major profitability indicators which include return on asset, return on equity, return on capital employed and net interest margin separately and found out that there is a positive relationship between GDP and bank performance. Kosmidou (2008) also supports the positive relationship by examining the relationship between bank specific and macroeconomic characteristics over bank profitability and found a positive relationship. They both used panel data analysis. 2.4.2.3 Further Empirical Studies of Determinants of Market Share There are other factors that affect market share. These include customer service quality, capital adequacy and industry specialisation. Customer service quality According to Kumar and Reinartz (2006), traditionally, customer satisfaction is a key mediator which leads to greater retention or loyalty, in turn resulting in greater profit for a firm. Organizations that make specific efforts to meet customers’ needs retain the old customers and gain new customers (Martin, 1989). These specific efforts include both human and non- human factors. These studies found a positive relationship but used different methodologies. University of Ghana http://ugspace.ug.edu.gh 38 Wiele et al. (2001) found a significant positive relationship between customer service quality and banks performance using. Their focus is to relate the customer satisfaction data gathered in 1998 with data on business performance in 1998 and 1999 with the use of primary data in their work and found out that there is a positive relationship between customer service quality and market share. Masukujjaman et al., (2010) also examine empirically the determinants of service quality in Bangladesh. A questionnaire for such purpose is designed and different statistical methods are applied to analyse the collected data. Again Tran et al. (2010) analyse empirical research on customer satisfaction and bank performance and found out that there is a positive relationship between these two variables using the linear regression model. Capital adequacy The ratio of equity to total asset is employed as a measure of bank capital adequacy. This measures the percentage of the total asset that is financed with equity capital. Capital adequacy therefore describes the sufficiency of the amount of equity that can absorb shocks that banks may experience. It is expected that the higher the equity to asset ratio, the lower the need for external funding and therefore the higher the profitability of the bank. In addition, well-capitalised banks face a lower cost of going bankrupt which reduces their cost of funding (Kosmidou, 2008). Nwankwo (1991) states that, adequate capital is the quantum of funds which a bank should have or plan to maintain in order to conduct its business in a prudent manner. It is the capital function of preventing bank failure by absorbing losses. These losses are related to University of Ghana http://ugspace.ug.edu.gh 39 the risks which banks undertake as a natural corollary of their efforts to serve the legitimate credit needs of the community. Adequate capital provides the ultimate protection against insolvency and liquidation arising from the risk in banking business. Any company or bank with inadequate capital faces hidden constraints. Berger (1995); Olalekan and Adeyinka (2013) found a positive relationship between capital adequacy and return on equity. The former used Granger's causality test while the latter used primary data analysis. This is in contrast with studies by (Berger, 1995; Demirguc-Kunt and Huizinga, 1999; Kosmidou, 2008; Pasiouras et al., 2006; Gul, Irshad and Zaman, 2011) which shows a significant negative relationship between bank profitability and capital adequacy by the use of pooled panel data for banks. Industry specialization Dunbar (1999) examines the effect of several factors on the market share of established banks between 1984 and 1995 using the Herfindahl Hirschman Index (HHI). Unranked banks are assigned the lowest ranking conferred in each year. Industry specialisation is negative for large banks and positive for smaller banks. He adds that, the large banks spend a lot of resources during industralisation as compared to the smaller banks which therefore drain the profitability of the large banks. University of Ghana http://ugspace.ug.edu.gh 40 2.5 Conclusion This chapter presented the relevant theoretical and empirical literature on the relationship between market share and profitability of banks as well as the determinants of market share. The theoretical review also included definition of concepts as well as theories associated with market share and profitability. The empirical review on the relationship between market share and profitability, with particular attention to the literature that attempted to establish the relationship between market share and profitability as well as the determinants of market share were also analysed. University of Ghana http://ugspace.ug.edu.gh 41 CHAPTER THREE OVERVIEW OF THE BANKING INDUSTRY 3.1 Introduction This chapter presents the overview of the banking industry, structure, discussion of the banking industry’s market performance in terms of total operating assets, industry deposit and industry advances from between 2004 and 2013 and the trend analysis of profitability between the period of 2004 and 2013. 3.2 Overview of the Banking industry Before the adoption of the Universal Banking Business Licence (UBBL), the banking sector has traditionally been segmented into merchant, commercial (retail) and development banks. While merchant banks have been restricted to corporate clients, the commercial and development banks have traditionally had customers across the entire financial market segments. It is against this backdrop and the need to create a level playing field for all banks that the idea of universal banking was adopted (Bank of Ghana, 2004). The aim is to allow all banks that comply with the prescribed capital requirements, the freedom to engage in permissible banking business without restrictions and thereby eliminate the compartmentalization (Hinson et al., 2006). Hinson et al. (2006) noted that, “before the passage of the universal banking law, banking was conducted along such narrow scopes as commercial, developmental or merchant banking”. In February 2004, Bank of Ghana formally introduced the Universal Banking Business Licence (UBBL), which is expected to bring more competition within the industry. For the banks to operate under the UBBL, existing banks must have a minimum net worth of University of Ghana http://ugspace.ug.edu.gh 42 GH¢70billion (excluding statutory reserves), and new banks should have a paid-up capital of GH¢70billion. Banks are required to hold 9% of the cedi and forex deposit base with the Bank of Ghana on daily basis as primary reserves and 35% of their deposit base in cedi denominated assets as secondary reserves (Banking Act, 2004 Act 673). With the passage of the universal banking law however, all types of banking can be conducted under a single corporate banking entity and this greatly reorganises the competitive scopes of several banking products from banks throughout the country are constantly seeking unique ways of differentiating their offering (Hinson et al., 2006). The willingness and ability of managers in banks to respond to changes in the economy will determine whether their own organizations survive and prosper. With so many changes occurring in the Ghanaian banking sector, including an expansion and intensification of competition and market share, the issue of market share and profitability have gained a considerable thought. 3.3 Structure of the banking industry According to the annual report of the Bank of Ghana (2013), the banking industry is made up of the universal banks, Association of Rural Banks (ARB) and the Non-Banking Financial Institutions (NBFIs). The universal banks consist of 15 foreign and 12 local banks. The ARB consist of 136 Rural and Community Banks (RCBs) which have been established throughout the regions in the country. The non-financial Institutions consist of finance house, savings and loans, leasing companies, mortgage company, micro finances and credit bureaux (Bank of Ghana, 2013). Hinson et al. (2006) defined universal banks as University of Ghana http://ugspace.ug.edu.gh 43 the commerce of accepting deposits and other repayable funds from the public, lending, investments in financial securities and money transmission services, the issuance and administration of means of payment, including credit cards, travelers’ cheques and bank drafts. Figure 3.1 depicts the structure of the banking Sector of Ghana which includes universal banks, ARB APEX banks and non-banking financial institution licensed by the Bank of Ghana. In August 2013, Bank of Ghana established the Banking Supervision Department (BSD) and Other Financial Institutions Supervision Department (OFISD) to provide close supervision of the Rural and Community Banks (RCBs), Micro Finance Institutions (MFIs) and forex bureaux. The BSD now supervises the Universal Banks, other NBFIs and credit reference bureaux. There were 27 Universal Banks, 57 NBFIs, 140 RCBs and 337 MFIs at the end of the year 2013. The number of credit reference bureaux remained at three (Bank of Ghana annual report, 2013). In 2013, the universal banks increased from 26 to 27 which was first capital plus bank. The Ghanaian owned Banks are 12 while as foreign Banks are now 15. Four Rural and Community Banks (RCBs) were licensed to bring the number to 140 since it was 136 in 2012. University of Ghana http://ugspace.ug.edu.gh 44 Figure 3.1: Summary of the structure of the Banking sector in Ghana Source: Bank of Ghana Annual Report (2013, pg. 21-24). Six new NBFIs were licensed in the year while one moved to universal bank, increasing the number of NBFIs from 52 to 57 in 2013. This consisted of 31 finance houses, 23 savings University of Ghana http://ugspace.ug.edu.gh 45 and loans companies, 2 leasing companies and 1 mortgage company. (Bank of Ghana annual report, 2013). 3.4 Market share analysis for the banks in Ghana from 2004 and 2013 The market share analysis of the selected banks is discussed in terms of the share of the operating asset, the share of industry deposit and the share of industry advance. 3.4.1 Share of operating asset Figure 3.2 below shows the trend analysis for industry operating asset for the period of 2004 and 2013 for the banks considered for the study. Ecobank (EBG) shows the highest market share of 13% in 2013 and has continued to be the dominant bank as it merged with The Trust Bank (TTB) in 2012 which has strengthened the customer base and branch network extension. EBG indicates about 5.26% increase from 2004 to 2013 in terms of the operational asset while Ghana Commercial Bank (GCB) had 8.79% decrease from 2004 to 2013. The banks that show an increase in the share of the operating asset are EBG, Stanbic, CAL, HFC and Bank Of Africa (BOA) with percentage increase of 5.26, 5.28, 1.66, 0.79 and 0.56 respectively. Stanbic is also growing due to the strategy of stemming loan losses and leverage on its corporate and investment to target the international clients. University of Ghana http://ugspace.ug.edu.gh 46 Figure 3.2: showing the trend analysis for industry operating asset between the period of 2004 and 2013 for the selected banks in Ghana. Source: Ghana Banking Survey, 2008 and 2014 Banks which include GCB, Standard Chartered Bank (SCB), Barclays Bank Ghana Limited (BBGL), Agricultural Development Bank (ADB), National Investment Bank (NIB), Société Générale Social Security Bank (SG-SSB), Prudential Bank Limited (PBL), First Atlantic Bank Limited (FABL), First Bank of Nigeria (FBN) and Universal Merchant Bank (UMB) have a reduction in the share of the operating asset with a percentage decrease of 8.97, 5.13, 9.04, 6.02, 1.09, 4.62, 0.55, 0.44, 0.18 and 4.43 respectively. Though GCB has the second largest share in 2013, the trend shows a decrease in the share. Among these banks, BBG has a decrease in the share more than the others while as FBN has the least decrease which shows that, FBN is the best among the non-performing banks. The decrease 0.00% 2.00% 4.00% 6.00% 8.00% 10.00% 12.00% 14.00% 16.00% 18.00% 20.00% sh ar e o f o p er at in g a ss et s in P er ce n ta g es Banks Trend analysis of industry operating asset 2004 2013 University of Ghana http://ugspace.ug.edu.gh 47 in operating assets for the banks is also due to fewer network branches and poor networking. 3.4.2 Share of industry deposit Figure 3.3 below shows the trend analysis of industry deposit for the period of 2004 and 2013 for the banks considered for the study. Figure 3.3: The trend analysis for industry deposit for the period of 2004 and 2013 Source: Ghana Banking Survey, 2008 and 2014 The banks that have an increase in the share of the industry deposit in 2013 are EBG, Stanbic, CAL, NIB, PBL, HFC and BOA with percentage increase of 5.5, 6.28, 0.65, 0.9, 0.23, 1.09 and 0.31 respectively. This shows that, Stanbic is increasing its share rapidly though EBG shows the highest share in 2013. 0.00% 5.00% 10.00% 15.00% 20.00% 25.00% sh ar e o f in d u st ry D ep o si t in P er ce n ta g es Banks share of industry deposit 2004 2013 University of Ghana http://ugspace.ug.edu.gh 48 EBG still leads the market due to its relation with the customer service as GCB is losing since the extensive branch does not seem to support in the mobilization of funds to the bank. Banks which include GCB, SCB, BBGL ADB, SG-SSB, FABL, FBN and UMB have a reduction in the share of the industry deposit with a percentage decrease 0f 10.13, 7.92, 9.93, 3.55, 3.74, 0.28, 0.61 and 4.88 respectively. Though GCB has the second largest share in 2013, the trend shows a decrease in the share. Among these banks, GCB has a decrease in the share more than the others while as FABL has the least which shows that FABL is the best among the non-performing banks. This decrease occurs when bank branches are not strategically located and also few numbers of branches. 3.4.3 Share of industry advances Figure 3.4 below shows the trend analysis for industry advances for the period of 2004 and 2013 for the banks considered for the study. From figure 3.4 below, the banks that have an increase in the share of the industry advances were EBG, STANBIC, CAL, PBL, HFC, BOA and FBN with percentage increase of 7.37, 5.08, 3.53, 0.69, 1.25, 1.38 and 0.29 respectively. These demonstrates that, EBG is increasing its share rapidly as compared to the first seven banks FBN been the least. Banks that have the highest share in advances are due to its reduction in lending to the services sector by 3% and increase its lending to communication, transport, mining and quarrying, construction and other sectors in the economy (Ghana Banking Survey, 2014). University of Ghana http://ugspace.ug.edu.gh 49 Figure 3.4: showing the trend analysis of industry advances Source: Ghana Banking Survey, 2008 and 2014 Banks which include GCB, SCB, ADB, BBGL, SG-SSB, NIB, FABL and UMB have a reduction in the share of the advances with a percentage decrease 0f 11.39, 7.32, 1.7, 12.52, 1.91, 3.61, 0.93 and 5.57 respectively. Though GCB has the third largest share in 2013, the trend shows a decrease in the share. Among these banks, BBG has a decrease in the share more than the others while as FABL shows the least decrease which implies that, FABL is the best among the non-performing banks. These reductions are due to the management style of the some of the banks and poor lending services. 