ADDIS ABABA UNIVERSITY

COLLEGE OF BUSINESS AND ECONOMICS

SCHOOLOFGRADUATESTUDIES

THE EFFECTS OF ASSET AND LIABILITY MANAGEMENT ON

THE FINANCIAL PERFORMANCE OF COMMERCIAL BANKS IN
ETHIOPIA

BY: SEBLE MIRUTSE

ADVISOR: ABEBAW KASSIE (PhD)

A THESIS SUBMITTED TO THE SCHOOLOFGRADUATESTUDIES,

COLLEGE OF BUSINESS AND ECONOMICS IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREEOF
MASTERS OF SCIENCE IN ACCOUNTING AND FINANCE.

NOVEMBER 2020
ADDIS ABABA, ETHIOPIA
i
 Declaration

I, Seble Mirutse, declare that thisThesis entitled “the effects of asset and liability management on the
financial performance of commercial banks in Ethiopia” is my original work carried out with the help
of my research advisor andall sources of materials usedfor thepurposeof thisThesis havebeen
dulyacknowledged.

Submitted by:Seble Mirutse

Signature: _______________
Date: _________________

Advisor: Abebaw Kassie (PhD)

Signature: _____________________

Date: _________________

ii
 ADDIS ABABA UNIVERSITY

COLLEGE OF BUSINESS AND ECONOMICS

SCHOOLOFGRADUATESTUDIES

Approval

This is to certify that the researchpaper prepared by Seble Mirutse on the title: “the effects of asset and
liability management on the financial performance of commercial banks in Ethiopia” submitted in
partial fulfillment of the requirements for the Degree of Master of Science in Accounting and Finance
complies with the regulations of the University and meets the accepted standards with respect to
originality and quality.

Signed bytheExamining Committee:

Advisor: Abebaw Kassie (PhD) Signature __________________ Date: ________________

Internal Examiner: ________________Signature: ______________Date: ________________

External Examiner: _________________Signature: _____________Date: _________________

Chair of Department or Graduate Program Coordinator

iii
 Acknowledgement

I am most indebted to God Almighty, the one and only provider of knowledge, compassion and grace
for his provision and protections throughout my life and the period of this study. I would like to show
my enormous appreciation to my Advisor Abebaw Kassie (PhD) for his guidance, encouragement and
comments.

My immense gratitude goes to my dearest husband Tadesse Legesse for his moral and financial
support. My beloved children, it is the best blessing to have four of you in my life and it makes me to
do more. May God bless you all in Jesus name!

Finally, my utmost appreciation goes to my Mother Tsedale Gebreslasea for her special
encouragement, prayers and support and my sister Almaz Mirutse for her unreserved support.

iv
 TABLE OF CONTENTS

Contents

Declaration ................................................................................................................ ii
Approval................................................................................................................... iii
Acknowledgement ....................................................................................................iv
List of Tables ......................................................................................................... viii
List of Figures ...........................................................................................................ix
List of acronyms......................................................................................................... x
Abstract ....................................................................................................................11
CHAPTER ONE ........................................................................................................ 1
INTRODUCTION ..................................................................................................... 1
1.1. Background of the study .............................................................................................1
1.2 Statement of the problem .........................................................................................2
1.3. Objectives of the study................................................................................................4
1.3.1. General Objective ...........................................................................................................4
1.3.2. Specific Objectives .........................................................................................................4
1.4. Hypothesis of the study............................................................................................4
1.5. Significance of the Study .........................................................................................5
1.6. Scope and Limitation of the study ...........................................................................5
1.7. Organization of the study.........................................................................................5
CHAPTER TWO ....................................................................................................... 6
LITERATURE REVIEW .......................................................................................... 6
2.1. Conceptual Literature..................................................................................................6
2.1.1 Asset, Liability, Management and financial performance of commercial banks ............6
2.2. Theoretical Framework ...............................................................................................9
2.2.1 Liquidity Preference Theory ......................................................................................9
2.2.2 The Portfolio Theory..................................................................................................9
v
 2.2.3 Liability Management Theory .................................................................................10
2.2.4 Commercial Loan Theory ........................................................................................10
2.3. Factors Influencing Bank Financial Performance .....................................................11
2.3.1 Effect of Capital Adequacy on financial performance ............................................12
2.3.2 Effect of Asset Quality on Financial Performance ..................................................13
2.3.3 Effect of Management Efficiency on financial performance ...................................13
2.3.4 Effect of Earning Quality on financial performance................................................14
2.3.5 Effect of Liquidity on financial performance ..........................................................14
2.3.6 Effect of Sensitivity on financial performance ........................................................14
2.4 Empirical Review ...................................................................................................15
2.5 Summary of Literature Review ..............................................................................21
2.6 Conceptual frame work ..........................................................................................22
CHAPTER THREE.................................................................................................. 23
RESEARCH METHODOLOGY............................................................................. 23
3.1 Introduction ............................................................................................................23
3.2 Study area ...............................................................................................................23
3.3 Research Approach ................................................................................................23
3.4 Research Design .....................................................................................................24
3.5 Target Population ...................................................................................................24
3.6 Sample Size and Sampling Method .......................................................................24
3.7 The StudyVariables ................................................................................................24
3.7.1 Dependent Variable .................................................................................................24
3.7.2. Independent Variables ..................................................................................................25
3.8 Model Specification of the study ...........................................................................27
3.9 Data Collection ......................................................................................................29
3.10 Data Validity and Reliability .................................................................................29
3.11 Data Analysis ...........................................................................................................29
CHAPTER FOUR .................................................................................................... 30
DATA PRESENTATION, ANALYSES AND DISCUSSIONS OF RESULTS ... 30
4.1 Introduction ............................................................................................................30

vi
 4.2 Findings of the study ..............................................................................................30
4.2.1 Trend Analysis .........................................................................................................30
4.2.2. Descriptive statistics ...................................................................................................37
4.2.3. Correlation Analysis .....................................................................................................39
4.2.4. Model Selection and Test of CLRM assumptions ........................................................40
4.2.4.1 Specification Test .......................................................................................................40
4.2.4.2 Test for Normality ......................................................................................................41
4.2.4.3 Test for Heteroskedasticity .....................................................................................42
4.2.4.4 Test for Multicollinearity............................................................................................43
4.2.4.5 Test for Autocorrelation .............................................................................................44
4.2.5. Analysis andInterpretationof Regression Result ...........................................................45
CHAPTER FIVE ......................................................................................................51
SUMMARY CONCLUSION AND RECOMMENDATIONS .............................. 51
5.1. Introduction ...............................................................................................................51
5.2Summary .....................................................................................................................51
5.3Conclusion...................................................................................................................53
5.4. Recommendations .....................................................................................................54
5.5Suggestion for further research ...................................................................................54
REFERENCES......................................................................................................... 55
Appendices ...............................................................................................................59

vii
 List of Tables

Table 4.1 descriptive statistics of variables ............................................................................................ 37

Table 4. 2 Correlation matrix of dependent and independent variables ................................................ 39

Table 4. 3 Hausman test .......................................................................................................................... 41

Table 4. 4 Shapiro-Wilk W test for normal data ..................................................................................... 42

Table 4. 5VIF Table ................................................................................................................................ 43

Table 4. 6 Correlation analysis among independent variables ............................................................... 44

Table 4. 7 Regression result .................................................................................................................... 45

viii
 List of Figures

Figure 2. 1 Conceptual frame work for ALM and financial performance based on literature. ............. 23

Figure 4.1: Trends of return on equity and capital adequacy ................................................................. 31

Figure 4.2 : Trends of return on equity and asset quality ....................................................................... 32

Figure 4.3: Trends of return on equity and management efficiency ....................................................... 33

Figure 4.4: Trends of return on equity and earning quality .................................................................... 34

Figure4.5: Trends of return on equity and liquidity of banks ................................................................. 35

Figure 4.6: Trends of return on equity and sensitivity ............................................................................ 36

Figure 4.7 Kernel density estimate ......................................................................................................... 42

ix
 List of acronyms

AIB Awash International Bank S.C

ALMAsset Liability Management
AQ Asset Quality
BoA Bank ofAbyssinia S.C
CA Capital Adequacy
CAMELS Capital Adequacy, Asset Quality, Management Efficiency, Earning
Quality,Liquidity, Sensitivity
CBE Commercial Bank of Ethiopia
CBO CooperativeBank ofOromia S.C
CLRM ClassicalLinear Regression Model
DB DashenBank S.C
EQ Earning Quality
LDCs Least Developed Countries
LIBLion International Bank
LIQ Liquidity
ME Management Efficiency
NBE National Bankof Ethiopia
NIB NibInternational Bank S.C
OIBOromia International Bank
ROE Return on Equity
UB United Bank
WB Wegagen Bank S.C

Notation:

S Sensitivity

x
 Abstract

The purpose of this study was to identify the effect of ALM on the Financial Performance of
Ethiopian commercial banks with in the period from 2010 to 2019 using panel data. ALM is
theprocess bywhichaninstitutionmanagesitsStatementof
FinancialPositioninordertoallowforalternative interestrateandliquiditystates.ALM system has
various functions to manage risks such as liquidity risk management, market risk management,
trading risk management, funding and capital planning, profit planning and growth projection
which helps for the survival and growth of the banks. The model used was CAMELS model.
Profitability was measured by ROE. Nine commercial banks were purposively selected. The reasons
for focusing on commercial banks were tomaintain similarityofdata. The study employed secondary
data collected from the National Bank of Ethiopia;
dataselectionwasdonebasedonthemeasurementsofthebankspecificvariablesunderinvestigation. The
study employed Quantitative research approach and explanatory design. The data was analyzed
using Stata software version 14.2. Multiple linear regression model was used where random effect
regression model was applied to study the impact and relationship of the independent variables
with the dependent variable. The result of the study displays that capital adequacy, asset quality
and management efficiency are statistically significant and have negative effect on profitability of
Ethiopian commercial banks. Conversely, liquidity is found to be statistically significant with a
positive effect. Whereas, earning quality and sensitivity to risk are found to be statistically
insignificant. Thus, the study recommends to bank managers that due emphasis on the significant
variables have a paramount importance to enhance profitability.

Key words: Asset Liability Management, Financial Performance, Commercial Banks

xi
 CHAPTER ONE

INTRODUCTION

1.1. Background of the study

As Banks serve primarily as a medium which bridges the gap between surplus and deficit units in an
economy, they are very vital organizations which help in the accomplishment of socioeconomic
activities undertaken by individuals, business organizations and even sovereign states (Mabvur et
al., 2012). Thus, the role of commercial banks in the resource allocation of a country is crucial
(Ongore and Vincent, 2013).

Effective performance and reliability of its banks is one of the most important factors influencing
consistent economic development of any country. The evaluation of soundness and stability of the
commercial banks and the stability of the financial system of a country are closely related (Miletic,
2009). Banksactasthebackboneofeconomicgrowthand prosperitybyactingasacatalystintheprocessof
development.Theytrainthehabitofsavingand mobilizefundsfromnumeroussmallhouseholdsand
businessfirmsspreadoverawidegeographicalarea. Thefundsmobilizedareusedforproductivepurposes
inagriculture, oiland gas, industry and trade(Abayomi &Shalem, 2001).

As thebankingsectoristhelife lineoftheeconomy, activity of banks should

betreatedwithutmostcare(Chandani et al.,
2014).Thebankingindustrycontributesdirectlytonationalincomeanditsoverallgrowth.
Asthebankingsectorhasamajor
impactontheeconomyasawhole,evaluation,analysis,andmonitoringofitsperformanceisvery
important(Dash & Das, 2009).Financial performance is highly influenced by management efficiency
(Abdu,2018).

Theabilitytoobtainmaximumprofits dependsonthefeasiblesetofassetsandliabilities
determinedbythemanagementandtheunitcosts incurred bythe bankforproducingeach component of
assets(Atemnkeng,2006).

Asset liability management is aprocess which enablesinstitutions tomanagetheirStatementof

FinancialPositioninordertoletalternative interestrateandliquiditystates.Itisthepracticeof

1
managingriskscaused by mismatchesbetween theassetsandliabilitiesofthebank.ALM
isanapproachthatprovidesinstitutions withmechanismsthat
makessuchriskacceptable.Ensuringliquidity;whileprotectingtheearnings is theshort- term
objectiveofALMinacommercialbank,maximizingtheeconomicvalue ofthebank is itslong-
termgoal(Vaidyanathan, 1999).

ALM is a dynamic process of planning, organizing, coordinating and controlling combinations of

costs and volumes, maturities, yields of assets and liabilities for the accomplishment of
predetermined objectives. The purposes of ALM system is to manage risks such as liquidity risk
management, market risk management, trading risk management, funding and capital planning,
profit planning and growth projection (Kosmidou & Zopounidis, 2004). Asset liability management
enables the firm to balance between its liabilities and assets. This in turn minimizes financial risks
and hence improves profitability.Thestrategyofkeepingbankassetandliability
allowsforachievingbankingharmonywhichreflects insoundperformancethatactualizesprofit
maximization and attainment of desired liquidity preference (Angele,2008). Perception about ALM
helps to identify the risk and reward that an institution can be involved (FabozziandKanishi,1995).

One of the most widely-used frameworksforbank performance evaluation is the CAMELmodel. The
model originally hadfiveparameters which wereCapitalAdequacy,Asset
Quality,ManagementEfficiency,Earning Quality,andLiquidity. Inorderto make thismethod more
comprehensive, thesixthparameter, SensitivitytoMarketRisk,wasaddedtothe former parameters;
which makes the model’s name CAMELS. These components arecriticalforthe survivalofbanks.
Inadequacyinanyparameterwouldresultinincreasedlikelihoodofbank failure (Sahajwala and Bergh,
2000).Therefore, this model, as improved with the sixth component Sensitivity to risk, was
employed in this study.

1.2 Statement of the problem

The role of commercial banks for the growth and sustainability of a country’s economy is
indisputable. The financial performance of banks highly depends on propermanagementofassets
andliabilitieswhich guaranteesasmoothandefficientfunctioningofthebankingsectorinamanner that it
accommodateschanges in theexternalenvironment.SinceALMplaysacriticalroleinriskmanagement,itis
a must forbanks to recognizethe importance of asset liability
managementandapplyeffectiveriskmanagementpractices (Angelopoulos etal., 2001).

2
Due to theconcurrent unstable financial markets and changes in interest rates, a wise portfolio
management should give a great concern for asset liability management (Mahail, 2009).Since ALM
has a direct effect on the financial performance of banks, it is wise to have an effective ALM process
within banks that closely monitor and match both the assetsand liabilities management. The entire
size and complexity of the economy increases theimportance of ALM and this is why it has to be
considered (Vossen,2010).

Various studies had been conducted on the financial performance of banks globally. Looking at the
findings of some of the studies conducted outside Ethiopia to indicate the gap both globally and
among the researchers conducted in Ethiopia; Aguenaou, Alahrech and Bounakaya (2017) suggest
that
Capitaladequacy,Assetquality,Earningsperformance,andliquidityhaveapositiveimpactonbanks’efficie
ncywithcapitaladequacyhavingthemostsignificantimpact. Boateng (2019)depicted that Earning stood
out as the highly significant factor that affects the performance of banks. Then again, Zafar et al.
(2017) cited in Boateng (2019) showed that only Asset Quality, Earning ability and Liquidity were
found to affect significantly bank performance; Where earning is the most highly significant
parameter. Accordingly, capital adequacy, management efficiency and sensitivity were positive and
insignificant.

On the other hand, though different researches had been done in relation to financial performance of
commercial banks of Ethiopia using CAMEL model, theresearchers are very inconsistent in their
findings. For example, taking only the studies that used ROE as a measure of profitability;
concerning Asset quality, Mulalem (2015), Abdu (2018) and Melaku (2017) put that asset Quality is
insignificant for profitability. Quite the opposite, Anteneh (2018)put that Asset Quality is significant
for profitability. Then again, regarding earning quality, Adamu and Kenenisa (2017) and Anteneh
(2018) found out that earning quality is insignificant for profitability. On the contrary, Mulualem
(2015) found that earning quality significantlyaffected profitability. Therefore, this study aimed to
fill this knowledge gap.

On top of that, asfarastheresearcher’sknowledge is concerned,thereis n o researchthatw a s

conducted on the effect of asset liability management on financial performance of Ethiopian
commercialbanksby taking Sensitivity to risk as an independent variable.Butthis study included an
explanatory variable which was not used by the above stated researchers i.e. Sensitivity to

3
risk.(Sahajwala and Bergh,( 2000) put that the components of the CAMELS Model arecriticalforthe
survivalofbanks and that inadequacyinanyparameterwouldresultinincreasedlikelihoodofbank failure.
Especially, since the component “Sensitivity to risk” is related to the main operation of Banks, it is
obvious that it must be addressed.

