i

Statement of Declaration

I, Mulualem Gizaw declared that this thesis with a topic of “Measuring Technical and Scale
Efficiency of Private Commercial Banks in Ethiopia: Using Data Envelopment Analysis (DEA)”
is my original work and has not been presented to any other university to earn a degree or
diploma.

___________________________

Mulualem Gizaw
 ii

Statement of Certification

This is to certify that the thesis prepared by Mulualem Gizaw, entitled: “Measuring Technical
and Scale Efficiency of Private Commercial Banks in Ethiopia: Using Data Envelopment
Analysis (DEA)” and submitted in partial fulfillment of requirements for Degree of Master of
Business Administration in Finance compiles with the regulation of the University and meets the
accepted standard with respect to originality and quality under the supervision of Habtamu
Berhanu (PhD)

Advisor Habtamu Berhanu (PhD) Signature ____________ Date________

Examiner Signature _____________Date_______

Examiner Signature _____________Date_______

 iii

Abstract

As banks are the main actors in the financial sector of Ethiopia, their efficiency should be
measured appropriately. The main objective of this study is to measure technical and scale
efficiency of private commercial banks in Ethiopia for 10 years period from 2009 to 2018. And
the study used secondary source of data observes from all 16 private commercial banks annual
report and from the National bank of Ethiopia. To measure technical and scale efficiency of the
banks, the study used Data Envelopment Analysis method (DEA). Form the period of 10 years
under the study only in two years in 2012 and in 2013 from the entire banks 50% of the sample
banks fully efficient. The finding of the study revealed that ADIB, AIB, BB, DGB, EB and ZB
were 100% efficient banks and OIB is the least efficient bank in resource utilization and WB
were scale inefficient from the entire sample banks. The mean efficiency of PCBs for OTE, PTE,
and SE is 91%, 92.33% and 98.56% respectively. Also the finding shows that all inefficient
banks need improvement in both input variables, for DEPO the banks slacks values is lies
between 3.90% and 23.70%, in case of NON-IE the slack value is lies between 2.70% and
23.70%. For NON-II (output variable) only five banks need improvements which are BOA,
BUIB, CBO, NIB and UB with a percentage of 16.85%, 9.23%, 11.53%, 19.31% and 18.48%
respectively. According to the outcome of the study mostly the cause of overall technical
inefficiency is managerial inefficiency (inefficient utilization of resource) rather than scale
inefficiency. The researcher recommends that inefficient banks should learn from the efficient
banks especially from ZB and ADIB and banks need to giving updated training to their
manager’s.
 iv

Acknowledgements

First of all to lord Jesus I say thank you lord Jesus not only for this but for all of my life.

Honestly I would like to thank my adviser Habtamu Brehanu (PhD) for his valuable contribution,
for his comment and improvement.

I thank my family specially my wife for supported me psychologically, covering my home duty
and prayed for me.
 v

Table of Contents
List of Contents Page

Statement of Declaration..................................................................................................................i
Statement of Certification................................................................................................................ii
Abstract..........................................................................................................................................iii
Acknowledgements........................................................................................................................iv
List of tables..................................................................................................................................vii
List of figures...............................................................................................................................viii
Abbreviation and Acronyms...........................................................................................................ix
1. INTRODUCTION.......................................................................................................................1
1.1 Background of the study........................................................................................................1
1.2 Overview of banking in Ethiopia...........................................................................................2
1.3 Problem statement..................................................................................................................4
1.4. Objective of the study...........................................................................................................5
1.4.1. General objective............................................................................................................5
1.4.2. Specific objective...........................................................................................................5
1.5. Research question.................................................................................................................6
1.6. Significance of the study.......................................................................................................6
1.7. Scope of the study.................................................................................................................6
1.8. Organization of the study......................................................................................................6
2. LITERATURE REVIEW............................................................................................................7
2.1 Introduction............................................................................................................................7
2.2 Efficiency: Concepts and Definitions....................................................................................7
2.3 Efficiency classification.........................................................................................................8
2.3.1. Technical Efficiency.......................................................................................................9
2.3.2. Scale Efficiency............................................................................................................10
2.4 Measuring bank efficiency...................................................................................................11
2.5 Data Envelopment Analysis (DEA).....................................................................................11
2.6 Empirical Review.................................................................................................................12
2.7. Conceptual framework........................................................................................................17
3. RESEARCH METHODOLOGY..............................................................................................18
 vi

3.1 Introduction..........................................................................................................................18
3.2 Research Design and Approach...........................................................................................18
3.3 Population and sample.........................................................................................................19
3.4 Data source and collection...................................................................................................19
3.5 Method of data analysis.......................................................................................................19
3.5.1 Constant return to scale (CRS) model:..........................................................................20
3.5.2 Variable return to scale (VRS) Model:..........................................................................21
3.6 Model variables (input and output)......................................................................................21
3.6.1. Production approach:....................................................................................................21
3.6.2. Intermediate approach:.................................................................................................22
3.7 Model specification..............................................................................................................22
4. Empirical result and interpretation.........................................................................................24
Introduction................................................................................................................................24
4.1 descriptive statistics of input and output variables..............................................................24
4.2 Input Output correlation analysis.........................................................................................26
4.3. Efficiency score of Ethiopian private commercial bank from 2009 to 2018..................27
4.4 Scale efficiency of Ethiopian private commercial banks.....................................................37
4.5. Mean efficiency score of overall technical, pure technical and scale for each bank..........38
4.6. Slack value..........................................................................................................................46
5. SUMMARY OF FINDING CONCLUSION AND RECOMMENDATION...........................48
5.1 Introduction..........................................................................................................................48
5.2 Summary of findings............................................................................................................48
5.3 Conclusion...........................................................................................................................49
5.4 Recommendation.................................................................................................................50
5.5 Suggestion for feature research............................................................................................51
References......................................................................................................................................52

List of tables
 vii

Pages

Table 4.1 descriptive statistics of input and output variables………………………..….....24

Table 4.2 input output correlation analysis………………………………………………....26

Table 4.4 Scale efficiency of Ethiopian private commercial banks………………………..37

Table 4.5 mean OTE, PTE and SE…………………………………………………………39

Table 4.6 Slack value……………………………………………………………………….46

List of figures
 viii

Page

Figure 2.1 Technical and Allocative Efficiencies…………………………………………....9

Figure 2.2 Scale Efficiency………………………………………………………………......10

Figure 4.1: efficiency of PCBs in CRS and VRS models for the year 2009…………………27

Figure 4.2: efficiency of PCBs in CRS and VRS models for the year 2010…………………28

Figure 4.3: efficiency of PCBs in CRS and VRS models for the year 2011………………....29