3.5 Trend of profitability analysis for the banks in Ghana (2004 and 2013). The trend of the profitability of the banks has shown an increase for some banks from the period of 2004 to 2013 while others have experienced deterioration within the period. 0.00% 5.00% 10.00% 15.00% 20.00% sh ar e o f in d u st ry a d v an ce s in p er ce n ta g es Banks Share of industry advances 2004 2013 University of Ghana http://ugspace.ug.edu.gh 50 3.5.1 Trend analysis of Return On Average Asset (ROA) of the selected banks Figure 3.5 below shows the trend analysis for the selected banks based on the Return on Asset. While some banks have experienced an increase in the ROA, others have suffered from the period of 2004 to 2013. This shows that, the largest banks do not always have the highest profit. Figure 3.5: Trend of ROA Source: Annual report of individual banks, 2004 and 2013 From figure 3.5, banks that have experience increase in their ROA include BBG, EBG, SCB, STANBIC, CAL, HFC and FBN with a percentage increase of 0.7, 0.007, 0.02, 0.005, 0.5, 0.22 and 0.001 respectively. This shows that in terms of profitability, BBG is the leading bank. Banks with a declining ROA include GCB, ADB, UMB, SG-SSB, NIB, PBL, BOA and FABL with a percentage decrease of 0.73, 0.01, 0.04, 0.03, 0.008, 0.00071, -0.2 0 0.2 0.4 0.6 0.8 1 A v er ag e R et u rn o n A ss et i n P er ce n ta g es Banks Trend of profitability based on the Return On Avergae Asset 2004 2013 University of Ghana http://ugspace.ug.edu.gh 51 0.00506 and 0.0023 respectively. GCB has suffered in the ROA from 2004 to 2013 which may be based on the management of the bank. 3.5.2 Trend analysis of Return on Average Equity (ROE) of the selected banks Figure 3.6 below shows the trend analysis for the selected banks based on the Average Return on Equity. While some banks have experienced an increase in the ROE, others have writhed from the period of 2004 to 2013. Figure 3.6: Trend of ROE Source: Annual report of individual banks, 2004 and 2013 From figure 3.6, banks that have experienced increase in their ROE include GCB, EBG, PBL and FBN with a percentage increase of 0.017, 0.38, 0.3 and 0.29 respectively. This shows that in terms of ROE, PBL is the leading bank. Banks with a declining ROE include -0.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 A v er ag e R et u rn o n E q u it y i n P er ce n ta g es Banks Trend of profitability based on Return On Average Equity (ROE) 2004 2013 University of Ghana http://ugspace.ug.edu.gh 52 BBG, SCB, ADB, UMB, STANBIC, SG-SSB, NIB, BOA, CAL, FABL and HFC with a percentage decrease of 0.64, 0.09, 0.36, 0.28, 0.3, 0.11, 0.07, 0.1, 0.02 and 0.579 respectively. UMB and BOA have suffered in the ROE from 2004 to 2013 which may be due to the management style of the bank. 3.6 Conclusion This chapter provided an overview of the banking industry of Ghana. It discussed the structure of the banking structure, issues in market share analysis for the various banks which include share of operating assets, industry deposits and industry advances and the trend of profitability indicators for the selected banks. University of Ghana http://ugspace.ug.edu.gh 53 CHAPTER FOUR METHODOLOGY 4.1 Introduction The purpose of this chapter is to provide the analytical framework for assessing the relationship between market share and profitability as well as the determinants of market share. The first part explains the econometric model adopted and the rationale behind it. The second part elaborates on two econometric techniques employed for assessing the relationship between market share and profitability as well as the determinants of market share. The chapter also provides definitions for the various variables used in the model as well as the description of the type and sources of data employed by the study including the sampling criteria. The final section provides concluding remarks. 4.2 The Theoretical Framework The discussion in literature review revealed that market share as a predictor of profitability might be a negative, positive or no relationship exist between them with supporting theories. These competing theories suggest that, one would have to employ econometric models to show the relationship between them. The objective of this thesis is to determine the factors that affect market share and use the determinants to establish the relationship between market share and profitability. For this purpose, an explanatory model appears more appropriate, and hence the theoretical model presented by (Bourke, 1989; Greene, 2003; Park 2005 and Dougherty, 2006) is selected. University of Ghana http://ugspace.ug.edu.gh 54 In these models, market share can be expressed as a function of several variables such as liquidity, asset quality, bank age, number of branches etc. This can be express as MS = f (L, AQ, BA, NB, U) where, MS = Market Share, L= Liquidity, AQ = Asset Quality, BA = Bank Age, NB = Number of Branches and U= other control variables. And also the relationship between market share and profitability can be expressed as profitability as a function of market share and other control variables as shown below. P = f (MS, U) where P = Profitability, MS = Market Share and U= other control variables. Three major reasons underline this choice: the model is a panel data which are commonly used because it has the advantage of giving more information as it consists of both the cross sectional information, which captures individual variability, and the time series information, which captures dynamic adjustment. Secondly, its logic in modelling the relationship between market share and profitability within the context of strategic decisions on financial indicators as well as the recognition of the impact of macro-economic variables through the use of linear functional form and lastly, in panel data modelling, several data points are used which improves the degrees of freedom. The collinearity among the explanatory variables is also reduced, thus the efficiency of economic estimates is improved. 4.3 The model Firstly, Market Share (MS) is considered as a function of both external and internal variables and the relationship can be written as 𝑀𝑆 = 𝑓(𝐸𝑉, 𝐼𝑉) ………………………………………………………………………. (1) University of Ghana http://ugspace.ug.edu.gh 55 In equation (1), the external variables include GDP per capita and inflation and the internal determinants include asset quality, bank size, operating efficiency, liquidity, bank age, ownership structure and number of branches. More precisely, the functional relationship between market share and the other variables can be expressed in the form 𝑌 = 𝐵𝑋 + 𝑈 …………………………………………………………………………… (2) Where Y is an nx1 vector of dependent variables, X is a nxk matrix of explanatory variables, B is a kx1 coefficient vector and U is an nx1 error vector. The exact model that is used to investigate the determinants of market share and the relationship between market share and profitability is identical to the one used by Dougherty (2006) and can be specified as follows: 4.3.1 Model specification for the determinants of market shares Industry Operating Asset (IOA) as a measure of Market Share (MS) is determined by Liquidity (L), Asset Quality (AQ), Bank Size (BS), Operating Efficiency (OE), Ownership Structure (OS), Bank Age (BA), Number of Branches (NB), Gross Domestic Product (GDP) and Inflation (INF) as shown in the equation (3) below IOAit = μαi + β1 1Lit + β2 1AQit + β3N 1 BSit + β4 1OEit + β1 1 OS + 𝛽1 1𝐵𝐴𝑖𝑡 + β1 1 NB + β2 2GDPit + β3 2INFit…………………………………………………………………….. (3) University of Ghana http://ugspace.ug.edu.gh 56 Industry Deposit (ID) as a measure of Market Share (MS) is determined by Liquidity (L), Asset Quality (AQ), Bank Size (BS), Operating Efficiency (OE), Ownership Structure (OS), Bank Age (BA), Number of Branches (NB), Gross Domestic Product (GDP) and Inflation (INF) as shown in the equation (4) below IDit = μαi + β1 1Lit + β2 1AQit + β3N 1 BSit + β4 1OEit + β1 1 OS + 𝛽1 1𝐵𝐴𝑖𝑡 + β1 1 NB + β2 2GDPit + β3 2INFit…… ……………………………………………………………… (4) Industry Advances (IA) as a measure of Market Share (MS) is determined by Liquidity (L), Asset Quality (AQ), Bank Size (BS), Operating Efficiency (OE), Ownership Structure (OS), Bank Age (BA), Number of Branches (NB), Gross Domestic Product (GDP) and Inflation (INF) as shown in the equation (5) below IAit = μαi + β1 1Lit + β2 1AQit + β3N 1 BSit + β4 1OEit + β1 1 OS + 𝛽1 1𝐵𝐴𝑖𝑡 + β1 1 NB + β2 2GDPit + β3 2INFit…………………………………………………………………….. (5) The equation on the determinants of market share can further be specified into one single equation where the dependent variable is the composite index of the bank’s market share which provides a useful statistical measure of the overall market or sector performance over time (Kakra and Ameyaw, 2010). The Composite Index (CI) as a measure of Market Share (MS) is determined by Liquidity (L), Asset Quality (AQ), Bank Size (BS), Operating Efficiency (OE), Ownership Structure (OS), Bank Age (BA), Number of Branches (NB), Gross Domestic Product (GDP) and Inflation (INF) as shown in the equation (6) below University of Ghana http://ugspace.ug.edu.gh 57 CI = μαi + β1 1Lit + β2 1AQit + β3N 1 BSit + β4 1OEit + β1 1 OS + 𝛽1 1𝐵𝐴𝑖𝑡 + β1 1 NB + β2 2GDPit + β3 2INFit……………………………………………………………………. (6) Taking the logarithmic transformation of equations give us: NIOAit = μαi + β1 1NLit + β2 1NAQit + β3 1NBSit + β4 1 NOEit + β1 1 OS + 𝛽1 1𝑁𝐵𝐴𝑖𝑡 + β1 1 NNB + β2 2NGDPit + β3 2NINFit………………………………………………….. (7) N IDit = μαi + β1 1𝑁Lit + β2 1NAQit + β3 1NBSit + β4 1NOEit + β1 1 NOS + 𝛽1 1𝑁𝐵𝐴𝑖𝑡 + β1 1 NNB + β2 2NGDPit + β3 2NINFit…………………………………………………… (8) NIAit = μαi + β1 1𝑁Lit + β2 1NAQit + β3 1NBSit + β4 1NOEit + β1 1 NOS + 𝛽1 1𝑁𝐵𝐴𝑖𝑡 + β1 1 NNB + β2 2NGDPit + β3 2NINFit…………..……………………………………….. (9) NCI = μαi + β1 1𝑁Lit + β2 1NAQit + β3 1NBSit + β4 1NOEit + β1 1 NOS + 𝛽1 1𝑁𝐵𝐴𝑖𝑡 + β1 1 NNB + β2 2NGDPit + β3 2NINFit……………………………………………............ (10) 4.3.2 Model Specification for the relationship between market shares variables and profitability The Return on Average Asset (ROA) as a measure of profitability is determined by measures of market share which are the Composite Index (CI), Industry Operating Asset (IOA), Industry Deposit (ID) and Industry Advances (IA) as shown in equation (1) below ROAit = μαi + β1 1CIit + β2 1IOAit + β3 1ID + β4 1IA……………………………………… (1) University of Ghana http://ugspace.ug.edu.gh 58 The Return on Average Equity (ROE) as a measure of Profitability is determined by measures of market share which are the Composite Index (CI), Industry Operating Asset (IOA), Industry Deposit (ID) and Industry Advances (1A) as shown in equation (2) below ROEit = μαi + β1 1CIit + β2 1IOAit + β3 1ID + β4 1IA…………………………………….. (2) In order to check for the significance of the relationship between market share and profitability, the determinants of market share are added to the profitability equation as a control variables as shown below NROAit = μαi + β1 1NCIit + β2 1NIOAit + β3 1NID + β4 1NIA + β1 1NLit + β2 1NAQit + β3 1NBSit + β4 1 NOEit + β1 1 OS + β1 1NBAit + β1 1 NNB + β2 2NGDPit + β3 2NINFit…… (3) NROEit = μαi + β1 1NCIit + β2 1NIOAit + β3 1NID + β4 1NIA + β1 1NLit + β2 1NAQit + β3 1NBSit + β4 1 NOEit + β1 1 OS + β1 1NBAit + β1 1 NNB + β2 2NGDPit + β3 2NINFit…… (4) Taking the logarithmic transformation of equation (3) and (4) gives NROAit = μαi + β1 1NCIit + β2 1NIOAit + β3 1NID + β4 1NIA + β1 1NLit + β2 1NAQit + β3 1NBSit + β4 1NOEit + β1 1 NOS + β1 1NBAit + β1 1 NNB + β2 2NGDPit + β3 2NINFit…… (5) NROEit = μαi + β1 1NCIit + β2 1NIOAit + β3 1NID + β4 1NIA + β1 1NLit + β2 1NAQit + β3 1NBSit + β4 1NOEit + β1 1 NOS + β1 1NBAit + β1 1 NNB + β2 2NGDPit + β3 2NINFit…… (6) University of Ghana http://ugspace.ug.edu.gh 59 Where: CIit : Industry Composite Index for the banks IOAit : Industry Operating Asset with subscript i for banks (i=1…15) and t for years (2004, 2005….2013). IDit : Industry Deposit IAit : Industry Advances ROAit: Average Return on Asset ROEit : Average Return on Equity Lit : Liquidity of the banks (ratio of the bank advances to customer deposit). AQit : Asset Quality of the banks (ratio of bad debt to bank advances) NBSit : Natural logarithmic of Bank Size (Total asset) OEit : Operating Efficiency which is measured with the cost to income ratio OSit : Ownership Structure BAit : Bank Age NBit : Number of Branches GDPit : Gross Domestic Product Growth for the year t INFit : Annual Inflation Rate for the year t N: Natural Logarithmic 𝛽 = 1…n, are the coefficients of the various measurement. University of Ghana http://ugspace.ug.edu.gh 60 The logarithmic transformation is done with two main objectives. Firstly, some of the estimates for the variables are larger values whiles others are in decimals which sometimes makes working with the data quite difficult. So the logarithmic transformation is done to make analysis and data handling simpler. Secondly, the logarithmic transformation means that the coefficients are elasticities and it provides a basis for comparison with other studies on the subject. 4.4 The measurement of the variables The measurement of the variables include market share measures, profitability measures, external and internal determinants of market share. 4.4.1 Market share measures According to the Ghana Banking Survey (2014), market share measured by the operating assets, industry advances and industry deposit. With respect to the work, all the three measures were used to do the analysis. Operating assets are defined here to include resources that are directly deployed to generate interest income. These include cash assets, liquid assets, investments, net loans and advances (Ghana Banking Survey, 2014). From the literature review, operating assets can have a positive, negative or no relationship with profitability. Industry advances are defined as the loans that the bank offers to customers and other banks (Ghana Banking Survey, 2014). From the literature review, industry advances can have a positive, negative or no relationship with profitability. University of Ghana http://ugspace.ug.edu.gh 61 Industry deposits are the total payments made to the bank by customers and other banks. These include current account, savings and time deposit, deposits from other banks (Ghana Banking Survey, 2014). From the literature review, industry deposit can have a positive, negative or no relationship with profitability. The composite index of the banks refers to the average joint impact of the industry advances, operating asset and industry deposit of the banks used which provides a useful statistical measure of the overall market or sector performance over time (Kakra and Ameyaw, 2010). From the literature review, the composite index can have a positive, negative or no relationship with profitability. 