Determinants of Capital Structure of Commercial Banks in Ethiopia

Weldemikael Shibru

A Thesis Submitted to

The Department of Accounting and Finance

Presented in Partial Fulfillment of the Requirements for the

Degree of Master of Science (Accounting and Finance)

Addis Ababa University

Addis Ababa, Ethiopia

June 2012
 Addis Ababa University

School of Graduate Studies

This is to certify that the thesis prepared by Weldemikael Shibru, entitled: Determinants

of Capital Structure: an Empirical Study on Ethiopian Banking Industry and submitted in

partial fulfillment of the requirements for the degree of Degree of Master of Science

(Accounting and Finance) complies with the regulations of the University and meets the

accepted standards with respect to originality and quality.

Signed by the Examining Committee:

Examiner Dr. Laxmikantham P. Signature _______________ Date ______________

Examiner Dr. Venkati P. Signature _______________ Date ______________

Advisor Dr. Wollela A. Signature ________________ Date ______________

__________________________________________________________________

Chair of Department or Graduate Program Coordinator

1
 ABSTRACT

Determinants of capital structure: An Empirical Study on Ethiopian Banking Industry

Weldemikael Shibru

Addis Ababa University, 2012

Determining the optimal capital structure is one of the most fundamental policy decisions

faced by financial managers. Since optimal debt ratio influences firm’s value, different

firms determine capital structures at different levels to maximize the value of their firms.

Thus, this study examines the relationship between leverage and firm specific

(profitability, tangibility, growth, risk, size and liquidity) determinants of capital

structure decision, and the theories of capital structure that can explain the capital

structure of banks in Ethiopia. In order to investigate these issues a mixed method

research approach is utilized, by combining documentary analysis and in-depth

interviews. More specifically, the study uses twelve years (2000 - 2011) data for eight

banks in Ethiopia.

The findings show that profitability, size, tangibility and liquidity of the banks are

important determinants of capital structure of banks in Ethiopia. However, growth and

risk of banks are found to have no statistically significant impact on the capital structure

of banks in Ethiopia. In addition, the results of the analysis indicate that pecking order

theory is pertinent theory in Ethiopian banking industry, whereas there are little evidence

to support static trade-off theory and the agency cost theory. Therefore, banks should

give consideration to profitability, size, liquidity and tangibility when they determine

their optimum capital structure.

iii 2
Acknowledgements

First of all, I would like to extend my deep indebtedness to my advisor, Wollela

Abehodie Yesegat (PhD) for her invaluable comments, encouragements and guidance at

various stage of the study.

I gratefully acknowledge Addis Ababa University; particularly Department of

Accounting and Finance for the financial and administrative support provided to me

during my thesis work. My thanks also go to Jimma University for giving me financial

support and study leave.

My heartfelt thanks are also extended to the management and staff members of the

Ethiopian Commercial Banks and the National Bank of Ethiopia for their support in

providing me all the necessary data required for the study.

I would also like to convey my sincere thanks to my parents, specially my father Shibru

Bekelcha and my mother Bizunesh Ejigu whose unconditional love and silent prayers

encouraged me throughout my tenure at Addis Ababa University.

Last but not the least, my special thanks goes to my friend Helen Abraham and for those

who helped me in any form of assistance.

iv
3
 Table of Contents

List of Figures ................................................................................................................... vii

List of tables ..................................................................................................................... viii

List of Acronyms ............................................................................................................... ix

Chapter one: Introduction ............................................................................................... 1

1.1. Statement of the problem ..................................................................................... 3

1.2. Objective, research questions and hypotheses ..................................................... 5

1.3. Research methodology ......................................................................................... 9

1.4. Scope and limitation of the study ....................................................................... 10

1.5. Significance of the study .................................................................................... 10

1.6. Organization of the thesis ................................................................................... 11

Chapter Two: Review of related literature ................................................................. 12

2.1. Theoretical review .................................................................................................. 12

2.1.1. Modigliani and Miller (MM) theory................................................................ 12

2.1.2. Static Trade-off Theory ................................................................................... 13

2.1.3. Pecking order theory ........................................................................................ 15

2.1.4. Agency cost theory .......................................................................................... 18

2.2. Empirical studies on the determinants of capital structure .................................... 23

2.3. Conclusion and knowledge gap.............................................................................. 33

v
4
Chapter Three: Research Design and methodology .................................................... 35

3.1. Objective, Hypotheses and research questions ...................................................... 35

3.2. Research Approaches ............................................................................................. 36

3.3. Methods adopted .................................................................................................... 38

3.3.1. Research methods: quantitative aspect ............................................................ 39

3.3.2. Research method: qualitative aspect ................................................................ 40

3.4. Conclusion and relation between research questions/ hypotheses and data sources

....................................................................................................................................... 45

Chapter Four: Results and Analysis ............................................................................. 47

4.1. Research hypotheses and questions........................................................................ 47

4.2. Results .................................................................................................................... 48

4.2.1. Documentary analysis...................................................................................... 48

4.2.4. Results of Regression analysis ........................................................................ 56

4.2.2. In-depth interview result .................................................................................. 58

4.3. Discussions of the Results ...................................................................................... 60

Chapter Five: Conclusions and Recommendations ..................................................... 67

5. 1. Conclusions ........................................................................................................... 67

5.2. Recommendations .................................................................................................. 70

References ........................................................................................................................ 71

5
List of Figures

Figure 2.1: The-static-tradeoff theory of capital structure ………………………………14

Figure: 3.1: Rejection and Non-Rejection Regions for DW Test………………………..46

Figure 4.1: Normality test ……………………………………………………………….54

vii

6
List of Tables

Table 3.1: Variable-Indicator List………..………………………………………………44

Table: 3.2 Relationships between research question, hypotheses and different data

source…………………………………………………………………………………….47

Table 4.1: Summary of descriptive statistics for dependent and explanatory variable….51

Table 4.2: Correlation (Pearson) matrix……………………………………...………….53

Table 4.3: Correlation matrix between explanatory variables…………….……..………54

Table 4.4: Heteroscedasticity Test: White test………………………..…………………55

Table 4.5: Correlated Random Effects - Hausman Test……...……….…………………56

Table 4.6: Fixed effect model estimates…………………………...…………………….58

viii 7
List of Acronyms

CLRM Classical Linear Regression Model

DW Durbin Watson

GR Growth

LEV Leverage

LQ Liquidity

NBE National Bank of Ethiopia

OLS Ordinary least Square

PR Profitability

RS Risk

SZ Size

TA Tangibility

ix
8
9
 Chapter one: Introduction

Capital structure refers to several alternatives that could be adopted by a firm to get the

necessary funds for its investing activities in a way that is consistent with its priorities.

Most of the effort of the financial decision making process is centered on the

determination of the optimal capital structure; where the firms’ value is maximized and

cost of capital is minimized. Capital structure theory suggests that firms determine what

is often referred to as a target debt ratio; which is based on various trade-off between the

costs and benefits of debt versus equity. The modern theory of capital structure was first

established by Modigliani and Miller (1958). Following the seminal work of Modigliani

and Miller (1958), a vast theoretical literature developed, which led to the formulation of

alternative theories, such as the static trade off theory, pecking order theory and agency

cost theory.

Trade- off theory proposes that the optimal debt ratio is set by balancing the trade-off

between the benefit and cost of debt. According to this theory, the optimal capital

structure is achieved when the marginal present value of the tax shield on additional debt

is equal to the marginal present value of the financial distress cost on additional debt

(Myers 1984). Pecking Order Theory emphasizes the information asymmetry between the

firm insiders and the outside investors suggesting that firms use debt only when the

internal financing is not available (Myers and Majluf 1984). Agency Cost Theory

predicts the capital structure choice is based on the existence of agency cost. This theory

investigates the relationship between the manager of the firm, and the outside equity and

debt holders (Jensen and Meckling 1976).

1
Starting with Modigliani and Miller (1958), the literature on capital structure has been

expanded by many theoretical and empirical contributions. For non-financial firms the

empirical literature has generally converged on particular variables that have been found

to be consistently correlated with leverage such as: age, size, growth, profitability,

market-to-book ratio, collateral value and dividend policy. On the other hand, the capital

structure of banks is still a relatively under-explored area in the banking literature.

Currently, there is no clear understanding on how banks choose their capital structure and

what factors influence their corporate financing behavior (Amidu 2007).

In Ethiopia as to the knowledge of the researcher there were few studies which relate

with this title these are, Ashenafi (2005) a case study in Addis Ababa Small and Medium

enterprises, Amanuel (2011) evidence from manufacturing share companies of Addis

Ababa city and Bayeh (2011) evidence from Ethiopian insurance company. Therefore,

given the unique financial features of banks and the environment in which they operate,

there are strong grounds for a separate study on capital structure determinants of banks in

Ethiopia.

Therefore, the main purpose of this study was to examine the relationship between

leverage and determinants of capital structure decision of banks in Ethiopia. This will

equip financial managers with applied knowledge of determining their capital structure,

and play role in filling gap in understanding of the capital structure decision.

The remainder of this chapter is organized as follows. Section 1.1 presents the statement

of the problem. Section 1.2 presents objective, research questions and hypothesis of the

study. Section 1.3 presents research methodology used. Section 1.4 presents the scope of

2
the study. Section 1.5 presents significance of the study. Finally, Section 1.6 presents

organization of the study.

1.1. Statement of the problem

While the choice of capital structure is one of the most important strategic financial

decisions of firms, it has been the subject of considerable debate and investigation. The

debate on what drives capital structure decisions is still open. Following the seminal work

of Modigliani and Miller (1958), a vast theoretical literature developed, which led to the

formulation of alternative theories, such as the static trade off model, pecking order

theory and agency cost theory. These theories point to a number of specific factors that

may affect the capital structure of firms such as (profitability, size, tangibility, growth,

risk, liquidity, age, dividend payout). However, the empirical evidence regarding the

alternative theories is still questionable (Rajan and Zingales 1995). For example, Static

trade off-theory assumes a firm’s optimal debt ratio is determined by a trade-off between

the bankruptcy cost and tax advantage of borrowing, holding the firm’s assets and

investment plans constant. According to this theory, higher profitability lower the

expected cost of distress, therefore, firms increase their leverage to take advantage from

tax benefits. Which means in other word profitability is positively related with leverage.

As well agency theory supports this positive relation because of the free cash flow theory

of Jensen (1986). But, pecking order theory Myers and Majluf (1984) throws doubt on

the existence of target capital structure, suggesting that firms use debt only when the

internal financing is not available. For this reason profitability is expected to have

negative relation with leverage.

3
The determinants of capital structure have been debated for many years and still represent

one of the most unsolved issues in corporate finance literature. Indeed, what makes the

capital structure debates so exciting is that only a few of the developed theories have

been tested by empirical studies and the theories themselves lead to different, not

mutually exclusive and sometimes opposed result and conclusion (Rajan and Zingales

1995). Morri and Beretta (2008) explained many theoretical studies and much empirical

research have addressed those issues, but there is not yet a fully supported and commonly

accepted theory; and the debate on the significance of determinant factors is still

unfolded.

Besides, although earlier studies have tremendous contributions to the theory of capital

structure, they were limited to developed financial system and restricted to non-banks.

Less developed countries like, Ethiopia, received little attention in the literature.

According to Octavia and Brown (2008) the capital structure of banks are still a relatively

under-explored area in the banking literature and the special nature of the deposit

contract, the degree of leverage in banking and the regulatory constraints imposed on

banks have meant that banks (and financial institutions in general) have been excluded in

previous empirical studies on standard capital structure choice. Nevertheless,

understanding the determinants of capital structure is as important for banks as for non-

banks firms. According to Amidu (2007) currently, there is no clear understanding on

how banks choose their capital structure and what factors influence their corporate

financing behavior. Thus, the lack of agreement about what would qualify as optimal

capital structure and lack of literature in the case of Ethiopia has motivated this study.

4
Therefore, this study tried to find out the relationship between leverage and firm specific

determinants of capital structure decision.

1.2. Objective, research questions and hypotheses

The main objective of this study was to examine the relationship between leverage and

firm specific (profitability, tangibility, growth, risk, size and liquidity) determinants of

capital structure decision and to understand about the theories of capital structure that can

explain the capital structure of the Ethiopian banking industry.

Based on the broad research objective, the following research questions and hypotheses

were developed.

Research questions (RQ)

RQ1. What determine the capital structure of banks in Ethiopia?

RQ2. Which theory explians the financing behavior adopted by Ethiopian

banking industry?

Hypotheses (HP)

To achieve the objective of this study, in addition to the research questions presented

above six hypotheses concerning the determinants of capital structure choice on the

Ethiopian banking industry were tested

Profitability:

Capital structure theories have different views on the relationship between leverage and

profitability. The trade-off theory argues that firms generally prefer debt for tax

5
considerations. Profitable firms would, therefore, employ more debt because increased

leverage would increase the value of their debt tax shield (Myers 1984).

In addition to the tax advantage of debt, agency and bankruptcy costs may encourage

highly profitable firms to have more debt in their capital structure. This is because highly

profitable firms are less likely to be subject to bankruptcy risk because of their increased

ability to meet debt repayment obligations. Thus, they will demand more debt to

maximize their tax shield at more attractive costs of debt. For these considerations, the

trade-off theory predicts a positive Relationship between leverage and profitability.

However, the pecking order theory of Myers and Majluf (1984) predicts the opposite. It

predicts a negative association between leverage and profitability because high profitable

firms will be able to generate more funds through retained earnings and then have less

leverage. Therefore, it is expected that there is negative relationship between profitability

and leverage ratio.

Hypothesis 1: There is a negative relationship between leverage ratios and profitability.

Growth:

According to pecking order theory firms with high growth will tend to look to external

funds to finance the growth. Myers (1977) confirms this and concludes that firms with a

higher proportion of their market value accounted for by growth opportunity will have

debt capacity. Therefore, it is expected that there is a positive relationship between

growth and leverage ratio

Hypothesis 2: There is a positive relationship between leverage ratios and growth.