1.3.Objectives of the study

1.3.1. General Objective

This study was intended to find out the effect of asset liability management on financial
performance of commercial banks in Ethiopia.

1.3.2. Specific Objectives

The specific objectives of this study were:
1. To explore the impact of capital adequacy on the Ethiopian commercial banks’ financial
performance.
2. To examine the outcome of Asset quality on the Ethiopian commercial banks’ financial
performance.
3. The investigate the consequence of management efficiency on the Ethiopian commercial
banks’financial performance.
4. To explore the outcome of earning quality on the Ethiopian commercial banks’ financial
performance.
5. To determine the result of liquidity on the Ethiopian commercial banks’ financial performance.
6. To examine the effect of sensitivity to risk on the Ethiopian commercial banks’ financial
performance.

1.4. Hypothesis of the study

Accordingly, the followinghypotheses were developed
basedonexistingtheoriesandempiricalstudies and weretested.
1. There is significant and positive relationship between Capital adequacy and performance
2. There is significant and negative relationship between Asset quality and performance
3. There is significant and negative relationship between Management efficiency and
performance
4. There is significant and positive relationship between Earning quality and performance
4
 5. There is significant and positive relationship between Liquidity and performance
6. There is significant and negative relationship between sensitivity to risk and performance

1.5. Significance of the Study

The financial sector, banking industry, is the backbone of a country’s economy. Since Asset
liability mix management is very important, ALM should be given great attention.Assets and
liabilities are vital in wise portfolio management due to the concurrent unstable financial markets
and changes in interest rates. Therefore, this study is significant because it had pointed out the
effect of ALM on Ethiopian commercial banks’ financial performance. The study is expected to
have its implications for banks by indicating factors that can affect bank performance. The study is
believed to contribute for further research as per the suggestions given by this study in this regard.

1.6. Scope and Limitation of the study

The study was restricted to Ethiopian commercial banks which are currently under operation.Nine
private banks were used as a sample. The study covered a period of ten years starting from 2010 to
2019. Qualitative research approach and descriptive and explanatory research design was used. A
dependent variable and 6 independent variables were considered in this study. Thestudywasbased on
secondary data, financial statements of the sample Banks. Thus, is comprehensive only to
thedegreeofaccuracyofthedataobtainedfromthesecondary source.

1.7. Organization of the study

With regard to the organization, the thesis has five chapters. The first chapter is the introductory part
and it contains background of the study, statement of the problem, general and specific objectives of
the study, hypothesis, significance of the study, and scope and limitation of the study. Chapter two
contains review of related literature and conceptual framework in the subject matter. The
thirdchapter includes different elements of methodology, such as study area, research approach,
research design, data collection, population and sampling, research instruments and data analysis.
Chapter four deals with theanalysisoftheresults anddiscussion. At last,the conclusions and policy
implicationare presented under chapter five.

5
 CHAPTER TWO

LITERATURE REVIEW

This chapter presents the review of both theoretical and empirical studies related to asset liability
management and financial performance of banks, in order to put the study within the context of the
existing literature.

2.1. Conceptual Literature

2.1.1 Asset, Liability, Management and financial performance of commercial

banks
As Banks serve primarily as a medium which bridges the gap between surplus and deficit units in an
economy, they are very vital organizations which help in the accomplishment of socioeconomic
activities undertaken by individuals, business organizations and even sovereign states (Kamoyo et
al., 2012).The importance of bank financial performance can be appraised at the micro and macro
levels of the economy. At the micro level, profit is the essential prerequisite of a competitive
banking institution and the cheapest source of funds. At the macro level, a sound and profitable
banking sector is better able to withstand negative shocks and contribute to the stability of the
financial system (Aburime,2008). Effective performance and reliability of its banks is one of the
most important factors influencing consistent economic development of any country(Miletic,2009).

As banks grant advances of long-term maturities for small number of debtors by collecting short
term maturity deposits from number of individuals, these conversion activities expose banks to
credit, interest rate and liquidity risks (Aburime, 2008).

Asset liability management is aprocess which enablesinstitutions tomanagetheirStatementof

FinancialPositioninordertoletalternative interestrateandliquiditystates.Itisthepracticeof
managingriskscaused by mismatchesbetween theassetsandliabilitiesofthebank.ALM

6
isanapproachthatprovidesinstitutions withmechanismsthat makessuchriskacceptable.
Ensuringliquidity;whileprotectingtheearnings is the short-
termobjectiveofALMinacommercialbank,maximizingtheeconomicvalue ofthebank is itslong-
termgoal(Vaidyanathan,1999).

Themostimportantthingforbank
managementistomanagemarketliquidityriskandinterestraterisk.Hencebanksneeda
frameworkwhichenablesthemtocombattheserisksandhelpsthemtooptimizethe
performanceofthebanks.Inthisscenario ALMisveryusefulandhelpfultooltoanalyzethe
performanceofbanks(Kumarand Dhar, 2014).It has been also said that ALMisconcerned
withanattemptto match assetsandliabilitiesin termsofmaturityandinterestrate sensitivity to minimize
interest rateand liquidityrisks(Zawalinska, 1999).

Due to the importance of Asset liability management,

commercialbankshallestablishanAsset&LiabilityManagementCommittee(ALCO)to
manageitsassets,liabilitiesandoff-balancesheetitemssoastofullymeetthebank’s
contractualcommitments. Thetermsofreference attached with NBE Directive No. SBB/57/2014 can
be used by itsALCO(Yesuf ,2016).

The balanced portfolio theory has added better awareness regarding performance of banks
(Nzongang and Atemnkeng, 2006). It was with the application of the Market power and Efficiency
Structure theories that investigations on bank performance started in late 1980s. (Athanasoglou et al,
2005).

Financial performance of banks dependsonthefeasiblesetofassetsandliabilities

determinedbythemanagementandtheunitcosts incurred bythe bankforproducingeach component of
assets.Mismatch between assets and liabilities of firms can be managed by a strong ALM process
which can minimize risks associated with failures that would be caused due to such mismatches
(Atemnkeng,2006).

A firm's overall financial well-being over a given period of time is measured by its financial
performance. Some of the models used to assess banksperformanceareBalancedscore card
(BSC),DataEnvelopment Analysis(DEA), Z score, Efficiency ratio (ER), ROE and ROA and CAMELS
model. Balancescorecard(BSC) isdevelopedby KaplanandNorto.BSC deals with
7
multiplemeasuresofperformanceand
offersastrategicframework,whichspecificallyinspirestheuseofbothfinancialand non-
financialmeasuresalongfourperspectivestomeasurefirmperformance which are
financial,customers,internalbusiness process,andlearningandgrowth(Kaplan&Norton,1996b).
Considering BSCtohavealltheanswersforchoosingthemost
appropriatemeasuresofcompanyperformancewhicharegovernedbytheorganization’sstrategic
orientation andexternalcompetitive environment, it has received muchattention astool
withinmanyindustriesincludinghospitality,health,manufacturing andbanking(Ashton,1998).

DataEnvelopment Analysis(DEA)isoneofthemeasuresofbankperformancewhichisusedtomeasurethe
productionorperformancefunctionofdecision-makingunits (DMUs).ThestandardDEAmodelshave
aninputandoutput
orientation.Aninputorientationidentifiestheefficientconsumptionofresourceswhileholding
outputsconstant.Anoutputorientationidentifiestheefficientlevelofoutputgivenexisting
resourceconsumption (R.Hoque,2012). Whereas, theZscore method
observesliquidity;profitability;reinvestedearningsandleverage
whichareintegratedintoasinglecompositescore( RoliPradhan, 2014). It was
developedbyProfessorEdwardAltmanin1968 toprovidea moreeffective financial assessment tool for
credit risk analystsand lenders.

Efficiency is also used to measure performance. It wasfirstintroducedinEdgeworth’s(1881)and

Pareto’s(1927)works
asaperformancemeasurementandwasfirstempiricallyimplementedinShephard’s(1953)book. Itis
definedasthefirm’sabilitytoconvertitsresourcesintorevenues;
theabilitytogeneratethehighestamountofoutputoutofthe lowestamount of input.
Efficiencyiscalculatedby dividingthefirm’stotalexpenseexcludinginterestexpensesoveritstotalrevenue
wherethelower the ratio,the greaterthefirm’s efficiencyand viceversa (Hays, Stephen, Arthur,2009).

ROE and ROA are also among those used to measure financial performance.ROE reflectstheability
of abank’smanagementtogenerateprofitsfromthebank’sassets. It is a measure of the company’s
efficiency at generating profit from every single unit of shareholders equity (Prasanna, Chandra
2011). Whereas, ROAreflectstheabilityof
bank’smanagementtogenerateprofitsfromthebank’sassets,althoughitmaybebiaseddue to off balance
sheetactivities(AhmedandKhababa, 1999).

8
The other model used to assess the financial performance of banks is the CAMELS model. This
model was first developed inthe U.S. and thenstarted to be used by
everybankandcreditunionintheU.S.andoutsideU.Sbydifferentbankingsupervisoryregulators.The
components of this model at the initial were Capitaladequacy,Assetquality,Management
efficiency,Earning quality,andLiquidity which form the model’s name CAMEL.Then in1997
Sensitivitytomarketrisk was added and the model’s name becomes CAMELS.

2.2. Theoretical Framework

The theories regarding asset liability management and financial performanceare discussed below.

2.2.1 Liquidity Preference Theory

Liquidity preference theory whichis also known as liquidity preference hypothesis was first
articulated by Keynes (1989). This theory describes that Investors need higher interest rate on
securities with long-term maturities because they would prefer to hold cash, which involves less
risk. When the investment is more liquid it is easy to sell it or easily converted in to cash with
minimum risk and also the demand of money increase and decrease depends on the interest rate,
when the interest rate decrease people demand more money to hold until the interest rate increase
and vice versa.

The implication of this theory is that firms should maintain a best level of liquidity not to miss an
opportunity that promise higher return in the future and Firms should work towards the balance
through proper management of their asset and liability to achieve capital requirement and future
higher return investment.

Credit risk is the risk that arises when debtors are not able to pay the money on time. It constrains
the firm from investing in profitable investment that promises higher returns in future. Therefore, it
is a must for financial institutions to assure the debtor’s ability to pay the debt on time before giving
out credit to the firm; it is important to know the credit worthiness of the firm to decrease the credit
risk. (Jappelli and Pagano, 2002)

2.2.2 The Portfolio Theory

Theportfoliotheoryapproachdenotesportfoliodivergence. It describes that

thepreferredportfolioarrangementofcommercial
9
banksareresultsofdecisionstakenbythebankmanagement.On top of that,theabilityto
obtainmaximumprofits dependsonthefeasiblesetofassetsandliabilitiesdeterminedby
themanagementandtheunitcostsincurredbythebankforproducingeachcomponentof assets(Nzongang
and Atemnkeng, 2006).

Theportfoliotheoryapproach isthemostapplicable
theoryandplaysasignificantroleinbankperformancestudies(NzongangandAtemnkeng,2006).According
tothis model,
theoptimumholdingofeachassetinawealthholder’sportfolioisafunctionofpolicydecisionsdeterminedby
anumberoffactorssuchasthevectorofratesofreturnonallassetsheldintheportfolio,avectorofrisksassociate
dwiththeownershipofeachfinancialassetsandthesizeoftheportfolio.

2.2.3 LiabilityManagementTheory
TheLiability management theorystates asabankcanholdreservest h r o u g h
liabilitiesfromdifferentsources such as;issuingtimecertificatesof
deposit,borrowingfromothercommercialbanks, borrowingfromCentralBank,raisingCapitalfunds
throughissuingsharesandbypluggingbackof profits.Thistheoryrecognizesthefactthatassetstructures
ofabankhasaprominentroletoplayinprovidingit withliquiditythatitneeds.Asitenhances
fundraisingopportunitiesforexecutionofattractive investments, thisapproachisconsidered
moreaggressivethan theothermethods.Banksin U.S.A.adoptedthisstrategybysourcingfor
potentialdepositorsmoreaggressivelybycreating marketingdepartmentstobeabletoremainprofitable
business starting from1960 Osifisan(1993) cited in Ogbeifun etal 2018).

According to this theory, banks can satisfy liquidity needs by borrowing in the money and capital
markets. So, there is no need to follow old liquidity norms like maintaining liquid assets, liquid
investments etc. currently, banks have focused on liabilities side of the balance sheet. The central
role of this theory was to consider both sides of a banks’ balance sheet as sources of liquidity
(Emmanuel, 1997).

2.2.4 Commercial Loan Theory

The commercial loan or the real bills doctrine theory which originated in England during the 18th
century states that a commercial bank should forward only short-term self-liquidating productive

10
loans to business organizations. This is because, since they acquire liquidity, they can automatically
liquidate themselves, as they mature in the short run and are for productive ambitions, there is no
risk of their running to bad debts and such loans are high on productivity and earn income for the
banks. Loans meant to finance the production, and evolution of goods through the successive phases
of production, storage, transportation, and distribution are considered as self-liquidating
loans(Emmanuel, 1997).

The commercial loan theory has certain defects. First, if a bank declines to grant loan until the old
loan is repaid, the disheartened borrower will have to minimize production which will ultimately
affect business activity. If all the banks pursue the same rule, this may result in reduction in the
money supply and cost in the community. As a result, it makes it impossible for existing debtors to
repay their loans in time. On the other hand, this theory believes that loans are self-liquidating under
normal economic circumstances. But if there is depression, production and trade deteriorate and the
debtor fails to repay the debt at maturity. In addition to this, this theory disregards the fact that the
liquidity of a bank relies on the salability of its liquid assets and not on real trade bills. It assures
safety, liquidity and profitability. The bank need not depend on maturities in time of trouble. At last,
the general demerit of this theory is that no loan is self-liquidating. A loan given to a retailer is not
self-liquidating if the items purchased are not sold to consumers and stay with the retailer.
(Guthua,2013).

2.3. Factors Influencing Bank Financial Performance

The performance variances between banks show differences in the market served and differences in
philosophy of the management; a function of external and internal influences (Koch 1995 cited in
Olweny2011).

AccordingtoOngore
VincentOkoth(2013),bankperformanceishighlyinfluencedbybothinternalandexternalfactors.Theintern
alfactorsarewithinthescopeofthebankandareeasytobe manipulatedanddifferfrombanktobank. Bourke
(1989) also putthattheprofitabilityofcommercialbank canbe affected by determinants of two types;
theinternaldeterminantswhicharemanagement controllable and the external
determinantswhicharebeyond the control ofmanagement.

11
Management verdicts and policy aims like the level of liquidity, provisioning policy, capital
adequacy, expense management and bank size, which are bank specific issues and ownership,
market attentiveness, stock market growth and other macroeconomic factors which are external
issues and are related to industrial structure have great impact of banks’ financial performance
(Athanasoglou et al2005).

Themainobjectiveofregulatoryandsupervisoryratingsystemsistomeasure
thebankperformanceatinternallevelanditscompliancewithregulatoryrequirementstokeep
thebankonrighttrack.Theseratingsarehighlyconfidentialandareonlyavailabletothebank
management.External credit ratingagencies examine and evaluate the banks and issue ratingsfor
thegeneralpublicandinvestorsinparticulars( Haseeb, 2011). It has been also said that bankprofitabilityis
usuallymeasuredbyinternaldeterminantswhichincludebankspecificvariables (AndreasandGabrielle,2009).

While evaluating banks’ performance,

CAMELmodelisaneffectivetoolandhelpsavoidbanks’failureandinefficienciesby
anticipatinganypotentialrisksandtakingpreventiveactions(Uyen, 2011).CAMELmodelisalso the
model recommendedbyBasleCommitteeonBankSupervisionandIMF(Baral,2005).The CAMELS
model isthe mostappropriateperformance evaluation approach as its
componentscoverandtouchmultipleaspectsofabank;
itleavesnoroomforothervariablestodisturbitsresults.(Kongiri, 2012).

2.3.1 Effect of Capital Adequacy on financial performance

TürkerKaya( 2001)citedinSerhat Yuksel,HasanDincerandUmitHacioglu (2015) put that
capitaladequacy reflectswhetherthebankhasenough capitaltobearunexpectedlossesarisinginthefuture;
whetherbankshaveadequatecapitalinordertomeetthewithdrawal demandofitscustomersincrisisperiod.
MaintainingaSignificantlevelofcapitaladequacy is necessarytopreventthebankfromfailure(Chen,
2003).Capital adequacyis
obviousindicatorsofthefinancialhealthofabankingsystem.Itisveryvaluableforabankto
preserve&safeguardstakeholders’confidenceandpreventingthebankfrombeing bankrupt.