Figure 4.4: efficiency of PCBs in CRS and VRS models for the year 2012………………....30

Figure 4.5: efficiency of PCBs in CRS and VRS models for the year 2013………………....31

Figure 4.6: efficiency of PCBs in CRS and VRS models for the year 2014………………....32

Figure 4.7: efficiency of PCBs in CRS and VRS models for the year 2015…………………33

Figure 4.8: efficiency of PCBs in CRS and VRS models for the year 2016…………………34

Figure 4.9: efficiency of PCBs in CRS and VRS models for the year 2017………………....35

Figure 4.10: efficiency of PCBs in CRS and VRS models for the year 2018………………..36

Figure 4.1 mean OTE, PTE and SE graph……………………………………………………38

Abbreviation and Acronyms

 ix

AB ABAY BANK

ADIB ADDIS INTERNATIONAL BANK

AWB AWASH BANK

BB BERHAN BANK

BCC BANKER CHARNES CUPER

BOA BANK OF ABYSSINIA

BUIB BUNNA INTERNATIONAL BANK

CBO COOPERATIVE BANK OF OROMIA

CRS CONSTANT RETURN TO SCALE

DB DASHEN BANK

DEA DATA ENVELOPMENT ANALYSIS

DEPO DEPOSIT

DFA DISTRIBUTION FREE APPROACH

DGB DEBUB GLOBAL BANK

DMU DECISION MAKING UNIT

EB ENAT BANK

FDH FREE DISPOSAL HULL ANALYSIS

FRA FINANCIAL RATIO ANALYSIS

LIB LION INTERNATIONAL BANK

MAX MAXIMUM

MIN MINIMUM

NIB NIB INTERNATIONAL BANK

NON-IE NON-INTEREST EXPENSES

NON-II NON-INTEREST INCOME

OBS OBSERVATION
 x

OIB OROMIA INTERNATIONAL BANK

OTE OVER ALL TECHNICAL EFFICIENCY

TFA THICK FRONTIER APPROACH

PCBS PRIVATE COMMERCIAL BANKS

PTE PURE TECHNICAL EFFICIENCY

SD STANDARD DEVIATION

SE SCALE EFFICIENCY

SFA STOCHASTIC FRONTIER ANALYSIS

TOL TOTAL LOAN

UB UNITED BANK

VRS VARIABLE RETURN TO SCALE

WB WEGAGEN BANK

ZB ZEMEN BANK
 1

CHAPTER ONE

1. INTRODUCTION

1.1 Background of the study

A financial sector which allocates resources efficiently is the engine that drives economic growth
of any country (Kamau, 2011). Strong financial system promotes investment by financing
productive business, mobilizing savings, facilitating trade activities and the financial sector as a
whole plays a key role in allocating the economy’s financial resources (Kizito, 2012). Banking
industry is one of significant sectors of the financial system in most countries (San & Heng,
2013). Commercial Banks plays an imperative part within the economic improvement of the
nations by allocates resource and by interfacing financial specialists with savers (Okoth et al.
2013 as cited in Kokobe & Birhanu 2015). Evaluating economic performance of banks is
important to society because if the financial institutions operate more efficiently, they will earn
greater profit and increase liquidity into the economy (Nguyen, 2007).

Well-functioning banking sector facilitate economic progresses, whereas poorly functioning

banking sector is problem to economic development and aggravate poverty (Rajha 2016). The
banking sector is a fundamental component of the financial system and its efficiency is important
for promoting access to financial services as well as stability of the economy (Kamau, 2011).
The sector plays an important role in the mobilization and allocation of savings. It plays the role
of mediator between the net savers and net borrowers and the gains to the real sector, is depend
on how the financial sector performs their function of intermediation efficiently (Kumar &
Singh, 2015). The effective financial intermediation mechanism distributes the credit to more
profitable segments in ideal way. In addition, this well-organized financial intermediation
mechanism also encourages innovations, because of high return on investment, with positive
implications for economic development (Luccheti, 2000).

Banking industry has significant contribution in development of the economies of developing

countries (Takbiri et al., 2015). As Kablan, (2010) mentioned in his study, in sub-Saharan Africa,
banks are the most important element of the financial system. In many countries, other financial
structures are underdeveloped or almost nonexistent. In the case of Ethiopia, banks, insurance
 2

companies and micro-finance institutions are the major financial institutions and banks dominate
the financial sector.

In Ethiopia banks are the only important formal organization which can provide finance for
firms. And Bank efficiency in Ethiopia is compelling agenda of concern for most of investor
invests in the industry and for the government and other organizations also (Gamachis, 2016).
Efficiency in the banking system shows improved profitability, ensures stability, and enhances
public confidence. Also, it increases the volume of funds intermediated, allocates resources
efficiently, induces liquidity, and facilitates better quality services for customers (Sufian &
Chong, 2008). The intensive and continuously increasing competition in the financial services
market creates a need for an access to information that would allow evaluating commercial banks
operating in this market. Such evaluations are really essential to both bank owners and customers
who expect high-level financial profits (Wozniewska, 2008). In much less monetized countries,
like Ethiopia, while monetary area is dominated by using banking industry, efficient and
effective functioning of the Banks has vast position in accelerating financial growth (Fentaw &
Sharma, 2017).

As banks dominate the financial sector in Ethiopia, ensuring the financial health of these
institutions is likely going to ensure the health of the performance of the financial sector of the
country (Abebaw and Kapur, 2012).

Therefore this paper intends to measure the intermediation efficiency of Ethiopian private
commercial banks, interims of technical and scale efficiency which shows can the banks
produced a maximum outputs from a minimum quantity of inputs or not and optimal activity
level of the firm respectively. Data Envelopment Analysis (DEA) method is used for the purpose
of classifying banks as relative efficient and inefficient. Because DEA is a powerful optimization
tool used to measure the efficiency of any sectorial unit in terms of both technical and scale
efficiency (Chandrasekar, et al., 2017).