6
 Tangibility:

Tangibility is an important determinant of the capital structure of a firm. The trade-off

theory predicts a positive relation between tangibility and debt levels. As the value of

intangible assets disappears (almost entirely) in the cases of bankruptcies, the presence of

tangible assets is expected to be important in external borrowing as it is easy to

collateralize them. Tangible assets often reduce the costs of financial distress because

they tend to have higher liquidation value (Titman and Wessels 1988; Harris and Raviv

1991). Pecking order theory of Myers and Majluf, (1984) conclude that issuing debt

secured by property, avoids the costs associated with issuing shares. This suggests that

firms with more collateralized assets (fixed assets) will be able to issue more debt at an

attractive rate as debt may be more readily available. This results in a positive association

between leverage and tangibility. Therefore, it is expected that there is a positive

relationship between tangibility and leverage ratio.

Hypothesis 3: There is a positive relationship between leverage ratios and tangibility.

Risk

Given agency and bankruptcy costs, there are incentives for the firm not to utilize the tax

benefit of debt within the static framework model. Firms with high earnings volatility

face a risk of the earnings level dropping below their debt servicing commitments,

thereby incurring a higher cost of financial distress. Accordingly, these firms should

reduce their leverage level to avoid the risk of bankruptcy. Therefore, the trade-off theory

predicts a negative relationship between leverage and earning volatility of a firm’s. The

pecking order theory allows the same prediction. Empirical evidence suggests that there

7
is a negative relationship between risk and leverage (Titman and Wessels, 1988). Hence,

risk is expected to have negative impact on leverage ratio.

Hypothesis 4: There is a negative relationship between leverage ratios and risk.

Size

According to trade-off theory, firm size could be an inverse proxy for the probability of

the bankruptcy costs. Larger firms are likely to be more diversified and fail less often.

They can lower costs (relative to firm value) in the occasion of bankruptcy. Larger firms

are more likely to have higher debt capacity and are expected to borrow more to

maximize the tax benefit from debt because of diversification (Titman and Wessels

(1988). Therefore, size has a positive effect on leverage. Size can be regarded as a proxy

for information asymmetry between managers and outside investors. Large firms are

subject to more news than small firms because the investment community would be more

concerned with gathering and providing information about large firms. This makes large

firms more closely observed by analysts and less subject to information asymmetry than

small firms. Thus, they should be more capable of issuing equity which is more sensitive

to information asymmetry and have lower debt (Rajan and Zingales, 1995). This suggests

that pecking order theory predicts a negative association between leverage and the size of

firm.

Hypotheses 5: There is a positive relationship between leverage ratios and size.

Liquidity

There are two different opinions on the association between liquidity and capital

structure: First view implies a positive significant relation that is consistent with trade off

8
theory. Companies with more liquidity (more current assets) tend to use more external

borrowing, because of their ability in paying off their liabilities. Second view points to a

negative significant relation that is consistent with the pecking order theory, arguing that

companies with more liquidity will decrease external financing, relying on their internal

funds. Thus, liquidity ratios may have a mixed effect on the capital structure decisions.

Most of the previous studies, confirm the negative relation, (Ahmed et al., 2010, and

Najjar and Petrov 2011). Hence, liquidity is expected to have negative impact on leverage

ratio

Hypothesis 6: There is a negative relationship between leverage ratios and liquidity.

1.3. Research methodology

In order to achieve the objective stated in the preceding section, considering the nature of

the problem and the research perspective this study used mixed research approach. A

mixed methods approach was chosen as it increases the likelihood that research generates

more accurate results than is the case if a single method had been adopted. As noted in

Creswell (2009) mixed research is an approach that combines or associates both

qualitative and quantitative research methods. It is also more than simply collecting and

analyzing both kinds of data, it involves the use of both approaches in tandem so that the

overall strength of a study is greater than either qualitative or quantitative research. As a

result, mixed methods provide a more accurate picture of the phenomena being

investigated.

The method adopted consists of structured document reviews and in-depth interviews to

collect the necessary data. Accordingly, the data related to a documentary analysis which

9
is necessary to undertake this study were gathered from the financial statements of eight

banks and NBE for twelve consecutive years (2000-2011) and the data was the audited

financial statements particularly balance sheet and income statement. Beside, in-depth

interview with five finance managers of the selected banks were utilized to gain a greater

insight into the findings from documentary analysis. Finally, the study analyses the

results obtained from the above mentioned data sources using both descriptive as well as

inferential statistics.

1.4. Scope and limitation of the study

The scope of this study was limited to the relationship between leverage and determinants

of capital structure decision of Ethiopian banks over the period 2000 to 2011. To this end,

this study was limited to firm specific determinant of capital structure (profitability,

tangibility, growth, risk, size and liquidity) and theories of capital structure that can

explain the capital structure of Ethiopian banking industry. The major limitations that

hamper the study were resource constraint and unavailability of active secondary market

which forced the researcher to measure the dependent variable i.e. measures of leverage

as well as the proxies of the independent variables in terms of book values rather than

market values.

1.5. Significance of the study

Some studies investigated the determinants of capital structure of firms in Ethiopia.

However, to the best knowledge of the researcher no single study has focused on the

banking industry of Ethiopia. Thus, this study will have significant role to play in filling

gap in understanding of the capital structure decision for banks in Ethiopia. Such an

10
understanding is important, because it equips financial managers with applied knowledge

of determining their capital structure. As an appropriate capital structure is important to a

firm as it will help in dealing with competitive environment within which the firm

operates, and which will maximize the return of the stockholders by increasing the value

of the firm. Additionally, this study will be used as an input to researchers for further

research on determinant of capital structure.

1.6. Organization of the thesis

This study is organized into five chapters. Chapter one presents research introduction,

statement of the problem, objective of the study, research question and hypothesis, scope and

limitation, and significance of the study. Following on this, chapter two of the study presents

review of theoretical and empirical literatures on determinants of capital structure. Chapter

three presents the research methodology. Then, chapter four present results and analysis of the

study and finally, chapter five present conclusions and possible recommendations.

11
 Chapter Two: Review of related literature

Capital structure refers to several alternatives that could be adopted by a firm to get the

necessary funds for its investing activities in a way that is consistent with its priorities.

Two major sources of financing that are available to firms are debt and equity. The

mixture of debt and equity is called capital structure. Most of the effort of the financial

decision making process is centered on the determination of the optimal capital structure;

where the firms’ value is maximized and cost of capital is minimized. This chapter

presents the theoretical and empirical literature review over the capital structure theme.

Section 2.1 covers theoretical literature review, section 2.2 covers reviews of prior

empirical studies including those conducted in Ethiopia and section 2.3 provides

conclusions and knowledge gap.

2.1. Theoretical review

The literature shows the existence of different theories related to capital structure. These

theories include Modigliani and miller (MM), static trade-off theory, pecking order

theory, and agency cost theory. The purpose of this section is, hence, to review these

theories of capital structure in an orderly.

2.1.1. Modigliani and Miller (MM) theory

Modigliani and Miller (1958) argued that capital structure is irrelevant to the value of a

firm under perfect capital market conditions with no corporate tax and no bankruptcy

cost. This implies that the firm’s debt to equity ratio does not influence its cost of capital.

A firm’s value is only determined by its real asset, and it cannot be changed by pure

12
capital structure management. Consequently, it means that there is no optimal capital

structure.

However, there is a fundamental difference between debt financing and equity financing

in the real world with corporate taxes. Dividends paid to shareholders come from the

after tax profit. By contrast, interest paid to bondholders comes out of the before-tax

profits. Thus, Miller and Modigliani (1963) argued that in the presence of corporate

taxes, a value-maximizing company can obtain an optimal capital structure. In other

words, if the market is not perfect, as result of, say, the existence of taxes, or of

underdeveloped financial markets, or of inefficient case, firms must consider the costs

entailed by these imperfections. A proper decision on capital structure can be helpful to

minimize these costs.

2.1.2. Static Trade-off Theory

Trade-off theory claimed that a firm’s optimal debt ratio is determined by a trade-off

between the bankruptcy cost and tax advantage of borrowing, holding the firm’s assets

and investment plans constant (Myers, 1984). The goal is to maximize the firm value for

that reason debt and equity are used as substitutes. According to this theory, higher

profitability decreases the expected costs of distress and let firms increase their tax

benefits by raising leverage; therefore, firms should prefer debt financing because of the

tax benefit. As per this theory firms can borrow up to the point where the tax benefit from

an extra dollar in debt is exactly equal to the cost that comes from the increased

probability of financial distress (Ross, 2002, p.586).

13
Due to the net tax advantage to corporate debt financing, the firm’s optimal capital

structure will involve distinctions in firm-specific characteristics, target leverage ratios

will vary from company to company. Institutional differences, such as different financial

systems, tax rate and bankruptcy law etc, will also lead the target ratio to differ across

countries. The trade-off theory predicts that safe firms, firms with more tangible assets

and more taxable income to shield should have high debt ratios. While risky firms, firms

with more intangible assets that the value will disappear in case of liquidation, ought to

rely more on equity financing. In terms of profitability, trade-off theory predicts that

more profitable firms should mean more debt-serving capacity and more taxable income

to shield, Therefore, based on this theory, firms would prefer debt over equity until the

point where the probability of financial distress starts to be important. This is illustrated

by figure 2.1

Figure 2.1: The-static-tradeoff theory of capital structure

Market value of
Firm

Cost of financial distress

D
PV of interest tax shield C
A B
Firm value under all equity financing

Debt
Source: Myers (1984) Optimum

14
In figure 2.1 the straight line AB shows the market value of the firm under the Modigliani

and Miller (1958) regime, in which the value of the firm is irrelevant and the capital

structure is equal to the value of an all-equity firm. If a firm uses debt in their capital

structure they have to pay interest which is generally tax deductible. Interest payments act

as a tax shield and allow the firm to increase its value. As the firm takes more debt its

value increases (curve AC). However, after a certain proportion of debt (the optimum

level) the value of the firm starts to decrease as the costs of debt start to outweigh the

benefits of debt. Curve AD illustrates how the costs of financial distress rise as firms use

increasing amounts of debt in their capital structure. At higher levels of debt the interest

payments of firms increase to cover for the potential risk of financial distress. Firms

trade-off the tax benefits that may be gained through using debt with costs of financial

distress and agency costs to maintain an optimal level of debt in their capital structure as

shown in figure 2.1.

The general results of various work in this aspect of leverage choice is that if there are

significant leverage-related costs, such as bankruptcy costs, agency costs of debt, and loss

of non-debt tax shields, and if the income from equity is untaxed, then the marginal

bondholder’s tax rate will be less than the corporate rate and there will be a positive

trade-off between the tax advantage of debt and various leverage-related costs.

2.1.3. Pecking order theory

Pecking Order Theory is developed by Myers and Majluf (1984) which stated that capital

structure is driven by firm's desire to finance new investments, first internally, then with

low-risk debt, and finally if all fails, with equity. Therefore, the firms prefer internal

15
financing to external financing. The pecking order theory discussed the relationship

between asymmetric information and investment and financing decisions. According to

this theory, informational asymmetry, which firm’s managers or insiders have inside

information about the firm‘s returns or investment opportunities, increases the leverage of

the firm with the same extent. So due to the asymmetric information and signaling

problems associated with external financing, the financing choices of firms follow an

order, with a preference for internal over external finance and for debt over equity.

Myers and Majluf (1984) argued that the capital structure can help to mitigate

inefficiencies in a firm’s investment decision that are caused by information asymmetries.

They demonstrate that if there is an asymmetry of information between investors and

firm insiders, then the firm’s equity may be underpriced by the market. As a result, new

equity, which is used to finance new investment projects, will be also under-priced.

Therefore, if management has favorable inside information and acts in the best interest of

the existing shareholders, then management will refuse to issue equity even if it means

passing up positive net present value projects because the net loss to existing

shareholders (due to under-pricing problem) might outweigh the project’s Net present

values. On the other hand, passing up positive net present value projects is contrary to the

wealth maximization. Using financial sources that may not be undervalued by the market,

particularly internally generated funds could solve this under-investment problem.

Accordingly, the existence of sufficient internal finance allows firms to accept desirable

investments without relying on costly external finance.

16
Myers and Majluf (1984), also argued that firms are most likely to generate financial

slack (i.e. liquid assets such as cash and marketable securities) to be used for internal

funding. Thus, in order to protect present shareholders, firms with financial slack and in

the presence of asymmetric information, will not issue equity, even though it may involve

passing up a good investment opportunity. If investors realize this point, then the market

will take the decision not to issue shares as good news. On the other hand, if management

does offer a new share issue, it will be interpreted as a bad news, and the firm’s share

issue will be under-priced. This adverse selection problem has an influence on the choice

between internal and external financing. This choice leads to the Pecking Order theory,

which Myers (1984) summarized as follows: Firms prefer internal finance. Firms adjust

their target dividend payout ratios to their investment opportunities, although dividends

are sticky and target payout ratios are gradually adjusted to shifts in available investment

opportunities. Sticky dividend policies as well as unpredictable fluctuations in both

profitability and investment opportunities mean that internally generated funds are more

or less than investment outlays. If internally generated cash flow is less than investment

outlays, the firm first exhausts its cash balances or marketable securities portfolio. If

external financing is required, firms will resort to the safest security first. They start with

debt, then hybrid securities such as convertible bonds and finally equity as a last resort. A

single optimal or target debt-equity ratio does not exist in the pecking order theory since

financing decision does not rely on the trade- off between marginal benefits and costs of

debt.

Moreover, Myers (1984) introduced implication similar to the pecking order theory

known as the modified pecking order theory. In this framework, both asymmetric

17
information and costs of financial distress are incorporated. Myers argued that as firm

climbs up the pecking order it faces higher probability of both incurring costs of financial

distress and passing up future positive net present value projects. Thus, firm may

rationally decide to reduce these costs by issuing stock now though new equity is not

needed immediately to finance real investment, just to obtain financial slack and move

the firm down the pecking order. Therefore, when issuing new capital, those costs are

very high, but for internal funds, costs can be considered as none. For debt, the costs are

in an intermediate position between equity and internal funds. Therefore, firms prefer

first internal financing (retained earnings), then debt and they choose equity as a last

option.