Among the ratios usedby several researchers to express capital adequacy are:
12
CapitalAdequacyRatio(CAR)which measurestheabilityofthebankinabsorbinglosesarisingfrom
riskassets.Thehighertheratiorepresentsbetterperformanceofthebank.ItshallbecomputedastierIcapital+ti
erIIcapital/riskweightedasset.TierIcapital
representsforEquityShareCapital+DisclosedReservesandTierIIcapitalisthesumof
UndisclosedReserves+GenerallossReserves+Subordinatetermdebts(Jayantak.2012).Advancetoassetra
tio is also used and itshowsthe
portionofloansandadvancesdeployedtothetotalfunds.Highertheratiobetteristhe
availabilityoffundsforloansandadvancesoutoftheirtotalassetsandviceversa.( Jayantak. 2012).

2.3.2 Effect of Asset Quality on Financial Performance

Banks’ asset quality which depends on exposure to specific risks, trends in non-performing loans,
and the well-being and success of bank borrowers, plays a great role in regulating the level of the
credit risk which is one of the factors that affect the health of a bank (Baral, 2005).The level of risk
related with the asset must be taken in to account while banks make decisions to avoid risks and
maximize financial performance (Aburime, 2008).

Since it captures the anticipation of management with regard to the performance of loans, the ratio
of loan loss reserve to gross loans is a good measure of credit risk or asset quality (Koch, 1995) as
cited on Olweny (2011). Demonstrating the trade -off between risk and return, returns of banks that
assume more risk with high loans growth are higher (Hempel et al 1994 cited in Jonathan 2013).

2.3.3 Effect of Management Efficiency on financial performance

Management efficiency is
usedasanindicatorofmanagement’sabilitytocontrolcostsandisexpectedtohavea
positiverelationwithprofits,sinceimprovedmanagementoftheseexpenseswillincrease
efficiencyandthereforeraiseprofits(Flaminietal,2009).Poorexpensesmanagementisthemaincontributor
stopoorprofitability; management efficiencyisoneofthekeydriversofprofitabilitythatis
examined(SufianandChong,2008).

Thoughtherelationshipbetweenexpenditureandprofitsseemsstraight forward
implyingthathigherexpensesmeanlowerprofitsandtheopposite,thismaynotalways
bethecase.Thereasonisthathigheramountsofexpensesmaybeassociatedwithhigher
volumeofbankingactivitiesandthereforehigherrevenues.Inrelativelyuncompetitive
13
marketswherebanksenjoymarketpower,costsarepassedontocustomers;hencethere
wouldbeapositivecorrelationbetweenoverheadscostsandprofitability(Flamini et al,2009).

Thecosttoincomeratioisusedtomeasurebanks operationalefficiency.
Costtoincomeratioshowstheoverheadsorcostsofrunningthebank,includingstaffsalariesandbenefits,occ
upancyexpensesandotherexpensessuchasofficesupplies.Thecosttoincomeratioiscalculatedbydividingt
heoverheadcosts(costsofoperatingabank)totheincomegeneratedbeforeprovisions(Kosmidou
etal,2004).

2.3.4 Effect of Earning Quality on financial performance

Earning
Qualitybasicallydeterminestheprofitabilityofbankandexplainsitssustainabilityandgrowth.Itindicatestheab
ilityofbanksingeneratingrevenuebyusingtheasset,
shareholdersequityandusingtheproportionofgrossincome toassessthe
earningsperformanceofbanks.The quality of earnings is the ability of earnings that are of high quality
if it reflects the current operating performance of the company. Moreover, it reflects a good indicator
of future operational performance and provides a good measure of the value of the company.
InterestincometoTotalRevenue(II/TR) is one of the ratios used to calculate earning quality. It
showstheshareofbanksinterest income earned fromadvances out of the total revenue(Hamad, 2015
cited in Al-Zubaid etal 2018).

2.3.5 Effect of Liquidity on financial performance

Liquidityindicatestheabilityofthebanktomeetitsfinancialobligationsinatimely handeffectivemanner.
Though therearedifferencesamongscholarswithrespecttothe measurementratios,totalloanto
customerdeposits ratio is themostcommonfinancialratiosthatreflecttheliquidityposition ofabank.
Themostcommonfinancialratiosthatreflecttheliquidityposition
ofabankarecustomerdeposittototalassetandtotalloanto customerdeposits(Said, 2003).

2.3.6 Effect of Sensitivity on financial performance

Sensitivity is the sixth component added to the previous five components of the CAMEL model
making the model’s name CAMELS. It measures how particular risk exposure can cause mess to the
bank’s profitability. It specially reflects the degree to which interest rate changes can affect earnings

14
and hence the bank’s capital (Suresh and Paul, 2014). It also shows the degree to which changes in
exchange rates, commodity price and equity price can affect earnings.

Sensitivity can be measured by GAP analysis, a tool used to evaluate a bank’s earnings exposure to
interest rate movements. A bank’s GAP over a given period is the difference between the value of
its assets that mature during that period and the value of its liabilities that mature during the same
period. If the difference is significant, it means interest rate changes have a great impact on net
interest income. If the amounts of maturing assets exactly offset by the reprising liabilities, a
balanced position will occur, the ratio will be 1.0. If the GAP ratio is less than 1.0, it means that the
bank’s liability matures earlier than its assets (an indication that the bank is liability sensitive). If the
GAP ratio is greater than 1.0, it means that the bank’s assets mature earlier than liabilities (an
indication that the bank is asset sensitive). GAP is the difference between risk sensitive assets and
risk sensitive liabilities. The GAP ratio is expressed as the ratio of risk sensitive assets to risk
sensitive liabilities.
GAP= RSA
RSL
Where the sum of net advances, net investments and money at call (short term loan payable
immediately on demand) is taken as rate sensitive assets and the sum of deposits and borrowings of
the bank is taken as risk sensitive liabilities (Suresh and Paul, 2014).

2.4 Empirical Review

This section presents someofthestudiesconductedtoexaminetherelationshipbetweenassetliability
managementandprofitability of banks.

ThestudyofKosmidou&Zopounidi(2004) examines2000yearsbalance
sheetofoneoftheGreekbankswiththeaidofGoal programmingfoundoutthatthebestcombinationof
variables(bond,depositandfacilities)withhighest returnwasselected.A
multicriteriamethodologyforbank’sassetand liabilitymanagement was
used.Giokas&Vassiloglou(1991)concludedthatmanagementof banksshouldpurse goal
ofprofitmaximizationandfinestallocationofrisks intheir capitalby applyingagoalprogramming model
intheirstudy,using largebanksinGreek ascasestudy.

BordeleauandGraham(2010) intheir studyoftheimpactofliquidity onbanks’ profitabilityin

15
theCanadianandAmericancontext fora timeframeextending from 1997to2009
observedthattheprofitabilityofbanksinCanada and America was increasing as their levelofliquidity
holdingswas increasing aswell to reach aninflection
pointwhereprofitabilitystartedincreasingbutatadecreasingrate,andsubsequently,it
decreased.Thisresearchconfirmsthatreachingacertainlevelof liquidityholdings diminishes banks’
profitabilitybecause it’s an opportunitycostof profitable investments.

Mihail (2009) did a study by taking a panel data of over 30 banks across Europe for the period from
2004 to 2011 on how asset liability management affects profitability of Banks. The vital goal of this
study was to examine ALM in banks. He used canonical correlations analysis and linear dependency
between two variables, i.e. (the structure of assets and liabilities.) was tested. It was concluded that
attention must be given to the management of assets and liabilities, the risk level, earnings, liquidity,
profit, solvency, the level of loans and deposits in order to be effective in banks,

Abdulazeez, Asish, and Rohani (2017)assessed the profitability of Saudi banks using the parameters

of the capital adequacy, asset quality, and management quality, earning ability and liquidity

framework for the period 2000-2014. Pooled ordinary least square and fixed effect model was used.

Their results showed that domestic banks are more profitable than foreign banks; and foreign banks

carry more credit risk in their portfolio. In contrast to domestic banks, operating expenses to total

income for foreign banks is significant but negatively related to profitability, indicating that cost

management inefficiency adversely affect the profitability of this group. Their results also indicated

that banks with larger size are less profitable. The study failed to identify the effect of banks’ capital

adequacy, liquidity, and asset risk portfolio on their asset quality, efficiency and profitability on a

stand lone basis between foreign and domestic banks rather focusing on earnings’ ability of the

studied banks.

MohamedandChithra(2016)havestudiedthefinancialperformanceofselectedConventional
andIslamicBanksintheKingdomofBahrainusingCAMELmodel.Basedontheirstudy
therearesignificantdifferencesintheperformanceacrossthebanksthoughtheywork
16
underthesamesocio,economic-political-legalandregulatoryframework.

TheresearchofAl-Tamimi(2010)usedtheCAMELratingmodelto
investigatethedeterminantsofconventionaland Islamicbanks’performanceinUAE andrevealed that
liquidityis amajor determinantof conventionalbanks’performance.

Lizabeth M. Samuel (2018) evaluated the performance of selected commercial banks in India by
using the CAMELS rating model; five years Data was used. The non-parametric analysis of the
banks showed that all the selected banks conform to the capital adequacy requirement as per the
Basel norms. In addition, all banks had sound asset quality and management efficiency.Yet, earnings
capacities as well as the liquidity of the banks were not satisfactory.

Sheela and
Bastray,(2014)studiedtheinfluenceofAssetLiabilityManagementontheCommercialbanks’profitability
inIndian. A case study ofonePublicSectorBanknamelyUnionBankofIndiaand
oneprivatesectorbanknamelyICCIbank.GapAnalysisTechnique, thestatistical
tools,themultivariatestatisticaltechniqueandRatioanalysiswasusedtointerpretthe
financialstatementsandanalyzethedatathisstudyprimarilybasedonsecondarydata,and
endeavorstoevaluatetheinterestraterisk that both the banks are exposed to, spread over a period from
2009 to 2014. The finding indicated that it is obvious that both banks are performing satisfactorily in
terms of profitability and adequacy,butthere is aneedtoaddressthe immediateconcernofliquidity.
OneoftheirfindingwasalsoIndianbankismore profitablewithgoodAsset-
LiabilityManagementstrategy,ThereforeInvestorswouldbe motivated to invest ina bank which has
high profitabilityratio.

Mohammed Kamru Ahsan (2016)measured financial performance using CAMEL on selected

Islamic Banks in Bangladesh. All the selected Islamic banks were found to be on strong position on
their composite rating system. The banks were found to be sound in every respect.

Zafar et al (2017) examined the performance of banks in Pakistan using the CAMELS ratio. Capital
adequacy, management efficiency and sensitivity were found to be positive and insignificantly
related to bank performance. On the other hand, asset quality, liquidity and Earning quality were
found be negatively related with performance and was significant; where Earning quality found to
be the most highly significant parameter. The fixed effect panel data was used for analysis.

17
Ahmad, Ahmad, and Adeel (2016)assessall listed banks of Pakistan Stock Exchange during the time
period of 2010-2015 to investigate the trade-off between liquidity and profitability in the banking
sector. The key research method adopted to gather secondary data for the study was document
investigation. The study showed that the banks’ management should maintained optimum liquidity
that will maximize profitability.

Nurazia and Evans (2005) examined whether the CAMELS ratio would be used to forecast bank
failure. The resultproposed that adequacy ratio, asset quality, management, earnings, liquidity and
banks size are statistically significant in explaining bank failure. Olweny and Shipo (2011) found
that the poor quality and low level of liquidity are the two major causes of bank failure.Ongore Kusa
(2013) concluded that the financial performance of commercial banks in Kenya was driven mainly
by board and management decision, while macroeconomic factors have insignificant contribution.
Alabede (2012) concluded that in the presence of the effect of global financial condition only assets
quality and market concentrationare significant determinants of the Nigerian bank’s performance.
The study recommended that reducing nonperforming assets and introducing a policy to inspire a
reasonable competition among the banks.

TheresearchconductedbyKamau(2009)inKenyanusingthenonparametric
approach(DEA)foundthatsurplusliquidityandefficiencyratioaresignificantlypositively
related,meaningthatexcessliquidityleadstoinefficiency. Gweyi, Tobias, and Oloko (2016)assess the
influence of liquidity risk on financial performance of deposit taking savings and credit co-
operatives (DT-Saccos) in Kenya. The study adopted a descriptive research design. The target
population for this study was 164 deposit taking Sacco societies licensed to undertake deposit-taking
Sacco business in Kenya. The study adopted census and considered all the Deposit Taking Saccos
for study. Secondary data was collected from 135 deposit taking Sacco’s audited financial statement
which represented 82.32% success rate. Data was analyzed using both descriptive and inferential
statistics. The result indicates as liquidity risk has a negative and significant influence on financial
performance.

Angele(2008)inherresearchonanalysis ofchosenstrategiesofassetandliabilitymanagement
incommercialbanks,,foundoutthatthecoreproblem inassetandliabilitymanagementisthefactthatthe
mainassetofcommercialbankcreditscannotalways beliquid,especiallyif thecountry’seconomy isindeep
recession.

18
The major findings of the study conducted by Gyekyi (2011) on the effect of asset liability
management on profitability of National Investment Bank in the New Juabeng Municipality in
Ghana was that reduction in assets value leads to increase in banking profitability. In addition,
profitability was found to increase or decrease directly with liability. The study used the goal
programming method.

Tuffour, Owusu, and Boateng, (2018) evaluated internal and external determinants of bank
profitability in Ghanaian banking industry. A panel data of 6 banks listed on the Ghana Stock
Exchange was analyzed over the period 2010-2015. Pooled regression models were used. The
statistical results revealed that capital adequacy, liquidity, total assets and real interest rate were the
major determinants of bank profitability in Ghana. Liquidity was found to have significant negative
effect on both profitability measures. while operating efficiency has negative and significant
influence on only return on equity. Further, capital adequacy was positive and significant for
determining both return on assets and return on equity, total assets found to have positive and
significant influence on only return on assets. The study recommended that the bank management
and regulator should stipulate optimum level for profitability determinants.

Onyekwelu, Chukwuani, and Onyeka (2018) using Ex-post facto research design
evaluatedliquidity’s effect on financial performance of deposit money banks in Nigeria forsample of
five banks.Ten years period; 2007-2016 secondary data were collected from the firms and were
analyzed using multiple regression analysis. The results of this study showed that Liquidity has
positive and significant effect on banks’ profitability, that is, return on capital employed (ROCE)
and recommended that the management should established liquidity level that will assurefinest
financial performance.

Saheed (2018)studied the effect of capital adequacy and operational efficiency on profitability of
deposit money banks (DMBs) in Nigeria for the period of 2008‐2016 using panel data of 15 listed
banks drawn from the Nigerian stock exchange. Correlational research design was used to examine
the effect of the bank specific factors on bank profitability. The study utilized feasible generalized
least square (FGLS) and found that capital adequacy has a positive and significant relationship with

19
bank profitability while operational efficiency has a negative and significant relationship with bank
profitability.

Tamiru (2013) in his study carried outon asset liability management and commercial Banks
profitability in Ethiopia, examined the effect of ALM on commercial banks profitability.
Profitability was measured by ROA as a function of balance sheet. The SCA model was used.
Macroeconomic variables the real growth rate (GDP) and the general Rate of inflation were also
incorporated in this study. The study covers the time period from 2005 to 2010 taking eight
commercial banks as a sample.The pooled OLS method was to estimate the regression result. The
study showed that the profitability of commercial banks in Ethiopia is positively affected by assets
management, except for fixed assets; which is negatively affected by liability management. The
macroeconomic variables have negative effect on commercial banks profitability. The study
concludes that assets management positively and liability management negatively affects
profitability of Ethiopian commercial banks.

Mulalem (2015), on his study used CAMEL approach to analyze financial performance of
commercial banks of Ethiopia using panel secondary data for the period 2010- 2014. Multiple
regression model was used for analysis. ROE and ROA was used to measure financial performance.
The study showed that CA, AQ and ME have negative relation while EQ and LiQ shows positive
relationship with both profitability measures with strong statistical significance. But CA was found
to be insignificant for ROA and AQ for ROE.

Abdu (2018) studied internal factors that can affect financial performance of Ethiopian private
commercial banks, bank size, liquidity, management efficiency, asset quality and capital adequacy,
by taking 6 banks as a sample for the period 2011-2017. Descriptive statistics was employed and
ROE, ROA and net interest margin were used to quantity performance. The study declares that CA,
ME and size of banks have positive and significant effect on all measures of performance. But
liquidity had negative and significant impact on ROE while asset quality was not significant
determinant of performance.

Anteneh (2018) conducted a study using CAMEL model by using sample of 11 commercialbanks for
the period2011to2016.Banksizeandnetinterestmarginwere alsoused together with the CAMEL
variables and returnonassetandreturnonequity were used to measure profitability. Paneldataand
descriptivestaticswere used and it was shown that capital
20
adequacy,assetquality,managementefficiency,liquidity,andnetinterestmarginweresignificantvariablest
oexplainROE,butearningqualityandsizeofthebankwerenotsignificantvariabletoinfluenceROE.