4.4.2 Profitability measures Based on the arguments of Rose et al., (2005), this study used the ratio of Return on Average Assets (ROA) and Return on Average Equity (ROE), as measures of bank’s profitability. Return on assets is the profit after tax divided by average total assets and it indicates the returns generated from the assets financed by the bank. Average assets are being used in this study, in order to capture any differences that occurred in assets during the fiscal year. Return on average equity is the ratio of the net profit after tax to the average total equity for the fiscal year. With respect to this work, ROA and ROE are used to measure profitability. University of Ghana http://ugspace.ug.edu.gh 62 4.4.3 Determinants of market share In broad, the determinants of market share are divided into two main categories, specifically, the bank specific factors (internal determinants) and external determinants. The succeeding deliberations give justification for the variables selected. 4.4.3.1 Internal determinants of market share Internal drivers of bank performance or profitability can be defined as factors that are influenced by a bank’s management decisions. Such management effects will definitely affect the operating results of banks. Although a quality management leads to a good bank performance, it is difficult, if not impossible, to assess management quality directly. In fact, it is implicitly assumed that such a quality will be reflected in the operating performance. As such, it is not uncommon to examine a bank’s performance in terms of those financial variables found in financial statements, such as the balance sheet and income statement (Krakah and Ameyaw, 2010). According to this study, the variables which are used as internal factors include: the ratio of bank’s advances to deposits, the ratio of provision for bad debt to advances, the bank’s total assets and expense to income ratio which are proxies to liquidity, asset quality, bank size and operating efficiency respectively as well as the year of establishment (bank age), number of branches and ownership structure of the banks. The choice of liquidity One important decision that the managers of banks must pay attention to is the liquidity management and specifically to the measurement of their needs related to the process of University of Ghana http://ugspace.ug.edu.gh 63 deposits and loans. For that reason the ratio of bank’s advances to deposits is used as a measure of liquidity. From the literature review, Molyneux et al., (1992) and Guru et al. (1999) discovered that negative correlation exists between the level of liquidity and market shares, which is highly correlated with profitability. However, Bourke (1989), and Kosmidou et al., (2005) found a significant positive relationship between liquidity and market share. Thus the relationship between liquidity and market share is uncertain. Asset quality The ratio of provision for bad debt to advances indicates how much of the total portfolio has been provided for but not charged off and is used as a measure of bank’s asset quality. This variable is incorporated into the regression model as a proxy for asset quality. Poor asset quality and successively credit risk can have rippling effect and thus lead to liquidation. From the literature review, the higher the ratio, the poorer the quality and therefore the higher the risk of the loan portfolio will be (Bessis, 2002). The risk-return hypothesis implies a positive relationship between risk and profits. According to Kosmidou (2008) bad asset quality may have an adverse effect on bank productivity by reducing interest on revenue earned by increasing the provision costs which as well decrease their market shares. For these reasons the sign of the quality of asset may be negative or positive. Bank size In most studies of bank performance determinants, the total asset is used a measure for bank size. Bank size is usually used to account for potential economies or diseconomies of scale in the banking sector. Economies of scale reduces the cost of gathering and processing information (Boyd et al., 1993) so that a positive effect of bank size is associated with a University of Ghana http://ugspace.ug.edu.gh 64 performance which also affect market shares. Bikker et al., (2002) and Goddard et al., (2004) have all linked bank size to capital ratios, which they found it to be positively related to bank size. So the expected sign is positive. Operating Efficiency This is measured by the ratio of expense to income of the bank. The major element of the bank expenses are the salary to the employees and administrative cost. It is used to measure the impact of efficiency on banks performance. A negative relationship is expected to occur between banks’ performance and the ratio since a higher cost to income ratio means there is a high cost and low income which means low profit indicating poor performance and vice versa. However, this may not be the case as higher amounts of operating cost could also reflect higher volume of banking activities. (Kosmidou et al, 2005). Ownership Structure The structure of the ownership of banks is subject to various regulations aiming to protect the rights of depositors, shareholders and to make banks operate more efficiently. Since many parties participate in bank activities (Setiyono and Tarizi, 2014). For this reason, ownership structure is used as a dummy variable which include local and foreign banks. Local banks are banks with majority shareholders from the indigenous and foreign banks with majority shareholders from overseas. From the literature, Cao et al., (2010); Setiyono and Tarizi, (2014) discovered that, foreign banks perform better that than local banks while the reverse is supported by Zouari and Taktak, (2014). University of Ghana http://ugspace.ug.edu.gh 65 The Age of the Bank The age of the bank is the number of years that the bank has been in existence which is equivalent to the year of establishment (Arjan, 2012). This is incorporated into the regression model to assess if bank age has any influence on its market performance. From the literature, Göbel and Zwick (2009); Ruis and Scholman discovered a negative relationship between age of the bank and its market performance. But Uschi and Veen, (2009) found a positive relationship between age and performance. Number of Branches The number of branches variable is incorporated in the model to ascertain whether there is a link between the banks with the largest number of branches and their market performance or not. From the literature, Samad et al., (2006) found no significant relationship between them but Shukla and Sinha (2013) establish a positive relationship between them. 4.4.3.2 External determinants These are factors that are beyond the control of a bank’s management. They represent events outside the influence of the bank. However, the management can anticipate changes in the external environment and try to position the institution to take advantage of anticipated developments. The two major components of the external determinants are macroeconomic factors and financial structure factors (Krakah and Ameyaw, 2010). The environments in which banks operate can influence their performance and can impact on their strategic positioning. These include the Gross Domestic Product (GDP) and inflation. University of Ghana http://ugspace.ug.edu.gh 66 Gross Domestic Product (GDP) growth The GDP is the quantity of entire economic activity within the economy and it is normally used as an economic indicator. In this study GDP growth is used. The GDP growth is the annual change in the GDP. According to Bikker et al. (2002) there is a positive association between economic growth and financial sector productivity, which also affect market shares positively. Again Kanwal and Nadeem (2013) also found a negative relationship between GDP growth and industry performance. Therefore there is an anticipation that there is both a positive and negative correlation between GDP growth and market shares. Inflation (INF) The variable inflation is used as a proxy for percentage change in aggregate price levels. Staikouras et al., (2003) point out that inflation may have direct effects and indirect effects on the profitability of the banks. The effect of inflation on bank profitability depends on whether inflation is anticipated or unanticipated (Perry, 1992). If inflation is anticipated the manager can increase the rates on loan faster than the rate at which operating cost is increasing so that inflation favourably impacts on profitability as well as their market shares. If inflation is unanticipated, it affects the performance of the banks negatively, which will also affect market shares adversely. The variables that are used to achieve the objectives are summarized in the table below in table 4.1 with their measurement and expected sign according to the various literature that has been reviewed. University of Ghana http://ugspace.ug.edu.gh 67 Table 4:1 summary of description of variables Variable Description Measurement Prior expectation Measures of profitability ROA Return on average assets Ratio of net profit to average total assets of banks N/A ROE Return on average equity Ratio of net profit to average total equity N/A Measures of market share IOA Industry Operating Asset Total Assets of the banks +/- ID Industry Deposits Total Deposits of the banks +/- IA Industry Advances Total Advances of the banks +/- CI Composite Index Average of IOA, ID and IA +/- Internal determinants of market share L Liquidity Ratio of Banks advances to customer deposits +/- AQ Asset quality Ratio of bad debt to bank advances +/- BS Bank size Total asset as a proxy + OE Operating Efficiency Cost to income ratio - OS Ownership structure Dummy variable (local or foreign) -/+ BA Bank Age Year of Establishment -/+ NB Number of Branches + External determinants of market share GDP GDP Growth Total goods and services in the economy +/- INF Inflation Annual inflation rate (percent) +/- Source: Author’s own. 4.5 Diagnostic test To ensure efficiency and precise prediction of the model to be used, certain diagnostic tests should be done to ensure that the data conforms to stated assumptions. Tests for heteroscedasticity, multicollinearity and autocorrelation should have been done but left out because the estimation technique to be used caters for such problems. The unit root test is the only diagnostic test done for this study. This is because the GMM takes first difference University of Ghana http://ugspace.ug.edu.gh 68 to eliminate unobserved bank-specific effects and use lagged instruments to correct for simultaneity in the first-difference equations. (Blundell and Bond, 1998). 4.5.1 Stationarity/Unit root test A unit root test tests whether a time series variable is non-stationary using an autoregressive model. In running regression for panel data, all the variables are expected to be stationary. Thus, all the variables used in the model should not have unit root. “A stochastic process is said to be stationary if its mean and variance are constant over time and the value of the covariance between two time periods and not at the actual time at which the covariance is computed” (Gujarati, 1995 p713). If the time series involved are unit root processes, a naive application of regression analysis may yield nonsense results (Breitung, 2000). The Fisher unit root test is employed in this study to test the existence of unit roots in the data. This test is chosen above the other tests for unit root because it can be carried out for any unit root test and the test does not require similarity adjustment factors that are specific to the sample size and specification. The Fisher test is also useful when there is a combination of both non-stationary and stationary variables and the test suffers the least when time is large and N sample is small. Also in the Fisher test, there is no restriction of the sample sizes for different samples, thus they can vary according to data availability. The Fisher test is chosen due these benefits it has over other unit root tests. University of Ghana http://ugspace.ug.edu.gh 69 4.6 Estimation technique In ensuring robustness of the results across different estimation techniques, two different estimation methods are used for estimation. The first method to be used is the robust random effect because of its wide usage especially with panel data and they control for time invariant omitted variables. However, according to Kinney and Dunson (2007), a less discussed issue is that random effect models as usually implemented might be insufficiently flexible. Secondly, the GMM estimation technique by Hansen (1982) is being used to estimate the parameters. It is believed that the GMM is superior to all the other techniques in the sense that it yields consistent and efficient estimates even when the data generating process exhibit the following characteristics.  When there exist specific fixed effects which are randomly distributed  When there is a specific serial correlation and heteroscedasticity in the stochastic error term  When there is no correlation between the stochastic error terms across banks  When the time is small and number of banks is large. Thus the GMM is preferred over the other techniques for two main reasons.  It eliminates the problem of information loss in cross-sectional regression as it allows for multiple observations for each bank  It gives consistent estimators even when the time period is small and the number of banks is large. University of Ghana http://ugspace.ug.edu.gh 70 4.7 Data This study employs secondary data which are mainly annual accounting data, of individual banks and macroeconomic data drawn from the period 2004-2013. The datasets are subsequently stacked as panels for analysis and are basically obtained from the Ghana Banking Survey (GBS) by PricewaterhouseCoopers in collaboration with the Ghana Association of Bankers, Annual Report and Accounts of individual banks and World Development Indicators (WDI). 4.8 Sampling Criteria All universal banks institutions existing in the banking industry from 2004 to 2013 are sampled. The sampling criteria are based on the adoption of the universal banks in 2004 (Bank of Ghana, 2004) in Ghana and also the availability of data and as such yielded a balanced dataset of fifteen (15) banks. This ensures that all entities are represented in the sample. 4.9 Conclusion This chapter discussed the theoretical basis of the study and subsequently developed a model which will be later estimated. The methodology used in estimating the results and the diagnostic test used are all clearly explained. Two econometric techniques; the random effect and the Generalised Method of Moment (GMM) are identified in achieving the determinants of market share and the relationship between market share and profitability. All the variables used in the study are well explained as well as their prior expectations. The sample size and the sample period are also outlined in this chapter. University of Ghana http://ugspace.ug.edu.gh 71 CHAPTER FIVE DISCUSSION OF RESULTS 5.1 Introduction  This chapter aims at using two econometric techniques, two econometric techniques – the random effect and the Generalised Method of Moment (GMM) to empirically establish the relationship between market share and profitability as well as to know the determinants of market share in Ghana. These methods are used in ensuring robustness of the results. It eliminates the problem of information loss in cross-sectional regression as it allows for multiple observations for each bank. It gives consistent estimators even when the time period is small and the number of banks is large and also eliminate endogeneity problems. This chapter begins with presentation and discussion of the results from the descriptive statistics of the selected variables, the correlation matrix and finally the empirical model. The chapter closes with a summary of the results presented. All analysis were carried out with STATA 13. 