2.1.4. Agency cost theory

Agency theory focused on the costs which are created due to conflicts of interest between

shareholders, managers and debt holders. According to Jensen and Meckling (1976),

capital structures are determined by agency costs, which includes the costs for both debt

and equity issue. The costs related to equity issue may include:

 The monitoring expenses of the principal (the equity holders)

 The bonding expenses of the agent (the manager)

 Reduced welfare for principal due to the divergence of agent’s decisions

from those which maximize the welfare of the principal.

Besides, debt issue increases the owner-manager’s incentive to invest in high-risk

projects that yield high returns to the owner-manager but increase the likelihood of

failure that the debt holders have to share if it is realized. If debt holders anticipate this,

higher premium will be required which in turns increase the costs of debt. Then, the

18
agency costs of debt include the opportunity costs caused by the impact of debt on the

investment decisions of the firm; the monitoring and bond expenditures by both the

bondholders and the owner-manager; and the costs associated with bankruptcy and

reorganization. Since both equity and debt incur agency costs, the optimal debt-equity

ratio involves a trade-off between the two types of cost.

Jensen and Meckling (1976) introduced two types of conflicts that are a major source of

agency costs and these are: agency costs that arise due to the conflicts of interest between

managers and shareholders and agency costs that arise as a result of the conflicts of

interest between shareholders and debt holders. The subsequent discussions present

shareholders-managers conflicts and shareholder-bondholder conflicts in an orderly

manner.

2.1.4.1. Shareholders-managers conflicts

This kind of conflict stems from the separation of ownership and control. If managers do

not own 100% of the firm, they can only capture a fraction of the gain earned from their

value enhancement activities but they need to bear the entire costs of these activities. The

shareholders-managers conflicts take several distinct forms:

 According to Jensen and Meckling (1976) managers prefer to make use of less

effort and have greater perquisite levels, such as luxuriant office and corporate

jets, different from the shareholder’s interest of firm value maximization. In

this case, increasing the managers’ equity holdings will help to align the

interests of shareholders and managers. Or, keeping managers equity

investment constant, increasing the debt level also helps to mitigate the loss of

19
 conflicts between shareholders and managers. Since debt forces managers to

pay out cash, reducing the free cash flow managers can waste on the

perquisites.

 According to Masulis (1988) conflict may arise because managers may prefer

short-term projects, which produce results early and enhance their reputation

quickly, rather than more profitable long-term projects.

 According to Harris and Raviv (1991) managers want to stay in their positions, so

they wish to minimize the likelihood of employment termination. As this

increases with changes in corporate control, management may resist

takeovers, irrespective of their effect on shareholder value. On operating

decisions, managers and shareholders may also have different preferences:

Harris and Raviv (1991) observed that managers will typically wish to

continue operating the firm even if liquidation is preferred by shareholders.

A special case of the conflicts between shareholders and managers is the over investment

problem. Jensen (1986) argued that, instead of working under shareholders interests to

maximize firm’s value, managers prefer to increase firm’s size to enjoy the benefit of

control. In this case, managers have incentives to cause their firm to grow beyond the

optimal size and accept negative net present value (NPV) projects. Jensen (1986) argued

that the overinvestment problem can be motivated by more free cash flow and less

growth opportunities. Issuing debt helps to mitigate agency problems that arise from

managerial behavior under divergent interests between shareholders and managers. For

example, the overinvestment problem can be mitigated by issuing debt since debt

obligates firm to pay out cash so prevents managers from investing in negative NPV

20
projects. Jensen (1986) refers to the non-discretionary nature of debt as the disciplining

role of debt. As Hunsaker (1999) pointed out, an increase in debt also increases the risk

of bankruptcy, therefore limits management’s consumption of perquisites. Besides, issue

convertible debt also helps to discipline managers’ behavior because they give managers

a chance to share in a firm’s profits in case of good performance and thus reduces the

monitoring costs.

2.1.4.2. Shareholder-bondholder conflicts

The typical phenomenon of these conflicts is that the shareholders or their representatives

make decisions transferring wealth from bondholders to shareholders. Certainly, the

bondholders are aware of the situations in which this wealth expropriation may occur,

therefore, will demand a higher return on their bonds or debts. Different fundamental

sources of equity-holders and debt-holders conflicts have been identified in the agency

cost literature;

 The direct wealth-transfer from bondholders to shareholders (Smith and Warner

1979): Shareholders can increase their wealth at the expense of bondholders’

interests by increasing the dividend payment; the issuance of debt with higher

priority will expropriate wealth from current bondholders.

 Asset-substitution is another source of the conflicts (Jensen and Meckling 1976):

When signing debt contracts, bondholders demand an interest rate according to

the riskiness of the firm’s investment activities. While debt contracts gives

shareholders an incentive to invest in risky projects because if it succeeds the

returns above the face value of debt will be owned by shareholders and in case of

21
 failure, the consequence is mainly born by bondholders because of shareholders’

limited liability. This excessive return from risky projects makes safe projects

less attractive to shareholders since returns from the safe projects are sufficient to

pay the bondholders. If bondholders can anticipate shareholders incentive of

substituting safe projects by risky projects, they will ask for a higher risk

premium. Also the anticipation of wealth expropriation will lead to the increase

in risk premium. The increased costs of debt are then born by shareholders since

they are residual claimants of the firm.

 Underinvestment problem is another agency problem results in shareholder-

bondholder conflicts Myers (1977): Underinvestment problem mostly incurs in

financial distress. The extension of debt decreases the shareholders incentives to

invest in new projects (even the projects with high growth opportunities will be

passed through) because the profits from these projects will be exhausted in debt

repayment.

One way to minimize these conflicts is that firms with high growth opportunities should

have lower leverage. The conflicts can also be mitigated by adjusting the properties of the

debt contracts, for example, the adjustment can be done by including covenants such as

adding limits on the dividends payment or setting restrictions on the disposition of assets.

Alternatively, debt can be secured by collateralization of tangible assets in the debt

contracts.

Determinants of capital structure

As shown in chapter one theoretically there are a large number of potential factors that may

have an impact on leverage ratio. These factors include size of the firm, tangibility,

22
profitability, risk, growth, and liquidity. However, there is a significant disagreement

among the capital structure theories, in particular, between the trade-off and the pecking

order theories about the influence of some factors on the firm’s capital structure, hence,

the issue remains as question to be addressed.

2.2. Empirical studies on the determinants of capital structure

Since the pioneering work of Modigliani and Miller (1958), the question of what

determines firms’ choices of capital structure has been a major field in the corporate

finance literature. Since then, several studies have been conducted in developing and

developed countries to identify those factors that have an effect on firms’ choice of

capital structure. Given the time constraint and the amount of empirical literature

available on the topic of this research it would have been quite difficult to present the

results of all the studies. Thus, the review of the empirical studies in this section on the

determinants of capital structure decision has a particular focus on those that have been

conducted since the 1988s.

Titman and Wessels (1988) studied the determinant of capital structure choice by

examining them empirically. They extended empirical work on capital structure theory in

three ways. First, they examined a much broader set of capital structure theories, many of

which have not previously been analyzed empirically. Second, since the theories have

different empirical implications in regard to different types of debt instruments, the

authors analyzed measures of short-term, long-term, and convertible debt rather than an

aggregate measure of total debt. Third, they used a factor-analytic technique that

mitigates the measurement problems encountered when working with proxy variables.

23
Titman and Wessels (1988) specifically tested how asset structure, non-debt tax shields,

growth, uniqueness, industry classification, firm size, earnings volatility and profitability

can affect the firm’s debt-equity choice. Their results indicated that debt levels are

negatively related to the uniqueness of a firm’s line of business. The short-term debt ratio

was negatively related to firm size. Besides that, a strong negative relationship was noted

between debt ratios and past profitability which is consistence with pecking order theory

Myers and Majluf (1984). However, they did not provide strong empirical support on

variables like non-debt tax shields, volatility, collateral value and future growth.

In a comparative study, Rajan and Zingles (1995) investigated whether the capital

structure in other developed countries is related to factors similar to those influencing the

US companies for the period of 1987-1991. Tangible assets, market to book ratio, firm

size and profitability are suggested as determinants of capital structure in these countries.

They find that firms with more collateralized assets are not highly levered. In addition,

they found that profitability and market to book ratio are negatively related to leverage.

However, they argue that the negative relationship with leverage appeared to be driven by

firms with high market to book ratio rather than by firms with low market to book ratio.

The study provides no evidence supporting the effect of the firm size on leverage.

Finally, the findings were not varied across the G-7 countries so they concluded that

capital structure in other countries was affected by factors similar to those that influence

the US companies.

24
Booth et al. (2001) assessed whether capital structure theory is portable across

developing countries with different institutional structures. The sample firms in their

study are from Malaysia, Zimbabwe, Mexico, Brazil, Turkey, Jordan, India, Pakistan,

Thailand, and Korea. Booth et al. (2001) use three measure of debt ratio; total debt ratio,

long-term book debt ratio, and long-term market debt ratio with average tax rate, assets

tangibility, business risk, size, profitability, and the market to book ratio as explanatory

variables. The study showed that the more profitable the firm, the lower the debt ratio,

regardless of how the debt ratio was defined. It also showed that the more the tangible

assets, the higher the long-term debt ratio but the smaller the total debt ratio.

Booth et al. (2001) concluded that the debt ratio in developing countries seemed to be

affected in the same way by the same types of variables that were significant in

developed countries. However, they pointed out that the long-term debt ratios of those

countries are considerably lower than those of developed countries. This finding may

indicate that the agency costs of debt are significantly large in developing countries or

markets for long term debt are not effectively functioning in these countries. Finally,

Booth et al. (2001) argued that their results are in line with Rajan and Zingales (1995)

except for the tax and the market-to-book ratio.

Bevan and Danbolt (2002) who extended the work of Rajan and Zingales (1995) tested

the determinants of capital structure in the United Kingdom non-financial firms by using

four measures of financial leverage. They used non-equity liabilities to total assets, total

debt to total assets, total debt to capital (where capital is defined as total debt plus

25
common shares with preferred shares), and adjusted debt to adjusted capital. All the

measures were regressed on market-to-book value, natural logarithm of sales (size),

profitability, and tangibility of assets. They found that determinants of capital structure

were significantly changed with respect to each measure of debt used. With the same

leverage definition as Rajan and Zingales, Bevan and Danbolt (2002) reported similar

results.

In their later paper, Bevan and Danbolt (2004) analyzed the determinants of capital

structure of 1054 UK Companies from the period 1991-1997. Secondly, they also

investigated the extent to which the influence of these determinants is affected by time-

invariant and firm specific heterogeneity. Bevan and Danbolt (2004) as Bevan and

Danbolt (2002) use market-to book value, natural logarithm of sales (size), profitability,

and tangibility of assets as determinants of capital structure. In addition to the time

invariant and firm specific heterogeneity, the focus was on the variety of long - run and

short run debts components rather than on the aggregate measures. They found that large

firms use long and short term debt more than small ones. Tangibility is found to be

positively related to both short and long-term debt, while profitability is found to be

negatively related. However, they find that profitable firms tend to use short-term debt

more than less profitable one.

The paper of Deesomsak et al. (2004) investigated the determinants of capital structure of

firms operating in the Asia Pacific region, in four countries with different legal, financial

and institutional environments, namely Thailand, Malaysia, Singapore and Australia.

26
OLS estimation model was used to analyze sample data included 294 Thailand, 669

Malaysian, 345 Singaporean, and 219 Australian firms for the period 1993-2001. Overall

they found leverage to be positively related to firm size and growth opportunities, non-

debt tax shields, liquidity to be negatively related to leverage. Moreover, they also found

that capital structure decision is not only the product of the firm’s own characteristics but

is also affected by the specific corporate governance, legal structure and institutional

environment of the countries.

The paper of Huang and Song (2005) employed regression model to document the

determinants of capital structure of Chinese listed companies. The data included market

and accounting figures of more than 1200 companies for the time period 1994-2003.

They find that leverage (long-term debt ratio, total debt ratio, and total liability ratio)

decreases with profitability, non-debt tax shield and managerial shareholdings, while it

increases with firm size and tangibility. In addition, the tax rate positively affects long-

term debt ratio and total debt ratio. Furthermore, they find a negative relationship

between leverage and firm growth opportunities.

Buferna et al. (2005) provided further evidence of the capital structure theories pertaining

to a developing country and examined the impact of lack of a secondary capital market

by analyzing a capital structure question with reference to the Libyan business

environment. They developed four explanatory variables that represent profitability,

growth, tangibility and size to test which capital structure theories best explained Libyan

companies’ capital structure. The results of cross-sectional OLS regression showed that

27
both the static trade-off theory and the agency cost theory are pertinent theories to the

Libyan companies’ capital structure whereas there was little evidence to support the

asymmetric information theory. The lack of a secondary market may have an impact on

agency costs, as shareholders who are unable to offload their shares might exert pressure

on management to act in their best interests.

A study made by Amidu (2007) examined the determinants of capital structure of

Ghanaian banks by employing panel regression model. Amidu (2007) has highlighted the

importance of distinguishing between long and short forms of debt while he made

inferences about capital structure. Amidu (2007) specifically tested the significance of

bank size, profitability, corporate tax, growth, asset structure, and risk in determining

bank capital structure. The result showed that short-term debt of banks is negatively

related to banks profitability, risk, and asset structure and positively related to bank size,

growth and corporate tax. On the other hand, the long-term debt of the banks is positively

related to banks’ asset structure and profitability and inversely related to bank risk,

growth, size and corporate tax. Generally, the variables examined were consistent with

the static trade-off and pecking order arguments, with the only exception being risk.

Gropp and Heider (2009) approached the issue from a different perspective. Using a

sample of banks from developed countries, they specifically tested the significance of

size, profitability, market-to-book ratio, asset tangibility, and dividend paying status in

determining bank leverage. In the process, they made a stark distinction between bank

book and market leverage as well as controlled for asset risk and macroeconomic factors.

28
They further examined whether asset risk captures the effect of risk adjustments on the

minimum capital required or it rather represents factors pertaining to the standard capital

structure theories. Overall, their results provided strong support for the relevance of

standard determinants of capital structure on bank capital.