Melaku (2017) analyzed performance of private commercial banks in Ethiopia using CMEL model
descriptive study. Six private banks were used as a sample for the period starting from 2007 to 2016.
Fixed asset to total asset, net profit per employee, total loan per number of branches and total
deposit per number of branches were also used on top of the CAMEL variables. The study showed
that the CAMEL variables except asset quality were found to be significant.

Abdu (2018) studied bank specific factors that can affect financial performance of private
commercial banks in Ethiopia, bank size, liquidity, management efficiency, asset quality and capital
adequacy,by taking 6 banks as a sample for the period 2011-2017. Descriptive statistics was
employed and return on equity, return on asset and net interest margin were used to measure
performance. The study found that capital adequacy, management efficiency and size of banks have
positive and significant effect on all measures of performance. But liquidity had negative and
significant impact on ROE while asset quality was not significant determinant of performance.

2.5 Summary of Literature Review

This study was conducted to identify the effect of asset liability management on the financial
performance of Ethiopian commercial banks. The literature review and the empirical studies show
that since ALM has a direct impact on the financial performance of banks, having an effective AML
process within banks is unquestionable to closely monitor and equalize the management of assets
and liabilities. The studies addressed in the above literature showed how asset liability management
using the CAMELS model affected profitability. Considering the researches conducted in Ethiopia,
there is inconsistency in findings.For example, taking only the studies that used ROE as a measure
of profitability; concerning Asset quality, Mulalem (2015), Abdu (2018) and Melaku (2017) put that
asset Quality is insignificant for profitability. Quite the opposite, Anteneh (2018) put that Asset
Quality is significant for profitability. Then again, regarding earning quality, Adamu and Kenenisa
(2017) and Anteneh (2018) found out that earning quality is insignificant for profitability. On the
contrary, Mulualem (2015) declare that earning quality is significant for profitability. On top of that,
there is no research which used the variable sensitivity to risk to assess the financial performance of
commercial banks. This study therefore, aims to find out what the effect of the CAMELs model
components on the financial performance of Ethiopian commercial banks really is.

21
2.6 Conceptual frame work

Capital
Adequacy

Asset
Sensitivity
Quality

ROE

Liquidity
Management
Efficiency

Earning
Quality

22
Figure 2. 1 Conceptual frame work for ALM and financial performance prepared based onliterature
(The CAMELS Model).

CHAPTER THREE

RESEARCH METHODOLOGY

3.1 Introduction
This chapter of the studypresents the overall research methods used in the study. These include the
study area, research approach, research design, population and sample,
variabledefinition,measurements,model specification,data collection, data validity and reliability,
and data analysis.ThestudyusedtheCAMELSmodeltoidentifytherelationshipbetweenAsset
liabilitymanagementandprofitabilityofcommercialbanksin Ethiopia

3.2 Study area

Thestudyfocusesonthe effect ofAsset Liability Management (ALM) onfinancialperformance of
Ethiopian Commercial Banks; the studyareai s Ethiopianbankingindustry. Ethiopian Banking
industry consists of 17 commercial banks, one Government owned and 16 private commercial banks
and one Development Bank which is owned by Government.

3.3 Research Approach

Researchapproachrevealsthe researcher’sattitudeastohowknowledgeisconstructedandgives
information as to what method isto beemployedinthestudy. Therearethreeresearchparadigms;
quantitativeresearch,qualitativeresearch,andmixedresearch(Creswell2017).Theessence
ofquantitativeresearchistouseatheory toframe andthus understandtheproblemathand (Jonker and
Pennink 2010).Quantitativeresearchisameansfortestingobjective
theoriesbyexaminingtherelationshipamongvariables(Creswell2017). Therefore, quantitative research

23
approach has been used in this study to investigate the effect of CAMELS components on the
financial performance of Ethiopian commercial banks.

3.4 Research Design

A research design is a plan or a millstone of the study used to gain answers to research questions or
problems. The choice ofresearchdesigndependsonobjectivesthattheresearcherswantto
achieve.Baridam(2001)suggestedthatthechoiceofadesignisinfluencedbythepurposeofthestudy,thestud
ysetting,unitofanalysisandtimehorizon. As this study targeted to investigate the statistical association
of the CAMELS components and financial performance of commercial banks, the research design
used for this study was explanatory research design. Explanatory research design is a research
design used to describe the dependent and independent variable and then make analysis between
them.

3.5 Target Population

Thetargetpopulationof this study wasallcommercialbanksregisteredbyNBEandunderoperation.
Currently,thereare17commercialbanksinEthiopia:onegovernment-owned and 16 private banks.

3.6 SampleSize and Sampling Method

Thestudywas conductedon9Ethiopian commercial banks.
Sinceitallowsfocusingonparticularcharacteristicsofapopulationthat areofinterest,purposivesampling
was used. It isatypeofsamplinginwhichparticularsettingsaredeliberately selected. Accordingly, the
selected banks as a sample for this study wereAIB, DB, WB, OIB, BOA, UB, NIB, CBO and LIB.

3.7 The StudyVariables

3.7.1 Dependent Variable

The dependent variable used for the study was the financial performance of the banks and measured
by return on equity. It is of great interest to the shareholders. ROE is a measure of the company’s
efficiency at generating profit from every single unit of shareholders equity. ROE which is also
referred to as net worth, allows investors to understand how their money is being put to productive
use. Prasanna Chandra (2011)put that ROE is the most critical measure of performance in an

24
accounting sense.Return on equity it is calculated as the ratio of net profit after tax to the
stakeholders’ equity.

3.7.2. Independent Variables

The independent variables used for the study were the CAMELS components i.e. Capital Adequacy,
Asset Quality, Management Efficiency, Earning Quality, Liquidity and Sensitivity to risk.

CapitalAdequacy
This component of the CAMELSmodel is a measurement that defines the solvency of a bank.
Adequate capital reserve helps banks to expand and increase the confidence of depositors and
regulators. The measure also provides a cushion for potential loan losses and other unanticipated
problems. In effect, capital adequacy improves the stability and efficiency of the bank (Parvesh
and Sanjeev, 2016). For the purpose of this study, the total equity to total assets ratio has been
used as a measure of capital adequacy.
AssetQuality
This ratio is a measure of the degree of the financial strength of a bank. Measurement of asset
quality is very significant since it depicts the profitability of the bank. The study used the ratio of
allowancefordoubtful loa ns to loans outstanding as a measure for assets quality, since loans and
advances make up thelarge portion of bank assets.(Jayantak.2012)put that itisthemost
standardmeasure to judgethe assets quality.
ManagementEfficiency
This is another critical measure as it assures the survival and growth of a bank. The management
efficiency ratio shows the obedience to the norms and regulations set, leadership and administrative
competence, and the ability to stand any changing operational environment. Operating cost to
operatingincomeratio was used for this study as a measure for management efficiency.
Thelowertheratioisthe betterperformanceofthe management. (Sangmi and Nazir, 2010)put as
operating cost to operating income ratio can be used to measure management efficiency.
[

EarningCapacity
This measure shows the bank’s capacity to create proper returns in order to be in a position to
expand, retain competitiveness and accumulate worthy retained earnings. High earnings quality
reflects the Bank’s current operating performance and its capacity of future operating performance.
Higher income level is expected to minimize the likelihood of risk of a bank

25
(ColeandGunther,1998).The ratio interest income to totalincomewas used as a measure of earnings
quality.Totalincomeofabankisdividedintotwoparts. Incomefromcoreactivity
i.e.incomefromlendingoperations)andincomegeneratedbynon-
coreactivitieslikeinvestments,treasuryoperations,corporateadvisoryservicesetc. (Jayanta 2012).
Liquidity
Liquidity means the fund available with a bank to meet its credit demands and cash flow
requirements. Banks with larger Volume of liquid assets are perceived safe since these banks will
not be affected by unexpected withdrawals. However,custodyof a larger volume of liquidity reduces
management’s capacity to pledgeconsistently to an investment strategy that protects investors’
interest, which is considered as its negative effect.According to Parvesh and Sanjeev (2016) banks
can maintain an adequate liquidity position by either increasing current liability or quickly
converting their assets in to cash.For the purpose of this study, the ratio of total loan to total deposit
was used as ameasure for liquidity where totaldepositsincludedemanddeposits,savingsdeposits,term
depositsandotherdeposits. The higher the ratio, the more credit the bank generates from its deposit
received from customers.
Sensitivity
Sensitivity is the sixth component added to the previous five components of the CAMEL model
making the model’s name CAMELS. It measures how particular risk exposure can cause mess to the
bank’s profitability. It specially reflects the degree to which interest rate changes can affect earnings
and hence the bank’s capital (Suresh and Paul, 2014).It also shows the degree to which changes in
exchange rates, commodity price and equity price can affect earnings.

Sensitivity can be measured by GAP analysis, a tool used to evaluate a bank’s earnings exposure to
interest rate movements. A bank’s GAP over a given period is the difference between the value of
its assets that mature during that period and the value of its liabilities that mature during the same
period. If the difference is significant, it means interest rate changes have a great impact on net
interest income. If the amounts of maturing assets exactly offset by the repricing liabilities, a
balanced position will occur, the ratio will be 1.0. If the GAP ratio is less than 1.0, it means that the
bank’s liability matures earlier than its assets (an indication that the bank is liability sensitive). If the
GAP ratiois greater than 1.0, it means that the bank’s assets mature earlier than liabilities (an
indication that the bank is asset sensitive). GAP is the difference between risk sensitive assets and
risk sensitive liabilities. The GAP ratio is expressed as the ratio of risk sensitive assets to risk
sensitiveliabilities.

26
GAP= RSA
RSL
Where the sum of net advances, net investments and money at call (short term loan payable
immediately on demand) is taken as rate sensitive assets and the sum of deposits and borrowings of
the bank is taken as risk sensitive liabilities.(Suresh and Paul, 2014). Therefore, This Ratio was used
for this study to measure sensitivity.

3.8 Model Specification of the study

To determine the relationship between the CAMELS components and the performance measure
(ROE), a regression model was adopted. Because the independent variables are more than one,
multiple regression model was deemed appropriate for the analysis of the data
Brook(2008).Thegeneral multivariate regression model withKindependent variables as Brook put
is:

Yi=β0+β1X1i+β2X2i+…+βkXKi+εi(i1, 2, 3…,n)

WhereYiistheithobservationofthedependentvariable,X1i,…Xkiaretheithobservationof

theindependentvariables,β0,…βkaretheregressioncoefficients,εiistheithobservationofstochastic
errorterm,andnisthenumberofobservations.Therefore,theeffectofasset liability management on
profitability based on the CAMELS model in both the FEM and REM are presented here under.

3.8.1 Linear Panel Data Model

To determine the relation betweendependent and independent in this study the fixed effects model
and the random effects model which are the most common linear panel data analysis models were
used (where the REM was chosen over the FEM for the data in the study). Return on equity a as
function six explanatory variables. The fixed effect model is given as.

𝑹𝑹𝑹𝑹𝑹𝑹𝒊𝒊𝒊𝒊 = 𝜷𝜷𝟎𝟎 + 𝜷𝜷𝟏𝟏 𝑪𝑪𝑪𝑪𝒊𝒊𝒊𝒊 + 𝜷𝜷𝟐𝟐 𝑨𝑨𝑨𝑨𝒊𝒊𝒊𝒊 + 𝜷𝜷𝟑𝟑 𝑴𝑴𝑴𝑴𝒊𝒊𝒊𝒊 + 𝜷𝜷𝟒𝟒 𝑬𝑬𝑬𝑬𝒊𝒊𝒊𝒊 + 𝜷𝜷𝟓𝟓 𝑳𝑳𝑳𝑳𝑳𝑳𝒊𝒊𝒊𝒊 + 𝜷𝜷𝟔𝟔 𝑺𝑺𝒊𝒊𝒊𝒊 + εit

And the random effect model is provided as:

𝑹𝑹𝑹𝑹𝑹𝑹𝒊𝒊𝒊𝒊 = 𝜷𝜷𝟎𝟎 + 𝜷𝜷𝟏𝟏 𝑪𝑪𝑪𝑪𝒊𝒊𝒊𝒊 + 𝜷𝜷𝟐𝟐 𝑨𝑨𝑨𝑨𝒊𝒊𝒊𝒊 + 𝜷𝜷𝟑𝟑 𝑴𝑴𝑴𝑴𝒊𝒊𝒊𝒊 + 𝜷𝜷𝟒𝟒 𝑬𝑬𝑬𝑬𝒊𝒊𝒊𝒊 + 𝜷𝜷𝟓𝟓 𝑳𝑳𝑳𝑳𝑳𝑳𝒊𝒊𝒊𝒊 + 𝜷𝜷𝟔𝟔 𝑺𝑺𝒊𝒊𝒊𝒊 + εit + 𝜇𝜇𝑖𝑖𝑖𝑖
27
𝑊𝑊ℎ𝑒𝑒𝑒𝑒, 𝜇𝜇𝑖𝑖𝑖𝑖 is random effect of the bank specific error y i, and for ℇ it + 𝜇𝜇𝑖𝑖𝑖𝑖 =𝛾𝛾𝑖𝑖𝑖𝑖 the model is given
as:

𝑹𝑹𝑹𝑹𝑹𝑹𝒊𝒊𝒊𝒊 = 𝜷𝜷𝟎𝟎 + 𝜷𝜷𝟏𝟏 𝑪𝑪𝑨𝑨𝒊𝒊𝒊𝒊 + 𝜷𝜷𝟐𝟐 𝑨𝑨𝑨𝑨𝒊𝒊𝒊𝒊 + 𝜷𝜷𝟑𝟑 𝑴𝑴𝑴𝑴𝒊𝒊𝒊𝒊 + 𝜷𝜷𝟒𝟒 𝑬𝑬𝑬𝑬𝒊𝒊𝒊𝒊 + 𝜷𝜷𝟓𝟓 𝑳𝑳𝑳𝑳𝑳𝑳𝒊𝒊𝒊𝒊 + 𝜷𝜷𝟔𝟔 𝑺𝑺𝒊𝒊𝒊𝒊 + 𝛾𝛾𝑖𝑖𝑖𝑖

Where:
Subscriptirefertofirmi,andsubscripttreferstoyear t.
ROEit=denotesthedependentvariable(profitability ofBankiattime t)measuredbyReturnon Equity (net
profit aftertax /total equity)
𝜷𝜷𝟎𝟎 =Intercept /Constant term
𝑪𝑪𝑪𝑪𝒊𝒊𝒊𝒊 =Capitaladequacyofbankiattime t(measured by ration of totalequitytototal asset)
𝑨𝑨𝑨𝑨𝒊𝒊𝒊𝒊 =Assetqualityofbankiattimet(measured by ration ofallowance
fordoubtfulloantoloansoutstanding)
𝑴𝑴𝑴𝑴𝒊𝒊𝒊𝒊 =ManagementefficiencyofBankiat timet(measured by ration of operating costto
operatingincome)
𝑬𝑬𝑬𝑬𝒊𝒊𝒊𝒊 =EarningsqualityofBankiattimet(measured by ration of interest incometototalincome)
𝑳𝑳𝑳𝑳𝑳𝑳𝒊𝒊𝒊𝒊 =LiquidityofBankiattimet(measured by ration of totalloan tototal deposit)
𝑺𝑺𝒊𝒊𝒊𝒊 =Risk Sensitivity of Bank i at time t (measured by ration ofRisk sensitive assets toRisk sensitive
liabilities)
𝜷𝜷𝟏𝟏 –𝜷𝜷𝟔𝟔 =Coefficientsofparameters
𝛆𝛆𝐢𝐢𝐢𝐢 =The unobservable errortermwhereiiscrosssectionalandtis timeidentifier
Yit,the gamma it, denotes the composite error.

Summing up, the random effects estimator and fixed effects estimator, both are models designed to
handle the specific structure of panel data. That is, unobservable individual heterogeneity is taken in
to account by both models. The difference between the two models is whether the individual
specific time invariant effects are correlated with the repressors or not. Moreover, given that the
fixed effects model is valid, the random effects estimator still produces consistent estimates of the
identifiable parameters.

28
As it often might be unlikely to believe that the individual specific effects are uncorrelated with the
relevant covariates, it is appealing to prefer the random effect estimator to the fixed effects
estimator.

3.9 Data Collection

Secondary data Collected from Ethiopian National Bank which is the Central Bank of the
Country was used for the study. Published financial statements of the sample banks were also
used from the banks’ websites. The study covered a 10 years period from the year 2010 to
2019.
Dataselectionwasdonebasedonthemeasurementsofthebankspecificvariablesunderinvestigation.
Theforemostreasonthatthisstudyfocusedon commercialbankswastomaintain similarityofdata.

3.10 Data Validity and Reliability

Publishing of any information has to attest to the validity and reliability and ensure that the
statements show a true and fair view of the banks financial position. Therefore, to ensure validity
and reliability of thedata collection mechanism, only published data in the form of financial
statements which is a requirement by law was used. The same financial statements were also
obtained from the national bank of Ethiopia (NBE) Therefore, the data will be considered valid and
reliable as it is collected from the attested documents.