1.2 Overview of banking in Ethiopia

The history of banking in Ethiopia is beginning from 1906. But, earlier than the introduction of
the modern banking system, ‘Equb’ and ‘Idir’ which are a kind of traditional financial group
helped to growing saving habits and insure the financial want of the society. In 1906 emperor
 3

Menelik II and Ma Gillivray, representative of the British owned National Bank of Egypt, signed
an agreement and this agreement has made true the opening of the first financial institution of
Ethiopia referred to as Bank of Abyssinia in 1906 (Muhabie, 2015). Until 1931 the financial
sector in Ethiopia was dominated by foreign ownership, in 1931 by nationalizing Abyssinian
bank and renamed as bank of Ethiopia it become the first nationally owned bank for the country
and for the continent also (Belay, 1990, Befekadu, 1995, as cited by Tony and Alemayehu, 2001)
In Italian invasion of 1935 put a different complexion on the evolution of banking in Ethiopia.
The operations of Bank of Ethiopia were replaced by subsidiaries of the Italian parent banks
Banco d' Italia, Banco di Roma Banco, di Napoli and Banco Nazionale del Lavaro. After the end
of the fascist occupation, it was the turn of a British-based bank, Barclays Bank, to set up shop
on Ethiopian soil in 1941; but it was shortly afterwards (1943) replaced by the state Bank of
Ethiopia. (Tekle Birhan, 2007).

The State Bank of Ethiopia operated as both a commercial and central bank until 1963. After
banking proclamation issued in 1963 it divided into central and commercial banking as the
National Bank of Ethiopia and the Commercial Bank of Ethiopia. Also the establishment of
private local banks and the entry of foreign banks was permitted in the new proclamation
(Alemayehu, 2006 & Solomon, 2011, cited in Habtamu, 2015).

All of this changed with the 1974 revolution. Under the Derg regime in January 1 1975 all
privately owned financial institutions were nationalized which includes three commercial banks,
thirteen insurance companies and two non-bank financial intermediaries (Befekadu, 1995,
Harvey 1996 as cited by Tony and Alemayehu, 2001). And the competitive banking
circumstances that begun to develop in the 1960s and 1974s was changed by the Derg rule over
the 1974-1991 periods. After the change of government in 1991, the command economic policy
was changed and the financial markets were deregulated (Aderaw & Singh, 2016).

Proclamation No. 84/1994 that allowed the private sector to engage in the banking business
marked the beginning of a new era in Ethiopian banking. Now days there are 16 privet
commercial banks operating in the country.
 4

1.3 Problem statement

An efficient banking system plays a great role for a progressive economic growth of any country
(Kumar & Singh, 2014). In the financial system of the country efficiency of the banking industry
is an important issues since the success of the whole monetary system and the stability of the
banking sector manly affected by such efficiency (Yilmaz, 2013 as cited in Kocisova, K. 2013).
Many researchers conclude that for the performance of the whole economy, for effective
implementation of the monetary policy and for effective payment system of any country the
efficiency of banking sector has a significant role (Gulati, 2011). A well-functioning financial
sector facilitates efficient intermediation of financial resources. Banks are the highest financial
intermediaries in the economy. So efficiency analysis is essential for the evaluation of banks’
performance (Wozniewska, 2008).

Traditional financial ratio analysis (FRA) and frontier analysis method like data envelopment
analysis (DEA) and stochastic frontier analysis (SFA) are mostly used methods to studying bank
efficiency (Mousa, 2015). In Ethiopia many of studies on bank efficiency was done in financial
ratio analysis (FRA) analysis such as, Dakito, (2015), Melaku, (2017), Gudata, (2015), Adamu
and Kenenisa, (2017), and Rahel and Maru, (2015).

Financial ratio analysis regularly helps to examine the financial soundness of the bank and its
management quality. Bank regulators, for example, national bank of Ethiopia use financial ratio
analysis to evaluating a bank’s performance. According to Stainer (1997), as cited in Yannick et
al. (2016) this ratio analysis face a fundamental problem if there is external factors which affect
their computation and FRA have no relationship to efficient resource usage.

FRA has its own importance for measuring bank efficiency but its major disadvantage is the
reliance on benchmark ratios which could be arbitrary and may mislead an analyst (Yeh (1996),
as cited in Yannick et al. 2016). Financial Ratios Analysis can be misleading because it’s
restricted to measure the complete efficiency of banks (Rao and Tekeste, 2012). The other major
limitations of FRA are its univariate nature, because of this drawback it’s difficult to measures
and predict efficiency of firms using such analysis (Mousa, 2015). But DEA allows measurement
of efficiency from multiple inputs and multiple outputs within multiple DMUs. DEA is a most
accurate technique to measure efficiency given limited number of DMUs (i.e., banks, hospital,
airports and so on) (Othman et al., (2016).
 5

Data envelopment analysis (DEA) method is popular in measuring efficiency in the countries
with developed banking systems, in our country also there are very few studies in DEA analysis
such as, Tesfaye (2014), Fasika (2016), Gamachis (2016), and Yedersal (2018), those studies
include commercial bank of Ethiopia in their study, in this study it is not include because,
‘‘commercial bank of Ethiopia get favorable support from the government in creating easy
market for deposit, loans and Forex which has contributed a lot in reducing the cost of fund and
boosting both interest and non-interest income’’ (Tesfaye, 2014).

Many studies made before focused on deposit mobilization, profitability, and liquidity. Hence
detail critical analysis of the efficiency of the banks should have been done. Also analysis at
individual banks level will give clear and pertinent information to potential investors in other
word “looking on the tree than the forest” is good and timely scenario in the banking sector.

The purpose of this study is to investigate the relative technical and scale efficiency of private
commercial banks in Ethiopia using data envelopment analysis (DEA), and studying on the issue
of technical and scale efficiency of Ethiopian private commercial banks using DEA is important
to know how the banks are using their input mix to produce a given output.

1.4. Objective of the study

1.4.1. General objective

The main objective of the study is to measure the relative efficiency levels of private
commercial banks (PCBs).

1.4.2. Specific objective

1. Measuring the technical efficiency scores for private commercial banks operating
in Ethiopia.
2. Measuring the scale efficiency scores for private commercial banks operating in
Ethiopia.
3. Comparing the mean efficiency scores of private commercial banks operating in
Ethiopia.
4. Identifying the proper improvement required on input and output by inefficient
private commercial banks operating in Ethiopia.
 6

1.5. Research question

1 Is private commercial banks (PCBs) in Ethiopia utilizes their resources efficiently?
2 Are private commercial banks (PCBs) producing at their most productive scale?
3 What is the mean efficiency of each private commercial banks operating in the
country for the period of 10 years?
4 What are the inputs and outputs to be increased or decreased by inefficient bank to
become efficient?

1.6. Significance of the study

The study can increase the horizons of the bank manager which may help them in improving the
banks’ efficiency and help the policy makers in their endeavor to improve the efficiency of
banking sector and identify the need for future reforms in private banks. Also assist academics in
their search for knowledge and theory and serve as a reference point for further future research.

1.7. Scope of the study

This study conducted to evaluate the efficiency of private banks in Ethiopia. All private
commercial banks in Ethiopia included in the study and the data set was limited to audited
financial statement of private commercial banks from the period of 2009 to 2018.