5.2 Descriptive analysis The statistics present the summary of descriptive statistics of the variables captured in the regression model. These statistics are generated to give overall description of the data used in the model and enable the researcher screen the data for any suspicious figure. The key descriptive measures are the mean, standard deviation, the minimum and the maximum values of the variables over the period under consideration. Key highlights as discuss in the ensuing discussion. University of Ghana http://ugspace.ug.edu.gh 72 Table 5.1 below shows the descriptive statistics of the variables. Variable Mean Standard deviation Minimum Maximum Liquidity 1331903 1.63e+07 .149 1.98e+08 Asset quality .8725818 .058962 .641 1.053 Bank size 7.13e+08 8.15e+08 334910.8 4.62e+09 Operating efficiency .605 .305 .158 3.638 Gross Domestic Product .076 .028 .040 .150 Inflation .286 .306 .107 .916 Bank age 35.633 29.856 5.000 117 Number of branches 34.302 37.761 2.000 161 Composite index .060 .049 0.000 .2961 Industry operating asset .059 .045 .009 .264 Industry deposit .062 .073 .006 .744 Industry advances .066 .063 .006 .548 Return on average asset 0.389 0.086 -0.048 0.883 Return on average equity 0.244 0.297 -0.458 1.200 Source: Author’s computation from Annual Reports and Accounts of individual banks from 2004 to 2013. The mean indicates the average values of the variables used in the study (Pindyck et al, 1991). From table 5.1, it shows that on average liquidity (as measured by the ratio of bank advances to deposit) is GH ¢1331903 in Ghana from 2004 to 2013. It also shows that the average GDP growth is 0.076. Asset quality on average is 0.873, operating efficiency on average is .6053445, inflation on average is 0.2863, bank size on average is 7.13e+08, Age of bank on average is 35.633, Number of branches of banks on average is 34.302, composite index of the banks on average is 0.060, return on average asset is 0.389 and return on average equity is 0.244. University of Ghana http://ugspace.ug.edu.gh 73 The standard deviation is a measure of how the variables are spread out around their various means (Pindyck et al, 1991). The difference between the minimum and the maximum values of the variables give the range of the variables used in the study (Pindyck et al, 1991). The range is an indicator of the level of fluctuations in the variables. The larger the range values the higher the level of fluctuations in a variable. From table 5.1 it is apparent that bank size had the largest standard deviation which reveals that the size of banks had more significant variance than other variables over the period. However, the variable GDP growth had the least standard deviation which suggests that economic growth has been stable when compared to the other variations over the period. The average industry operating asset stands at .0587 with 0.009 and 0.264 as minimum and maximum values respectively. Average industry deposit is 0622384 with the minimum value of .006 and the maximum value of .744. Average industry advances is 0.066 with minimum value and maximum value 0.548. The average industry profit stands at 0.389 with -0.0477 and .883 as minimum and maximum values for Return On Average Assets (ROA) whiles Return On Average Equity Average (ROE) is 0.244 with -0.458 and 1.200 as the minimum and maximum values respectively. 5.3 Correlation matrix The table 5.2 presents the correlation matrix for all the variables incorporated into the model. The coefficient of correlation provides an index of the direction and the magnitude of the relationship between two sets of scores without implying causality (Pindyck et al, 1991). The sign of the coefficient is an indication of the direction of the relationship. The absolute value of the coefficient indicates the magnitude. University of Ghana http://ugspace.ug.edu.gh 74 Correlation matrix is useful to the extent that it reveals whether there are elements of Multicollinearity in the data (Pindyck et al, 1991). Multicollinearity is the situation when some or all of the explanatory variables are highly related making it difficult to tell which of them is influencing the dependent variable (Pindyck et al, 1991). The severity of Multicollinearity would be manifested in a situation where all p-values of regression coefficients are insignificant, but overall model having significant F statistic. Table 5.2.1: Correlation matrix of determinants of market share L AQ BS OE GDP INF BA NA OS _CONS L 1 AQ -0.094 1 BS 0.010 -0.001 1 OE -0.017 -0.007 0.123 1 GDP 0.034 -0.00 -0.094 -0.07 1 INF BA NA OS -0.024 0.018 0.050 0.09 0.166 0.006 -0.066 0.008 -0.141 -0.167 -0.469 -0.169 0.012 0.163 -0.131 -0.014 -0.69 -0.001 0.027 0.025 1 0.049 -0.002 0.008 1 -0.328 -0.123 1 0.24 1 CONS 0.068 -0.965 0.012 -0.113 -0.177 -0.066 -0.070 0.040 -.042 1 Source: Author’s computation from Annual Reports and Accounts of individual banks from 2004 to 2013. Table 5.2.2: Correlation matrix of determinants of profitability IOA ID IA CI _CONS IOA 1 ID 0.6041 1 IA 0.620 0.907 1 CI -0.732 -0.9593 -0.952 1 _CONS -0.162 0.164 0.093 -0.184 1 University of Ghana http://ugspace.ug.edu.gh 75 Source: Author’s own computation from Annual Reports and Accounts of individual banks from 2004 to 2013. 5.4 Diagnostic results Diagnostic tests are conducted to ensure the robustness of the developed model. They are done to determine whether the estimates are consistent, reliable and unbiased among others. 5.4.1 Unit root test A unit root test is undertaken on all the variables to examine their Stationarity conditions. The results of the unit root tests on all the variables are presented in the table 5.3 below. Using the Choi (2001) criterion, the null hypothesis is rejected at 1 percent in favor of the alternative hypothesis for all the variables on the basis of the inverse χ2 test. Choi (2001) argues that when the number of panels is finite, the inverse χ2 test is applicable and powerful. This statistic has a χ2 distribution with 2N degrees of freedom, and large values are cause to reject the null hypothesis. As a rule of thumb, when the probability of the inverse χ2 of a variable is less than the significance level, we reject the null hypothesis and conclude that the variable is stationary. It can therefore be concluded that for all the variables in the study at least one panel is stationary at 1 per cent hence spurious regressions are unlikely to occur in the study. The inverse normal, the inverse logit and the modified inverse χ2 tests also corroborate the inverse χ2 test, but the inverse normal is stationary at 10 percent. University of Ghana http://ugspace.ug.edu.gh 76 Table 5.3 showing unit root test results Variable Inv. chi-squared Inv. Normal Inv. Logit M. Inv. Chi squared Stats P-value Stats P-value Stats P- value Stats p-value IOA 94.302 0.000 -1.385 0.083 -4.502 0.000 8.301 0.000 ID 181.050 0.000 -4.242 0.000 -10.443 0.000 19.501 0.000 IA 64.110 0.000 -3.249 0.001 -3.379 0.001 4.404 0.000 ROAA 79.917 0.000 -2.541 0.006 -3.602 0.000 6.444 0.000 ROAE 114.044 0.000 -5.484 0.000 -7.262 0.000 10.850 0.000 L 68.381 0.000 -2.790 0.003 -3.458 0.000 4.956 0.000 AQ 184.707 0.000 -3.071 0.001 -9.383 0.000 19.973 0.000 BS 1006.338 0.000 -29.362 0.000 -72.017 0.000 126.045 0.000 OE 126.557 0.000 -6.293 0.000 -8.482 0.000 12.466 0.000 GDP 53.766 0.005 -3.748 0.000 -3.460 0.000 3.068 0.001 INF 58.207 0.002 -4.121 0.000 -3.837 0.000 3.642 0.000 CI 99.642 0.000 -5.822 0.000 -6.625 0.000 8.991 0.000 BA 1.000 0.000 - - - - -3.873 0.000 NA 52.358 0.007 -2.604 0.005 -2.7946 0.003 2.886 0.002 OS 1.000 0.000 - - - - -3.873 0.010 𝐇𝟎: all panel contain unit root 𝐇𝐚: at least one panel is stationary Source: Author’s computation from Annual Reports and Accounts of individual banks, from 2004 to 2013. 5.5 Regression results The regression results of the determinants of market share are presented in the tables below using the random effect and the Generalised Method of Moments (GMM) estimation technique. Random effect results are presented together with the GMM results. The coefficients of the variables are shown below and all the variables are in natural logarithm. 5.5.1 Industry Operating Asset (IOA) regression results as a proxy for market share: The table 5.4 below summarises the empirical results for the model industry operating asset as a proxy for the market share. It indicates the share of industry operating assets in the University of Ghana http://ugspace.ug.edu.gh 77 market among the fifteen banks in Ghana. The industry operating asset model is a multiple regression equation comprising of nine explanatory variables out of which seven are bank- specific and two external factors. Table 5.4: Estimation results for the determination of market share (Industry Operating Asset). Variables Random Effect GMM Liquidity (L) 0.004 (.017) -0.037 (.101) Asset Quality (NAQ) 0 .005 (.533) -1.348*** (.394) Bank size (BS) 0.062*** (.021) 0.332 *** (.056) Operating Efficiency (OE) -0.242*** (.0903) -0.189* (.108) Bank Age (BA) 0.011*** (.003) 0.002 (.002) Number of Branches (NB) 0.008*** (.002) 0.001* (.002) Ownership Structure (OS) -.025 (.142) 0.209 (.181) Gross Domestic Product(GDP) -0.361** (1.434) -0.001*** (.0612) Inflation (INF) 0.043*** (.076) 0.030** (.021) _CONS 3.265 (.040) 1.575 (.944) DIAGNOSTICS R Squared 0.6411 F Statistics Wald chi2 90.76 108 Obs 139 235.72 ***p<0.01, **p<0.05,*p<0.10 Figures in parenthesis are robust standard errors Source: Author’s computation from annual reports and accounts of individual banks from 2004 to 2013. University of Ghana http://ugspace.ug.edu.gh 78 The internal or bank specific factors include the bank size, liquidity, operating efficiency, bank age, ownership structure, asset quality and Number of branches of the banks and the external include GDP and Inflation. From the regression results, the total asset is used as a proxy for bank size in the regression model according to the studies of Boyd et al., (1993). The result indicates, that bank size determines market share positively with coefficient of 0.062 and is significant at 1%. This implies that, a 1% increase in banks size leads to 0.0623% increase in industry operating asset which means that, industry operating asset is not highly responsive to changes in bank size. This further implies that, a 1% increase in bank size leads to 0.332% increase in industry operating asset from the Generalised Method of Moment (GMM) results which is also significant at 1%. This signifies that, in the Ghanaian banking sector, bank size leads to economies of scale, thereby making larger banks have the highest share in the market in terms of the operational asset. Economies of scale will reduce the cost of gathering and processing information. The larger the bank size, the more the bank performs well. Also, most of the banks are associated with diversification which may impact favourably on risk and product portfolio. This result is consistent with the findings of (Molyneux et al., 1992; Bikker et al., 2002 and Goddard et al., 2004). Operating efficiency has a negative relationship with industry operating asset with coefficient of - 0.242 which is significant at 1%. This means that, a 1% increase in operating efficiency leads to a decrease in industry operating asset by 0.242% across time. This shows that, industry operating asset is responsive to changes in operating efficiency. University of Ghana http://ugspace.ug.edu.gh 79 This further shows that, a 1% increase in operating efficiency leads to a 0.189% decrease in industry operating asset from the GMM result which is significant at 10%. A chunk of bank’s expense is composed of salary expense and administrative cost. Thus, most of most of the banks in the economy spend more on administrative expenses which are not being translated proportionately into profitability. This result is consistent with the studies by (Antonio, 2010 and Ayanda eta al., 2013). Bank age shows a positive relationship with industry operating asset with coefficient of 0.011 which is significant at 1%. Thus, a 1% increase in bank age leads to 0.011% increase in industry operating asset. This shows that, the old banks have skill advantage, experience and larger networks as compared to the new companies which makes them have increased in operating asset as compared to the new banks. This result is consistent with studies by Uschi and Veen (2009) and in contrast with studies by (Göbel and Zwick, 2009; Ruis and Scholman, 2012) which states that, most of the older banks are less innovative as compared to the younger banks and the younger banks perform better than older banks. In relationship to the banks in the country, the old banks perform better than the new banks in terms of the industry operating asset (Ghana Banking Survey, 2014). Number of branches has a positive relationship with industry operating asset with coefficient of 0.008 which is significant at 1%. This implies that, a 1% increase in number of branches leads to 0.008% in industry operating asset. This also shows that, a 1% increase in the number of branches lead to increase industry operating asset by 0.01% from the University of Ghana http://ugspace.ug.edu.gh 80 GMM result significant at 10%. The result implies that, the number of branches show the number of workers and assets in the sector and these workers are being trained to acquire knowledge and skills to affect productivity and is consistent with studies by (Shukla and Sinha, 2013; Hameed et al., 2014). This is true since most banks in the economy that have more branches have more assets as compared to the banks with smaller branches and also banks that build their branches at strategic locations also perform better. Asset quality has a negative relationship with industry operating asset with coefficient of - 1.348 when GMM is used at 1% significant level. This implies that, a percentage increase in asset quality leads to a decrease in industry operating asset by 1.348% which shows that, a change in asset quality is highly responsive to industry operating asset. The study suggests that, the higher the provision for bad debt to advances ratio, the higher the credit risk and the higher the accumulation of unpaid loan and interest. This study is consistent with studies made by (Acaravci and Calim, 2013 and Heffernan, 1996). The Gross Domestic Product (GDP) variable is incorporated into the regression model to analyse the impact of economic activity on bank market share. The result indicates, that GDP has a negatively significant influence on bank market performance with coefficient of -0.361. This shows that, a 1% increase in