Caglayan and Sak (2010) studied the determinants of capital structure of banks in

Turkish. The objective of the study was to examine the relationship between the leverage

level and a set of explanatory variables (Tangibility, Size, Profitability, and Market to

book ratio) by using panel data analysis to establish the determinants of capital structure

of bank over the period 1992-2007. The main results of their study reveal that size and

market to book have positive and statistically significant impact on the book leverage

while the variables of tangibility and profitability have negative and significant impacts

on the book leverage. These findings strongly confirm the pecking order theory; except

the relationship with tangibility which weakly confirms the agency cost theory

Khrawish and Khraiwesh(2010) examined the determinants of capital structure; evidence

from Jordanian industrials companies over the period (2001- 2005). Using panel data,

Leverage ratio, Long-term debts/total Debts and five explanatory variables that represent

size, tangibility, profitability, long-term debt and short-term debt were calculated. Based

on the statistical analysis, they found that there was a significant positive relationship

between leverage ratio and size, tangibility, long-term debt and short-term debt and there

was a significant negative relationship between leverage ratio and Profitability of the

firm.

29
Ahmed et al., (2010) investigated the impact of firm level characteristics on capital

structure of life insurance companies of Pakistan over the period of seven year from 2001

to 2007. For this purpose, leverage was taken as dependent variable while profitability,

size, growth, age, risk, tangibility of assets and liquidity were selected as independent

variables. The result of OLS regression model indicated that profitability, liquidity, risk

and age have negative relationship with leverage which follows pecking order theory

while size indicated positive relationship with leverage which shows consistency with

trade- off theory. The results also indicated that leverage has statistically insignificant

relationship with growth and tangibility of assets.

Gurcharan (2010) analyzed the determinants of capital structure in four countries of the

ASEAN members, namely Malaysia, Indonesia, Philippine and Thailand, for the period

from 2003 to 2007 with a sample of 155 main listed companies from four selected

ASEAN stock exchange index-links. Based on the empirical result, he found that

profitability and growth opportunities for all selected ASEAN countries reveal statistical

significant with inverse relationship with leverage. Whereas non-debt tax shield has

significant negative impact on leverage mainly for Malaysia index link companies only.

Firm size shows a positive significant relationship for Indonesia and Philippine index link

companies. Also he found that country-effect factors; stock market capitalization and

GDP growth rate show significant relationship with leverage while bank size and

inflation indicate insignificant impacts on leverage.

30
Recently, Najjar and Petrov (2011) studied capital Structure of insurance companies in

Bahrain. The goal of the study was to investigate the effect of specific firm characteristics

on capital structure. They used panel data derived from annual reports and financial

statements of five insurance companies listed on the Bahrain Stock Exchange for the

period of 2005-2009 and apply multiple linear regression analysis using SPSS to identify

those effects. Najjar and Petrov identified a strong relationship between firm

characteristics, such as tangibility of assets, profitability, firm size, revenue growth, and

liquidity. The results of their study reveal that tangibility and size shows a positive

significant relationship with the debt ratio which confirms the static trade off theory

while liquidity shows negative significant relationship with debt ratio which confirms the

pecking order theory. But, profitability and revenue growth are not statistically

significant and require further research.

More recently, the paper of Olayinka (2011) examined the determinants of capital

structure of 66 firms listed on the Nigerian stock Exchange during the period 1999-2007

using panel data. He used six independent variables namely; sales growth, tangibility,

profitability, liquidity, size and business environment and leverage as dependent variable.

The OLS results showed that there is a negative relationship between leverage and

growth opportunities which is consistent with the prediction of the trade off theory. In the

same way, leverage is negatively related with tangibility and profitability which is

consistent with the predictions of pecking order theory but positively related to liquidity

as well as size which is consistent with trade-off theory. In general as per Olayinka, three

of the variables are not significant determinants of capital structure.

31
In the case of Ethiopia, there have been a few studies on determinants of capital structure.

These studies include Ashenafi (2005), Amanuel (2011) and Bayeh (2011). Ashenafi

(2005) approached the question of capital structure using data from medium firms in

Ethiopia. Ashenafi (2005) took variables like non-debt tax shield, economic risk, age of

firms, size of firms, tangibility, profitability and growth were regressed against leverage.

The results showed that non-debt tax shield, economic risk, profitability, growth,

tangibility, and age showed a negative coefficient of correlation with debt to equity ratio.

Amanuel (2011) studied determinants of capital structure of manufacturing share

companies in Addis Ababa, Ethiopia for the period over 2003-2010. The objective of the

study was to examine the relevance of theoretical internal (firm level) factors determine

capital structure of manufacturing share companies in Addis Ababa, Ethiopia. Amanuel

(2010) used seven explanatory variables: tangibility, non-tax shield, growth, earning

volatility, profitability, age and size, and three dependent variables: total debt ratio, short

term ratio and long term ratio to establish the determinants of capital structure of

manufacturing companies in Ethiopia. In connection to this, samples of 12 companies

were taken and secondary data was collected from audited financial statement of the

selected companies. The results of OLS regression showed that tangibility, non debt tax

shields, earning volatility, profitability, and size of the firm variables are the significant

determinants of capital structure of Addis Ababa manufacturing share companies at least

for one of the model out of the three models employed in the study. But no clear and

statistical proved relations were obtained for the variables growth of the firm and age of

the firm in any of the capital structure models.

32
Bayeh (2011) investigated empirically the determinants of capital structure in the case of

insurance industry in Ethiopia. In connection to this, nine insurance companies were

included in the sample for the period over 2004 - 2010. Bayeh (2011) also used seven

explanatory variables: liquidity, tangibility, growth, business risk, profitability, age and

size, and three dependent variables: total debt to equity ratio, total debt ratio and long

term ratio to test the determinants of capital structure of insurance companies in

Ethiopia. The OLS result showed that growth, profitability, and age of the firm were

found to have significant influence on capital structure of Ethiopian insurance companies.

However, tangibility and size of the firm were found to have significant influence on

capital structure of Ethiopian insurance companies.

2.3. Conclusion and knowledge gap

This chapter reviewed the literature on determinants of capital structure decision, starting

with the famous irrelevance theory of Modigliani and Miller (1958). However, by

relaxing the theory of Modigliani and Miller’s (1958) assumptions of perfect capital

markets, several theoretical frameworks have been developed to explain the firm’s capital

structure such as static trade-off theory, pecking order theory, and agency theory.

Static trade off-theory assumes that a firm’s optimal debt ratio is determined by a trade-

off between the bankruptcy cost and tax advantage of borrowing, holding the firm’s

assets and investment plans constant. Whereas, pecking order theory is another

dimension of the capital structure theories. According to this theory capital structure is

driven by firm’s desire to finance new investments, first internally, then with low-risk

33
debt, and finally if all fails with equity. Therefore, the firms prefer internal financing to

external financing.

Agency theory focuses on the costs which are created due to conflicts of interest between

shareholders, managers and debt holders. According to this theory capital structures are

determined by agency costs, which includes the costs for both debt and equity issue. This

shows that theories of capital structure have been resulting in different conclusions.

Similarly, the findings of prior empirical studies have provided varying evidence related

to the determinants of capital structure. For instance, Caglayan and Sak (2010) have

studied the determinants of capital structure of banks in Turkish and provided evidence

that pecking order theory is pertinent theory to Turkish banks. Beside, Buferna et al.

(2005) provided evidence that trade of theory and agency are pertinent theories of the

capital structure to a developing country. On the other hand, Amidu (2007) on Ghanaian

banks supports the static trade-off and pecking order argument. However, in the context

of Ethiopia as to the knowledge of the researcher there is no study conducted on the

capital structure of the banking industry. In addition, most empirical work on capital

structure has predominantly relied on quantitative analysis of secondary data to examine

the determinants of capital structure. Therefore, this study will fill the gap by examining

the determinants of capital structure in the context of Ethiopian banking industry.

34
 Chapter Three: Research Design and methodology

The preceding chapter tried to present the literature review along with the knowledge gap

that this study will be filling in. The purpose of this chapter is to discuss the research

methodology along with the detailed methods planned to be used in the study. The

chapter is organized in four sections. The first section 3.1 presents the research objective,

research questions and hypotheses; this is followed by the research approaches in section

3.2. Section 3.3 presents the methods adopted including the data collection tools and

methods of data analysis. Finally, conclusions and the relation between research

questions/ hypotheses, and the different data sources are presented in section 3.4.

3.1. Objective, Hypotheses and research questions

As shown in chapter one this study intended to examine the relationship between

leverage and bank specific (profitability, tangibility, growth, risk, size and liquidity)

determinants of capital structure decision and to understand about theories of capital

structure that can explain the capital structure of banks in Ethiopian. In order to achieve

the objective of the study, six hypotheses (HP) and two research questions (RQ) were

developed as shown below.

HP 1: There is a negative relationship between leverage ratios and profitability.

HP 2: There is a positive relationship between leverage ratios and growth.

HP 3: There is a positive relationship between leverage ratios and tangibility.

HP 4: There is a negative relationship between leverage ratios and risk.

HP 5: There is a positive relationship between leverage ratios and size.

35
HP 6: There is a negative relationship between leverage ratios and liquidity.

In addition, the following two research questions were developed.

RQ1. What determine the capital structure of banks in Ethiopia?

RQ2. Which theory explians the financing behavior adopted by Ethiopian

banking industry?

3.2. Research Approaches

As noted in Creswell (2009) in terms of investigative study there are three familiar types

of research approaches to business and social research namely, quantitative, qualitative

and mixed methods approach. Therefore, the following discussion briefly presents the

basic nature of quantitative, qualitative and mixed research approaches along with their

respective merits and demerits.

Quantitative research is a means for testing objective theories by examining the

relationship among variables (Creswell 2009, p.4). In quantitative research approach

there are two strategies of inquiries namely, survey design and experimental design. The

chief advantage of this approach is that numbers are easy to work with, data are readily

collected, coded, summarized and analyzed (Dunn 1999, p. 37). Further quantitative

research approach has the advantage of being able to make generalizations, for a broader

population, based on findings from the sample. Apart from of its advantages, as noted by

Dunn (1999) quantitative research approach has the following disadvantages. For

example, the sample selected may not represent the total population and the researchers

36
know much about the collective or average experience of research participants, but not

their individual experiences (Dunn 1999).

Qualitative research approach is one in which the investigator often makes knowledge

claims based primarily on the multiple meanings of individual experiences, socially and

historically constructed meanings, participation in issues, collaboration or change

oriented with an intent of developing a theory or pattern (Creswell 2003, p. 18). As noted

in Sarantakos (2005, p. 45 cited in Yesegat 2009, p. 73) qualitative research approach

uses strategies of inquiry such as narratives, ethnographies, grounded theory studies, or

case studies. The key advantage of qualitative research design is that it discloses the

richness of human experience (Lincoln and Guba 1985, cited in Dunn, 1999, p.37).

Moreover, qualitative research design has advantages like flexibility and emergent

without being constrained by standardized procedures (Liamputtong and Ezzy 2005, p.

204, cited in Yesegat 2009, p. 74). A part from the above mentioned advantages,

qualitative research design has also its own weaknesses. As noted in Dunn (1999) the

demerits of this approach includes; absence of quick response, difficulty, inefficiently,

and lack of generalization among others.

Mixed research is an approach to inquiry that combines or associates both qualitative and

quantitative forms (Creswell, 2009). As a major advantage, when the investigator uses

this approach he can learn more about the research problem (Leedy and Ormorod, 2005

cited in Semu 2010, p. 44). In connection to this, Greene et al. (1989, p. 256, cited in

yesegat, 2009, p. 75) also emphasized that as all methods have inherent biases and

limitations, so use of only one method to assess a given phenomenon will inevitably yield

biased and limited results. Besides, as an additional merit, the approach is not limited to

37
one method or the researcher is not committed to only one method which means the

investigator is flexible. Considering the research problem and objective along with the

philosophy of the different research approaches, mixed research approach was found to

be appropriate for this study. The following section hence presents the methods adopted

in the study.

3.3. Methods adopted

Research designs are plans and the procedures for research that span the decisions from

broad assumptions to detailed methods of data collection and analysis (Creswell 2009, p.

3). Therefore, in order to achieve the objective stated in the preceding section,

considering the nature of the problem and the research perspective this study used mixed

research approach. Accordingly, the quantitative method was mainly used to investigate

determinants of capital structure of banks in Ethiopia, and the financial data were

collected through structured survey of documents. Following this, the qualitative method

was used to support the quantitative findings and to gain additional insight into the

factors that may affect the capital structure of banks in Ethiopia. The qualitative data

were collected through in-depth interviews with the finance managers of Ethiopian

commercial banks.

A mixed methods approach was chosen as it increases the likelihood that the research

generates more accurate results than is the case if a single method had been adopted. As

noted in Creswell (2009) mixed research is an approach that combines or associates both

qualitative and quantitative research methods. It is also more than simply collecting and

analyzing both kinds of data, it involves the use of both approaches in tandem so that the

38
overall strength of a study is greater than either qualitative or quantitative research. As a

result, mixed methods provide a more accurate picture of the phenomena being

investigated. The subsequent discussions provide the quantitative aspect of the study, the

qualitative aspect, in-depth interviews in particular, and data analysis methods. .

3.3.1. Quantitative aspect of the study

The quantitative aspect of the research method intends to obtain data needed to generalize

about the determinants of capital structure of banks in Ethiopia. Specifically, the current

study employed a survey design that was administered through structured review of

documents. Hence, the next section presents the survey design.

3.3.1.1. Survey design: structured review of documents

The purpose of survey research is to generalize or makes claim from the sample to the

population so that inferences can be made about some characteristic, attitude or behavior

of the population (Creswell 2009). In connection to this, Leedy and Ormord ( 2005,

p.183 cited in semu. 2010, p. 45) also noted that survey research involves acquiring

information about one or more group of people perhaps about their characteristics,

opinions, attitudes, or previous experiences-by asking them questions and tabulating their

answers. Generally, the ultimate goal is to learn about a large population by surveying a

sample of that population.

The researcher selected survey design because surveys are relatively inexpensive and it

enables to gather enough information, which may not available from other sources.