1.8. Organization of the study

The remaining part of this study is organized as follows: Chapter two contain theoretical and
empirical review of related literature on bank efficiency and efficiency measures. Chapter three
focuses on methodology of the study. Chapter four contains data analysis and interpretation.
The final chapter includes summary of findings, conclusion and recommendation.
 7

CHAPTER TWO
2. LITERATURE REVIEW

2.1 Introduction
Banks are the most important financial institutions in the financial system and in the economy.
Therefore, an efficient and stable banking system is a prerequisite to facilitating economic
growth and avoiding financial crises. Banking sector efficiency is important for promoting
access to financial services as well as stability of the banking sector as integral component of the
financial system. According to (Ikhide, 2009, as cited in Kamau, 2011,) Banks play essential role
in the proper functioning of payments systems and their efficiency is directly related to improved
productivity in the economy.

Efficiency in banking sector has been attracting the attention of a larger number of researchers.
In this part of the thesis, both theoretical and empirical literature on bank efficiency will be
reviewed

2.2 Efficiency: Concepts and Definitions

Efficiency has been defined differently by scholars. As per Wikipedia efficiency can be
described as the extent which, time, effort or cost is well used for intend task or purpose. It is
often used with specific purpose of relaying the capability of specific application of effort to
produce specific outcome effectively with minimum amount or quantity of waste, or unnecessary
effort. Efficiency is related to the ability to produce a result with minimum effort or resources. It
measures how close a production unit gets to its production possibility frontier, which is
composed of sets of points that optimally combine inputs in order to produce one unit of output
(Kablan, 2010). Effectiveness is referring to the ability of the bank to set and achieve its goals
and objectives, while efficiency refers to ability of the bank to produce output with minimal
resources or input, or commonly defined as the ratio of outputs over inputs (Sherman and Zhu
(2006) cited in Othman, et al., (2016). According to (Usman et al., 2010) bank efficiency is the
best level of output reached without any changes to the amount of input. Efficiency of a bank
refers to the ability of the bank to provide its service with the minimum possible resources, or
producing maximum possible products and services using limited amount of inputs (Yidersal,
2018).
 8

From a general point of view, effectiveness describes the capability of an individual, a group or a
system to achieve the assigned goals with the disposable resources (Yannick et al. 2016). The
idea of efficiency of a production unit was first introduced by Farell (1957), under the concept of
“input oriented measure”. According to Farell, a technical efficiency measure is defined by one
minus the maximum equi-proportionate reduction in all inputs that still allows continuous
production of given outputs. Technical efficiency is linked to the possibility of avoiding wasting
by producing as much outputs as the use of input allows it (output oriented measure), or by using
as less as input that the production objective plans it (input oriented measure). This efficiency is
measured by comparing observed and optimal values of production, costs, revenue, profit or all
that the production system can follow as objective and which is under appropriate quantities and
prices constraints. Therefore, we can analyze technical efficiency, in terms of deviation
compared with an idealistic production frontier isoquant (Kablan, 2007). In practice, the value of
technical efficiency ranges from 0 to 1, a value of 1 indicating the bank is the most efficient, and
as it approaches to 0 means inefficiency is increasing relative to the competitors (Yidersal,
2018).

2.3 Efficiency classification

According to Sherman and Zhu (2006) cited in Othman, et al., (2016), overall productivity of a
bank depends on four components of efficiency classification and they are:

1. Technical efficiency: Also known as global efficiency measures the ability of banks to
produce actual outputs with fewer inputs, or less resource used indicates higher efficiency;

2. Scale efficiency: Refers to the optimal activity volume level whereby inefficiency may arise if
goods or services are produced above or below optimal level that resulted in added fixed
cost;

3. Price efficiency: Bank could increase its efficiency if it could purchase the inputs (human
capital and material) at lower price without sacrificing the quality;

4. Allocative efficiency: Measure the optimal mix of several inputs in order to produce products
or services, such as banks incorporate automatic teller machines (ATM) and Internet banking
for capital labor tradeoffs to increase efficiency.
 9

2.3.1. Technical Efficiency

According to (Farrel, (1957) as cited in Emrouznejad, & Cabanda, (2015) defined technical
efficiency as the ability of a firm to obtain optimal output from given inputs. He illustrated this
by using two inputs (x1 and x2) to produce a single output (q), under the constant returns to
scale. Technical efficiency focuses more on the physical relationship between the levels of inputs
to the level of outputs; it requires inputs and outputs without price (Bauer et al. 1998). Technical
efficiency, the most common of the efficiency measure, reflects the ability of the firm to obtain
maximum output from a set of inputs. That is, it refers to the use of productive resources in the
most technologically efficient manner (Worthington, 2004).

Figure 1: Technical and Allocative Efficiencies (Emrouznejad, & Cabanda, p. 4, 2015)

As Coelli et al. (2005) illustration cited in Emrouznejad, & Cabanda, (2015) in the figure above,
“if a firm uses quantities of inputs, defined by the point P, to produce a unit of output, then the
distance QP represents the technical inefficiency of that firm, which is the amount by which all
inputs can be proportionally reduced while the output remains constant. This is represented by
the ratio of QP/0P by which all inputs can be reduced to achieve an efficient production. Thus,
technical efficiency (TE) of a firm is expressed as the ratio TE = 0Q/0P, which is equal to one
minus QP/0P. It takes an interval value between zero and one as an indicator of the degree of
technical efficiency of a firm. A firm is fully technically efficient when a value of one is
obtained. In the Figure, point Q is technically efficient because it lies on the efficient frontier in
which case TE = 1”
 10

2.3.2. Scale Efficiency

Measures the efficiency of a DMU at a given point with regard to what it could accomplish if it
works at the most productive scale size, where the average production reaches a maximum level
(Kounetas and Tsekouras, 2007). The scale efficiency of an organization can be determined by
comparing the technical efficiency scores of each service producer under constant returns to
scale and variable returns to scale. Thus, when efficiency is assessed under the assumption of
variable returns to scale, the efficiency scores for each organization indicate only pure technical
inefficiency (Pasiouras 2006). According to Adongo et al, (2005) as cited in Tesfaye (2014)
Scale Effectiveness frequently emerges from the capacity of expansive firms to distribute settled
costs such as promotion costs or cost of technology across a greater volume of output. It also
shows whether the decision-making units (e.g. banks) operate at the minimum of their long run
average cost curve. It focuses on technical efficiency, which is the ability of a bank to produce
maximal output from a given set of inputs over a certain time period