3.11 Data Analysis

Thestudyisbasedonpaneldataofselectcommercialbanks in Ethiopia. The data constitutes cross-

sectional and time series elements; where the cross-section relates to the selected banks and the
timeseries with the 10 years data taken. Panel datawas used,becauseithelps toidentify acommongroupof
characteristics and
controlsforindividualheterogeneity,lesscolinearityamongvariablesandpathstendenciesin
thedatawhich could not be achieved viatimeseriesandcross-sectionaldata(Baral,2005). Thedata
was analyzed using the statistical software known as Stata version 14.2. Trend Analysis, Descriptive
Statistics, correlation analysis and regression analysis were applied to achieve aforementioned
research objectives. Tostudytherelationshipbetweenthedependentvariableand the independent
variables, correlationmatrixwas used.P value and t-statistics were usedto study the significance of
29
independent variables and Regression coefficient todetermine how each explanatory variable
influencefinancialperformance.DiagnosticstestsofCLRMassumptionsincludingModel specification,
Multicollinearity,Heteroskedasticity, autocorrelationandnormalitytestswere alsoconducted to
ensuresafe application of multiple linearregressionmodel; since
theassumptionsunderlyinginCLRMmustbe fulfilled to saytheregressionmodelisacceptable.

CHAPTER FOUR

DATA PRESENTATION, ANALYSES AND

DISCUSSIONSOF RESULTS

4.1 Introduction
This chapter presents analysis of panel data of the selected banks in Ethiopia for the period from
2010 up to 2019 and the findings thereof. Thevariablesverifiedinthestudy
includes:Capitaladequacy,Asset quality,Management Efficiency,EarningQuality,Liquidity,
Sensitivity to risk and Return on Equity which was taken as a measure of profitability. The chapter is
organized into two main parts, introduction and findings of the study, where part two is organized in
twofive subparts; sub part one, trend analysis, sub part two, descriptive statistics, sub part three
correlation analysis, sub part fourmodel selection and test for classical linear regression model
assumptions and at last sub part five analysis of regression results.

4.2 Findings of the study

4.2.1 Trend Analysis

Return on equity and capital adequacy
Return on equity measures profitability of the banks; how much profit a company generates with the
money shareholders have invested.

30
Source: Own computation using the data from 2010 to 2019, Stata 14.2 output

Figure 4.1: Trends of return on equity and capital adequacy

Figure 4.1 shows the profitability of banks as measured by return on equity and capital adequacy
measured as the ratio of total equity to total asset. The trend is presented for nine commercial banks
under the study period from 2010 to 2019. As clearly shown in the graph above, the trend of capital
adequacy under the study period is more or less constant trend for all banks around 14 percent level.
The trends of return on equity showed a variation across banks under the study period. Since 2016
return on equity or profitability for Awash international bank, cooperative bank of Oromia and for
united bank shows an increasing trend from 21.5% to 41.1%, 3 % to 23% and 18 % to 22%. But the
level of return on equity of Dashen bank, Oromia international bank and Bank of Abyssinia shows a
downward trend since 2014 from 30.69 % to 16%, 34.2 % to 23.7% and 33.94% to 16.9%
respectively. The trend of return on equity for Lion international bank, Nib international bank and
Wegagen bank were stagnant under the study period around 19 %.

31
Return on equity and asset quality

Source: Own computation using the data from 2010 to 2019, Stata 14.2 output

Figure 4.2: Trends of return on equity and asset quality

As per Figure 4.2 above, the trend of return on equity and asset quality for nine commercial banks of
Ethiopia under the study period is depicted. More specifically, in average term the trends of asset
quality for all banks it seems stagnant around zero but there is slight difference across banks. The
trends of asset quality for cooperative bank of Oromia, Lion international bank and Wegagen bank
showed increasing trend since 2015 from 0.0% to 3.41 %, 0.0% to 2.48% and 0.0% to 2.16%
respectively. But the trend of asset quality form reaming commercial banks have almost stagnant
trend around 1 percent during the study period.

32
Return on asset and management efficiency

Source: Own computation using the data from 2010 to 2019, Stata 14.2 output

Figure 4.3: Trends of return on equity and management efficiency

Figure 4.3 above have showed that the trend of return on equity and management efficiency for all
sample commercial banks under the study period. Management efficiency of banks measured as the
ration total operating cost to operating income. As clearly shown in the graph above the trend of
management efficiency for bank of Abyssinia, Nib international bank, Dashen bank and Wegagen
bank have an increasing trend from the period 2014 to 2019 i.e 55.82% to 76.12%, 54.10% to
72.16%, 55.36% to 76.11% and 66.55 % to 70.09% respectively. However, for cooperative bank of
Oromia, united bank and Oromia international bank have showed a downward trend from the
period 2016 to 2019 i.e 96.54% to 79.34%, 102.69% to 78.10% and 78.62 % to 69.01 %. The trend
of management efficiency for Awash international bank and Lion international bank was almost
constant around 60 percent.

33
Return on equity and earning quality

Source: Own computation using the data from 2010 to 2019, Stata 14.2 output

Figure 4.4: Trends of return on equity and earning quality

As per figure 4.4 the trend of return on asset and earning quality for all samples under the study
period is presented. Earning quality is measured as the ratio of interest income to total income. As
presented in the graph above overall trends of earning quality for all commercial banks registered an
upward trend since 2014. More specifically, Nib international bank, united bank and bank of
Abyssinia had registered higher increment as compared to other commercial banks from the period
2014 to 2019 i.e 67.03% to 85.7 %, 66% to 85% and 61.96 % to 81.71% respectively. But during the
study period the lowest increment on earning quality was recorded on Oromia international bank and
Addis international bank i.e 73.3 % and 74.2 % in 2019 respectively.

34
Return on equity and liquidity

Source: Own computation using the data from 2010 to 2019, Stata 14.2 output

Figure 4.5: Trends of return on equity and liquidity of banks

Figure 4.5 presented the trend of return on equity and liquidity of banks for all sample commercial
banks under the study period. Liquidityindicatesthat
theabilityofthebanktomeetitsfinancialobligationsinatimely handeffectivemanner. Liquidity of the
banks was measured as the ratio of total loan to total deposit. As per the graph above the trends of
liquidity for Awash international bank, bank of Abyssinia and united bank showed increasing trend
from 2015 to 2019 i.e 61 % to 79.3%, 53.11 % to 73.83% and 58.10 % to 74.7 % respectively.
However, cooperative bank of Oromia recorded downward trend from 2015 to 2018 i.e 89 % to
61%. For the remaining commercial banks, the liquidity trend ware almost stagnant since 2015
around 61 %.

35
Return on equity and sensitivity

Source: Own computation using the data from 2010 to 2019, Stata 14.2 output

Figure 4.6: Trends of return on equity and sensitivity

The trend of return on equity and risk sensitivity of banks is clearly presented in Figure 4.6 above.
Risk sensitivity of banks measured as the ratio of risk sensitive asset to risk sensitive liabilities. As
clearly indicated in the above figure the trend of risk sensitivity for Awash international bank,
Dashen bank, bank of Abyssinia and Lion international bank is recorded in upward trend from the
period 2014 to 2019 i.e 87.08 % to 98.55 %, 76.63 % to 102.9%, 81.77% to 96.7 % and 84.36 % to
97.1 % respectively. But the trend for Cooperative bank of Oromia, Nib international bank and
Wegagen bank is showed as downward trend from the period 2015 to 2019 i.e 110.32% to 82.3%,
110.02 % to 94.67% and 104.26 % to 98 % respectively. The trend for Oromia international bank
and united bank is more or less constant trend over time.

36
4.2.2. Descriptive statistics
Table 4.1 below summarizes by mean,standard deviation,minimum and maximum values
thedescriptivestatisticsofthedependent variable, ROEand theindependentvariables; capital adequacy,
asset quality, management efficiency, earning quality, liquidity and sensitivity to risk
forthesamplebanksfor the period from2010to2019.

Table 4.1descriptive statistics of variables

Variables Observation Mean Std. Dev. Min Max

ROE 90 22.43 6.96 3 41.1
CA 90 14.73 3.17 7.04 22.23
AQ 90 1.51 1.36 0 7.41
ME 90 63.41 11.29 43.78 102.69
EQ 90 63.19 11.79 37.9 85.7
LIQ 90 61.09 8.52 40 89
S 90 84.27 14.81 48.43 110.32
Own computation using Stata 14.2 output

Return on equity: As shown in table 4.1, ROE which is the measure of profitability shows
that during the study period; 2010-2019, the average profitability of commercial banks of
Ethiopia was 22.43 % with a range of minimum 3% and maximum 41%; which depicted that
41% was earned by the most profitable Ethiopian commercial banks during the study period.
The standard deviation of ROE which infers the variation of Ethiopian commercial banks’
profitability from the mean; being 6.9% showed that there waslow variation.

Capital Adequacy: The measure of capital adequacy provides a cushion for potential loan losses
and other unanticipated problems. In effect, capital adequacy improves the stability and efficiency
of the bank (Parvesh and Sanjeev, 2016). For the purpose of this study, the total equity to total
assets ratio has been used as a measure of capital adequacy.The mean value of
capitaladequacybeing 14.73% which is above the minimum requirement of 12%(NBE minimum
capital requirement directives No MFI/27/2015) indicates thataround 15% of the total assets of the
banking sectorwere financed by shareholders’ equity. The stated mean value with a range of
minimum 7% and maximum 22.23% and standard deviation of 3.1% showed that the capital
adequacy of the banks across the study period do not have significant difference.
37
Asset Quality:This ratio is a measure of the degree of the financial strength of a bank. Measurement
of asset quality is very significant since it depicts the profitability of the bank. The study used the
ratio of allowancefordoubtful l oa ns to loans outstanding as a measure for assets quality,Theoutput
for asset quality of meanvalue1.5%with a range of minimum 0%, maximum 7.4%
andstandarddeviation 1.36% showedthattherewas almost no variation of asset quality among
Ethiopian commercial banksduring the study period. It also showed that current credit risk
management practices being exercised before lending are less strict as the maximum value of 7.4%
change from the minimum 0% showed.

Management Efficiency:The management efficiency ratio shows the obedience to the norms and
regulations set, leadership and administrative competence, and the ability to stand any changing
operational environment. Operating cost to operatingincomeratio was used for this study as a
measure for management efficiency. Thelowertheratiois the betterperformanceofthe
management(Sangmi and Nazir, 2010).Theaverage value of management
efficiencywas63.41%,anindicatorthatoverheadsarehighintheEthiopianbankingsectori.e. banks
consume in average 63 cents to generate one-birr revenue.Related to this, the minimum value 43%
and the maximum value 102.69% together with the 11.29% standard deviation showed that there
wasvariation among management efficiency of the sample Ethiopian commercial banks during the
study period.

Earning Quality:A high earnings quality reflects the Bank’s current operating performance and its
capacity of future operating performance. Higher income level is expected to reduce the likelihood
of risk of a banks (ColeandGunther,1998).In this study, the ratio interest income to totalincomewas
used as a measure of earnings quality.The mean value of earning quality being 63.19% together with
the minimum value of 37.9, maximum value 85.7% and standard deviation of 11.79% depicted that
there was variation of earning quality among the sample commercial banks of Ethiopia during the
study period.

Liquidity:According to Parvesh and Sanjeev (2016) Banks can maintain an adequate liquidity
position by either increasing current liability or quickly converting their assets in to cash.For the
purpose of this study, the ratio of total loan to total deposit was used as a measure for
liquidity.Themeanvalueofliquiditybeing 61% displayed that the banking sector was very liquid
during thestudy period,which is fourtimesmorethanthecurrentminimumregulatoryrequirementof
National Bank of Ethiopian which is15%(NBE DirectivesNo.SBB/57/2014). Even the minimum
38
value, 40%, is more than twice of the minimum requirement. The 61% mean value showed that
61% of the deposits of commercial banks were converted to loan during the study period. The
standard deviation of 8.52% showed that the liquidityratiodiffersfrom bank to bank and from time
to time within a specific bank. All sample bankshave maintained the minimum level of liquidity
requirement.

Sensitivity to Risk: Sensitivity to riskmean the degree to which changes in conversion rates,
commodity price and equity price can affect earnings quality of the banks. The mean value of
sensitivity was 84.27% with a minimum value of 48.43 and a maximum of 110.32% and standard
deviation of 14.8%. All together showed that there was high deviation from the mean in this regard;
which showed that Ethiopian commercial banks had different level of risk sensitivity during the
study period.

4.2.3. Correlation Analysis

In this section the correlation analysis between the dependent and explanatory variables were
presented.
Table 4.2Correlation matrix of dependent and independent variables

ROE CA AQ ME EQ LIQ S
ROE 1.00
CA -0.2197 1.00
AQ 0.3272 -0.1075 1.00
ME -0.6446 -0.3405 -0.4494 1.00
EQ -0.3713 -0.2655 -0.4281 0.6895 1.00
LIQ -0.1225 0.0016 -0.2831 0.3218 0.6084 1.00
S -0.2681 0.0673 -0.05210 0.4024 0.6691 0.07908 1.00
Own computation using Stata 14.2 output

As per the correlation table above, return on equity has negative relationship with capital adequacy,
management efficiency, earning quality, liquidity and sensitivity of banks. The correlations clearly
show that capital adequacy, management efficiency, earning quality, liquidity and sensitivity to risk
of banks moves in opposite direction to profitability; while it has positive relationship with asset
quality which implies that they move in the same direction.The correlation result also showed that
profitability is most highly correlated with management efficiency.

39
Since correlation analysis just show the relationship and direction of dependent and independent
variables but not the cause and effect association, regression analysis presented in the below part
was taken as a remedy to fill this gap.

4.2.4. Model Selection and Test of CLRM assumptions

Holding theCLRMassumptions helps
theestimatorsdeterminedbyOLStohaveanumberofdesirablepropertiesthatisconsistent,
unbiased,andefficient.Therefore, tests of CLRM wereconductedto
ensurewhetherthedatafitsthebasicassumptionsofclassicallinearregressionmodelornot.The
twotypesofpanelestimatorapproachesthat can be used in financial research according to Brook,
(2008) are the fixedeffectsmodels(FEM)and therandomeffectsmodel(REM).

4.2.4.1 Specification Test

In this study the researcher used the specification tests in order to choose the suitable econometric
model. In this regard, Hausman specification tests enable to determine which of the two models
(fixed effects model or random effects model) is best suited in comparison to the data. The Hausman
test (1987) can be used to many econometric problems that need specification. In this research case,
the tested hypothesis offers guidance for choosing between fixed and random effects.

And therefore, it was hypothesized that:

H O = Random effect model

H 1 = Fixed effect model

And the Hausman test was applied and the result is given as follows;

40
Table 4. 3Hausman test
. hausman fe re

Coefficients
(b) (B) (b-B) sqrt(diag(V_b-V_B))
fe re Difference S.E.

ca -.8976087 -1.091099 .1934904 .1408974

aq -.5191734 -.6806271 .1614537 .1393628
me -.6534199 -.6075555 -.0458644 .0213762
eq .1243775 .055691 .0686865 .0253578
liq .1972556 .1899193 .0073363 .0146311
s -.0620463 -.0600975 -.0019488 .0093746

b = consistent under Ho and Ha; obtained from xtreg

B = inconsistent under Ha, efficient under Ho; obtained from xtreg

Test: Ho: difference in coefficients not systematic

chi2(6) = (b-B)'[(V_b-V_B)^(-1)](b-B)
= 3.65
Prob>chi2 = 0.7237
(V_b-V_B is not positive definite)

Source: Own computation using Stata 14.2 output

The results from the Hausman test led to the acceptance of the null hypothesis and rejection of the
alternative hypothesis. The model used is therefore the random effects model.

4.2.4.2 Test for Normality

The normality assumption plays a crucial role in the validity of inference procedures, specification
tests and forecasting. Non-normal error components in the panel data affect the performance of
several tests, like the performance of panel heteroskedasticity tests severely affected. Among all the
CLRM assumption the zero- mean value of the error term is the first one to be addressed. Thus, in
order to test the normality of the data Shapiro-Wilk W test for normal data is used. According to
Shapiro-Wilk W test for normal data, the data is normal if the p value is greater than 0.05 and not if
p value is less than 0.05. In test Shapiro-Wilk W test the null hypothesis states that the error term of
the model is normally distributed and if the P value is greater than 0.05 then the null hypothesis will
be accepted.

41
Based on Shapiro-Wilk W test for normal data and residual plot below the error term is normally
distributed since p value is 0.058 which is above 0.05. In this case, we have enough evidence to say
error term of the model is normally distributed.