Accordingly, the survey was carried out by means of a document review. The data related

to a documentary analysis which is necessary to undertake this study were gathered from

39
the financial statements of eight banks and NBE for twelve consecutive years (2000-

2011), and the data was the audited financial statements particularly balance sheet and

income statement. This was done to avoid the risk of distortion in the quality of data.

Sampling design

The population of the study was all commercial banks registered by NBE. Currently, as

per NBE (2009/10) annual report 15 banks are operating in Ethiopia. For this study,

twelve years data (2000- 2011) were considered. Therefore, those Banks which were

established after 2001 and started to provide financial statement in the succeeding fiscal

year were not included in this study because this study incorporated only banks that have

financial statements for the year, 2000, and onwards. Therefore, only eight banks

information were used in this study to examine the determinants of capital structure.

3.3.2. Qualitative aspect of the study

Qualitative research approach is a means for exploring and understanding individuals or

groups scribe to a social or human problem (Creswell, 2009). Qualitative research is

typically used to answer questions of complex phenomena on which data can be collected

using instruments like structured and unstructured interviews, group discussions,

observation and reflection field notes, various texts like reflexive Journals, pictures, and

analysis of documents and literature. Thus, in the current study to gather the qualitative

data needed for addressing the research questions stated in the preceding section, in-

depth interviews with finance managers of some selected banks were employed. The next

section presents the in-depth interviews.

40
3.3.2.1. In – depth interview

In-depth interviews with some selected Ethiopian commercial banks finance managers

were utilized to gain a greater insight into the findings from documentary analysis. The

interviews were conducted with five finance managers of Ethiopian commercial banks

namely, Construction and Business Bank, Commercial Bank of Ethiopia, United bank,

Bank of Abyssinia and Wegagen bank. The finance managers were chosen as they are

believed to be the most knowledgeable parties about the determinants of capital structure.

Beside, the respondents were contacted once and each respondent was contacted at

different times. This allowed the interview time to be utilized effectively and improved

the efficiency of the interview process by helping the interviewees’ to save time.

Generally, based on the objective and research questions presented in the preceding

section a number of unstructured interview questions were asked to better understand the

determinants of capital structure.

3.3.3. Data analysis method

Survey data collected through document review was analyzed statistically using both

descriptive and inferential statistics. Descriptive statistics of the variables and different

percentiles of the dependent variable were calculated over the sample period. In addition,

Correlation matrix was used to identify the relationship of each variable among them and

with dependent variable. Then, using statistical package ‘EVIEW 6’ OLS (ordinary least

squares) multiple regressions and t-statistic were carried out to test the relationship

between leverage and their potential determinants. Multiple regressions were also used to

determine the most significant and influential explanatory variables affecting the capital

41
structure of banks in Ethiopia. In connection to this, the general model for this study, as is

mostly found in the existing literature is represented by;

= α+ +

The subscript i representing the cross-sectional dimension and t denote the time-series

dimension. The left-hand variable , represents the dependent variable in the model,

which is the firm’s debt ratios. Contains the set of independent variables in the

estimation model, is taken to be constant overtime t and specific to the individual cross-

sectional unit i. If is taken to be the same across units, then OLS provides a consistent

and efficient estimate of .

Therefore, the model for this study, was based on the one used by Amidu (2007) with

some modification to explain the relationships between leverage and determinants of

capital structure as shown below.

= + ( )+ ( )+ ( )+ ( )+ ( )+ ɛ

Where:

PR = profitability

TA = tangibility

GR = growth

RS = risk

SZ = size

LQ = liquidity

42
Table 3.1: Variable-Indicator List

Variables Indicator Expected sign

Dependent Variable
Leverage Total debt/total asset NA
Independent Variables
Profitability Ratio of EBIT to total asset -
Tangibility Fixed Assets / Total Asset +
Growth Annual change in total asset +
Risk Standard Deviation of -
Operating Income
Size Natural Logarithm of Total +
Asset
Liquidity Liquid assets/ deposits -

As noted in Brooks (2008) the model should follow the classical linear regression model

(CLRM) assumptions, which were required to show that the estimation technique, OLS,

had a number of desirable properties, and also so that hypothesis tests regarding the

coefficient estimates could validly be conducted. In doing so, different test of the CLRM

assumptions were made. Among them the major ones are: test for heteroscedasticity,

autocorrelation, multicollinearity and normality. Accordingly, to perform a

heteroscedasticity test the popular white test was used. In this test, if the p-value is very

small, less than 0.05 the null hypothesis of the variance of the residuals is homogenous

must be rejected. As noted in brooks (2008) this test is the most popular because it makes

few assumptions about the likely form of the heteroscedasticity. Gujarati (2004) indicates

that Heteroskedasticity is a systematic pattern in the errors where the variances of the

errors are not constant. Similarly, Brooks (2008) noted that if the errors do not have a

constant variance, they are said to be heteroscedastic.

43
On the other hand, multicollinearity means that there is linear relationship between

explanatory variables which may cause the regression model biased (Gujarati, 2004).

When there is strong correlation between variables it becomes difficult to identify the

impact of individual independent variables. Thus, in order to examine the possible degree

of multicollinearity among the explanatory variables, correlation matrixes of the selected

explanatory variables was used. Usually the multicollinearity exists if the correlation

between two independent variables is more than 0.75 (Malhotra, 2007).

Beside, the researcher tested assumption of no serial correlation of the residuals. This

assumption implies that the errors associated with one observation are not correlated with

the errors of any other observation. For this purpose, Durbin-Watson (DW) measure was

used. According to Brooks (2008), DW has 2 critical values: an upper critical value and a

lower critical value, and there is also an intermediate region where the null hypothesis of

no autocorrelation can neither be rejected nor not rejected. The rejection, non-rejection,

and inconclusive regions are shown on the number line in figure 3.1. So, the null

hypothesis is rejected and the existence of positive autocorrelation presumed if DW is

less than the lower critical value; the null hypothesis is rejected and an existence of

negative autocorrelation is presumed if DW is greater than 4 minus the lower critical

value; the null hypothesis is not rejected and no significant residual autocorrelation is

presumed if DW is between the upper critical value and 4 minus the upper limits; the null

hypothesis is neither rejected nor not rejected if DW is between the lower and the upper

limits, and between 4 minus the upper and 4 minus the lower limits.

44
Figure: 3.1: Rejection and Non-Rejection Regions for DW Test

Reject H0: Do not reject Reject H0:

Positive H0: No evidence of Negative

Autocorrelation of Autocorrelation Auto-correlation

Inconclusive Inconclusive

0 dL dU 2 4-dU 4-dL 4

In addition, it’s important that the residuals from the regression models should follow the

normal distribution. Normality assumption of the regression model was tested with the

Jarque- Bera measure. If the Jarque Bera value is greater than 0.05, the hypothesis of the

normality must be fail to rejected (Brooks, 2008). Finally, Hausman specification test

was used to test the fixed effects model against the random effects model.

3.4. Conclusion and relation between research questions/ hypotheses and data
sources

This chapter presented the research questions / hypotheses, and the method adopted to

address them. It also explains the quantitative, qualitative and mixed research approach

with the method adopted for this study. In connection to this, based on the underlying

principles of research methods and the research problem mixed methods approach has

been chosen as appropriate to this research. Beside, this chapter puts forward the

45
necessary information about the sampling design and the data collection instrument.

Finally, the analysis techniques used in this study were presented.

The relationship between research questions and hypotheses on the one hand and

different data sources on the other hand are summarized in table 3.2.

Table: 3.2 Relationships between research question, hypotheses and different data
source
Research questions and hypotheses Data sources

RQ1. What determine the capital structure In-depth unstructured interview

of banks in Ethiopia?

RQ2. Which theory explians the financing Data from Financial Statements of banks:
behavior adopted by Ethiopian balance sheet and income statement
banking industry?

HP 1: There is a negative relationship Data from Financial Statements of banks:

between leverage ratios and balance sheet and income statement
profitability.

HP 2: There is a positive relationship Data from Financial Statements of banks:

between leverage ratios and balance sheet and income statement
growth.

HP 3: There is a positive relationship Data from Financial Statements of banks:

between leverage ratios and balance sheet and income statement
tangibility.

HP 4: There is a negative relationship Data from Financial Statements of banks:

between leverage ratios and risk. balance sheet and income statement

HP 5: There is a positive relationship Data from Financial Statements of banks:

between leverage ratios and size. balance sheet and income statement

HP 6: There is a negative relationship Data from Financial Statements of banks:

between leverage ratios and balance sheet and income statement
liquidity

46
 Chapter Four: Results and Analysis

The preceding chapters presented the orientation of the study, literature review and the

research methodology adopted in the study. This chapter presents the research questions/

hypotheses, results and analysis of data from both documentary analysis and in depth

interview with financial managers of selected commercial banks of Ethiopia. The chapter

is organized into three sections. The first section 4.1 presents research hypotheses and

questions as presented in the previous chapter. This is followed by the results of both

documentary analyses (structured review of documents) and in-depth interview in section

4.2. Section 4.3 discusses the results of the study.

4.1. Research hypotheses and questions

As stated in the previous chapter this study intended to examine the relationship between

leverage and firm specific (profitability, tangibility, growth, risk, size and liquidity)

determinants of capital structure decision and to understand about theories of capital

structure that can explain the capital structure of banks in Ethiopian.

In addition, as noted previously, in order to achieve this broad objective the study

developed the following hypotheses and research questions.

HP 1: There is a negative relationship between leverage ratios and profitability.

HP 2: There is a positive relationship between leverage ratios and growth.

HP 3: There is a positive relationship between leverage ratios and tangibility.

HP 4: There is a negative relationship between leverage ratios and risk.

47
HP 5: There is a positive relationship between leverage ratios and size.

HP 6: There is a negative relationship between leverage ratios and liquidity.

The research questions were

RQ1. What determine the capital structure of banks in Ethiopia?

RQ2. Which theory explians the financing behavior adopted by Ethiopian

banking industry?

4.2. Results

This section discusses the results of the different data sources. Accordingly, the results of

the documentary analysis (structured reviews of documents) and in depth interviews were

presented in the following subsections.

4.2.1. Documentary analysis (structured review of financial records)

Documentary analysis was mainly used for this study, to investigate the determinants of

capital structure of banks in Ethiopia. To this end, the data related to banks which were

necessary to undertake the study was gathered from the financial statements of eight

banks and NBE for twelve consecutive years (2000-2011). Balance sheet and income

statement were predominantly used to analyze the determinants of capital structure of

Ethiopian banks.

Based on the above data source, the following discussions present the results of the

documentary analysis. Accordingly, the result of descriptive statistics, correlation

48
analysis, the test of CLRM assumption and result of the regression analysis are presented

in the following sub-sections.

4.2.1.1. Descriptive statistics

The study examined the determinants of capital structure for eight banks over the time

period from 2000-2011. The descriptive statistics of the dependent and explanatory

variables for the sample banks were summarized in table 4.1. The total observation for

the each dependent and explanatory variable was 96. Moreover, the table also shows the

mean, standard deviation, minimum, median and maximum values for the dependent and

independent variables.

The mean leverage (total debt to total asset) of banks was 88.9 percent with the standard

deviation of 3.5 percent. This means that more than 88.9 percent of the banks in Ethiopia

were financed by debts. This highlights that debt ratio was high in this study. Leverage

for the sample period was ranged from 80 percent to 96 percent with a standard deviation

of 4 percent.

Profitable firms are stronger to face financial distress and stronger to continue more than

unprofitable firms in the future. Profitability, given as the ratio of pre-tax profits plus

interest expense to total assets, registered a mean value of 5.03 percent indicating a return

on assets of 5.03 percent, and median of 5.2 percent with a standard deviation of 1.2

percent and profitability for the sample was ranged from 1.04 percent to 7.02 percent.

This shows the existence of great variation in profit among banks in Ethiopia. Growth

was measured as the annual percentage change in total asset and this shows a mean of

25.8 percent. This indicates that, on average, growth rate was 25.8 percent during the

49
twelve - year period and growth in total asset for the sample period were ranged from -

94.8 percent to 88.1 percent with standard deviation of 24.4 percent. This indicates the

existence of high variation in growth rate among banks in Ethiopia. Tangibility, measured

by fixed asset to total asset shows that on average, 1.7 percent of the firms’ assets were

fixed. The fixed assets to total asset for the sample were ranged from 0.55 percent to 6.6

percent with standard deviation of 1.06 percent. Concerning, the firms risk which was

presented by the standard deviation of operating income (volatility of earning). The mean

of this variable was 0.66 percent and the median was 0.49 percent with a standard

deviation of 0.65. Firms vary in adopting risk; for the study sample, risk was ranged

between 0.000 to 3.3 percent.

Table 4.1: Summary of descriptive statistics for dependent and explanatory variable
variables observation mean SD minimum median maximum
Lev 96 0.89 0.04 0.80 0.89 0.96
Pr 96 0.05 0.01 0.01 0.05 0.07
Gro 96 0.26 0.24 -0.95 0.26 0.88
Tang 96 0.02 0.01 0.01 0.01 0.07
Risk 96 0.01 0.01 0.00 0.01 0.03
Size 96 21.79 1.36 18.78 21.68 25.46
Lq 96 0.51 0.14 0.27 0.51 1.16
Note: LEV refers to total leverage. Profitability (Pr), growth (Gro), tangibility (Tang),
risk (Risk), Size (SIZE) and liquidity (Lq).
Source: Structured review of financial statements and own computations

The mean of the firms' size which was represented by the natural logarithm of total assets

was 21.79 and median was 21.68 with a standard deviation of 1.36. Natural logarithms of

total assets for the sample were ranged from 18.78 to 25.46. Besides, summary of test

statistic shows that the mean of liquidity was 50.6 percent and the median of 50.6 percent

with the standard deviation of 14.4 percent. This reveals as there was high variation in

50
liquidity among Ethiopian banks. Beside, for the study sample liquidity was ranged in

between 27.3 percent to 111.5 percent.

4.2.1.2. Correlation analysis

Table 4.2, shows the correlation between the explanatory variable and leverage in this

study. As noted in Brooks (2008), Correlation between two variables measures the degree

of linear association between them. To find the association of the independent variables

with the leverage, Pearson product moment of correlation coefficient was used. Values of

the correlation coefficient are always ranged between positive one and negative one. A

correlation coefficient of positive one indicates that a perfect positive association

between the two variables; while a correlation coefficient of negative one indicates that a

perfect negative association between the two variables. A correlation coefficient of zero,

on the other hand, indicates that there is no linear relationship between the two variables.