GDP leads to a decrease in industry operating asset by 0.361 % at 5% significant level. It further shows that, a 1% percent increase in GDP leads to 0.001% decrease in industry operating asset when the GMM is used and significant at 1%. This study supports the findings of Kanwal and Nadeem (2013) and University of Ghana http://ugspace.ug.edu.gh 81 Sufian (2011) who conclude that, there is a negative relationship between GDP and performance of the banks. Whereas, insignificant negative relationship of GDP is in contradiction to the theory which asserts that economic growth enhances profits and downturn adversely affects the interest income (Alper and Anbar, 2011). This opposite result may be due to other reasons which include the customer’s preference or choice of depositing excess funds, taking loans and informational asymmetry of customer and lack of information regarding economic changes in the country. This is true since when the GDP of the economy increases, the operating asset of the banks declines for the banks in Ghana. The result indicates that, inflation variable which is captured in the model has a positive influence on industry operating asset with coefficient of 0.043 at 1% significant level which shows that, a percentage increase in inflation leads to an increase in industry operating asset 0.043%. It again shows that, a percentage increase in inflation leads to an increase in industry operating asset by 0.030% from the GMM result at 5% significant level. This signals that, bank management in the economy are able to forecast accurately inflation and are proactive in managing anticipated inflation. By making accurate forecast of inflation, the manager can increase the rates on loan faster than the rate at which operating cost is increasing so that inflation favourably impacts on market share. This is true since as inflation increases, the industry assets of the banks also rises. This result is consistent with most studies (Bourke, 1989; Molyneux et al., 1992; Athanasoglou et al. 2005). University of Ghana http://ugspace.ug.edu.gh 82 5.5.2 Industry Deposits (ID) regression results as a proxy for market share The table 5.5 summarises the empirical results for the model industry deposit as a proxy for market. It indicates the share of industry deposit in the market among the fifteen banks in Ghana. The model is a multiple regression equation comprising of nine explanatory variables out of which seven are bank-specific and two external factors. Table 5.5: Estimation results for the determination of Industry Deposit (ID) Variable Random effect GMM Liquidity (L) 0.003 (.019) -0.197 * (.105) Asset Quality (AQ) 0.928 (.636) -0.329 (.437) Bank Size (BS) 0.057** (.025) 0.288*** (.059) Operating Efficiency (OE) -0.157* (.106) -0.188* (.114) Bank Age (BA) 0.001 *** (.004) 0.009* (.002) Number of Branches (NB) 0.012*** (.002) 0.002 (.002) Ownership Structure (OS) 0.145 (.207) -0.104 (.117) Gross Domestic Product (GDP) -0.319 *** (.090) -0.355*** (.065) Inflation (INF) 0.062 (.046) 0.051** (.051) _CONS 2.419 (1.689) 0.496 (1.062) DIAGONOSTICS R Squared 0.6204 F statistics Wald chi2 54.64 85.12 Number of Observation 139 108 ***p<0.01, **p<0.05,*p<0.10 Figures in parenthesis are robust standard errors Source: Author’s computation from annual reports and accounts of individual banks, from 2004 to 2013. University of Ghana http://ugspace.ug.edu.gh 83 From the regression results, bank size determines industry deposit positively with coefficient of 0.057 at 5% significant level. This implies that, a 1% increase in bank size leads to 0.057% increase in industry deposit which indicate that, industry deposit is not highly responsive to changes in bank size. This further shows that, a 1% increase in bank size leads to 0.288% increase in industry deposit from the GMM result at 1% significant level. This suggests that, in the Ghanaian backing sector, bank size leads to economies of scale, thereby making larger banks have the highest share in the market in terms of the deposits. Economies of scale will reduce the cost of gathering and processing information. The larger the bank size, the more the bank performs well. It could also mean that, bank size is associated with diversification which may impact favourably on risk and product portfolio. This result is consistent with the findings of (Molyneux et al., 1992; Bikker et al., 2002 and Goddard et al., 2004). Operating efficiency also shows a negative relationship with industry deposit with coefficients of - 0.157 at 10% significant level. This means that, a 1% increase in operating efficiency leads to a decrease in industry deposit by 0.157% across time. This shows that, industry deposit is responsive to changes in operating efficiency. This further shows that, a 1% increase in operating efficiency leads to a 0.188% decrease in industry deposit at 10% significant level. A chunk of bank’s expense is composed of salary expense and administrative cost. It is possible that, in the banking industry, high bank salaries and administrative expenses are not being translated proportionately into profitability. This result is consistent with the studies by (Antonio, 2010; Ayanda eta al., 2013). University of Ghana http://ugspace.ug.edu.gh 84 The bank age shows a positive relationship with industry deposit with coefficient of 0.001 at 1% significant level. Thus a 1% increase in bank age leads to 0.001% increase in industry deposit. Also, from the GMM results, a 1% increase in bank age leads to 0.009% increase in industry deposit. This shows that, the old banks have skill advantage, experience and larger networks than the new companies which makes them have increase in deposit as compared to the new banks. This result is consistent with studies by Uschi and Veen (2009) and in contrast with studies by (Göbel and Zwick, 2009; Ruis and Scholman, 2012) which states that the most of the older banks are less innovative as compared to the younger banks and the younger banks perform better than older banks. This is true in the banking system within the economy since the Ghana Banking Survey (2014) shows that, as the older banks perform better than new banks which may due to customer loyalty. The number of branches has a positive relationship with industry deposit with coefficient of 0.012 at 1% significant level. This implies that, a 1% increase in number of branches leads to 0.012% in industry deposit. The result implies that, the number of branches show the number of workers and assets in the sector and these workers are being trained to acquire knowledge and skills to affect productivity and to increase deposit by having good quality of customer service which motivate customers to deposit more and is consistent with studies by (Shukla and Sinha, 2013; Hameed et al., 2014). Liquidity also shows a negative relationship with industry deposit with coefficient of - 0.197 when GMM is used at 10% significant level. This implies that, a percentage increase in liquidity lead to a decrease in industry deposit by 0.197% which shows that, a change in University of Ghana http://ugspace.ug.edu.gh 85 liquidity is highly responsive to industry deposit. This conforms to the assertion that, holding assets in a highly liquid form tends to reduce income as liquid asset are associated with lower rates of return. For instance, cash which is the most liquid of all assets is a non- earning asset. It would therefore be expected that, higher liquidity would negatively correlates with banks performance. The results is consistent with studies by (Molyneux and Thornton, 1992; Guru et al., 1999 and Tabar et al., 2013). The Gross Domestic Product (GDP) variable is incorporated into the regression model to analyse the impact of economic activity on bank market share. The result indicates that, GDP has a negative significant influence on bank market performance with coefficient of -0.319. This shows that, a 1% increase in GDP leads to a decrease in industry deposit by 0.319 % at 1% significant level. It further shows that, a 1% percent increase in GDP leads to 0.35% decrease in industry deposit when the GMM is used at 1% significant level. This supports the findings of Kanwal and Nadeem (2013) and Sufian (2011) who conclude that there is a negative relationship between GDP and performance of the banks. Whereas, the insignificant negative relationship of GDP is in contradiction to the theory which asserts that economic growth enhances profits and downturn adversely affects the interest income (Alper and Anbar, 2011). This opposite result may be due to other reasons which include the customer’s preference or choice of depositing excess funds, taking loans and informational asymmetry of customer and lack of information regarding economic changes in the country. This is true since when the GDP of the economy increases the industry deposit of the banks declines in Ghana. University of Ghana http://ugspace.ug.edu.gh 86 Inflation has a positive influence on industry deposit with coefficient of 0.051 at 5% significant level which shows that, a percentage increase in Inflation leads to an increase in industry deposit 0.051% when the GMM is used. This signals that, bank managers are able to forecast accurately inflation and are proactive in managing anticipated inflation. By making accurate forecast of inflation, the manager can increase the rates on loan faster than the rate at which operating cost is increasing so that inflation favourably impacts on market share. This result is consistent with most studies (Bourke, 1989; Molyneux et al., 1992; Athanasoglou et al. 2005). This is true since during the period where inflation was high for the country, the industry deposit was also higher for the banks. 5.5.3 Industry Advances (IA) dependent variable regression results as a proxy for market share. The table 5.6 summarises the empirical results for the industry advances model as a proxy for market Share. It indicates the share of industry advances to its customers and other banks. The industry advances model is a multiple regression equation comprising of nine explanatory variables out of which seven are bank-specific and two external factors. From the regression results, bank size determines market share positively with coefficient of 0.036 at 5% significant level. This implies that, a 1% increase in bank size leads to 0.036% increase in industry advances which means that, industry advances is not highly responsive to changes in bank size. This further implies that, a 1% increase in bank size will lead to 1.007% increase in industry advances when the GMM is used at 5% significant level. This shows that, bank size leads to economies of scale in the Ghanaian banking sector, thereby making larger banks have the highest share in the market in terms of the University of Ghana http://ugspace.ug.edu.gh 87 advances. Economies of scale will reduce the cost of gathering and processing information. The larger the bank size, the more the bank performs well. Also, most of the banks are associated with diversification which may impact favourably on risk and product portfolio. This results is consistent with the findings of (Molyneux et al., 1992; Bikker et al., 2002 and Goddard et al., 2004). Table 5.6 Estimation results for the determination of market share (Industry Advances) Variable Random effect GMM Liquidity (L) -0.015 (.0259) 0.343 (.282) Asset Quality (AQ) 1.73** (.735) -2.039 (1.961) Bank Size (BS) 0.036** (.033) 1.007** (.495) Operating Efficiency (OE) -0.281** (1.387) -2.960** (1.305) Bank Age (BA) 0.007*** (.002) 0.061** (.0235) Number of Branches (NB) 0.006*** (.002) 0.009 * (.004) Ownership Structure (OS) 0.013 (.117) -7.524** (3.736) Gross Domestic Product (GDP) -0.257** (.117) 0.741 * (.442) Inflation (INF) 0.078 (.064) 0.092** (.042) _cons 1.936 (2.216) 0.038 (2.950) DIAGONOSTICS R Squared 0.4556 F Statistics Wald chi2 73.51 58.34 Number of Observation 139 108 ***p<0.01, **p<0.05,*p<0.10 Figures in parenthesis are robust standard errors Source: Author’s computation from annual reports and accounts of individual banks from 2004 to 2013. University of Ghana http://ugspace.ug.edu.gh 88 Operating efficiency has a negative relationship with industry advances with coefficient of - 0.281 at 5% significant level. This means that, a 1% increase in operating efficiency leads to a decrease in industry advances by 0.281% across time. This shows that, industry advances are less responsive to changes in operating efficiency. It further shows that, a percentage increase in operating efficiency leads to a decrease in industry advances by 2.960% when GMM is used at 5% significant level. A chunk of bank’s expense is composed of salary expense and administrative cost. This demonstrates that, high bank salaries and administrative expenses are not being translated proportionately into profitability in the country. This result is consistent with the studies by (Antonio, 2010; Ayanda eta al., 2013). Bank age has a positive relationship with industry advances with coefficient of 0.007 at 1% significant level. Thus, a 1% increase in bank age leads to 0.007% increase in industry advances. From the GMM result, a 1% increase in bank age leads to an increase of 0.061% at 5% significant level. This shows that, the old banks have skills advantage, experience and larger networks to the new companies which makes them have increase in operating asset as compared to the new banks. This result is consistent with studies by Uschi and Veen (2009) and in contrast with studies by (Göbel and Zwick, 2009; Ruis and Scholman, 2012) which states that, most of the older banks are less innovative as compared to the younger banks and the younger banks perform better than older banks. This is true for banks in the country where older banks perform better than new banks according to the survey presented by the Ghana Banking Survey (GBS). University of Ghana http://ugspace.ug.edu.gh 89 Number of branches variable indicates a positive relationship with industry advances of coefficient of 0.006 at 1% significant level. This implies that, a 1% increase in number of branches leads to 0.006% in industry advances. This also shows that, a 1% increase in the number of branches leads to increase industry advances by 0.009% when GMM is used at 10% significant level. The result implies that, the number of branches show the number of workers and assets in the sector and these workers are being trained to acquire knowledge and skills to affect productivity and is consistent with studies by (Shukla and Sinha, 2013; Hameed et al., 2014). Also, most of branches are built on strategic locations which enable banks to give out loans to customers and other banks in the economy. Asset quality has a positive relationship with industry advances with coefficient of 1.73 at 5% significant level. This implies that, a percentage increase in asset quality leads to an increase in industry advances by 1.73% which shows that, a change in asset quality is highly responsive to industry advances. The study suggests that, higher returns on assets are associated with larger bank size, activity diversification, and private ownership which affect the performance of banks and their market shares and is supported by Flamini et