Figure 2: Scale Efficiency (Emrouznejad, & Cabanda, p. 5, 2015)

As Emrouznejad, & Cabanda, (2015) illustrates in the above finger the scale efficiency, where
point D is depicted as technically inefficient firm (lies below the production frontier), and can be
improved further by moving from point D to point E under the VRS frontier and also from point
E to point F under the CRS frontier. Moreover, the scale efficiency (SE) of firm D is expressed
as SE = GF/GE, which represents the distance from technical efficiency of point E to the CRS
technology.
 11

2.4 Measuring bank efficiency

Efficiency measurement determine how firm can maximize its output and profit and at the same
time minimize its cost (Mokhtar et al., 2008). As Balcerzak, et.al, (2017) Stated, Evaluation of
efficiency is an integral part of rational behavior of the production units that aims to survive in a
challenging competitive environment in a long term. In practice it is possible to apply several
methods to verify the level of efficiency. Financial ratio analysis (FRA), data envelopment
analysis (DEA), and the stochastic frontier analysis (SFA) are frequently and widely used
approaches to analyze the financial performance of the banks. Financial ratio analysis is the
simplest technique in banking efficiency measurement; it has also been criticized to encounter
many limitations such as not controlling input prices or product mix. Then, a measure that can
incorporate all the available inputs and outputs of the bank is needed and the other two
techniques meet this demand (Ncube, 2009). In many literatures Data Envelope Approach (DEA)
and Stochastic Frontier Analysis (SFA) are the most preferred efficiency measurement
techniques form non-parametric and parametric analysis methods. (Usman et al., 2010)
concluded that there is no superior method exists to measure bank efficiency. However,
according to Nigmonov, (2010) the DEA has been found to have a number of advantages over
other methods. First, it does not require specification of any functional relationship between
inputs and outputs or a priori specification of weights of inputs and outputs. Second, it easily
accommodates multiple inputs and outputs for which are the norm for the banking sector. Third,
it is suitable for measuring the efficiency of firms that lack competitive prices as could be the
case of a concentrated banking sector (Erasmus & Makina, 2014). DEA is seen as the foremost
commonly utilized non – parametric technique in technical efficiency measurement, especially in
banking area (Banerjee, 2012).

2.5 Data Envelopment Analysis (DEA)

Efficiency analysis is essential for the evaluation of bank performance. The DEA is non-
parametric approach, which is most popular for evaluating efficiency in the banking sector.
There are two model of DEA method. The first method was developed by (Charnes et al., 1978)
which are based on Farrell’s (1957) efficiency measures and is it call CCR (Charnes, Cooper and
Rhodes) model. CCR model was developed under the assumption of constant returns to scale
(CRS). On the other hand, the second model is BCC (Banker, Charnes and Cooper) model,
introduced by Banker et al., (1984) as an extension of the CCR model. BCC model was
 12

developed under the assumption of variable returns to scale (VRS). The primary steps in
constructing a DEA method is selecting decision making units (DMU’s) that computes a
comparative ratio of outputs to inputs for each unit (Othman, et.al., 2016). Avkiran, (2006) stated
that: DEA identifies a DMU as either efficient or inefficient compared to other units in its
reference set. For evaluating the efficiency of bank performance DEA used two approaches. The
first approach is the intermediation approach where bank present oneself as a financial
intermediaries. In this approach from perspective of cost-revenue management, where bank’s
major business activity is to borrow funds from depositors and lends those funds to other for
spread. The second approach is production approach where usually as inputs are labor and
capital and outputs are loans and deposits. Avkiran, (2000) argued that for analyzing bank
efficiency it is better to use intermediation approach. The DEA technique will be considering
more detail on the next chapter of the study.

2.6 Empirical Review

Efficiency of a bank refers to the ability of the bank to provide its service with the minimum
possible resources, or producing maximum possible products and services using limited amount
of inputs. The efficiency of the banking system is the most important issues in the financial
market because it affects the stability of the banking industry and then, the effectiveness of the
nation’s monetary policy (Yilmaz, 2013 as cited in Gamachis, 2016). There are a number of
studies that evaluate the efficiency of commercial banks using DEA method; Reviews of these
various studies were presented as follows.

An article entitled, Technical, Scale, and Allocative Efficiencies in U.S. banking: An Empirical
Investigation, by Aly et al., (1990), applied DEA to explore various measures of efficiency for
sample 322 banks in 1986. The study employed three inputs (labor, capital, and loanable funds)
and five outputs (commercial and industrial loans, consumer loans, real estate loans, other loans,
and demand deposits). The result indicates a low level of overall efficiency. The main source of
inefficiency is technical in nature and on average the bank in the sample is scale efficient.

The Study that examine technical efficiency, pure technical efficiency and scale efficiency of
Russia’s commercial banks by Yadav, (2015), taking a sample of 131 using a non-parametric
approach (data envelopment analysis) form the period of 2007 to 2014. Found that Scores of
technical efficiency range from 31% to 51% which implies that banks need to reduce their inputs
 13

from 49% to 69% to be on efficiency frontier. Result also shows that commercial banks in the
sample are by and large operating at decreasing returns to scale and also shows that banks
underperform in the utilization of inputs (total expenses and deposits) to create optimum outputs
(loans and net investment).the study conclude that scores of scale efficiency are higher than the
pure technical efficiency, explains that the main reason for the inefficiency of commercial banks
in Russia is due to managerial inefficiency.

Karimzadeh, (2012), examines the efficiency of Indian commercial banks during 2000 – 2010 by
utilizing Data Envelopment Analysis (DEA). Based on the sample of 8 commercial banks, by
using intermediation approach the researchers used loans and investments as output variables and
fixed assets, deposits, and number of employees as Inputs the findings reveal that the mean of
cost (economic) efficiency, technical efficiency, and allocative efficiency are 0.991, 0.995, and
0.991 in VRS model and 0.936, 0.969 and 0.958 in CRR model, respectively. And he confirmed
that selected Public Sector Banks are more efficient than Private sectors during the study period
in India.

Thu Huong and Firoz, (2016), evaluating the efficiency of Vietnamese commercial banks using
data envelopment analysis during the period 2011 – 2014, with comparison among different
groups such as state owned vs. non–state owned banks, listed vs. unlisted banks, and large vs.
small banks. The findings indicate larger banks performed better than smaller banks in terms of
technical efficiency, but there was not much difference among the groups in terms of average
overall technical efficiency. State-owned and listed banks obtained higher efficiency levels than
non-state-owned and unlisted banks.