The correlation matrix in Table 4.2 shows that leverage (dependent variable) was

negatively correlated with profitability, growth, tangibility, risk and liquidity of the firm.

Which indicates that firm with higher leverage have less profitability, growth, tangibility,

risk and liquidity. However, only size have positive correlation with leverage. The result

also shows that leverage was correlated at -0.35 with profitability and had statistically

significant correlation. Similarly, leverage was correlated at 0.56 with size and had

statistically significant correlation. Besides, leverage was correlated at -0.14, -0.02, -0.19,

0.11 with growth, tangibility, liquidity, and risk respectively and had statistically

insignificant correlation.

51
Table 4.2: Correlation (Pearson) matrix
LEV PR GRO TANG RISK SIZE LQ
LEV 1.00
PR -0.35 1.00
GRO -0.14 0.13 1.00
TANG -0.02 -0.14 -0.05 1.00
RISK -0.11 -0.31 -0.43 0.00 1.00
SIZE 0.56 0.13 -0.03 -0.43 -0.29 1.00
LQ -0.19 -0.10 -0.20 -0.18 0.10 0.10 1.00
Source: Structured review of financial statements and own computations

4.2.1.3. Tests for the Classical Linear Regression Model (CLRM) Assumptions

Different tests were run to make the data ready for analysis and to get reliable output

from the research. These tests were intended to check whether the CLRM assumptions,

i.e. the OLS assumptions, are fulfilled when the explanatory variables are regressed

against the dependent variables. Accordingly, the following sub-section presents tests of

CLRM.

Test of Normality

The normality tests for this study as shown in figure 4.1 the kurtosis is close to 3, and the

Bera-Jarque statistic has a P-value of 0.412 which was greater than 0.05 implying that the

data were consistent with a normal distribution assumption.

52
Figure 4.1: Normality test

12
Series: Standardized Residuals
Sample 2000 2011
10 Observations 96

8 Mean -7.77e-19
Median -0.002062
Maximum 0.031852
6 Minimum -0.025995
Std. Dev. 0.012367
4 Skewness 0.130488
Kurtosis 2.388094

2 Jarque-Bera 1.770151
Probability 0.412683
0
-0.02 -0.01 -0.00 0.01 0.02 0.03

Source: Structured review of financial statements and own computations

Test of multicollinearity

In order to examine the possible degree of multicollinearity among the explanatory

variables, correlation matrixes of the selected explanatory variables were presented in

table 4.3. Usually the multicollinearity exists if the correlation between two independent

variables is more than 0.75 (Malhotra, 2007). As it appears in the correlation matrix table

4.3, there were no such high correlation between the explanatory variables. Thus, there is

no problem of multicollinearity for this study.

Table 4.3: Correlation matrix between explanatory variables

PR GRO TANG RISK SIZE LQ
PR 1.00
GRO 0.13 1.00
TANG -0.14 -0.05 1.00
RISK -0.31 -0.43 0.00 1.00
SIZE 0.13 -0.03 -0.43 -0.29 1.00
LQ 0.10 -0.20 -0.18 0.09 0.10 1.00
Source: Structured review of financial statements and own computations

53
Test of Heteroscedasticity

Table 4.4 presents three different types of tests for heteroscedasticity and then the

auxiliary regression in the first results table displayed. The test statistics give us the

information we need to determine whether the assumption of homoscedasticity is valid or

not, but seeing the actual auxiliary regression in the second table can provide useful

additional information on the source of the heteroscedasticity if any is found. In this case,

both the F- and χ2 versions of the test statistic give the same conclusion that there is no

evidence for the presence of heteroscedasticity, since the p-values are considerably in

excess of 0.05. The third version of the test statistic, ‘Scaled explained SS’, which as the

name suggests is based on a normalized version of the explained sum of squares from the

auxiliary regression, similarly suggests in this case that there is no evidence of

heteroscedasticity problem.

Table 4.4: Heteroscedasticity Test: White test

F-statistic 0.834728 Prob. F(27,68) 0.6930

Obs*R-squared 23.89741 Prob. Chi-Square(27) 0.6360

Scaled explained SS 17.64855 Prob. Chi-Square(27) 0.9139

Source: structured review of financial statements and own computations

Test for Assumption of Autocorrelation

As noted in Brooks (2008) this is an assumption that the covariance between the error

terms over time (or cross-section ally, for that type of data) is zero. In other words, it is

54
assumed that the errors are uncorrelated with one another. If the errors are not

uncorrelated with one another, it would be stated that they are ‘auto correlated’ or that

they are serially correlated.

Table 4.6 presents the Durbin-Watson test value for the autocorrelation of residual which

is 1.41. The relevant critical values for the test are dL= 1.40, dU = 1.66, and 4 - dU = 4-

1.66=2.34; 4 - dL = 4-1.40=2.30. Accordingly, Durbin-Watson test value is clearly

between the lower limit (dL) which is 1.40 and the upper limit which is 1.66 and thus the

null hypothesis is neither rejected nor not rejected.

Random Effect versus Fixed Effect Models

Table 4.5, presents the Hausman specification test which suggests the fixed effects model

was better than random effects model as the p-value (0.00), is less than 0.05 for

dependent variables which imply that the random effects model should be rejected and

thus, the analysis is based on the fixed effects estimates.

Table 4.5: Correlated Random Effects - Hausman Test

Prob.
Test Summary Chi-Sq. Statistic Chi-Sq. d.f.

Cross-section random 168.899377 6 0.0000

Source: structured review of financial statements and own computations

55
4.2.4. Results of Regression analysis

As shown in chapter three, the model used to find out and explain the association

between the dependent variable and the independent variables was:

= + ( )+ ( )+ ( )+ ( )+ ( )+ ɛ

Where:

LEV= leverage

PR = profitability

TA = tangibility

GR = growth

RS = risk

SZ = size

LQ = liquidity

This study used panel data models where the random effect and fixed effect models could

be used to estimate the relationships among variables. An appropriate model for this

analysis, testing random versus fixed effects models, was selected. To perform this

comparison, the character of the individual effects was tested through the Hausman's

specification test. According to Hausman test results shown in table 4.5, the fixed effects

were found to be more appropriate for the model at the 1 percent level. Thus, the

relationship between leverage and the explanatory variables were examined by the fixed

effects model in this study. The result obtained by the fixed effect model is reported in

Table 4.6.

56
Table 4.6: Fixed effect model estimates

Variable Coefficient Std. Error t-Statistic Prob.

C 0.769041 0.046189 16.65004 0.0000***

PR -1.208480 0.146528 -8.247453 0.0000***

GRO -0.011816 0.007409 -1.594788 0.1146

TANG -0.559071 0.235235 -2.376651 0.0198**

RISK -0.140258 0.273598 -0.512643 0.6096

SIZE 0.010113 0.002151 4.702570 0.0000***

LQ -0.049525 0.011398 -4.345242 0.0000***

R-squared 0.880471

Adjusted R-squared 0.861521

F-statistic 46.46359 Durbin-Watson stat 1.405285

Prob(F-statistic) 0.000000

**, significant at 5 percent

***indicate significant at the 1%,
Source: structured review of financial statements and own computations

The fixed effect result in table 4.6 indicates that profitability was strongly statistically

significant (p-value = 0.00) at 1 percent level and had negative relation with leverage

ratio. Similarly, liquidity was strongly statistically significant (p-value = 0.00) at 1

percent level and had negative relation with leverage ratio. In the same way, size was

statistically significant (p-value = 0.00) at 1 percent level and had positive relation with

leverage ratio. Beside, the fixed effect table 4.6 reveals that tangibility was statistically

57
significant (p- value = 0.019) at 5 percent level and had negative relation with leverage

ratio. But risk and growth do not have statistically significant relationship with leverage

with a p-value of 0.6096 and 0.1149 respectively. Furthermore the table 4.6 shows that

the adjusted R square is 0.86 which indicates that about 86 percent of the variability in

leverage is explained by the selected firm-specific factors (Profitability, Tangibility, risk,

Growth, liquidity, and Size).

4.2.2. In-depth interview results

In order to deeply understand how banks choose between different types of finance and

its determinants unstructured interviews were utilized with some selected Ethiopian

commercial banks finance managers. The interviews were conducted with five finance

managers of Ethiopian commercial banks namely, Construction and Business Bank,

Commercial Bank of Ethiopia, United bank, Bank of Abyssinia and Wegagen bank. The

finance managers were chosen as they are believed to be the most knowledgeable parties

about the determinants of capital structure. Beside, the interviews were conducted

independently with the official. They were asked unstructured interview questions in

relation to the financing of their company. These interview questions were designed to

find out the main factors that determine the capital structure of banks, the source of

finance that bank mainly use, factors responsible to make equity issue, factors responsible

to determine appropriate amount of debt and the influence of change in size on the source

of finance.

According to an interview with financial managers of the banks, the factors that can

determine the capital structure of banks in Ethiopia were; profitability, size, liquidity,
58
ownership structure, maintaining a target debt-to equity ratio, technology, and

government regulation. In addition to this the interviews result indicted that the main

sources of finance for their company were deposit, retained earnings, and equity.

Regarding factors responsible for making equity issue, the most important factor was to

fund a major expansion.

As per the interview with the financial managers, profitability increases the level of

leverage in Ethiopian banking industry. According, to the officials profitability increases

the goodwill of the bank in the eyes of the public which will increases their deposit this

means in other word profitability has a positive relationship with leverage ratio.

Furthermore, the official’s revealed banks with high liquidity ratios or more liquid assets

prefer to utilize these assets to finance their investments and discourage to raise external

funds. Thus, this indicates as liquidity has a negative relationship with leverage ratio.

Regarding to the size of banks the officials suggests that as the size of the banks become

large there levels of leverage ratio also become high which shows a positive relationship

between size and leverage ratio.

59
4.3. Discussions of the Results

The preceding sections present the overall results of the study. Thus, this section

discusses in detail the analyses of the results for each explanatory variable and their

importance in determining leverage ratio. In addition, the discussions analyses the

statistical findings of the study in relation to the previous empirical evidences. Hence, the

following discussions present the relationship between explanatory variables and

leverage ratio.

Profitability

The results of fixed effect model in table 4.6 indicated that profitability had a negative

relationship with leverage, and statistically significant (p-value = 0.00) at 1% level. Thus,

the result was in accordance with the expected sign. This implies that every one percent

change (increase or decrease) in bank’s profitability keeping the other thing constant has

a resultant change of 121 percent on the leverage in the opposite direction. This result

also shows that, higher profits increase the level of internal financing in Ethiopian

banking industry. Beside, the result revealed the suggestions that profitable banks

accumulate internal reserves and this enables them to depend less on external funds. Even

though, profitable banks may have better access to external financing, the need for debt

finance may possibly be lower, if new investments can be financed from accumulated

reserves.

The result of this study is consistent with the pecking order theory that suggests profitable

firms prefer internal financing to external financing. Beside, a negative relationship

between profitability and leverage was observed in the majority of empirical studies

60
Rajan and zingales (1995), Amidu (2007), and Caglayan and Sak (2010) were some of

them. However, Regardless of the above fact, the interview result revealed that

profitability had a positive effect on the leverage ratio. This comment suggested that

banks with higher profitability will have more leverage in their capital structure, which is

in contrary to the above findings of the regression result. This is may be because of the

increased goodwill that profitable banks have in the eyes of the public, which in turn

resulted in an increased level of deposits for them.

Size

The result of fixed effect model table 4.6 indicated that Size had positive relationships

with the leverage of banks, and statistically significant (p-value = 0.00) at 1% level. This

implies that every one percent change (increase or decrease) in the banks size keeping the

other thing constant had a resultant change of 1 percent on the leverage in the same

direction. The results also suggested that the bigger the bank, the more external funds it

will use. The possible reason is that, larger banks have lower variance of earnings, and

the providers of the debt capital are more willing to lend to larger banks as they are

perceived to have lower risk levels.

In addition, the results confirm the concept that large firms can borrow more easily, either

because of a better reputation or because of a perceived lower risk due to better

diversification. This is largely consistent with the Static Trade-off Theory and agency

cost theory. Beside, many previous studies indicated a similarly strong significant

positive relationship, for example Titman & Wessels, (1988), Rajan and Zingales, (1995),

Booth et al., (2001), Amidu (2007), and Caglayan and Sak (2010) were some of them.

61
The findings from interviews data were also provide further support for the findings of

the regression result which demonstrates a positive relationship between size and

leverage. Therefore, based on this finding the relationship between size and leverage was

in accordance with the expected sign.

Tangibility

The results of fixed effect model table 4.6 indicated that the relationship between

tangibility and leverage was found to be negative and statistically significant (p-value =

0.019) at 5% level. Therefore, the result was not in accordance with the expected sign.

The result also implies that every one percent change (increase or decrease) in the banks

tangibility keeping the other thing constant had a resultant change of 55.9 percent on the

leverage in the opposite direction. This significant negative relationship between

tangibility and leverage contradicts with various previous research findings like Rajan

and Zingales (1995), Amidu (2007), and Frank and Goyal (2009) which suggest that

firm’s borrowing capability depends upon collateralizable value of assets (tangibility) and

with theories (Static trade-off theory and asymmetric theory) which stated the positive

relation between leverage and tangibility.

The likely reason of this relationship might be that banking industry in Ethiopia had a

close relationship with creditors, because the relationship can substitute for collateral. In

contrary to the above findings of regression result, the findings from the interview result

indicates tangibility was not a proper determinant of bank capital structure. As per the

interviews with the finance managers of banks, one possible explanation for this was the

use of ownership structure and reputation, in which fixed asset are not used as security.

62
Risk

Risk was considered to be one of the key factors that can affect the capital structure of

banks in Ethiopia. Both theories i.e., static trade of theory and pecking order theory

predict a negative relationship between risk and leverage ratio for at least two reasons:

first, earnings volatility reduces investors ability to predict about future performance and

earnings; second, the higher volatility lead to higher probability of default.