al., (2009). The owner Structure that is incorporated into the model as a dummy shows that, holding all other factors constant, local banks have a percentage increase of 7.524 more than foreign banks. The result indicates that foreign banks with worse market performance borrow more and local banks get financial support from the government. This is consistent with a study by (Zouari and Taktak, 2012). University of Ghana http://ugspace.ug.edu.gh 90 The Gross Domestic Product (GDP) variable is incorporated into the regression model to analyse the impact of economic activity on bank market share. The result indicates that, GDP has a negatively significant influence on bank market performance with coefficient of -0.257 at 5% significant level. This shows that, a 1% increase in GDP leads to a decrease in industry advances by 0.257 %. It further shows that, a 1% percent increase in GDP leads to 0.749% increase in industry advances when GMM is used at 10% significant level. This supports the findings of Kanwal and Nadeem (2013) and Sufian (2011) who conclude that, there is a negative relationship between GDP and performance of the banks. Whereas, insignificant negative relationship of GDP is in contradiction to the theory which asserts that economic growth enhances profits and downturn adversely affects the interest income (Alper and Anbar, 2011). This opposite result may be due to other reasons which include the customer’s preference or choice of depositing excess funds and taking loans and informational asymmetry of customer and lack of information regarding economic changes in the country. The banking survey indicates that, as GDP rises, advances to customers by banks falls. Inflation shows a positive relationship with industry advances when the GMM is used of coefficient of 0.092 at 5% significant level. Thus, a 1% increase in inflation leads to 0.092% increase in industry advances. This indicates that, bank management in the economy are able to forecast accurately inflation and are proactive in managing anticipated inflation. By making accurate forecast of inflation, the manager can increase the rates on loan faster than the rate at which operating cost is increasing so that inflation favourably University of Ghana http://ugspace.ug.edu.gh 91 impacts on market share. This result is consistent with studies by (Bourke, 1989; Molyneux et al., 1992; Athanasoglou et al. 2005). 5.5.4 Composite Index results as a proxy for market share The table 5.7 summarises the empirical results for the composite index as a proxy for market share. The composite index indicates the average combination of share of industry operating asset, Industry Deposit and Industry advances in the market among the fifteen banks in Ghana. The composite index model is a multiple regression equation comprising of nine explanatory variables out of which seven are bank-specific and two external factors. From the regression results, bank size determines market share positively with coefficient of 0.003 at 5% significant level. This implies that, a 1% increase in bank size leads to 0.003% increase in composite index which means that, composite index is not highly responsive to changes in bank size. This further implies that, a 1% increase in bank size leads to 0.001% increase in the composite index when GMM is used at 5% significant level. This suggests that, bank size leads to economies of scale in the economy, thereby making larger banks have the highest share in the market in terms of the operating asset, advances and deposit. Economies of scale will reduce the cost of gathering and processing information. The larger the bank size, the more the bank performs well. It could also mean that, bank size is associated with diversification which may impact favourably on risk and product portfolio. This result is consistent with the findings of (Molyneux et al., 1992; Bikker et al., 2002 and Goddard et al., 2004). University of Ghana http://ugspace.ug.edu.gh 92 Table 5.7: Estimation results for Composite Index (Average of IOA, ID and IA) Variable Random effect GMM Liquidity (L) -0.001 (.002) -0.000 (.001) Asset Quality (AQ) 0.030 (.039) -0.055 (.045) Bank Size (BS) 0.003** (.002) 0.001** (.001) Operating Efficiency (OE) -0.020** (.008) -0.011 * (.006) Bank Age (BA) 0.004 *** (.002) 0.005 *** (.001) Number of Branches (NB) 0.007 *** (.000) 0.003 *** (.000) Ownership Structure (OS) 0.017*** (.005) -0.016 (.021) Gross Domestic Product (GDP) -0.025*** (.007) -0.013 *** (.004) Inflation (INF) 0.003 (.004) 0.002 (.002) _cons 0.528 (.131) 0.285 (.086) Diagnostics R Squared 0.600 F Statistics Wald Chi2 199.64 83.48 Number of observation 143 130 ***p<0.01, **p<0.05,*p<0.10 Figures in parenthesis are robust standard errors Source: Author’s computation from annual reports and accounts of individual banks from 2004 to 2013. Operating Efficiency shows a negative relationship with the composite index with coefficient of - 0.020 at 5% significant level. This means that, a 1% increase in operating efficiency leads to a decrease in the composite index by 0.020% across time. This shows that, composite index is less responsive to changes in operating efficiency. This further University of Ghana http://ugspace.ug.edu.gh 93 indicates that, a 1% increase in operating efficiency leads to a 0.011% decrease in composite index when GMM is used at 10% significant level. A chunk of bank’s expense is composed of salary expense and administrative cost. It is possible that high bank salaries and administrative expenses in the economy are not being translated proportionately into profitability. This result is consistent with the studies by (Antonio, 2010; Ayanda eta al., 2013). The bank age shows a positive relationship with the composite index with coefficient of 0.004 at 1% significant level. Thus, a 1% increase in bank age leads to 0.004% increase in composite index and further shows 0.005% increase in bank age when there is a 1% increase in bank age when GMM is used at 1% significant level. This shows that, the old banks in the economy have skills advantage, experience and larger networks to the new banks which makes them have increase in operating asset, deposit and advances as compared to the new banks. This results is consistent with studies by Uschi and Veen (2009) and in contrast with studies by (Göbel and Zwick, 2009; Ruis and Scholman, 2012) which states that most of the older banks were less innovative as compared to the younger banks and the younger banks perform better than older banks. Number of branches show a positive relationship with composite index with coefficient of 0.007 at 1% significant level. This implies that, a 1% increase in number of branches leads to 0.007% in composite index. This also shows that, a 1% increase in the number of branches will increase the composite index by 0.003% when GMM is used at 1% significant level. The result implies that, the number of branches shows the number of University of Ghana http://ugspace.ug.edu.gh 94 workers and assets in the Ghanaian banking sector and these workers are being trained to acquire knowledge and skills to affect productivity and also building bank branch at a strategic location which is consistent with studies by (Shukla and Sinha, 2013; Hameed et al., 2014). GDP has a negative significant influence on bank market performance with coefficient of -0.025 at 1% significant level. This shows that a 1% increase in GDP leads to a percentage decrease in Composite Index by 0.025 %. It further shows that, a 1% percent increase in GDP leads to 0.013% decrease in the composite index when GMM is used at 1% significant level. This supports the findings of Kanwal, S. and Nadeem, M. (2013) and Sufian, F. (2011) who conclude that there is a negative relationship between GDP and performance of the banks. Whereas, insignificant negative relationship of GDP is in contradiction to the theory which asserts that, economic growth enhances profits and downturn adversely affects the interest income (Alper and Anbar, 2011). This opposite result may be due to other reasons which include the customer’s preference or choice of depositing excess funds, taking loans and informational asymmetry of customer and lack of information regarding economic changes in the country. Owner structure is incorporated into the model as a dummy. It shows that, holding all other factors constant, foreign banks have a percentage increase of 0.007 more than local banks. The result indicates, that local banks with worse market performance borrow more and banks within strategic industries borrow less due to the alternative financial support from University of Ghana http://ugspace.ug.edu.gh 95 the government. This is consistent with studies by (Micco et al., 2004 and Setiyono and Tarizi, 2010). 5.6 Regression results for the relationship between market share and profitability The Hausman test is performed to determine the appropriateness of the model to be adopted, where the null hypothesis is that the preferred model is random effects and the alternative states that the fixed effects is preferred. As indicated by Hausman test chi2 (6), the preferred model is the random effect for both ROA and ROE. The author also used the Generalised Method of Moments (GMM). All variables are in natural logarithm. 5.6.1 Dependent variable: Return On Average Asset (ROA) as a proxy for profitability. From table 5.8 below, it shows that, the industry operating asset has a positive relationship with ROA with coefficient of 0.961at 1% significant level, which demonstrates that a 1% increase in industry operating asset leads to an increase in ROA by 0.961. It further shows that, a percentage increase in industry operating asset leads to 2.03% increase in ROA when GMM is used at 10% significant level. This implies that, as the banks increase their operating assets, return on asset will increase. Thus, a large market share of operating asset is a reward for providing better value and a means of realizing lower cost. This assessment is supported by the theoretical review that market share is a predictor of profitability (Prescott et al., 1983; O ‘Regan, 2002) and empirically supported by (Newton, 1986; Wernerfelt, 1986 and Venkatraman et al., 1990) which states that there is a positive relationship between market share and profitability. University of Ghana http://ugspace.ug.edu.gh 96 Table 5.8: Estimation results showing the relationship between market share and Average Return on Asset (ROA) Variable Random effect GMM Industry Operating Asset (IOA) 0.961*** (0.306) 2.030* (1.057) Industry Deposit (ID) 0.212 (.350) -0.890 (.802) Industry Advances (IA) 0.274* (0.187) 0.705*** (.230) Composite Index (CI) -1.867* (4.673) -19.168*** (6.584) Liquidity (L) 0.029 (0.044) .029*** (.005) Asset Quality (AQ) 0.490 (1.330) -.491 (7.776) Bank Size (BS) 0.019* (0.061) -0.017* (.046) Operating Efficiency (OE) -0.208*** (0.251) -0.964*** (.360) Bank age 0.003** (0.003) 0.071** (.033) Number of Branches (NB) .001* (0.003) -0.008* (.009) Ownership Structure (OS) 0.3055 (0.168) 4.930* (3.067) Gross Domestic Product (GDP) 0.253 (0.232) 0.253 (0.274) Inflation (INF) -0.098 (0.114) -.023 (.036) _Cons -5.300 (4.557) -54.626 (24.253) DIAGONOTICS R Squared 0.631 Wald chi2 198.35 445.34 Number of observation 130 92 ***p<0.01, **p<0.05,*p<0.10 Figures in parenthesis are robust standard errors Source: Author’s computation from annual reports and accounts of individual banks from 2004 to 2013. University of Ghana http://ugspace.ug.edu.gh 97 The average composite index also shows that, there is a negative relationship between composite index and ROA with coefficient of 1.867 at 10% significant level from the random effect results. This implies that, a percentage increase in composite index leads to 1.867% decrease in ROA. When GMM is used, it also shows a negative relationship with ROA with coefficient of – 19.168 which implies that, a percentage increase in composite index leads to 19.168% decrease in ROA at 1% significant level. The result explains that, market share has a negative relationship with profitability which is supported by Hagigie (1999). This means that, for an organisation to pursue higher market shares, it must sacrifice some profit meaning there is an inverse relationship between market share and profitability. However, from the determinants of market share, only four variables are significant from the random effect results which include bank size, operating efficiency, bank age and number of branches at 10%. 1%, 5% and 10% significant levels respectively. This implies that, the relationship between ROA and industry operating asset, composite index and advances are not significant. Also, based on the GMM results, four variables are significant which include liquidity at 1%, bank age at 5%, operating efficiency at 1% and ownership Structure at 10% significant levels. These show that, combining the results of market share findings, the relationship between ROA with composite index, advances and operating asset are not significant since the determinants of market share are not significant. Thus, requirements for effective and profitable performance go beyond the share position of the market. University of Ghana http://ugspace.ug.edu.gh 98 5.6.2 Dependent variable: Return On Average Equity (ROE) as a proxy for profitability. Table 5.9 (page 100) shows the estimation results for the relationship between market share and Return on Average Equity (ROE). From table 5.9, it shows that, the industry operating asset has a positive relationship with ROE with coefficient of 0.378 at 10% significant level which shows that, a 1% increase in industry operating asset leads to an increase in ROE by 0.378. It further shows that, a percentage increase in industry operating asset leads to 2.030% increase in ROE when GMM is used at 5% significant level. This implies that, as the banks increase their operating assets, return on equity will increase. Thus, a large market share of operating asset is a reward for providing better value and a means of realizing lower cost. This assessment is supported by the theoretical review that market share is a predictor of profitability (Prescott et al., 1983; O ‘Regan, 2002) and empirically supported by (Newton, 1986; Wernerfelt, 1986 and Venkatraman et al., 1990) which states that there is a positive relationship between market share and profitability. The industry deposit also shows a positive relationship with ROE with coefficient of 0.727 at 5% significant level which explains that a 1% increase in industry deposit leads to 0.727% increase in ROE. Thus, large market share of the deposit is a reward for providing improved value and a means of appreciating lesser cost. This assessment is supported by the theoretical review that market share is a predictor of profitability (Prescott et al., 1983; O ‘Regan, 2002) and empirically supported by (Newton, 1986; Wernerfelt, 1986 and University of Ghana http://ugspace.ug.edu.gh 99 Venkatraman et al., 1990) which states that there is a positive relationship between market share and profitability. The industry advances variable indicates a positive relationship with ROE with elasticities of 0.524 at 1% significant level. This implies that, a percentage increase in industry advances lead to 0. 