Baidya & Mitra (2012), the study was to measure and evaluate the technical efficiency of 26
Indian public sector banks from the financial year 2009–2010.data envelopment analysis (DEA)
models: CCR and Andersen and Petersen’s super-efficiency model is employed. The results
reveal that average technical efficiency of entire sample is 86.5% and that only seven banks
(23%) are found to be fully efficient. So, there is a scope of efficiency improvement of 19 public
sector banks in India. The study has found that, the banks which are using more labor for
providing their services are relatively more inefficient.
 14

Tahir et al., (2009) used the DEA approach to measure the overall, pure technical, and scale
efficiencies of Malaysian commercial banks from 2000-2006. They specified two inputs (total
deposits and total overhead expenses), and one output, total earning assets. They found that
domestic banks were relatively more efficient than foreign banks. Their results indicated that the
domestic banks’ inefficiency was attributable to pure technical inefficiency rather than scale
inefficiency. On other hand, 20 the inefficiency of foreign banks was attributed to scale
inefficiency rather than the pure technical inefficiency.

Kumar & Singh, (2015) studied Technical and Scale Efficiency of India Banks using Data
Envelopment Analysis (DEA) from 2006 to 2010. The study observed five private and five
public sector commercial banks. They indicate that deregulation of banking sector has led to an
increase in the efficiency of commercial banks in India. They show increase in efficiency of
banks in India is not only because of increase in pure technical efficiency but also due to increase
in its scale efficiency. Also shows that performance of private sector banks has been better than
public sector banks during the period and source of inefficiency is mainly due to its scale rather
than pure technical inefficiency.

Mongid & Tahir, (2010) this study estimates the technical and scale efficiency of rural banks in
Indonesia during the period of 2006 and 2007 by using the non-parametric approach – Data
Envelopment Analysis (DEA). They used intermediation approach to select input and output
(total deposit and total overhead expenses as input and total earning assets as output). The results
suggest that technical efficiency score is lower than scale efficiency score which indicates that
portion of overall inefficiency is due to producing below the production frontier rather than
producing at an inefficient scale.

Raphael, (2012) investigated the efficiency of commercial banks in Tanzania using a Data
Envelopment Analysis (DEA), over the period from 2008 to 2011. The study used three input
variables (deposit, interest expenses and operating expenses) and four output variables (loan,
investment, interest income and no interest income), the analysis result showed that most
commercial banks in Tanzania technically inefficient. In terms of size, large banks showed better
performance compared to small banks. As to the study, commercial banks should minimize the
use of input resources while maintaining the same level of output to improve technical
efficiency.
 15

Yannicka et al., (2016) Study Technical efficiency assessment of banking sector of Côte d’Ivoire
using data envelopment analysis, in the study 14 banks have been evaluated from 2008 to 2010.
The outcomes revealed that Ivorian banks do not operate efficiently in terms of credits assignment
moreover find that foreign ownership private banks are relatively more efficient than public
ownership ones. And Computation of scale efficiency scores proves that on average Ivorian
commercial banks scale Inefficiency is 38%. They do not use scale economy to improve their
outputs.

Kamau, (2011), using Data Envelopment Analysis (DEA) and Malmquist productivity index
(MPI), investigate intermediation efficiency and productivity in the banking sector in the post
liberalization period in Kenyan Commercial banks, and the results show that though the banks
were not completely effective in all perspective, under study period commercial banks efficiency
score was not less than 40% at any point. In terms of ownership and size, foreign banks were
found to be more efficient than local banks, and in local category local private were more
efficient than local public, large sized banks were more efficient than medium and small sized
banks. And the inefficiency is mostly because of inefficient use of deposit or keeping excess
liquidity.

In our country also there are few studies conducted using DEA method to measure bank
efficiency; Tesfaye, (2014) conducts assessment on the efficiency level of Ethiopian Banks for
the period 2008-2012 using the Data Envelopment Analysis. The study found that the industry
efficiency level is at modest level but the technical and scale efficiency of Banks is characterized
by both inter and intra group variations across different ownership and size. And CBE’s
efficiency score persistently at the frontier, banks that were recently emerged in the industry
were less efficient than the other group.

Fasika, (2016) with the objective of evaluating the technical efficiency of commercial banks in
Ethiopia, Employing DEA over the period 2011 to 2014 with the sample of 15 commercial banks
in Ethiopia. By using three input variables (interest expense, operating expenses and deposit) and
three output variables (interest income, noninterest income and loans) found that under constant
returns to scale (CRS), cooperative bank of Oromia (CBO), Berhan international bank (BrIB)
and Dashen bank (DB) were the most efficient commercial banks while commercial bank of
Ethiopia (CBE), united bank (UB), lion bank (LIB) and Buna international bank (BuIB) were the
 16

least efficient commercial banks. Under the variable returns to scale, BrIB, CBO and nib
international bank (NIB) were found to be more efficient banks while CBE, UB and BuIB were
the least efficient banks. Also found privately owned commercial banks in Ethiopia are more
efficient compared to government owned commercial banks considering the scale
efficiency/inefficiency score; CBO and DB were characterized as the most scale efficient
commercial banks In general, the study found that majority of commercial banks in Ethiopia
experienced relative inefficiency both under the CRS and VRS assumptions.

Gamachis, (2016) under the title of ‘‘Technical Efficiency and Productivity of Ethiopian
Commercial Banks’’ study adopts DEA to measure efficiency of banks and MPI to measure the
productivity gains of banks over time period of 2007 to 2011 by taking a sample of ten
commercial banks .and taking Fixed Assets and Labor as input and Total Deposits and Net Loans
and Advances as output variables. The study found that, on average, Ethiopian commercial banks
were relatively technically inefficient. And Scale inefficiency takes the leading contribution for
source of inefficiency.

Yidersal, (2018) by employing Data Envelopment Analysis (DEA) Conduct a study to measure
the relative technical, cost, revenue, and profit efficiency of the Ethiopian Commercial Banks
using data of 18 commercial banks for the period covered 2005 to 2016. The researcher found
that four banks namely Commercial bank of Ethiopia, Adis International bank, Zemen bank &
Enat bank are the most efficient banks in terms of Technical Efficiency, and are found to be on
the DEA frontier under both input & output orientations. Under Cost Efficiency, the giant
Commercial Bank of Ethiopia, Adis International Bank and Debub Global Bank are found to be
the most efficient ones, and on the DEA frontier. Finally, the research finding shows that
Commercial Bank of Ethiopia and Adis International Bank are 100% efficient compared to other
participants under both revenue & profit efficiency.
 17

2.7. Conceptual framework

The main objective of the study is to measure efficiency of private commercial banks. According
to camp, (2001) a conceptual framework is a structure which the researcher believes can best
describe the progress of the phenomenon to be studied. This conceptual framework shows that
the study used two inputs and two outputs to estimate the intimidation efficiency interims of
technical and scale efficiency using DEA as a tool for the analysis.