However, there was no support of risk influencing the level of leverage of banks in

Ethiopia. The coefficient for risk on leverage was negative and statistically insignificant

with the p-value of 0.61. Though, negative sign confirms that risky banks are expected to

have less leverage ratio which was consistent with Pecking Order Theory and trade-off

theory, but insignificant result indicates that risk was not considered as a proper

explanatory variable of leverage in Ethiopian banking industry. This insignificant result

was also consistent with the findings of Titman and Wessels (1988) and Amidu (2007).

Similarly, the findings from interviews data were also provide further support for the

findings of the regression result which demonstrates that risk was not a proper factor that

determine the capital structure of banks in Ethiopia.

Liquidity

The results of fixed effect model table 4.6 indicated that liquidity had a negative

relationship with leverage, and statistically significant (p-value = 0.00) at 1% level. Thus,

the result was in accordance with the expected sign which state that there is negative

relationship between leverage and liquidity. This negative sign shows the inverse

relationship between the liquidity and leverage. In other word it implies that every one

63
percent change (increase or decrease) in the bank liquidity keeping the other thing

constant had a resultant change of 4.9 percent on the leverage in the opposite direction.

The negative and statistically significant influence of liquidity in this study was

consistent with a theoretical analysis of pecking order and agency cost theory, which state

that high liquidity firms use internal resources instead of external to finance their

projects. Therefore, this negative effect of Liquidity on leverage was also largely

consistent with the empirical evidence of Deesomsak et al. (2004), Ahmed et al., (2010),

and Najjar and Petrov (2011). In the same way, the interview with the finance manager of

banks indicates banks with high liquidity ratios or more liquid assets were prefer to

utilize these assets to finance their investments and discourage to raise external funds.

Thus, the findings from interviews support the findings of the regression result which

demonstrates a negative relationship between liquidity and leverage.

Growth

According to the trade-off theory, firms holding future growth opportunities, which are

intangible assets, tend to borrow less than firms holding more tangible assets because

growth opportunities cannot be collateralized Myers (1977). However, the pecking order

theory of Myers and Majluf (1984) predicts that leverage and growth are positively

related. For growing firms, internal funds may be insufficient to finance their positive

investment opportunities and, hence, they are likely to be in need of external funds.

According to the pecking order theory, if external funds are required, firms will prefer

debt to equity because of lower information costs associated with debt issues. This results

in a positive relationship between leverage and growth opportunities.

64
The result of fixed effects estimation model table 4.6 revealed that there was a negative

and statistically insignificant relationship between leverage and growth of banks. The

negative coefficient of growth indicates a negative relationship between growth and

leverage. However, this negative relationship is found statistically insignificant with the

p-value of 0.11. Though negative sign confirms that growing banks are expected to have

less debt ratio which was consistent with trade of theory and previous empirical findings

of Huang and Song (2005) and Olayinka (2011) the insignificant result indicates that

growth was not considered as a proper explanatory variable of leverage in Ethiopian

banking industry. This insignificant result was also consistent with the previous empirical

findings of Titman and Wessels (1988), Ahmed et al., (2010) and Najjar and Petrov

(2011).

The possible reason may be that the measure (percentage change in total asset) used in

this study did not reflect the growth of banks fully. Other more significant results might

be obtained by using another measure (market-to-book ratio) for growth which was

difficult to use it for this study where there is no active secondary market. In the same

way, the findings from interviews data were also provide further support for the findings

of the regression result which demonstrates that growth was not a proper factor that

determine the capital structure of banks in Ethiopia.

This chapter discussed the results of the documentary analysis and in depth interview

results, and then presented the discussions of these results using the appropriate method.

Accordingly, the chapter discussed the descriptive analysis, correlations between the

65
variables and through the regressions analyses; it illustrates how the independent

variables influence the dependent variable. Thus, a discussion of the result indicates that

profitability, tangibility, size and liquidity were statistically significant factors that

determine the capital structure of banks in Ethiopia. However, discussions of the result

indicate that risk and growth were not an important explanatory variable of leverage in

Ethiopian banking industry. The next chapter presents conclusions and recommendations

of the study.

66
 Chapter Five: Conclusions and Recommendations

The preceding chapter presented the results and discussion, while this chapter deals with

the conclusions and recommendations based on the findings of the study. Accordingly

this chapter is organized into two sub-sections. Section 5.1 presents the conclusions and

section 5.2 presents the recommendations.

5. 1. Conclusions

Since the seminal work of Modigliani and Miller (1958), the issue of capital structure has

attracted intense debate in the field of financial management. The basic question is

whether there exists an optimal capital structure and what might be its determinants.

Extensive research has attempted to identify these factors; however, the findings of prior

empirical studies have provided varying evidence related to the impact of these factors on

capital structure. Furthermore, the majority of these studies have been conducted in

developed countries that have many institutional similarities.

In light of the above, the main objective of this study was to examine the relationship

between leverage and firm specific (profitability, tangibility, growth, risk, size and

liquidity) determinants of capital structure decision, and to understand about theories of

capital structure that can explain the capital structure of banks in Ethiopian. To achieve

the intended objective the study used mixed methods by combining quantitative and

qualitative approaches together. The quantitative data were collected through survey of

document reviews from a sample of eight banks over the time period from 2000-2011.

The collected data were analyzed by employing multivariate OLS model using statistical

package ‘EVIEW 6’. Beside, the qualitative data that were collected through in-depth

67
interviews was used to support the quantitative findings and to gain additional insight

into the factors that may affect the capital structure of banks in Ethiopia.

In order to conduct the empirical analysis, one dependent variable (at book value), and

six independent variables were selected from prominent previous research works on

capital structure; namely profitability, growth, tangibility, risk, size, and liquidity. The

results of the fixed effect estimation model showed the existence of the following

relationship between leverage and six independent variables.

Profitability had statistically significant negative relationship with leverage, which was in

line with prior expectation. This result also supports the pecking order theory and prefers

using internal finance before raising debt or equity. Similarly, liquidity had a negative

and statistically significant relationship with leverage, which was also in line the

expected sign. A negative sign suggests that banks with high liquidity ratios or more

liquid assets are prefer to utilize these assets to finance their investments and discourage

to raise external funds. Moreover, the result for liquidity clearly supports the pecking

order and agency theories.

Regarding to the effect of tangibility on the capital structure of banks in this study, the

result shows that as there was negative and statistically significant relationship with

leverage, which is in line with the extended form of pecking order theory. Beside, the

results of study indicated that bank size had statistically significant positive relationship

with leverage, which was consistent with trade- off theory and the expected sign. The

result also implies that the bigger the bank, the more external funds it will use.

68
Surprisingly, Growth and risk had a negative relationship with leverage, and statistically

insignificant. In addition to the findings of fixed effect regression results, interviews were

undertaken with the finance managers of selected banks to better investigate the

determinants of bank capital structure. Accordingly, the interview result also indicates

bank size, profitability, and liquidity were the main factors that determine the capital

structure of banks in Ethiopia.

In conclusion, the finding of the study suggests that profitability, liquidity, tangibility,

and bank size were important variables that influence banks’ capital structure. However,

there were no support of banks’ risk and growth influencing the level of leverage of

banks in Ethiopia. The results also, confirms that pecking order theory was pertinent

theory in Ethiopian banking industry, while there were little evidence to support static

trade-off theory and the agency cost theory.

69
5.2. Recommendations

In light of the major finding obtained from the results, the following recommendations

were made.

The analyses indicated that the variables of profitability, liquidity, size and tangibility

were significantly related to leverage ratio. Therefore, banks should pay greater attention

to these significant variables in determining their optimal capital structure.

The study also shows that, banks in Ethiopia mainly use debt as external source of

finance. Thus, the managements of Banks should place greater emphasis on the

facilitation of equity capital in order to obtain sufficient capital to expand their branch

network which in turn creates greater market share for them.

This study examined only firm specific determinants of capital structure of banks in

Ethiopia because of resource and time limitation. Thus, future researcher may address

these deficiencies by including external variable like inflation, GDP, interest rate and

ownership structure, in order to demonstrate the impact of both internal and external

variables on the choice of capital structure.

70
References

Ahmed, N Ahmad Z and Ahmed I 2010, ‘Determinants of capital structure: a case of life

insurance sector of Pakistan’, European Journal of Economics, Finance and

Administrative Sciences, vol.24, pp. 7-12

Amanuel, M 2011, ’The determinants of capital structure evidence from manufacturing

share companies of Addis Ababa city’, Master’s thesis, Addis Ababa University

Amidu, M 2007, ‘Determinants of capital structure of banks in Ghana’, Baltic Journal of

Management, Vol.2 No. 1, 2007, pp 67-79

Ashenafi, B 2005, ‘Determinants of capital structure in medium enterprises in Ethiopia’,

Master’s thesis, Addis Ababa University

Bayeh, A 2011, ‘Capital structure determinants: an empirical study on insurance industry

in Ethiopia’, Master’s thesis, Addis Ababa University

Bevan, A and Danbolt J 2002, ‘Capital structure and its determinants in the UK: a

decomposition analyses’, Applied Financial Economics, vol. 12, no. 3, pp. 159-

170.

Bevan, A and Danbolt J 2004, ‘Testing for inconsistencies in the estimation of UK capital

structure determinants’, Applied Financial Economics, vol. 14, no.1, pp. 55-66

Brooks, C 2008, Introductory Econometrics for Finance, 2nd ed., Cambridge University,

New Work

Booth, L Aivazian, V Demirguc-Kunt, V and Maksimovic, V 2001, ‘Capital structures in

developing countries’, Journal of Finance, vol. 56, No. 1, pp. 87– 130

Buferna, F Bangassa K and Hodgkinson L 2005,’Determinants of capital structure

evidence from Libya’, Research Paper Series, No. 2005/08, pp. 1-31

71
Caglayan, E and Sak, N 2010, ‘The determinants of capital Structure: evidence from the

Turkish Banks’, Journal of Money, Investment and Banking, ISSN 1450-288X

Issue 15 (2010)

Creswell, J W 2003, Research design: qualitative, quantitative and mixed methods

approaches, 2nd ed., Sage Publications, California.

Creswell, J W 2009, Research design: qualitative, quantitative and mixed methods

approaches, 3rd ed., Sage publications, United State of America.

Davies, B M 2007, Doing a successful research project: using qualitative or quantitative

methods. Palgrave MacMillan, UK.

Dunn, S D 1999, The practical researcher: a student guide to conducting psychological

research. 1st ed. McGraw, Hill, United State of America

Deesomsak R, Paudyal K and Pescetto G 2004, ‘The determinants of capital structure:

evidence from the Asia Pacific region’, Journal of Multinational Financial

Management, Vol.14, pp. 387–405

Flannery, M J and Rangan, K P 2006, ‘Partial adjustment toward target capital

structures’, Journal of Financial Economics, vol. 79, pp. 469–506

Gropp, R and Heider,F 2009, ‘The determinants of bank capital structure’, European

Central Bank Working Paper Series. No. 1096, September, pp. 1-50

Gujarati, D N 2004, Basic Econometrics, 4th ed., McGraw Hill, Boston

Gurcharan, S 2010, ‘A review of optimal capital structure determinant of selected

ASEAN countries’, International Research Journal of Finance and Economics,

ISSN 1450-2887 Issue 47 (2010.

72
Harris, M and Raviv, A 1991, ‘The theory of capital structure’, Journal of Financial

Economics, Vol. 41, no. 1, pp. 297-355

Huang, G and Song, F M 2005, ‘The determinants of capital structure: Evidence from

China’, China Economic Review, vol. 17, pp. 14-36.

Jensen, M 1986, ‘The agency costs of free cash flow: corporate finance and takeovers’,

American Economic Review, vol. 76, no. 2, pp. 323-329.

Jensen, M and Meckling, W 1976, ‘Theory of the firm: managerial behavior, agency

costs and ownership structure’, Journal of Financial Economics, vol. 3, no.4, pp.

305-360.

Khrawish, A and Khraiwesh A 2010, ‘The determinants of the capital structure: evidence

from Jordanian industrial companies’, JKAU: Econ. & Adm, Vol. 24, No. 1, pp:

173-196

Malhotra, N 2007, Marketing Research: An Applied Orientation, 5th ed., Phi, New Delhi.

Masulis, R 1998, The debt/equity Choice. Ballinger Publishing Company.

Modigliani, F and Miller, M 1958 ‘The cost of capital, corporation finance and the

theory of investment’, the American Economic Review, vol. 48, pp. 261-297

Modigliani, F and Miller, M 1963, ‘The Corporate Income Taxes and the Cost of

Capital’: A Correction, American Economic Review, vol. 53 no.3, pp. 433-443

Morri, S C and Beretta, C 2008, ‘The capital structure determinants of REITs, is it a

peculiar industry?’ Journal of European Real Estate Res; vol. 1, pp. 6-57.

Myers, S C 1977, ‘Determinants of capital borrowing’, Journal of Finance Economics,

Vol. 5, pp. 5147-75.

Myers, S C 1984, ‘The capital structure puzzle’, Journal of Finance, Vol. 39, pp. 575-92

73
Myers, S C and Majluf, N 1984, ‘Corporate financing and investment decisions when

firm have information that investors do not have’, Journal of Financial

Economics, Vol. 13, pp. 187-221.

Najjar,N and Petrov, K 2011, ‘Capital structure of insurance companies in Bahrain’,

national Journal of Business and Management Vol. 6, No. 11

Octavia, M and Brown, R 2008, ‘Determinants of Bank capital structure in developing

countries’, Research Paper Series, Department of Finance, The University of

Melbourne, Victoria 3010, Australia.

Olayinka, A 2011, ‘Determinants of capital structure: evidence from Nigerian panel

data’, African Economic and Business Review, Vol. 9, No. 1, Spring 2011 ISSN

1109-5609.

Rajan, R G and Zingales, L 1995, ‘What do we know about capital structure? Some

evidence from international data’, Journal of Finance, Vol. 50, pp. 1421-1460

Ross S A Westerfield, R W and Jaffe, J 2002, Corporate finance, McGraw – Hill, USA.

Semu, S Z 2010, ‘Impact of reducing loan by Ethiopian banks on their own

performance’, Master’s thesis, University of South Africa.