524% increase in ROE and it further shows that, a 1% increase in industry advances lead to 0.398 in ROE when GMM is used at 1% significant level. This assertion also conforms to the theory that market share is a predictor of profitability and for managers to increase their return on assets, market share has to increase and that there is a positive relationship between market share and ROE. The average composite index has a negative relationship with ROE with coefficient of – 8.673 at 5% significant level which implies that, a percentage increase in composite index leads to -8.673% decrease in ROE and further shows that, a 1% increase in composite index leads to a decrease in ROE by -7.347% when GMM is used at 1% significant level. The result explains that, market share has a negative relationship with profitability which is supported empirically by Hagigie (1999). This means that, for an organisation to pursue higher market shares, it must sacrifice some profit meaning there is an inverse relationship between market share and profitability. University of Ghana http://ugspace.ug.edu.gh 100 Table 5.9: Estimation results for the relationship between market share and profitability (ROE) Variable Random effect GMM Industry Operating Asset (IOA) 0.378* (.280) 2.030** (1.057) Industry Deposit (ID) .727** (.334) -0.890 (.802) Industry Asset (IA) .524*** (.167) 0.705*** (.230) Composite Index (CI) -8.673** (4.345) -19.168*** (6.584) Liquidity (L) .0208 (.037) 7.232* (4.715) Asset Quality (AQ) -.451 (1.284) -6.169 (8.772) Bank Size (BS) -.0100* (.052) 2.337* (1.472) Operating Efficiency (OE) -.193 (.219) 3.837 (3.016) Bank Age (BA) .007* (.004) .137*** (.046) Number of Branches (NB) .0048* (.004) .024* (.020) Ownership Structure (NOS) .160 (.216) -4.813 (4.174) Gross Domestic Product (GDP) -.122 (.202) -3.4829 (1.775) Inflation (INF) -.126* (.095) -.0200* (.085) _Cons 5.720 (3.976) 26.92 (12.018) DIAGONOTICS R Squared 0.719 F Statistics Wald chi2 139.31 202.32 Number of observation 130 92 ***p<0.01, **p<0.05,*p<0.10 Figures in parenthesis are robust standard errors Source: Author’s computation from annual reports and accounts of individual banks from 2004 to 2013. University of Ghana http://ugspace.ug.edu.gh 101 Nevertheless, the determinants of market share in relation to profitability shows that, three variables are significant from the random effect results which include bank age, number of branches and inflation at 10% significant levels respectively. These indicates that, the relationship of ROE with industry deposit, composite index and advances are not significant. Also, based on the GMM results, four variables of the determinants of market share are significant which include liquidity at 10%, bank age at 1% and bank size at 10% significant levels. This shows that, the relationship of ROE with composite index, advances and deposit are not significant. Thus, requirements for effective and profitable performance go beyond the share position of the market. 5.7 Conclusion This chapter analysed the determinants of market share and the relationship between market share and profitability. The results indicate that, bank size, operating efficiency, bank age, number of branches, GDP and Inflation significantly determine market share. The results also show that, there is no significant relationship between market share and profitability. Thus, requirements for effective and profitable performance go beyond the share position of the market. University of Ghana http://ugspace.ug.edu.gh 102 CHAPTER SIX SUMMARY, CONCLUSION AND POLICY RECOMMENDATIONS 6.0 INTRODUCTION This chapter presents the summary of findings, conclusions and recommendations based on the results. The summary presents a snapshot of the study, unfolding the various highlights of the study. The inference based on the empirical study is captured in the conclusion while the recommendations are proposed based on the conclusions. 6.1 Summary The role of banks in the economic development and growth of a country cannot be undervalued. A healthy financial system is essential for proper financial mediation leading to sustainable private investment and the promotion of entrepreneurship (Kutsienyo, 2011). The banking industry has witnessed many reforms and policies over the past decade. The competitive background and functioning environment have become vibrant. There is heightened pressure on banks to compete as banks have become more integrated into the global financial system. In order to withstand economic shocks and to maintain financial stability, it is imperative to identify the determinants that influence bank market share in Ghana and the association between market share and bank profitability. These are the objectives this study sought to accomplish. This study examines the determinants of bank market share and the relationship between market shares and profitability of banks in Ghana. These determinants are categorized into internal factors which are bank-specific characteristics and external factors. This University of Ghana http://ugspace.ug.edu.gh 103 categorization is in line with earlier studies (Bourke, 1989 and Molyneux et al., 1992). Four key measures of market share (dependent variables) are used in the study. These comprise of Industry Operating Asset (IOA), Industry Deposit (ID) and Industry Advances (IA) which are based on the ratios from the Ghana Banking Survey, 2014 and the average composite index. Also, two key measures of profitability are used in this study. These include Return on Average Asset (ROA) and Return on Average Equity (ROE). The variables used as factors that affect market are bank specific factors which include liquidity, asset quality, operating efficiency, bank size (bank total asset), bank age, ownership structure, number of branches and external variables which include inflation and Gross Domestic Products (GDP). The variables are specially chosen to determine the factors of market share. The relationship between market share and profitability is achieved by finding the relationship between the dependent variables of profitability and four key measures of market share including the determinants of market share. A panel data of 15 banks are analysed over a period of 2004 to 2013 due to availability of data and the adoption of the universal banking in 2004 since the banks used are universal banks. Specifically banks that existed from 2004 to 2013. 6.1.1 Main Findings of Determinants of Market Share The results indicate that, out of the nine variables studied, determinants of market share include bank size, operating efficiency, bank age, number of branches, Gross Domestic Product and inflation. The study finds a significant positive relationship between bank size and the market share. Thus, as the total assets increases, market share also increases. University of Ghana http://ugspace.ug.edu.gh 104 Operating efficiency also shows a negative relationship with market share, which implies that, a chunk of bank’s expense is composed of salary expense and administrative cost. It is possible that high bank salaries and administrative expenses are not being translated proportionately into profitability and therefore have a negative relationship to market performance. Bank age indicates a positive relationship with market share, which implies that the older banks have experience and skills that makes them perform better the new banks. Number of branches of banks finds a positive relationship with market share, which indicates that, the number of branches of banks signifies the size of the bank which leads to increase in their operating asset. GDP shows a negative relationship with market share, demonstrating that, an increase in GDP leads to a decrease in market share which may be as a result of managerial decision of the banks or information asymmetry by the customers. Inflation has a positive relationship with market share, showing that managers are able make accurate forecast of inflation to increase their share of operating asset. University of Ghana http://ugspace.ug.edu.gh 105 6.1.2 Relationship between market share and profitability Using ROA and ROE as a measure of profitability to show the relationship between profitability and market, the results indicate that, there is no significant relationship between market share and profitability. 6.2 Conclusion The study examines the determinants that influence market share and the relationship between profitability and market share. Two main determinants of bank market share are identified; the internal determinants and the external determinants. The internal determinants are bank-specific and are controlled by the bank manager. The external determinants are outside the control of banks, although banks can strategically be positioned to exploit the opportunities in these environments or mitigate threats from this environment. The external environment defines the sociological, political, technological and the economic scenes in which banks operate. The results indicate that, bank size, operating efficiency, bank age, number of branches, GDP and Inflation significantly determine market share. The results also show that, there is no significant relationship between market share and profitability. 6.3 Policy Recommendations Based on the results of the study, the following policy implications and recommendations are suggested. These recommendations generally aim at improving market share and profitability of banks, especially in Ghana. University of Ghana http://ugspace.ug.edu.gh 106 Findings from the study reveals that, operating efficiency measured as expenses over income have a negative relationship. Thus, efficient management of bank operations can alleviate the high operational cost that erodes bank profits. Bank occupancy cost and salaries are major components of operational cost. It is therefore recommended that, banks must employ more technologies to automate their service delivery. These technologies would enable banks to explore new markets without maintaining a physical presence. It would reduce the amount of staff costs, occupancy cost, paper cost and queuing times in the banking halls. Managerial cost and other expenses should be at optimal level and consistent with profit maximisation objectives of shareholders. The findings indicate that, bank size has a positive relationship with market share through economies of scale and scope. Economies of scale and economies of scope derived from bank size and producing two or more products jointly play a crucial role in bank performance. The benefit of size would reflect on the ability to reach wider markets. It is therefore recommended that, banks should look beyond the local market and strategically expand their operations to other geographical markets and sectors of the economy. The agriculture and agro-processing sector is still a potential market for banks. Also, banks should consider diversification of their product portfolio. In this way, banks can leverage on their assets to offer other ancillary services and maximise returns. University of Ghana http://ugspace.ug.edu.gh 107 The number of branches of banks also have a positive relationship with bank’s market performance therefore, bank branches should be built at strategic locations since bank branches play a role in choosing a financial services provider by customers. Banks must take radical steps in building the capacity of its employees to attract more customers to deposit as they make advances to customers and other banks. Bank age shows a positive relationship with banks’ market share. Thus, the existing banks have skills advantage, experience and larger networks to the new banks which makes the old banks have market advantage to the new banks and therefore, the new banks can also perform better in the market through the Research and Development and also learning from the management style of the existing banks. The study also shows a negative relationship between market share and GDP. As indicated earlier, the insignificant negative relationship of GDP is in contradiction to the theory which asserts that economic growth enhances profits and downturn adversely affects the interest income. This opposite result may be due to other reasons which include the customer’s preference or choice of depositing excess funds, taking loans, informational asymmetry of customer and lack of information regarding economic changes in the country. Hence, the negative relationship between Gross Domestic Product (GDP) and bank performance. Thus, government policies on employment and investments should be intensified to increase the market performance of banks. In an economy where government University of Ghana http://ugspace.ug.edu.gh 108 stimulates the creation of jobs and creates the right investment climate for both local and foreign investors, banks are likely to thrive well. Also, proceeds from the oil sector can be invested in the banking areas where there seems to be resources but no capital to convert these resources into finished goods and services. Moreover, there should be enough education concerning the banking system in Ghana so that information would be reached out to all especially to those in the rural areas in the community. Although inflation seemed to have a positive influence on bank market performance, high inflation may generally be undesirable. The results suggest that probably, bank managers are accurately predicting inflation and are able to adjust their lending rates accordingly. Low inflationary regimes create a stable economy and a congenial investment climate for businesses, enabling businesses to pursue a long term project critical to their survival and growth. The study indicates that, there is no significant relationship between market share and profitability and therefore, before adoption of market share strategy, management need to understand the context of their environment. Organisations need to understand their market share position, what their strengths and weakness are and what drives profitability in their respective weakness. University of Ghana http://ugspace.ug.edu.gh 109 6.4 Limitations of the study and areas for future research There are certain equally important variables that were omitted from the study due to data availability. This data availability problem did not permit the author to predict what determines market share and the relationship between profitability and market share. These variables included capital adequacy and bank concentration. Notwithstanding these limitations which could be the basis for further research, the results of the current study are still valid and could be used as the basis for policy formulation. University of Ghana http://ugspace.ug.edu.gh 110 REFERENCES Acaravci, S. K. and Çalim, A. E. (2013).Turkish Banking Sector’s Profitability Factors International Journal of Economics and Financial Issues 3 (1), 27-41. Akuffo-Duah, D. (2011). Competition in the Banking Industry. Master’s Thesis of Business Administration. MBA programme. Kumasi, Ghana. Alper, D., Anbar, A. (2011), Bank Specific and Macroeconomic Determinants of Commercial Bank Profitability: Empirical Evidence from Turkey. Business and Economics Research Journal, 2(2), 139-152. Annual Reports and Accounts of individual banks in Ghana, 2004 to 2013. Antonio, T. P. (2012). What Determines the Profitability of Banks? Evidence from Spain. Accounting & Finance, 53, 561-586. Athanasoglou P., Delis, M., Staikouras, C. (2006). Determinants of Banking Profitability in the South Eastern European Region. Munich Personal RePEc Archive (MRPA) working paper No. 10274. (http://mpra.ub.uni-muenchen.de/10274/) (Assessed on May 6 2015). Awuah, L. (2011). An Evaluation of Strategies for Achieving Competitive Advantage in the Banking Industry: The Case of Ghana Commercial Bank Limited. Master’s Thesis in the Business Administration. Kumasi, Ghana. University of Ghana http://ugspace.ug.edu.gh 111 Ayanda, A.M., Ekpo, I. C. and Mustapha, A.M. (2013). Determinants of Banks’ Profitability in a Developing Economy: Evidence from Nigerian Banking Industry Interdisciplinary Journal of Contemporary Research in Business January 2013 Vol. 4, No 9. Baltagi, B. H. (2001). Econometric Analysis of Panel Data, 3rd Ed, John Wiley and Sons Ltd., Southern Gate, Chichester. Bank of Ghana Annual Report 2013. (www.annualreportsghana.com/Services/Reports.aspx) (Accessed on 26 November 2014). Banking Act, 2004 Act 673. (www.bu.edu/bucflp/files/2012/01/Banking-Act-No.