INPUT VARIABLE OUTPUT VARIABLE

 TOTAL
 DEPOSIT BANKS LOAN AND
 NON- INTERMEDIATION ADVANCE
INTEREST OPERATION  NON-
EXPANSES PROCESS INTEREST
INCOME

Efficiency
Estimation techniques
DEA

Technical and scale efficiency result

of DMU
 18

CHAPTER THREE

3. RESEARCH METHODOLOGY

3.1 Introduction
This chapter describes the methodology that used to address the research problem. The chapter
describes the research design and approach, population and sample, Data source and collection,
Method of data analysis, Model variables (input and output) and Model specification, of the
study.

3.2 Research Design and Approach

Research design refers to the overall strategy that one chooses to attack the problem which
requires integration of different components of the study in a coherent and logical way, thereby,
ensuring to solve the problem in efficient way. It constitutes the blueprint for the collection,
measurement, analysis of data, interpretation and reporting of conclusions. Research design is
necessary because it makes possible the smooth sailing of the various research procedures,
thereby creation research as professional as possible, yielding maximum information with a
minimum expenditure of effort, time and money (Islamia, 2016). According to Islamia, (2016)
research design generally categorized in to four group based on the purpose of the research;
Exploratory Research, Descriptive Research, Explanatory Research and Experimental Research.

This study was conducted using descriptive research. A descriptive study defines a subject by
constructing a profile of people, groups or events through tabulation and the collection of data on
the frequencies on study variables (Cooper & Schindler, 2007). Descriptive research describes
phenomena as they exist and the observer observes and describe what did he find? Descriptive
research answers the questions, what, who, where, how and when. It’s more common in the
social sciences, as in socio-economic survey and job and activity analysis (Islamia, 2016).

There are three types of research Approaches; Quantitative: approach of measurements and
numbers, Qualitative: approach of words and images, and Mixed Methods approach of
measurements, numbers, words and images. Creswell, (2014) mentioned that through
quantitative research, phenomena are being explained “by collecting numerical data that are
analyzed using mathematically based methods (in particular statistics). According to (Kothari,
 19

(2004) Quantitative research is based on the measurement of quantity or amount. It is applicable

to phenomena that can be expressed in terms of quantity. This study used secondary source of
data (financial reports) which is numerical, therefore quantitative approach was preferred.

3.3 Population and sample

All members who meet the particular condition specified for a research study can be known as a
target population (Alvi, m. 2016). All people or items in a given study called population whereas
a process of selecting part of the population for investigation known as sampling (Rahi, 2017).
For this study the target population is all private commercial banks in Ethiopia. According to
national bank of Ethiopia at the end of 2018 there were 16 private commercial banks.

In DEA method there are two important aspects that shows census is more preferable than
sampling, the first one is that the result found from a sample cannot generalized for the whole
population the other one is the analysis result of DEA is not absolute its relative. This indicates
that the efficient DMU score 100 per cent efficiency the other DMUs will be benchmarked
against the efficient (Sanjeev, 2006). Thus the study considered all the private commercial banks
which were operating in the country as a decision making units.

3.4 Data source and collection

Due to the nature of the study only secondary data is used. The data was found from the audited
financial statements of the banks for the period 2009 - 2018. And those data collected from the
published audited annual report of the all private commercial banks and from the records held by
National bank of Ethiopia. Specifically the data were gathered from the balance sheet and the
income statement of commercial banks covered in the study period. The data collected from
secondary source were analyzed and presented through Graphs and tables.

3.5 Method of data analysis

As stated in Karimzadeh, (2012) Data Envelopment Analysis (DEA) is a linear programming
based method developed by Charnes, Cooper and Rhodes in 1978 sometimes called frontier
analysis, which used to measurer performance and it’s used for evaluating the relative efficiency
of productive units, having the same multiple inputs and multiple outputs. DEA has been widely
used to measure efficiency performance of different financial institutions like banks, insurance
and mutual funds. Particularly in the banking sector, it has been applied to benchmark the
 20

efficiency performance of different banks or to study the efficiency estimates of different

branches of a particular bank (Debasish, 2006). As stated in previous chapters the study was used
Data Envelopment Analysis (DEA) to carrying out this study.

In 1978 Charnes, Cooper and Rhodes generalized Farrell’ singe input single output efficiency
measure to multiple - input multiple - output situations and operationalized it using mathematical
programming (Emrouznejad & Cabanda, 2015). This method is called Data envelopment
analysis. DEA is a nonparametric linear programming (LP) technique that permits evaluation of
the relative efficiency of decision-making units (DMUs). DEA is used to evaluate the relative
efficiency of a number of producers or decision-making unites (DMUs). It allows us to compare
the relative efficiency of DMUs which have multiple input and output by determining the
efficient DMU as a benchmark, and the efficiency score in the presence of multiple input and
output factors is defined as:
Efficiency = weighted sum of outputs / weighted sum of inputs
There are two types of efficiencies in DEA – input oriented and output oriented. Input oriented
efficiency aims at reducing input amounts as much as possible while keeping at least the present
output levels and output oriented technical efficiency maximizes the output level while using at
least the present input levels (Baidya & Mitra, 2012). In an attempt to use both input-oriented
and output-oriented models to calculate DEA efficiency score, (Ramanathan, 2007 as cited in
Othman, et.al, 2016) discovered that both models generated similar results. This suggested that
there is no obvious difference in efficiency score generated by both models. Thus, no misleading
interpretations of DEA score if either one model is chosen. As mentioned in the previous
chapter, there are also two assumptions under DEA model, the constant return to scale (CRS)
and the variable return to scale (VRS). On this study the researcher used input oriented both
constant return to scale (CRS) and the variable return to scale (VRS) models.