Smith, C and Warner, J 1979, ‘On financing contracting: an analysis of bond covenants’

Journal of Financial Economics, Vol. 7, pp. 117-161.

Titman, S and Wessels, R 1988, ‘The determinants of capital structure choice’, Journal of

Finance, Vol. 43, No. 1, pp. 1-19

Yesegat, W A 2009, ‘value added tax in Ethiopia: a study of operating costs and

compliance’, PhD thesis, University of New South Wales, Australia.

74
75
Appendices

Appendix 1: Heteroskedasticity Test

Heteroskedasticity Test: White

F-statistic 0.834728 Prob. F(27,68) 0.6930

Obs*R-squared 23.89741 Prob. Chi-Square(27) 0.6360
Scaled explained SS 17.64855 Prob. Chi-Square(27) 0.9139

Test Equation:
Dependent Variable: RESID^2
Method: Least Squares
Date: 05/30/12 Time: 21:02
Sample: 1 96
Included observations: 96

Variable Coefficient Std. Error t-Statistic Prob.

C -0.014464 0.038861 -0.372211 0.7109

PR 0.167193 0.174600 0.957573 0.3417
PR^2 0.656611 0.572159 1.147602 0.2552
PR*GRO -0.067153 0.045473 -1.476758 0.1444
PR*TANG -0.681842 1.036846 -0.657612 0.5130
PR*RISK 0.394197 1.541361 0.255746 0.7989
PR*SIZE -0.010222 0.007682 -1.330612 0.1878
PR*LQ 0.035405 0.057642 0.614223 0.5411
GRO 0.010952 0.012835 0.853263 0.3965
GRO^2 -0.001867 0.001557 -1.199028 0.2347
GRO*TANG 0.011950 0.050384 0.237180 0.8132
GRO*RISK -0.064869 0.084754 -0.765380 0.4467
GRO*SIZE -0.000266 0.000517 -0.513708 0.6091
GRO*LQ -0.002168 0.003353 -0.646702 0.5200
TANG 0.143976 0.312727 0.460390 0.6467
TANG^2 -0.782922 0.830071 -0.943199 0.3489
TANG*RISK 2.956747 1.856435 1.592702 0.1159
TANG*SIZE -0.003073 0.015120 -0.203265 0.8395
TANG*LQ -0.091952 0.112477 -0.817514 0.4165
RISK 0.127655 0.516930 0.246948 0.8057
RISK^2 -0.632274 2.679930 -0.235929 0.8142
RISK*SIZE -0.011830 0.022423 -0.527594 0.5995
RISK*LQ 0.181849 0.161638 1.125034 0.2645
SIZE 0.000762 0.003138 0.242772 0.8089
SIZE^2 -1.39E-06 6.40E-05 -0.021700 0.9828
SIZE*LQ -0.000406 0.000632 -0.643220 0.5222
LQ 0.007850 0.016624 0.472229 0.6383
LQ^2 -0.000431 0.004007 -0.107499 0.9147

R-squared 0.248931 Mean dependent var 0.000463

Adjusted R-squared -0.049287 S.D. dependent var 0.000610
S.E. of regression 0.000625 Akaike info criterion -11.67845
Sum squared resid 2.66E-05 Schwarz criterion -10.93052
Log likelihood 588.5658 Hannan-Quinn criter. -11.37613
F-statistic 0.834728 Durbin-Watson stat 1.233597
Prob(F-statistic) 0.693024

76
Appendix 2: Hausman Test

Correlated Random Effects - Hausman Test

Equation: Untitled
Test cross-section random effects

Chi-Sq.
Test Summary Statistic Chi-Sq. d.f. Prob.

Cross-section random 168.899377 6 0.0000

Cross-section random effects test comparisons:

Variable Fixed Random Var(Diff.) Prob.

PR -1.208480 -1.320870 0.007422 0.1920

GRO -0.011816 -0.022387 0.000013 0.0034
TANG -0.559071 0.530236 0.033115 0.0000
RISK -0.140258 -0.470665 0.006899 0.0001
SIZE 0.010113 0.018173 0.000003 0.0000
LQ -0.049525 -0.072462 0.000032 0.0001

Cross-section random effects test equation:

Dependent Variable: LEV
Method: Panel Least Squares
Date: 05/30/12 Time: 20:45
Sample: 2000 2011
Periods included: 12
Cross-sections included: 8
Total panel (balanced) observations: 96

Variable Coefficient Std. Error t-Statistic Prob.

C 0.769041 0.046189 16.65004 0.0000

PR -1.208480 0.146528 -8.247453 0.0000
GRO -0.011816 0.007409 -1.594788 0.1146
TANG -0.559071 0.235235 -2.376651 0.0198
RISK -0.140258 0.273598 -0.512643 0.6096
SIZE 0.010113 0.002151 4.702570 0.0000
LQ -0.049525 0.011398 -4.345242 0.0000

Effects Specification

Cross-section fixed (dummy variables)

R-squared 0.880471 Mean dependent var 0.889929

Adjusted R-squared 0.861521 S.D. dependent var 0.035771
S.E. of regression 0.013311 Akaike info criterion -5.666360
Sum squared resid 0.014530 Schwarz criterion -5.292393
Log likelihood 285.9853 Hannan-Quinn criter. -5.515196
F-statistic 46.46359 Durbin-Watson stat 1.405285

77
Prob(F-statistic) 0.000000

Appendix 3: Summary of raw data

YEAR BANK LEV GRO TANG pr risk size LQ

2000 CBE 0.935 0.137 0.012 0.051 0.011 23.710 0.440
2001 CBE 0.939 0.084 0.011 0.030 0.015 23.791 0.350
2002 CBE 0.963 0.031 0.010 0.041 0.008 23.821 0.430
2003 CBE 0.947 0.093 0.009 0.040 0.001 23.910 0.690
2004 CBE 0.947 0.156 0.008 0.027 0.009 24.055 0.740
2005 CBE 0.957 0.186 0.007 0.033 0.004 24.225 0.690
2006 CBE 0.958 0.081 0.006 0.040 0.006 24.303 0.750
2007 CBE 0.933 0.212 0.006 0.035 0.004 24.495 0.780
2008 CBE 0.940 0.160 0.006 0.048 0.009 24.644 0.470
2009 CBE 0.915 0.178 0.007 0.056 0.006 24.808 0.360
2010 CBE 0.955 0.249 0.008 0.048 0.006 25.030 0.290
2011 CBE 0.955 0.540 0.007 0.047 0.001 25.462 0.358
2000 DB 0.911 0.283 0.020 0.044 0.003 20.578 0.539
2001 DB 0.915 0.272 0.016 0.061 0.012 20.819 0.398
2002 DB 0.918 0.351 0.015 0.050 0.007 21.119 0.427
2003 DB 0.935 0.340 0.013 0.034 0.012 21.412 0.400
2004 DB 0.935 0.345 0.015 0.044 0.008 21.708 0.400
2005 DB 0.929 0.278 0.013 0.044 0.001 21.953 0.360
2006 DB 0.915 0.329 0.013 0.055 0.008 22.238 0.311
2007 DB 0.910 0.329 0.016 0.058 0.002 22.522 0.344
2008 DB 0.907 0.296 0.012 0.063 0.004 22.781 0.474
2009 DB 0.907 0.243 0.011 0.057 0.005 22.999 0.593
2010 DB 0.909 0.269 0.013 0.057 0.000 23.237 0.518
2011 DB 0.905 0.187 0.013 0.065 0.006 23.408 0.526
2000 AIB 0.876 0.416 0.030 0.057 0.003 20.448 0.465
2001 AIB 0.885 0.195 0.032 0.053 0.003 20.626 0.408
2002 AIB 0.882 0.226 0.034 0.043 0.007 20.829 0.433
2003 AIB 0.902 0.260 0.041 0.031 0.008 21.060 0.477
2004 AIB 0.912 0.263 0.042 0.037 0.004 21.294 0.508
2005 AIB 0.898 0.258 0.034 0.040 0.002 21.523 0.446
2006 AIB 0.897 0.327 0.030 0.054 0.010 21.806 0.362
2007 AIB 0.887 0.297 0.026 0.069 0.011 22.066 0.362
2008 AIB 0.876 0.258 0.027 0.064 0.004 22.296 0.477
2009 AIB 0.883 0.333 0.023 0.050 0.010 22.583 0.642
2010 AIB 0.882 0.237 0.029 0.064 0.010 22.796 0.662

78
2011 AIB 0.871 0.273 0.025 0.071 0.005 23.037 0.523
2000 BOA 0.859 0.851 0.008 0.053 0.001 20.392 0.334
2001 BOA 0.836 0.248 0.013 0.069 0.012 20.613 0.273
2002 BOA 0.877 0.275 0.011 0.037 0.023 20.856 0.479
2003 BOA 0.888 0.167 0.010 0.026 0.007 21.011 0.471
2004 BOA 0.878 0.189 0.012 0.052 0.018 21.184 0.493
2005 BOA 0.877 0.298 0.017 0.056 0.003 21.445 0.467
2006 BOA 0.858 0.378 0.013 0.058 0.001 21.765 0.359
2007 BOA 0.881 0.198 0.012 0.046 0.008 21.946 0.376
2008 BOA 0.902 0.257 0.015 0.027 0.013 22.175 0.415
2009 BOA 0.891 0.283 0.014 0.047 0.014 22.424 0.600
2010 BOA 0.891 0.147 0.012 0.051 0.003 22.561 0.576
2011 BOA 0.891 0.159 0.012 0.058 0.005 22.708 0.477
2000 WB 0.903 0.404 0.018 0.041 0.008 20.058 0.635
2001 WB 0.901 0.134 0.015 0.053 0.009 20.184 0.503
2002 WB 0.901 0.108 0.022 0.050 0.003 20.286 0.443
2003 WB 0.895 0.376 0.017 0.036 0.010 20.606 0.446
2004 WB 0.887 0.282 0.014 0.055 0.014 20.854 0.467
2005 WB 0.889 0.418 0.013 0.053 0.002 21.203 0.481
2006 WB 0.887 0.398 0.011 0.057 0.003 21.538 0.372
2007 WB 0.884 0.541 0.009 0.060 0.002 21.970 0.485
2008 WB 0.853 0.185 0.010 0.068 0.006 22.140 0.608
2009 WB 0.857 0.241 0.011 0.066 0.001 22.356 0.782
2010 WB 0.868 0.122 0.014 0.069 0.002 22.471 0.774
2011 WB 0.864 0.404 0.014 0.069 0.000 22.810 0.695
2000 UB 0.800 0.882 0.042 0.049 0.016 18.778 0.461
2001 UB 0.811 0.497 0.033 0.056 0.005 19.181 0.535
2002 UB 0.830 0.467 0.025 0.041 0.010 19.565 0.751
2003 UB 0.856 0.494 0.019 0.028 0.010 19.966 0.603
2004 UB 0.858 0.437 0.013 0.031 0.002 20.329 0.545
2005 UB 0.854 0.592 0.010 0.056 0.018 20.794 0.560
2006 UB 0.881 0.490 0.009 0.056 0.000 21.193 0.486
2007 UB 0.835 0.365 0.015 0.058 0.002 21.504 0.492
2008 UB 0.856 0.489 0.010 0.058 0.000 21.902 0.567
2009 UB 0.888 0.431 0.009 0.048 0.007 22.261 0.687
2010 UB 0.892 0.267 0.007 0.060 0.008 22.498 0.693
2011 UB 0.883 0.310 0.008 0.061 0.001 22.768 0.590
2000 NIB 0.832 -0.949 0.016 0.010 0.034 19.073 1.115
2001 NIB 0.842 -0.949 0.008 0.058 0.034 19.797 0.438
2002 NIB 0.815 0.349 0.009 0.060 0.001 20.096 0.484
2003 NIB 0.859 0.657 0.007 0.034 0.018 20.601 0.415

79
2004 NIB 0.861 0.409 0.006 0.051 0.012 20.944 0.398
2005 NIB 0.871 0.389 0.006 0.052 0.000 21.273 0.379
2006 NIB 0.859 0.170 0.015 0.056 0.003 21.430 0.300
2007 NIB 0.837 0.286 0.016 0.057 0.000 21.681 0.370
2008 NIB 0.836 0.400 0.012 0.061 0.003 22.018 0.540
2009 NIB 0.848 0.317 0.012 0.061 0.000 22.293 0.708
2010 NIB 0.846 0.242 0.012 0.063 0.001 22.510 0.743
2011 NIB 0.835 0.191 0.011 0.065 0.002 22.685 0.710
2000 CBB 0.929 0.000 0.038 0.046 0.000 20.697 0.274
2001 CBB 0.931 -0.006 0.037 0.049 0.012 20.691 0.283
2002 CBB 0.920 -0.010 0.040 0.046 0.002 20.680 0.297
2003 CBB 0.916 -0.017 0.038 0.039 0.005 20.664 0.359
2004 CBB 0.921 0.122 0.066 0.026 0.010 20.779 0.487
2005 CBB 0.942 0.733 0.019 0.026 0.000 21.329 0.586
2006 CBB 0.913 -0.019 0.022 0.057 0.022 21.309 0.576
2007 CBB 0.888 0.051 0.022 0.057 0.000 21.359 0.511
2008 CBB 0.892 0.266 0.021 0.066 0.007 21.595 0.627
2009 CBB 0.896 0.084 0.022 0.059 0.005 21.676 0.515
2010 CBB 0.898 0.220 0.020 0.061 0.001 21.874 0.530
2011 CBB 0.896 0.108 0.021 0.058 0.002 21.977 0.556

80
Appendix 4

Addis Ababa University

School of Business and Public Administration

Department of Accounting and Finance

Interview questions that were conducted on the determinants of capital structure of

commercial banks in Ethiopia.

1. What sources of finance does your company mainly use?

2. What are the main factors that can affect the capital structure of your company?

3. When your company considers issuing equity, what factors affect its decisions

about equity?

4. How does your company determine the appropriate amount of debt finance and

what are the possible factors that can affect it.

5. Does the sources of finance used by your company changed as your business has

grown in size?

If you have any opinion regarding the determinants of capital structure please specify…

81