Table 4. 4 Shapiro-Wilk W test for residuals

Shapiro-Wilk W test for normal data

Variable Obs W V z Prob>z

r 90 0.97314 2.032 1.564 0.05895

Source: Own computation using Stata 14.2 output

Source: Own computation using Stata 14.2 output

Figure 4.7 Kernel density estimate

4.2.4.3Test for Heteroskedasticity

The heteroskedasticity assumption states that the variance of the error term (𝑈𝑈𝑖𝑖𝑖𝑖 𝑈𝑈𝑗𝑗𝑗𝑗 ) across
observation is unchanged. If the variance of the error term is not constant is it side to be
heteroskedasticity (Wooldridge, 2004). In the presence of heteroskedasticity error terms in the
model, the regression coefficient, results will be consistent estimates but the estimates will not be
42
efficient. The loss of efficiency of the estimates will lead to invalid inference through biased
standard error (Gujarati, 2004). Modified Wald test for heteroskedasticity was used for testing
whether the error variances are constant or not. The decision rule behind this test for
Heteroskedasticity states that the significant result from the test is indicating the regression of the
residuals on the predicted values reveals significant Heteroskedasticity. Even if the problem of
Heteroskedasticity does not really matter in panel data approach, we must detect and give
appropriate estimate to avoid biased estimation. And hence the problem of Heteroskedasticity was
handled by using generalized least squares or using xtgls command in order to get unbiased estimators.
After adjusting the heteroskedasticity problem the test presented as follows.

Modified Wald test for groupwise heteroskedasticity

in cross-sectional time-series FGLS regression model

H0: sigma(i)^2 = sigma^2 for all i

chi2 (9) = 9.57

Prob>chi2 = 0.3860

Source: Own computation using Stata 14.2 output

4.2.4.4 Test for Multicollinearity

In the presence of multicollinearity, the explanatory variables correlated with each other and the
regression coefficients possess large standard errors (in relation to the coefficient themselves).
Because of the presence of multicollinearity in a given model, the coefficients cannot be estimated
with great precision or accuracy (Gujarati, 2004). To check the presence of multicollinearity or not
this study used Pearson pairwise correlation and variance inflation factor. Thus, as can be seen from
the result below, since the VIF, 2.60 is less than 10 and no correlation matrix value of greater than
0.8, there is no multicollinearity problem.

Table 4. 5VIF Table

Variable VIF 1/VIF

s 4.03 0.248358
eq 3.23 0.309182
liq 2.99 0.334756
me 2.27 0.440249
aq 1.71 0.583642
ca 1.34 0.747066

Mean VIF 2.60

43
Source: Own computation using Stata 14.2 output

Table 4.6 Correlation analysis among independent variables

. pwcorr ca aq me eq liq s

ca aq me eq liq s

ca 1.0000
aq -0.1075 1.0000
me -0.3405 -0.4494 1.0000
eq -0.2655 -0.4281 0.6895 1.0000
liq 0.0016 -0.2831 0.3218 0.6084 1.0000
s 0.0673 -0.5210 0.4024 0.6691 0.7908 1.0000
Source: Own computation using Stata 14.2 output

4.2.4.5 Test for Autocorrelation

The covariance between two consecutive error terms i.e (𝑈𝑈𝑖𝑖𝑖𝑖 𝑈𝑈𝑗𝑗𝑗𝑗 = 0) we can say that the error term
uncorrelated to each other or simply the error terms subject to autocorrelation Verbeek, (2000). The
classical error component panel data model assumes serially uncorrelated disturbances, where the
covariance between error terms over time is zero. If the error terms are correlated with one another,
it is said to be they are autocorrelated or that they are serially correlated. A number of tests for the
presence of autocorrelation in a fixed effects panel data model have been proposed in the literature.
The Wooldridge test for autocorrelation in our panel data was used for testing serial correlation
between error variances. From Wooldridge test for N = 90 observations and k = 6 number of
explanatory variables taken at 5% level of significance, accepted the null which states that there is
no first order autocorrelation.

Wooldridge test for autocorrelation in panel data

H0: no first-order autocorrelation
F( 1, 8) = 4.718
Prob > F = 0.0616

Source: Stata 14.2 out put

Test for cross-section dependency

Cross-sectional dependence arises because of the presence of common shocks and unobserved
components that ultimately become part of the error term. If there is a cross-sectional dependency in

44
the model, the standard fixed effect and random effect estimators are consistent but not efficient, the
estimated standard errors are biased. Lagrange multiplier (LM) test, developed by Breusch and
Pagan (1980) is widely used in cross-sectional dependency test. Therefore, this study used Breusch
Pagan cross-sectional dependency test. As per the test below, Breusch Pagan cross-sectional
dependency test does not reject the null hypothesis of cross-sectional independence because of the
p-value in both testes more than 0.05 i.e. 0.4012. Hence, there is enough evidence suggesting the
absence of cross-section dependency in the model.

Breusch-Pagan LM test of independence: chi2(36) = 37.477, Pr = 0.4012

Based on 10 complete observations
Own computation using Stata 14.2 output

4.2.5. Analysis andInterpretationof Regression Result

Table 4. 7 Regression result

Variables Coefficient Std. Err. t-statistic p-value
CA -1.185 0.1541 -7.69 0.000***
AQ -0.70 0.4061 -1.72 0.085*
ME -0.563 0.05648 -9.97 0.000***
EQ -0.0045 0.0645 -0.07 0.9444
LIQ 0.1809 0.08582 2.11 0.035**
S -0.0495 0.05735 -0.86 0.388
Cons 70.089 5.3186 13.18 0.0000
R2 within = 0.6616 Wald chi2(6) = 161.98
2
R between = 0.6668 Prob > chi2 = 0.0000
2
R overall = 0.6610 Observation = 90
Note: ***,**, & * indicates significance level at 1%, 5% & 10% respectively.
Source: Own computation using Stata 14.2 output

Therefore, according to the above regression result, the model developed to analyze the effect of
ALM on the Ethiopian commercial banks’ financial performance in this study was:

ROE it = 70.089-1.1850CA-0.7000AQit-0.5631MEit-0.0045EQ+0.1809LIQit-0.0495S

45
Random effect Regression analysis of ROE

The regression Table 4.7 demonstrates the relationship between dependent variable (ROE) and six
independent variables. As it is observed from the regression output, between and the overall R-
square shows a value 0.6668 and 0.6610 respectively, which means 66 % variation on ROE was
explained by independent variables;management efficiency, capital adequacy, asset quality, earning
quality, liquidity and sensitivity of the banks collectively. The reaming 34 % of the variation is
unexplained by the employed explanatory variables. From the total of six independent variables four
variables were significant i.e. capital adequacy, asset quality, management efficiency and liquidity.
But two variables were found not having significant effect on return on equity of Ethiopian
commercial banks i.e. earning quality and sensitivity. From the random effect regression model the
probability values of F-statistic tells us the overall significance the model (all independent variables
altogether). Accordingly, the probability value of F-statisticwas0.0000 with Wald chi2(6) of 161.98
the p-value was smaller than 0.05, therefore, we can say that overall the model is significant at 1%
and the explanatory variables were jointly significant.

Capital Adequacy (CA)

As per table 4.7 presented that, the coefficient of capital adequacy measured by total capital to total
asset is -1.185. The coefficient is negative and has statistically significant effect on return on equity
at 1 percent level of significance. On the average holding other independent variables constant,
when capital adequacy increased or decreased by 1 % then return on equity of banks in the study
sample was decreased or increased by 1.185 %. Therefore, the study failed to accept the alternative
hypothesis that capital adequacy has a positive effect on return on equity in private commercial
banks of Ethiopia. This means, there is no sufficient evidence to support the positive relationship
between return on equity and capital adequacy.

The negative relationship is not as expected and this negative relationship between capital adequacy
and return on equity might be becausea bank with high capital adequacy ratio has low finical
performance i.e. return on equity. The possible reason for the negative relationship could be argued
that high capital adequacy leads to low profits since banks with high level of capital ignores the

46
potentials investment opportunities or simply risk averse. The finding of this study is in line with the
findings ofMulalem (2015). But the result of this study is inconsistent with the findings of
Jayanta(2012), Chen (2003), Serhat Yuksel,HasanDincerandUmitHacioglu (2015), Lizabeth Samuel
(2018), Mohammed Kamru Ahsan (2016), Zafar et al. (2017), NuraziaandEvans(2005), Tuffour,
Owusu, and Boateng, (2018), Saheed (2018), Adamu and kenenisa (2017), Anteneh (2018), Melaku
(2017) and Abdu (2018). They argued that banks with adequate level of capital are in good position
since it is necessary to prevent from failure and bankruptcy and then improve the performance of
banks.

Asset quality (AQ)

As presented in the above table the coefficient of asset quality measured by non-performing loans to
total loans is -0.7. The coefficient is negative and has statistically moderate significant effect on
return on equity at 10 percent significance level. On the average holding other independent variables
constant when, asset quality increased by 1 % then return on equity of banks in the study sample
was decreased by 0.7 %. Therefore, the study failed to accept the null hypothesis that asset quality
has a positive effect on return on equity of private commercial banks in Ethiopia. This means, there
is enough evidence to support the positive relationship between return on equity and asset quality.
The relationship is negative as expected and this negative relationship between asset quality and
return on equity could be attributed to the fact that a bank which has high NPLhas low level of
finical performance i.e. return on equity. The possible reason for the significant negative
relationship between asset quality and return on equity could be the low collectivity of the loans
with their interest income according to the schedule in Ethiopian commercial banks.

The finding of this study is consistent with the findings of Baral (2005), (Aburime, 2008), Koch
(1995), Olweny (2011), Jonathan (2013), Lizabeth Samuel (2018), Mohammed Kamru Ahsan
(2016), NuraziaandEvans(2005), Mulalem (2015), Adamu and kenenisa (2017), Anteneh (2018),
Melaku (2017), Abdu (2018) they proposed that the quality of asset depends on the level of
doubtful loans and outstanding loans and hence asset quality affect performance of the banks. The
lower the ratio of total doubtful loans to total loans means that the higher asset quality this in turn
leads to the favorable for bank performance by minimizing risk related to loan. Poor asset quality is
major causes of bank failures as a result high non-performing loan associated with low financial
performance.

47
Management Efficiency

As per what table 4.7 presented, the coefficient of management efficiency measured by total
operating cost to total operating income is -0.563. The coefficient is negative and has statistically
significant effect on return on equity at one percent level of significance. The coefficient of
management efficiency interpreted as holding constant other explanatory variables when
management efficiency increased by one percent, the level of bank performance in terms of return
on equity decreased by 0.563%. Therefore, the study failed to accept the null hypothesis that
management efficiency has a positive effect on return on equity in Ethiopia private commercial
banks. This means, there is enough evidence to support the positive relationship between return on
equity and management efficiency measured as the ratio of total operating cost to total operating
income. The relationship is negative as expected and this negative relationship between management
efficiency and return on equity is because of a bank which has high total operating cost related to
total operating income has low finical performance i.e return on equity.

The findings of this study is in line with the findings of Flaminietal. (2009), Sufianand Chong,2008
and Kosmidou etal, 2004, Abdulazeez, Asish, and Rohani (2017), Lizabeth Samuel (2018),
SheelaandBastray,(2014), Mohammed Kamru Ahsan (2016), Ahmad, Ahmad, and Adeel (2016),
NuraziaandEvans(2005), Ongore and Kusa (2013), Angele(2008), Gyekyi (2011), Tamiru(2013),
Mulalem (2015), Adamu and kenenisa (2017), Anteneh (2018), Melaku (2017) and Abdu
(2018).They argued that management efficiency is one of the key determinants of profitability of
banks. The lower ratio of total operating cost to total operating income leads to the higher
management efficiency in turn this leads to profitability and performance of banks. But the result of
this study inconsistent with the findings of Zafar et al. (2017).

Earnings quality

As presented in table 4.7, the coefficient of earning quality measured as interest income to total
income is -0.0045. The coefficient is negative and has statistically insignificant effect on bank
performance measured as return on equity. The coefficient interpreted as on the average holding
48
other explanatory variables constant when earning quality increased by one percent then return on
equity decreased by 0.0045% but the effect is statistically insignificance. Therefore, the study failed
to accept the alternative hypothesis that earning quality has a positive effect on return on equity in
Ethiopia private commercial banks. This means, there is enough evidence to support the positive
relationship between return on equity and earning quality.

The findings of this study is inconsistent with the findings of Hamad (2015) and Zubaid et al.
(2018), Mihail (2009), Mohammed Kamru Ahsan (2016), Saheed (2018), Anteneh (2018) and
Melaku (2017) they argued that earning quality explains the sustainability and growth of banks. It
will also determine the performance of banks; the higher earning quality reflects the current and the
future operational performance of banks; move in good way.

Liquidity

As table 4.7 above presented, the coefficient of liquidity measured as total loan to total deposit is
0.1809. The coefficient is positive and has statistically significant effect at 5 percent level of
significance on bank performance measured in return on equity. The coefficient interpreted as on the
average holding other explanatory variables constant when liquidity of banks increased by one
percent then the level of bank performance measured as return on equity increased by 0.1809%.
Therefore, the study failed to accept the null hypothesis that liquidity of banks has a negative effect
on return on equity in Ethiopia private commercial banks. This means, there is enough evidence to
support the negative relationship between return on equity and liquidity of banks measured as the
ratio of total loan to total deposit. The relationship is positive as expected and this positive
relationship between liquidity of banks and return on equity could be attributed to the fact that a
bank which has high total loan relative total deposit has high finical performance i.e return on
equity.

The finding of this study is consistent with the findings of Said (2003), Graham(2010), Mihail
(2009), Al-Tamimi(2010), SheelaandBastray,(2014), Mohammed Kamru Ahsan (2016), Ahmad,
Ahmad, and Adeel (2016), NuraziaandEvans(2005), Olwenyand Shipo(2011),
Kamau(2009), Onyekwelu, Chukwuani, and Onyeka (2018) Gweyi, Anteneh (2018) and Melaku
(2017) they argued that the ratio of total loan to total deposit reflects the most common liquidity
position of banks. The higher level of liquidity in the banks reflects, higher interest income from the

49
loan and then increase the bank performance.But the result is inconsistent with the findings of
Tobias, and Oloko (2016), Tuffour, Owusu, and Boateng, (2018), Abdu (2018).

Risk sensitivity

As presented in the table, the coefficient of risk sensitivity of banks measured as the ratio of risk
sensitivity assets to risk sensitivity liabilities is - 0.049. The coefficient is negative and has
statistically insignificant effect on bank performance measured in return on equity. On the average
holding other explanatory variable constants when, risk sensitivity of banks increased by one
percent then the level of bank performance measured in return on equity decreased by 0.0495 % in
the study sample.But the effect is statistically insignificant. Therefore, the study rejects the
alternative hypothesis that there is negative relationship between risk sensitivity of banks and return
on equity. This means, there is enough evidence to support the positive relationship between risk
sensitivity of banks and return on equity in Ethiopia private commercial banks.The result of the
study is in line with the findings of Suresh and Paul, 2014) but inconsistent with the findings of
Zafar et al. (2017).

50
 CHAPTER FIVE

SUMMARY CONCLUSION AND RECOMMENDATIONS

5.1.Introduction
Thischapterpresentssummaryofresearch findings,conclusionandrecommendations together
withsuggestionsforfutureresearch. Accordingly, the first part deals with summary of the study
findings, the second part put conclusion of the study as per the result, the third part presents the
study recommendations. Finally, suggestions for further research has been given on the last part of
this chapter.

5.2. Summary

The main objective of the study was to investigate the effect of asset liability management on
financial performance ofEthiopian commercial banks based on the CAMELS model. The following
hypotheses which were formulated in line with the specific objectives were tested.
1. There is significant and positive relationship between Capital adequacy and performance
2. There is significant and negative relationship between Asset quality and performance
3. There is significantand negative relationship between Management efficiency and
performance
4. There is significant and positive relationship between Earning quality and performance
5. There is significant and positive relationship between Liquidity and performance
6. There is significantand negative relationship between sensitivity to risk and performance

The research used quantitative approach and descriptive design. Descriptive statistics, correlation
and multiple regression analysis were conducted to analysis the panel,secondary data of 9 Ethiopian
commercial banks taken as a sample for the period of 2010 to 2019. The study used Stata software
version 14.2. the model selected as appropriate for the data was Random effect model. Since the
51
CLRM assumption tested were all found to be valid, the model was statistically appropriate to the
data assuring that the data was represented consistently. The variables studied was ROE as a
measure of performance and the CAMELS components; capital adequacy, asset quality,
management efficiency, earning quality, liquidity and sensitivity to risk.