-673-of-2004.pd) (Accessed on 28 November 2014). Berger, A. N. and Bouwman, H. C. (2013). How Does Capital Affect Bank Performance During Financial Crises? Journal of Financial Economics, 109 (1), 146–176. Bessis, J. (2002), Risk Management in Banking, John Wiley and Sons, Chichester. 2nd ed. Bikker, J.A., Hu, H. (2002). Cyclical Patterns in Profits, Provisioning and Lending of Banks and Pro-cyclicality of the New Basel Capital Requirements. BNL Quarterly Review 221, 143-175. University of Ghana http://ugspace.ug.edu.gh 112 Blundell, R., and Bond, S. (1998). Initial conditions and moment restrictions in dynamic panel data models. Journal of econometrics, 87(1), 115-143. Bordeleau É. and Graham, C. (2010). The Impact of Liquidity on Bank Profitability Bank of Canada Working Paper 2010-38 December 2010. (www.bankofcanada.ca/wp- content/uploads/2010/12/wp10-38.pdf) (Assessed on 6 May 2015). Bourke, P. (1989). Concentration and Other Determinants of Bank Profitability in Europe, North America and Australia. Journal of Banking and Finance, 13: 65-79. Boyd, J. and Runkle, D. (1993). Size and performance of banking firms: testing the predictions of theory. Journal of Monetary Economics, Vol.31: 47-67. Boot, A.W.A. (2003). Consolidation and Strategic Positioning in Banking with Implications for Europe, Brookings-Wharton, Papers on Financial Services, Vol.(1), p. 37-83. Bossone, B., and Lee J.-K. (2004), In Finance, Size Matters: The "Systemic Scale Economies" Hypothesis, IMF Staff Papers, Vol. 51(1), p. 19-46. Breitung, J. (2000). The Local Power of Some Unit Root Tests for Panel Data. In Advances in Econometrics, Volume 15: Non stationary Panels, Panel Cointegration, and Dynamic Panels, Baltagi, B. H. (Ed.), 161–178. Amsterdam: JAI Press. Burja, C. (2011). Factors Influencing the Companies’ Profitability. Annales Universitatis Apulensis Series Oeconomica, 13(2). University of Ghana http://ugspace.ug.edu.gh 113 Choi, I. (2001), “Unit Root Test for Panel Data”, Journal of International Money and Finance, 20: 249-272. Cooper, G. L. and Nakanishi, M. (2010). Market-Share Analysis Evaluating Competitive Marketing Effectiveness. Boston: Kluwer Academic Publishers, 304 pg. (1988). (Web Edition). Davis, S. P., Schul, L.P., Babakus, E. And Pedrick, D. (1993). An Examination of the Role of Market Share Position in the Strategy-Performance Equation. Journal of Strategic Marketing 1 153-169. Depken, C. A., II (1999). Free-Agency and the Competitiveness of Major League Baseball, Review of Industrial Organization, 14, 205-217. Demirguc-Kunt, A. and Huizinga, H. (1999). Determinants of commercial bank interest margins and profitability: Some international evidence, World Bank Economic Review, 13, 379-408. Domeher, D., Frimpong, J. M., and Appiah, T. (2015). Adoption of financial innovation in the Ghanaian banking industry. African Review of Economics and Finance, 6(2), 88-114. Dougherty C. (2006). Introduction to Panel Data Models, Chapter 14: 408 (Website: http://www.oup.com/uk/orc/bin/9780199280964/dougherty_chap14.pdf) (Accessed 2011 January 30). Flamini, V., McDonald, C., Schumacher, L. (2009). The Determinants of Commercial Bank Profitability in Sub-Saharan Africa. IMF Working Paper. University of Ghana http://ugspace.ug.edu.gh 114 Finlay, P. (2000). Strategic Management: An Introduction to Business and Corporate Strategy, Prentice Hall, London. Fries, S. and Taci, A. (2005), “Cost Efficiency of Banks in Transition” Gale, B. T. and Buzzell, R. D. (1993). Market position and competitive strategy. In Day et al. (Eds). The interface of marketing and Strategy, JAI Press, Greenwich, CT. Ghana Banking Survey, (2008). PricewaterhouseCoopers and Ghana Association of Bankers. (Website:http://www.pwc.com/en_GH/gh/pdf/ghana-banking-survey- 2008.pdf). (Accessed 2014 September 15). Ghana Banking Survey, (2014). PricewaterhouseCoopers and Ghana Association of Bankers. (Website:http://www.pwc.com/en_GH/gh/pdf/ghana-banking-survey- 2014.pdf). (Accessed 2014 September 15). Gobel, C. and Zwick, T. (2009). Age and Productivity: Evidence from Linked Employer Employee Data. Germany, pp. 1-35. Available: https://ftp://ftp.zew.de/pub/zew- docs/dp/dp09020.pdf (2015, May 30). Goddard, J., Molyneux, P., Wilson, J.O.S., (2004). ‘‘The profitability of European banks: a cross-sectional and dynamic panel analysis’’. Manchester School 72 (3), 363-381. Greene, W.H. (2003). Econometric Analysis, Prentice Hall, 5th Edition. Gujarati N. D. (1995). Basic Econometrics (3rd edition), Singapore: McGraw-Hill. University of Ghana http://ugspace.ug.edu.gh 115 Gul, S., Irshad, F. and Zaman K. (2011). Factors affecting bank Profitability in Pakistan. The Romanian Economic Journal, 39, 61-87. Guru B.K., Staunton J. and Balashanmugam B. (1999). ‘‘Determinants of Commercial Bank Profitability in Malaysia’’, Paper presented at the 12th Annual Australian Finance and Banking Conference, Sydney, Australia, 16-17 December. Hagigi, M., Manson, G. I., Mascarenhas, B. (1990). Increase assets efficiency to gain multinational market share. Management international Review, 39(3) 205. Hansen, L. P. (1982). Large sample properties of generalized method of moments estimators. Econometrica: Journal of the Econometric Society, 1029-1054. Heffernan, S., (1996). Modern banking in theory and practice, John Wiley and Sons, Chichester. Hinson, R., Amidu, M. and Mensah, R. (2006). Determinants of Ghanaian Bank Service Quality in a Universal Banking Dispensation. Banks and Bank Systems / Volume 1, Issue 2, Page 3. Howard, B. B. and Upton, M. (1991). Introduction to Business Finance. New York; McGraw Hill Book. Huerta, T. (2008). Competition in Banking: The Role of Orderly Entry and Orderly Exit, Speech on Banking Sector, FSA Chatham House Conference, London. University of Ghana http://ugspace.ug.edu.gh 116 Hughes, J. P. and Mester, L. J. (2008). Efficiency in Banking: Theory, Practice, and Evidence Working Paper No. 08-1. Kanwal, S. and Nadeem, M. (2013). The Impact of Macroeconomic Variables on the Profitability of Listed Commercial Banks in Pakistan European. Journal of Business and Social Sciences, Vol. 2, No.9, Pp 186-201, December 2013. P.P. 186 – 201’ Kinney, S. K., & Dunson, D. B. (2007). Fixed and random effects selection in linear and logistic models. Biometrics, 63(3), 690-698. Kosmidou, K. (2008). The determinants of banks' profits in Greece during the period of European Union financial integration. Managerial Finance 34(3): 146-159. Kosmidou, K., Pasiouras, F. and Tanna, S., (2005). Determinants of Profitability of Domestic UK Commercial Banks: Panel Evidence from the Period 1995-2002. In Money Macro and Finance (MMF) Research Group Conference (Vol. 45). Kotler, P. and Keller K. L. (2006). Marketing Management, 6th edition. Upper Saddle River: Pearson Prentice Hall. Krakah, A.K. and Ameyaw, A. (2010). The Determinants of Bank’s Profitability in Ghana, The Case of Merchant Bank Ghana Limited (MBG) and Ghana Commercial Bank (GCB). Master’s Thesis in Business Administration. MBA programme. Accra, Ghana. University of Ghana http://ugspace.ug.edu.gh 117 Kumar, V. and W. J. Reinartz (2006). Customer Relationship Management-A databased approach. USA, John Wiley & Sons, Inc. Kutsienyo, L. (2011). The Determinant of Profitability of Banks in Ghana. Master’s Thesis of Business Administration. MBA programme. Kumasi, Ghana. Mamatzakis, C. and Remoundos, C. (2003). Determinants of Greek Commercial Banks Profitability, 1989-2000. pp 53(1): 84-94. Mamman, A. and Yusuf A. H (2013). Impact of private sector credit on the real sector of Nigeria International Journal of Business and Social Research (IJBSR), Volume - 3, No.-5. Masukujjaman, M.D. and Akter, A. (2010). Quality of Categorized Service and Customer Satisfaction in Banking Industry: An Empirical Study on Private Commercial Banks in Bangladesh Volume– V, Issue– 02, July-December, 2010. Micco, A., Panizza, U. and Yañez, M. (2004). Bank Ownership and Performance Inter- American Development Bank Research Department Working Paper Number 518. Molyneux, P. and Thornton, J. (1992). “Determinants of European Bank Profitability: A Note”. Journal of Banking and Finance, 16: 1173-1178. Mutshinyani, M. (2009). A Study of the Relationship between market shares and profitability in South African listed retail companies. Master’s Thesis in Business Administration. MBA programme, South Africa. University of Ghana http://ugspace.ug.edu.gh 118 Newton, J.K. (1983). Market Share- A Key To Higher Profitability? Long Range Planning 16(1) 37-41. Nwankwo, G.O. (1991). Bank Management, Principles and Practice, Malthouse Press Ltd. Lagos. O’Regan, N. (2002. Market Share: The Conduit for Future Success? European Business Review, 14(4) 287-293. Olalekan, A. and Adeyinka, S. (2013). Capital Adequacy and Banks' Profitability: An Empirical Evidence from Nigeria American International Journal of Contemporary Research Vol. 3 No. 10; October 2013. Park, H.M (2005). Linear Regression Models for Panel Data Using SAS, Stata, LIMDEP, and SPSS, Indiana University. Perry, P. (1992). ‘‘Do banks gain or lose from inflation’’, Journal of Retail Banking, Vol. 14 No. 2: 25-40. Pindyck, R. S. and Rubinfeld, D. L. (1991). Econometric Models and Economic Forecasts, N.Y.: McGraw-Hill pp1-54. The Netherlands. Available: https://www.entrepreneurship-sme.eu (2015, May 30). Prescott, J. E. and Venkatraman, N. (1986). The Market Share- Profitability Relationship: an empirical assessment of major assertions and contradiction. Strategic Management Journal 7(4) 377-394. University of Ghana http://ugspace.ug.edu.gh 119 Reininger, T., Schardax, F. and summer, M., (2002). “Financial System Transition in Central Europe: The First Decade”, SUERF Studies, n° 16, Vienna. Ruis, A. and Scholman, G. (2012). Ageing and entrepreneurship: An empirical study of the relationship between entrepreneurial age and objectives, strategy and performance. EIM Research Reports H, 201206. Sala-i-Martin, X., (1997). “I just ran two million regressions”, American Economic Review, 87, 178-183. Samad, A., Glenn, M. L. and Miah, F. (2006). Inter-Temporal Performance: Does Bank- Size Matter? Banks and Bank Systems, Vol 1, Issue 2, Pp 137 -143 Scholtens, B. (2000). Competition, Growth, and Performance in the Banking Industry. Center for Financial Institutions, Working Papers 00-18, Wharton School Center for Financial Institutions, University of Pennsylvania. Setiyono, B. and Tarizi, A. (2014). Disclosure, ownership structure and bank risk: Evidence from Asia. Pp. 1-47, France. Available: https://hal-unilim.archives- ouvertes.fr/hal-00947590 (2015, May 30). Shukla, S. and Sinha, A. (2013). Employee Turnover in banking sector: Empirical evidence from IOSR Journal of Humanities and Social Science (IOSR-JHSS) Volume 11, Issue 5 (May. - Jun. 2013), PP 57-61. University of Ghana http://ugspace.ug.edu.gh 120 Staikouras, C. K., & Wood, G. E. (2004). The determinants of European bank profitability. International Business & Economics Research Journal, 3(6), 57-68. Sufian, F. (2011). Profitability of the Korean Banking Sector: Panel Evidence on Bank- Specific and Macroeconomic Determinants. Journal of Economics and Management, 7(1), 43-72. Tabari1, N. A. Y and Ahmadi, M. Emami, M. (2013). The Effect of Liquidity Risk on the Performance of Commercial Banks International Research Journal of Applied and Basic Sciences Vol, 4 (6): 1624-1631. Tran, A. T. O., Kok-Boon and Van Anh L. Q. (2010). Customer satisfaction impacts on bank performance: an empirical study from a developing country. Journal of International Business and Economic publisher: International Academy of Business and Economics Audience: Academic Sept, 2010 Source Volume: 10 Source Issue: 3. Uschi, B. G. and Veen, S. (2009). The Impact of Aging and Age Diversity on Company Performance. ISU Working Paper 78, University of Zurich, Switzerland. Available: http://www.zora.uzh.ch (2015, May 31). Venkatraman, N. and Prescott, J. E. (1986). The Market Share- Profitability Relationship: an empirical assessment of major assertions and contradiction. Strategic Management Journal 7(4) 377-394. University of Ghana http://ugspace.ug.edu.gh 121 Venkatraman, N. And Prescott, J.E. (1990. The Market Share - Profitability Relationship: Testing Temporal Stability across Business Cycles. Journal of Management 16(4) 783-805. Walter, I. (2003). Strategies in Financial Services, the Shareholders, and the System: Is Bigger and Broader Better? Brookings-Wharton, Papers on Financial Services, Vol. (1), p. 1-36. Wernerfelt, B. (1986). The Relationship between Market Share and Profitability. Journal of Business Strategy 67-74. Wiele, T. V. D., Boselie, P. and Hesselink, M. (2001). Empirical Evidence for the Relation between Customer Satisfaction and Business Performance Paper for QMOD'2001 Conference in Linkoping/Sweden, 12-14 September 2001. Yannopoulos, P. (2012). The Market Share Effect: New Insights from Canadian Data. (www.jgbm.org/page/31Peter%20Yannopoulos.pdf) (Assessed on 14 April 2015). Zouari, S. And Taktak, N. B. (2012). Ownership Structure and Financial Performance in Islamic Banks: Does Bank Ownership Matter? Economic Research Forum. Working Paper Series Working Paper 713. University of Ghana http://ugspace.ug.edu.gh 122 APPENDIX Banks considered for the study 1. Standard Chartered Ghana (SCB) 2. Ghana Commercial Bank (GCB) 3. Société Générale Social Security Bank (SG-SSB) 4. National Investment Bank (NIB) 5. Agricultural Development Bank of Ghana (ADB) 6. Prudential Bank Limited (PBL) 7. Merchant Bank (Ghana) Limited now Universal Merchant Bank (UMB) 8. Ecobank Ghana (EBG) 9. CAL Bank formerly Continental Bank (CAL) 10. HFC Bank (HFC) 11. Stanbic Bank (Stanbic) 12. First Atlantic Bank Limited (FABL) 13. Bank of Africa Ghana Limited (BOA) formerly Amalgamated Bank Limited 14. Barclays Bank of Ghana (BBGL) 15. International Commercial Bank (ICB) now First Bank of Nigeria (FBN) University of Ghana http://ugspace.ug.edu.gh