The findings of the descriptive study depicted thatduring the study period; 2010-2019, the average
profitability of commercial banks of Ethiopia was 22.44 %. The standard deviation of ROE being
6.9% showed that there was low variation.The mean value of capitaladequacy being
14.73%indicates that around 15%ofthetotalassetsofthebankingsector
werefinancedbyshareholders’equity. The standard deviation of 3.1% showed that the capital
adequacy of the banks across the study period do not have very significant difference.The output for
asset quality of meanvalue1.5%andstandarddeviation 1.36% showedthattherewas almost no
variation of asset quality among Ethiopian commercial banks during the study period. Theaverage
value of management efficiencywas63.41%,anindicatorthatoverheadsarehighin
theEthiopianbankingsectorwhere the 11.29% standard deviation showed that there was a significant
variation among management efficiency of the sample Ethiopian commercial banks during the study
period.The mean value of earning quality being 63.19% and standard deviation of 11.79% depicted
that there was variation of earning quality among the sample commercial banks of Ethiopia during the
study period.Themeanvalueofliquidity being 61% displayed that the banking sector was very liquid
during the study period, The 61% mean value showed that 61% of the deposits of commercial banks
were converted to loan during the study period. The standard deviation of 8.52% showed that the
liquidityratiodiffersfrom bank to bank and from time to time within a specific bank. The mean value
of sensitivity was 84.27% and standard deviation of 14.8%. showing that there was very high
deviation from the mean in this regard; which showed that Ethiopian commercial banks had different
level of risk sensitivity during the study period.

Whereas, the regression result showed that the between and the overall R-square shows a value
0.6668 and 0.6610 respectively, which means 66 % variation on ROE were explained by
independent variables are management efficiency, capital adequacy, asset quality, earning quality,
liquidity and sensitivity of the banks collectively. The reaming 34 % of the variation is unexplained
by the employed explanatory variables. Capital adequacy, asset quality, management efficiency and
liquidity were found significant. Butearning quality and sensitivity were not significant. From the

52
random effect regression model the probability values of F-statistic tells us the overall significance
the model (all independent variables altogether). Accordingly, the probability value of F-
statisticwas0.0000 with Wald chi2(6) of 161.98 the p-value was smaller than 0.05, therefore,the
model overall is significant at 1% and the explanatory variables were jointly significant.

Regarding the relationship of the explanatory variables with the explained variable,Capital
adequacy, asset quality, management efficiency, earning quality and sensitivity were found to have
negative relationship with profitability while liquidity was found to be the only variable having
positive relationship with profitability.

5.3 Conclusion

As depicted by the R2and F-statistics,ROE was66% explained by independent variablesi.e earning

quality, capital adequacy, asset quality, management efficiency, liquidity and sensitivity of the banks
collectively and only 34% by other factors, this shows that asset liability management has vert great
effect on profitability of Ethiopian commercial banks.

Capital adequacy found to be statistically significant at 1% level of significance, has very great
effect on profitability witha negative relationship.Asset quality also has moderate effect on
profitability; it is found to be statistically significant at 10% level of significance having
negative relationship.Management efficiency is found to be the factor having the utmost effect
on profitability being significant at 1% level of significance and great correlation with
financial performance; have negative relationship. Earning quality and profitability have
negative relationship but earning quality is found to be statistically insignificant. Liquidity is
found to be the only factor that have positive relationship with financial performance with
statistical significance at 5% significance level. Sensitivity and profitability have negative
relationship but sensitivity is found to be statistically insignificant.

Based on the research findings, it has been concluded that capital adequacy, management efficiency
and liquidity are the most important factors that affect the financial performance of Ethiopian
commercial banks since they are significant at 1% and 5% significance level; where management
efficiency is the utmost important factor having greater correlation with profitability of Ethiopian
commercial banks. Asset quality is also one of the factors that affect financial performance of

53
commercial banks of Ethiopia; it was significant at 10% level of significance.

5.4. Recommendations

Since capital adequacy, asset quality, management efficiency and liquidity are the important factors
that affect the financial performance of Ethiopian commercial banks, bank managers should give
them due attention. Specially, great attention should be given for management efficiency to
maximize profitability.

1. The research findings revealed thatnow a days there are less strict credit risk management
practices.Thus,
Ethiopiancommercialbanksneedtoadvancetheircreditriskmanagementdimensions.
2. Since a significant variation among management efficiency of Ethiopian commercial banks
were observed and that management efficiency was found to be the utmost important factor
affecting profitability, banks should pay great attention to qualified and dedicated man power
and adherence of internal policies in this regard.
3. To use the opportunity of very rewarding investments and at the same time meet withdrawal
and credit needs of customers, commercial banks of Ethiopia must manage their liquidity
level.

5.5 Suggestion for further research

It will be of a great contribution to observethe financial performance of Ethiopian commercial banks
by incorporating some other explanatory variables. One of the limitations of this study was that it
focusedonly on bank specific factors. Though, there are different ratios that can be used in this
regard, it is difficult to consider all at a time. So, this study used only some of the ratios available.
Therefore, investigating the financial performance of Ethiopian commercial banks using different
ratios may indicate a different result. The findings of this study base only on secondary data
obtained from the NBE and is dependable only to the level of accuracy of the financial statements
used. Therefore, investigating the issue by supporting with primary data, will play a paramount
importance. At last since the CAMELS model can also be used to investigate financial performance
of Insurances and other financial institutions, considering this will have its significance for
managers of such institutions.

54
 REFERENCES

Abayomi,O.O.,&ShalemJ.F.(2001).AssetliabilitymanagementinNigerianBanks,TheJournaloftheNi
gerianinstituteofBankers,74(1).

Aburime,U.(2005)DeterminantsofBankProfitability:Company-
LevelEvidencefromNigeria.Nigeria: University ofNigeria,EnuguCampus.

Ahmed,A.M.&Khababa,N.(1999).PerformanceofthebankingsectorinSaudiArabia,JournalofFinanc
ial Management Analysis, 12 (2), 30-36.

Al-Tamimi,H.,Hassan,A.(2010)FactorsInfluencingPerformanceoftheUAEIslamicand
ConventionalNationalBanks..DepartmentofAccounting,FinanceandEconomics,
CollegeofBusinessAdministration,UniversityofSharjah. AndOff-
SiteMonitoringSystems”.JournalofFinancialServicesResearch,13(2),p.

Angele,L.(2008).AnalysisofchosenstrategiesofassetandliabilitymanagementincommercialBanks,J
ournalofEconomicsofEngineeringDecision,2,57.

Angelopoulos,P.&Mourdoukoutas,P.(2001).BankingRiskManagementinaGlobalizing
Economy,Greenwood PublishingGroup, Westport, CT.

Anteneh T. (2018) FinancialPerformanceanalysisofCommercialBanksinEthiopia:

ACAMELApproach. (Masterthesis).AddisAbebaUniversity.

Ashton,C.1998.BalancedScorecardbenefitsNatWestBank.HumanResource
ManagementInternationalDigest,6(3).

Aspal,P.K.andSanjeev,Dh.(2016).CamelsRating

Athanasoglou,P.P.,Sophocles,N.B.,Matthaios,D.D.(2005)Bank-specific,industry-specificand
macroeconomicdeterminantsofbankprofitability.Workingpaper,BankofGreece.1(1),3-4.

Baral,J(2005),HealthCheck-upofCommercialBanksintheFrameworkofCAMEL,ACase
StudyofJointVentureBanksinNepal.TheJournal ofNepaleseBusinessStudiesVolII.
(Noone)

BaselCommittee,(2006).InternationalConvergenceofCapitalMeasurementandCapitalStandards:AR
evisedFramework-ComprehensiveVersion.BaselCommitteeonBankingSupervision.

Bordelau,E.,Graham,C.(2010).TheImpactofLiquidityonBankProfitability.FinancialStabilityDepar
tmentBankofCanada.12-17.

55
Bourke,P.(2001).ConcentrationandotherdeterminantsofbankprofitabilityinEurope,North
Americaand Australia. Journal of Banking andFinance, 1(2)1-7

Brooks, C., 2008.IntroductoryEconomitricsForFinance. 2nded. NewYork.

CAMELApproach.(Masterthesis).AddisAbebaUniversity.

ChandaniA,MehtaM,ChandrasekaranKB.2014.AWorkingPaperontheImpactofGenderof
Chen,J.(2003).CapitaladequacyofChinesebanks:Evaluationandenhancement.

Cole,RebelA.andGunther,Jeffery(1998).“PredictingBankFailures:AComparisonofOn-

Creswell,JW2009,“ResearchDesign:quantitative,qualitativeandmixedmethodsapproaches,”3rdedn
,SagePublications,California.

DashM,DasA.ACAMELSAnalysisoftheIndianBankingIndustry.RetrievedMay24,2011,
fromSocial Science Research, 2009. 5

DoumposM,ZopounidisC.Amulticriteriadecisionsupportsystemforbankrating.Decision

Edgeworth,F.(1881)MathematicalPsychics:An Essay on theApplication of MathematicstotheMoral

Sciences.London.

Emmanuel, N.R,(1997).CommercialBankinginanEraofDeregulation, Greenwood

PublishingGroup, 3rd Ed.

FabozziF.J.andAtsuoK.,(2003),AssetandLiabilityManagement.S.Chandandcompany: Delhi.

Fabozzi,F.J.,&Konishi,A.(1995).Asset-liabilitymanagement. NewDelhi: S Chand&Co.

Flamini,V.(2009).TheDeterminantsofCommercial BankProfitabilityinSub-SaharanAfrica.IMF
workingpaper

Gujarati,DN2004,BasicEconometrics, 4thedn,McGrawHill,Boston

Gulgzoztorul(2011)“performanceevaluationofbanksandbankinggroups:Turkeycase,thesis,
MiddleEasttechnicaluniversity, 2011, turkey.

Guthua,A.M.(2012).Theeffectofassetliabilitymanagementonliquidityriskofcommercialbank in
kenya.

HaseebZamanBabar&GulZeb,2011,“CAMELSRatingSystemforBankingIndustryinPakistan”Mast
erThesis.UmeaSanrol ofBusiness

56
Hays,F.,Lurgio,S.,Gilbert,A.(2008).EfficiencyRatiosandCommunityBankPerformance.JournalofF
inanceandAccountancy,5(2),1-15.

Jappelli, T. &Pagano, M. (2002).Information sharing, lending and defaults:

crosscountryevidence, Journal ofBanking&Finance, 26, 17-45.

Jonker,J.andPenninkB.(2010).TheEssenceofResearchMethodology:AconciseguideforMastersand
PhDstudents in management science.journalof international bankingregulation, 4 (4),
320-327.

Kamau,A.W.(2009).EfficiencyintheBankingSector:AnEmpiricalInvestigationofCommercial
Banks in Kenya. UniversityofNairobi

Kaplan,R.S.&Norton,D.P.(1996b).UsingtheBalancedScorecardasaStrategic
ManagementSystem.HarvardBusinessReview,(January-February,1996).

Kongiri,A.(2012).EffectsofCamelVariablesonBankEfficiency:aPanelAnalysisofkenyanCommerci
alBanks.TheJournalofFinance,7-27.

Kosmidou, K. & Zopounidis, D. (2004). Goal Programming Techniques for Bank Asset Liability
Management. Volume 90. ISBN: 978-1-4020-8104-0 (Print) 978-1-4020-8105-7

Kosmidou, K. (2008).The determinants ofbanks’profits in Greece during the period

ofEUFinancial integration.Journal ofManagerial Finance.5(1),85-88.

Kumar,A.andDhar,J.(2014).AnEmpiricalAnalysisandComparativeStudyofLiquidityRatios and
Asset-LiabilityManagement ofBanksOperatinginIndia InternationalJournal ofSocial,
Behavioral, Educational, Economic, Business andIndustrial Engineering,Vol:8, No:1.

KwadwoBoateng(2018).
DeterminantsofBankProfitability:AComparativeStudyofIndianandGhanaianBanks.Interna
tional Journal of Emerging Technologyand InnovativeResearch, Vol.5Issue

LeaderontheFinancialPerformanceoftheBank:ACaseofICICIBank.Procedia
EconomicsandFinance,11, 459-471.

Mabvure, T.J., Gwangwava, E., Faitira, M., Mutibvu, C. &Kamoyo, M. (2012). Non- Performing
loans in Commercial Banks: A case of CBZ Bank Limited in Zimbabwe.
Interdisciplinary Journal of Contemporary Research in Business 4(7), 1-22.

Melaku,A.(2017).FinancialPerformanceAnalysisofPrivateCommercialBanksofEthiopia:CamelRatings
.InternationalJournalofScientificandResearchPublications,7(10),ISSN2250-3153.

57
Miletic,I.,(2009),“MacroeconomicandMicroeconomicCausesfortheInstabilityofBanks”,EconomicResearch,
22(1):47-59. ModelforEvaluatingFinancialPerformanceofBanking Sector: ATheoretical
Perspective.InternationalJournalofSystemModelingandSimulation,1(3).

MohammadKamrul Ahsan(2016).
MeasuringFinancialPerformanceBasedonCAMEL:AstudyonSelectedIslamicBanksin
Bangladesh. Asian Business Review Vol. 6No.1Issue 13

MulualemGetahun,2015“AnalysisofFinancialPerformanceofCommercialBanksinEthiopia,”
MasterThesisDept. ofAccounting andFinance

Mulualem,M.(2015).AnalyzingFinancialPerformanceofCommercialBanksinEthiopia: NBE
DirectivesNo.SBB/57/2014

NBE minimum capital requirement directives No MFI/27/2015

Nzongang,T.,&Atemnkeng,J.(2006).MarketStructureandProfitabilityPerformanceintheBankingIn
dustryofCFAcountries:theCaseofCommercialBanksinCameroon.

Olweny,T.,Shipho,T.M.(2011)EffectsofBankingSectoralFactorsontheProfitabilityof Commercial
Banks in Kenya. EconomicsandFinanceReview, 1(5), 1-30.

Ongore,V.O.(2011)Therelationshipbetweenownershipstructureandfirmperformance:
AnempiricalanalysisoflistedcompaniesinKenya.AfricanJournalofBusinessManagement,
5(6), 2120-2128.

Pareto,V.(1927).ManualofPoliticalEconomy.London/Basingstoke,Macmillan.

Prasanna Chandra (2011). FinancialManagement: TheoryandPractice

8thEdition.TataMcGrawHillNewDelhi, India

R.hoque,2012,dataenvelopmentanalysisofbankingsectorinBangladesh,InstituteofStatisticalResear
ch andTraining, Universityof Dhaka, Dhaka-1000, Bangladesh

RoliPradhan2014,ZScoreEstimationforIndianBankingSector,InternationalJournalofTrade,
Economicsand Finance,Vol. 5, No. 6, December 2014

Sahajwala,R.,&vanderBergh,P.(2000).SupervisoryRiskAssessmentandEarlyWarning
Systems.BaselCommitteeonBankingSupervisionWorkingPapers,No.4,BankofInternational
Settlements.

Said,M.(2003):“liquidity,SolvencyandEfficiency:AnempiricalanalysisoftheJapanesebanks
distress.” Journal ofOxford, Vol.5, No.3, pp. 354–358.

58
SangmiandNazir2010,AnalyzingFinancialPerformanceofCommercialBanksinIndia:
ApplicationofCAMELModel:http://www.jespk.net/publications/28.pdfretrievedonApril1,
2014

SerhatYuksel2015”CAMELS-basedDeterminantsfortheCreditRatingofTurkishDepositBanks”
InternationalJournal ofFinance &Banking StudiesVol 4, No 4, 2015 ISSN: 2147-4486

Sheela,PandBastray,T.(2014).effectofassetliabilitymanagementoncommercialbank
profitabilityinIndianfinancialmarket.InternationalJournalofBusinessandAdministration,
Vol.1,Issue.6, July-Sep, 2014.

Sufian, F & Chong, R (2008), “Determinant of Bank profitability in a developing economy:

empirical evidence from the Philippines’, Asian Academy of Management Journal of
Accounting & Finance, Vol. 4, no. 2, pp. 91-112.

SushendraKumarMisa&PerveshK.Aspel,(2013)“ACamelModelAnalysisofStateBankGroup”Worl
dJournal ofSocial Science,Vo.3No.4Issue,pp36-55.

Tamiru,B.(2013).AssetLiabilityManagementandCommercialBanksProfitabilityinEthiopia,
Journal ofFinance and Accounting, 4, 10, 2222-1697

UyenDang,2011,“TheCAMELRatingSysteminBankingSupervisionAcaseStudy”International
Business,ArcadaUniversityofAppliedSciences.

Vaidyanathan,R.,(1999).Asset-
liabilitymanagement:IssuesandtrendsinIndiancontext.ASCIJournalofManagement,29(1):39
-48.

Vossen,B.(2010).AstudyonBankLiquidityManagement.UniversityatAlbany,StateUniversity
ofNewYork.

Zafar,S.,Haida,M.,AfiridiA.S.,UroogeS.(2017).AnEmpiricalInvestigationofBankingSectorPerfor
manceofPakistanby
UsingCAMELSRatioofFramework.EmergingIssuesinEconomicsandFinance.JournalofMa
nagerialScienceVol.XINumber 03.

Zawalinska,K.(1999).AssetandLiabilityManagement.TheInstitutionalApproachtoALMby
CommercialBanksinPoland:aSpecialFocusonRiskManagement,CASE-CenterforSocial
ScienceResearch,Warsaw.

Appendices

Appendices I: List of Commercial Banks operating in Ethiopia

59
S.No Name of the Bank Year of Establishment

1 Commercial Bank of Ethiopia 1974

2 Awash International Bank 1994

3 Dashen Bank 1995

4 Bank of Abyssinia 1996

5 Wogagen Bank 1997

6 United Bank 1998

7 Nib International Bank 1999

8 Cooperative Bank of Oromia 2004

9 Lion International Bank 2006

10 Oromia International Bank 2008

11 Zemen Bank 2008

12 Bunna International Bank 2009

13 Birhan International Bank 2009

14 Abay Bank 2010

15 Addis International Bank 2011

16 Debub Global Bank 2012

17 Enat Bank 2012

Source: www.nbe.gov.et

Appendices II: Ratios considered in the study

BANK YEAR ROE CA AQ ME EQ LIQ S

AIB 2010 29.30 7.04 4.71 48.91 44.20 51.50 68.29

60
AIB 2011 32.10 13.21 3.64 45.55 42.60 51.50 70.31
AIB 2012 27.00 13.83 2.70 52.23 60.20 59.80 88.14
AIB 2013 28.00 13.91 2.30 56.20 59.80 61.50 81.42
AIB 2014 27.20 12.97 2.27 56.87 56.70 61.00 87.08
AIB 2015 23.00 13.34 1.74 62.57 63.50 67.40 95.56
AIB 2016 21.50 13.29 1.53 65.09 68.10 67.70 90.27
AIB 2017 23.70 11.46 1.46 64.12 68.70 73.80 95.52
AIB 2018 31.30 11.75 0.82 63.72 77.70 72.00 91.19
AIB 2019 41.10 12.91 0.86 58.49 74.20 79.30 98.55
BOA 2010 25.45 10.96 7.41 58.13 55.84 61.36 56.82
BOA 2011 29.04 10.91 3.33 58.19 60.20 54.58 66.21
BOA 2012 27.60 12.96 2.57 60.09 68.80 57.56 77.68
BOA 2013 21.48 12.89 1.99 60.09 68.80 55.34 83.15
BOA 2014 33.94 15.36 0.00 55.82 61.96 55.64 87.34
BOA 2015 17.47 12.82 0.00 69.02 70.70 53.11 81.77
BOA 2016 18.33 14.29 0.00 71.37 67.21 58.76 86.00
BOA 2017 22.68 13.05 0.00 69.22 66.93 67.28 92.42
BOA 2018 15.74 14.40 1.17 76.61 83.12 69.75 94.43
BOA 2019 16.90 13.85 1.32 76.12 81.71 73.83 96.70
CBO 2010 14.50 11.81 2.53 71.87 58.30 52.60 53.00
CBO 2011 22.00 11.39 2.00 64.51 49.00 40.00 52.53
CBO 2012 31.00 13.52 1.44 54.02 57.00 49.00 70.25
CBO 2013 37.00 12.88 1.72 47.73 44.00 47.00 66.76
CBO 2014 38.00 18.43 0.00 46.14 48.00 67.00 83.07
CBO 2015 25.00 14.70 0.00 60.77 57.00 89.00 110.32
CBO 2016 3.00 11.76 0.00 96.54 77.00 70.00 91.40
CBO 2017 13.00 9.59 0.00 82.28 77.00 68.00 91.32
CBO 2018 23.00 8.63 0.00 76.65 74.00 57.00 78.20
CBO 2019 23.00 8.67 3.41 79.34 76.00 61.00 82.30
DB 2010 31.89 10.64 2.18 52.48 50.05 49.77 62.74
DB 2011 35.77 11.20 1.99 50.87 47.08 52.51 65.80
DB 2012 40.44 13.33 2.15 48.23 52.03 57.76 76.83
DB 2013 31.33 13.10 2.25 55.25 56.18 55.91 78.34
DB 2014 30.69 14.50 0.00 55.36 53.19 55.33 76.63
DB 2015 26.41 14.01 1.68 61.68 56.23 58.18 86.77
DB 2016 23.15 13.66 1.71 65.22 55.67 55.78 83.34
DB 2017 20.57 13.17 2.02 71.31 60.62 65.09 93.58
DB 2018 18.84 16.56 0.98 74.19 73.22 64.71 94.76
DB 2019 16.00 13.31 0.65 76.11 80.33 72.84 102.09
LIB 2010 18.43 17.73 1.62 54.92 50.77 57.39 56.56
LIB 2011 15.00 21.70 1.45 56.18 54.00 52.00 64.17

61
LIB 2012 19.00 20.62 1.55 52.32 53.00 56.00 75.11
LIB 2013 23.00 21.80 1.30 49.26 57.00 63.00 91.33
LIB 2014 17.00 19.85 0.00 62.81 61.00 57.00 84.36
LIB 2015 21.00 16.89 0.00 64.72 51.00 64.00 91.82
LIB 2016 20.74 15.80 0.00 68.23 58.92 67.95 94.86
LIB 2017 21.32 15.19 0.00 66.34 73.24 62.52 90.05
LIB 2018 23.99 14.21 2.48 68.13 77.82 64.96 92.04
LIB 2019 24.68 14.43 1.94 68.64 75.40 72.29 97.01
NIB 2010 24.42 15.46 3.90 48.75 47.85 61.69 71.48
NIB 2011 23.61 16.84 4.12 47.60 50.69 53.64 66.35
NIB 2012 21.21 18.94 2.71 48.71 57.10 63.53 87.53
NIB 2013 18.75 18.94 2.50 53.98 67.03 68.26 98.46
NIB 2014 16.38 18.87 0.00 54.10 67.03 68.25 102.98
NIB 2015 16.28 17.26 0.00 63.53 73.47 70.53 110.02
NIB 2016 16.60 16.61 0.00 65.63 79.74 60.47 95.03
NIB 2017 16.21 14.05 0.00 64.99 77.00 65.25 97.09
NIB 2018 16.26 13.91 0.00 73.55 83.67 62.44 91.59
NIB 2019 18.50 14.54 0.98 72.16 85.70 70.27 94.67
OIB 2010 12.00 18.95 1.14 73.20 37.90 44.90 48.43
OIB 2011 17.50 16.83 1.06 62.98 40.50 43.40 64.07
OIB 2012 13.50 16.90 1.29 71.59 53.40 48.20 81.74
OIB 2013 13.60 15.48 1.46 72.54 58.70 53.10 87.97
OIB 2014 23.70 14.04 0.00 62.33 59.90 50.60 73.56
OIB 2015 17.40 12.07 0.00 71.77 67.70 64.60 91.90
OIB 2016 13.50 13.33 0.00 78.62 73.60 55.30 81.38
OIB 2017 19.30 11.54 1.86 74.65 61.00 53.50 78.43
OIB 2018 34.20 12.50 0.79 62.00 66.50 58.60 82.87
OIB 2019 23.70 13.61 1.56 69.01 74.30 65.80 91.70
UB 2010 30.10 13.04 3.65 51.50 49.20 55.30 53.31
UB 2011 30.00 13.95 2.77 48.85 54.00 54.00 65.90
UB 2012 30.00 15.29 2.33 51.12 62.00 60.50 83.34
UB 2013 19.00 14.54 1.86 66.24 66.00 58.40 93.30
UB 2014 14.00 15.28 1.44 70.72 75.00 56.90 89.10
UB 2015 17.00 13.21 1.22 73.14 71.00 58.10 92.41
UB 2016 18.00 13.47 1.30 74.40 73.00 65.50 98.73
UB 2017 17.00 12.79 1.24 102.69 77.00 72.70 101.65
UB 2018 21.00 11.85 1.30 75.48 79.00 65.30 92.02
UB 2019 22.00 11.82 0.51 78.10 85.00 74.70 100.16
WB 2010 25.10 21.23 3.97 43.83 43.73 63.06 65.66
WB 2011 23.66 19.60 4.54 43.78 38.64 48.85 61.74
WB 2012 27.06 22.23 2.43 46.10 51.95 61.92 91.54

62
WB 2013 22.86 20.06 2.24 52.38 61.55 62.12 95.07
WB 2014 19.99 21.20 1.67 62.55 61.75 54.92 54.11
WB 2015 15.34 19.54 0.00 65.88 64.36 61.51 104.26
WB 2016 15.46 19.07 0.00 68.79 66.81 67.75 107.56
WB 2017 14.39 17.92 0.00 66.96 62.76 73.01 109.86
WB 2018 17.27 15.29 1.75 65.99 68.52 73.38 103.46
WB 2019 22.10 15.79 2.16 77.09 78.49 69.87 98.00

Appendices III: Descriptive statistics Stata 14.2 output

63
. xtsum roe ca aq me eq liq s

Variable Mean Std. Dev. Min Max Observations

roe overall 22.43922 6.960336 3 41.1 N = 90

between 3.480648 18.822 28.42 n = 9
within 6.128333 2.489222 37.79922 T = 10

ca overall 14.73467 3.177325 7.04 22.23 N = 90

between 2.52367 12.138 19.193 n = 9
within 2.090581 9.403667 21.02667 T = 10

aq overall 1.518 1.364375 0 7.41 N = 90

between .4329301 .916 2.203 n = 9
within 1.30117 -.358 7.149 T = 10

me overall 63.41989 11.29641 43.78 102.69 N = 90

between 4.758647 57.375 69.869 n = 9
within 10.35635 41.57489 96.88589 T = 10

eq overall 63.19511 11.79186 37.9 85.7 N = 90

between 4.379834 58.46 69.12 n = 9
within 11.03652 41.74511 85.06511 T = 10

liq overall 61.09356 8.524998 40 89 N = 90

between 3.368476 53.8 64.55 n = 9
within 7.9042 41.03356 90.03356 T = 10

s overall 84.27356 14.81003 48.43 110.32 N = 90

between 4.68282 77.915 91.52 n = 9
within 14.1289 49.25756 116.6786 T = 10

. sum roe ca aq me eq liq s

Variable Obs Mean Std. Dev. Min Max

roe 90 22.43922 6.960336 3 41.1

ca 90 14.73467 3.177325 7.04 22.23
aq 90 1.518 1.364375 0 7.41
me 90 63.41989 11.29641 43.78 102.69
eq 90 63.19511 11.79186 37.9 85.7

liq 90 61.09356 8.524998 40 89

s 90 84.27356 14.81003 48.43 110.32

64
Appendices IV: Correlation Matrix b/n dependent and independent variables
ROE CA AQ ME EQ LIQ S
ROE 1.00 -0.2197 0.3272 -0.6446 -0.3713 -0.1225 -0.2681
CA -0.2197 1.00 -0.1075 -0.3405 -0.2655 0.0016 0.0673
AQ 0.3272 -0.1075 1.00 -0.4494 -0.4281 0.2831 -0.05210
ME -0.6446 -0.3405 -0.4494 1.00 0.6895 0.3218 0.4024
EQ -0.3713 -0.2655 -0.4281 0.6895 1.00 0.6084 0.6691
LIQ -0.1225 0.0016 -0.2831 0.3218 0.6084 1.00 0.07908
S -0.2681 0.0673 -0.05210 0.4024 0.6691 0.07908 1.00

65
Appendices V: Random and Fixed Effect models’ output /Hausman test/
. xtreg roe ca aq me eq liq s, fe

Fixed-effects (within) regression Number of obs = 90

Group variable: bank1 Number of groups = 9

R-sq: Obs per group:

within = 0.6688 min = 10
between = 0.5286 avg = 10.0
overall = 0.6316 max = 10

F(6,75) = 25.25
corr(u_i, Xb) = -0.1353 Prob > F = 0.0000

roe Coef. Std. Err. t P>|t| [95% Conf. Interval]

ca -.8976087 .2343539 -3.83 0.000 -1.364466 -.4307518

aq -.5191734 .4440961 -1.17 0.246 -1.403858 .3655114
me -.6534199 .0635456 -10.28 0.000 -.7800092 -.5268307
eq .1243775 .0729163 1.71 0.092 -.0208791 .2696342
liq .1972556 .0882326 2.24 0.028 .0214873 .3730239
s -.0620463 .0587814 -1.06 0.295 -.1791448 .0550522
_cons 63.21088 6.674435 9.47 0.000 49.91472 76.50703

sigma_u 2.4766868
sigma_e 3.8417134
rho .29359374 (fraction of variance due to u_i)

F test that all u_i=0: F(8, 75) = 2.78 Prob > F = 0.0094

. estimate store fe

66
. xtreg roe ca aq me eq liq s, re

Random-effects GLS regression Number of obs = 90

Group variable: bank1 Number of groups = 9

R-sq: Obs per group:

within = 0.6616 min = 10
between = 0.6668 avg = 10.0
overall = 0.6610 max = 10

Wald chi2(6) = 161.98

corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0000

roe Coef. Std. Err. z P>|z| [95% Conf. Interval]

ca -1.091099 .187269 -5.83 0.000 -1.45814 -.7240585

aq -.6806271 .4216626 -1.61 0.106 -1.507071 .1458164
me -.6075555 .0598423 -10.15 0.000 -.7248443 -.4902668
eq .055691 .0683649 0.81 0.415 -.0783018 .1896838
liq .1899193 .087011 2.18 0.029 .0193808 .3604579
s -.0600975 .058029 -1.04 0.300 -.1738323 .0536373
_cons 68.02288 5.850727 11.63 0.000 56.55567 79.4901

sigma_u 1.5119284
sigma_e 3.8417134
rho .13411377 (fraction of variance due to u_i)

. estimate store re

. hausman fe re

Coefficients
(b) (B) (b-B) sqrt(diag(V_b-V_B))
fe re Difference S.E.

ca -.8976087 -1.091099 .1934904 .1408974

aq -.5191734 -.6806271 .1614537 .1393628
me -.6534199 -.6075555 -.0458644 .0213762
eq .1243775 .055691 .0686865 .0253578
liq .1972556 .1899193 .0073363 .0146311
s -.0620463 -.0600975 -.0019488 .0093746

b = consistent under Ho and Ha; obtained from xtreg

B = inconsistent under Ha, efficient under Ho; obtained from xtreg

Test: Ho: difference in coefficients not systematic

chi2(6) = (b-B)'[(V_b-V_B)^(-1)](b-B)
= 3.65
Prob>chi2 = 0.7237
(V_b-V_B is not positive definite)

67
Appendices VI: Cross-sectional Indpependence test
. xttest2

Correlation matrix of residuals:

__e1 __e2 __e3 __e4 __e5 __e6 __e7 __e8 __e9

__e1 1.0000
__e2 -0.1374 1.0000
__e3 0.6166 0.2605 1.0000
__e4 -0.5316 0.2846 0.1024 1.0000
__e5 0.4966 -0.2284 0.0481 -0.3949 1.0000
__e6 0.4786 -0.5904 -0.1067 -0.3525 0.1574 1.0000
__e7 0.1919 0.0061 0.1075 -0.1062 -0.0717 0.2915 1.0000
__e8 -0.2355 -0.1109 -0.5188 -0.1923 0.0748 -0.1708 0.2475 1.0000
__e9 0.6100 0.2040 0.5423 -0.3084 0.1562 0.1090 -0.3157 -0.3911 1.0000

Breusch-Pagan LM test of independence: chi2(36) = 37.477, Pr = 0.4012

Based on 10 complete observations

68
Appendices VII- Regression Output
Since there is a Heteroskedasticity problem in this model and to correct the problem it is
recommended that to use GLS (Generalized least square) regression analysis in panel data.
The output presented in the following form.
. xtgls roe ca aq me eq liq s

Cross-sectional time-series FGLS regression

Coefficients: generalized least squares

Panels: homoskedastic
Correlation: no autocorrelation

Estimated covariances = 1 Number of obs = 90

Estimated autocorrelations = 0 Number of groups = 9
Estimated coefficients = 7 Time periods = 10
Wald chi2(6) = 180.32
Log likelihood = -252.3308 Prob > chi2 = 0.0000

roe Coef. Std. Err. z P>|z| [95% Conf. Interval]

ca -1.185089 .1541514 -7.69 0.000 -1.48722 -.8829579

aq -.7000735 .4061453 -1.72 0.085 -1.496104 .0959567
me -.563192 .0564806 -9.97 0.000 -.6738919 -.4524921
eq -.0045142 .0645653 -0.07 0.944 -.1310597 .1220314
liq .1809124 .0858282 2.11 0.035 .0126922 .3491327
s -.04954 .057358 -0.86 0.388 -.1619595 .0628795
_cons 70.089 5.318691 13.18 0.000 59.66456 80.51344

69