3.5.1 Constant return to scale (CRS) model: The original DEA approach by Charnes et al.
(1978) assumed constant returns to scale of activities by DMUs. The CCR model is the
most widely used DEA model. It is used in frontier analysis when a constant return to
scale relationship is assumed between inputs and outputs. Being the first DEA model to
be developed, this model calculates the overall efficiency for each unit, where both pure
technical efficiency and scale efficiency are aggregated into one value. Factors such as
 21

imperfect competition, or limited financial resources, may prevent decision-making units

from operating at an optimal scale, this assumption is only appropriate when all DMUs
operate at an optimal scale. (Coelli, et. al., 1998). The reason is the technical efficiency
obtained from CCR DEA is composed of two constituents which are the result scale
inefficiency and pure technical inefficiency (Coelli, et. al., 2005).
3.5.2 Variable return to scale (VRS) Model: The first extension of basic CCR model is
called the DEA BCC model developed by BCC in 1984, with other criteria are the same
as CCR except it complement the equation to measure input excesses and output
shortfalls (Cooper et al., 2006; Ong et al., 2003as cited in Othman, et.al., 2016). The BCC
model is the DEA model used in frontier analysis when a variable returns to scale
relationship is assumed between inputs and outputs. This model focuses primarily on the
technological aspects of production correspondences, and can be used to estimate
technical and scale efficiency without requiring estimates of input and output prices.
Since the study assessing how efficient DMUs use inputs to produce outputs both CRS
and VRS assumptions are necessary.
The study used MAXDEA7 software to conduct the necessary analysis and to measure the
technical and scale efficiency of private commercial banks working in the country.

3.6 Model variables (input and output)

DEA is highly sensitive in the choice of input and output variables rather than unit of amount
and it is not required advanced assumption about the analytical form of production (Tesfaye,
2014). Regarding the appropriate inputs and outputs variables to be employed by DEA model for
banks, as mentioned in several studies, there are two main approaches that can be used to
determine the bank inputs and outputs. They are production approaches and intermediation
approaches (Mousa, 2015).
3.6.1. Production approach: which considers the bank as normal company or producer, and hence
the inputs are the physical elements such as labor and capital and all other assets and liabilities
are outputs, this approach argued that all deposits (which are assets) should be treated as output
since they are produced by capital and labor. According to Johnes et al. (2009, p.14) as cited in
Mousa, (2015) in production approach capital and labor inputs which is number of employees
and capital expenditures on fixed assets and Output are number of deposit accounts or
transactions and loans.
 22

3.6.2. Intermediate approach: by this approach the selection is made based on the bank's assets
and liabilities, bank assets including labor represent the inputs and liabilities represent the
outputs. On this approach banks are a mediator between borrowers and depositors that accept
deposits and offer loans and other investments. Output is measured by interest income, total
loans, total deposits and non-interest income, while inputs are usually represented by operating
and interest costs (Mousa1, 2015).
According to Avkiran (2000) for analyzing bank efficiency intermediation approach is better
one. According to Berger and Humphrey (1997) as cited by Yadav, (2015), intermediation
approach is well suited to analyzing firm level efficiency, whereas the production approach is
suited to measuring branch level efficiency. And this study followed the intermediation approach
to select two inputs and two outputs listed below
Input

I. Total deposit
II. Non-Interest expense
Output
I. Total loan and advance
II. Non-Interest Income

3.7 Model specification

The study measures the relative technical and scale efficiency score for the banks using the DEA
approach, the input-oriented CCR and BCC models were used. Based on Horvatova, (2018) the
input-oriented CCR-I (Charnes–Cooper–Rhodes) Input model can be written in the form of
linear programming problem as follows:
m
max z=∑ ui . yiq (1)
i=1

Under the conditions:

m r

∑ ui . yiq≤ ∑ vj . xjk ; k = 1, 2, … . . n (2)

i=1 j=1

m r

∑ ui . yiq−¿ ∑ vj . xjk ; ≤ 0 ¿ k = 1, 2, …. . n (3)

i=1 j=1
 23

∑ vj . xjk=1 (4)
j=1

ui ≥ 0, i = 1,2, … . . r (5)
vj ≥ 0, i = 1,2, … . . m (6)
Where:
Z = relative efficiency of the DMU
m = number of output produced by the DMU
r = number of inputs employed by the DMU
Yi, represent output data for DMU
Xj, represent input data for DMU
Ui = output weights
Vj, = input weight
K, represent number of DMU

Based on Horvatova, (2018) the input-oriented BCC-I (Banker–Charnes–Cooper– Input) model

can be written in the following form:
m
max z=∑ ui . yiq+ μ (7)
i=1

Under the conditions:

m r

∑ ui . yiq+ μ ≤ ∑ vj . xjk ; k = 1,2, … . . n (8)

i=1 j=1

∑ vj . xjk=1 (9)
j=1

ui ≥ 0, i = 1,2, … . . r (10)
vj ≥ 0, i = 1,2, … . . m (11)
Where:
Z = relative efficiency of the DMU
m = number of output produced by the DMU
r = number of inputs employed by the DMU
Yi, represent output data for DMU
 24

Xj, represent input data for DMU

Ui = output weights
Vj, = input weight
K, represent number of DMU

CHAPTER FOUR

4. Data Analysis and Interpretation

Introduction
This chapter deals with Empirical result and interpretation of the study. On the first subdivision
descriptive statistics of variables (input output) and correlation analysis among input output is
presented. On the other section the main objective of the study which is efficiency of private
commercial banks is discussed and the method that described in chapter three which is data
envelopment analysis were used to measured efficiency and inefficiency level of private
commercial banks in Ethiopia.

The data envelopment analysis software (MAX DEA 7) is used to measure overall technical
efficiency (OTE), pure technical efficiency (PTE) and scale efficiency (SE) of the banks and
STATA software (version 11) is used to check the correlation between variables.

4.1 descriptive statistics of input and output variables

DMU OBS INPUT OUTPUT
 
(16) 143 DEPO NON-IE TLAA NON-II
Min 263 15 158 7
Max 9566.089 539.817 5898.455 443.062
218.162
AB 8 Mean 3731.362 2302.611 180.5103
1
179.449
SD 3205.829 2000.037 143.1939
8
Min 211 21 154.488 23
Max 2970.377 207.384 2034.996 201.469
ADIB 7 100.069
Mean 1354.09 914.6033 113.4424
2
SD 983.9334 69.8343 683.1338 65.09123
 25

8
Min 4,962 155 2,713 201
Max 43,451 1,934 31,049 1,203
AIB 10 Mean 17,100 723 11,379 711
607.256
SD 12267.1 9312.124 335.0781
7
Min 4494.19 145 2708.96 128.92
Max 25794.54 1452.15 17780.96 789
BA 10 Mean 11131.97 498.74 6846.31 371.56
447.344
SD 7059.109 5073.402 206.6929
8
Min 238 12.441 153 2
Max 10889 694.298 10097.04 476.83
206.379
BIB 9 Mean 3590.536 2675.165 180.9079
9
236.969
SD 3633.958 3261.775 183.3072
4
Min 240 18 192 11
Max 9848.374 575.787 6841.603 359.861
199.976
BUB 9 Mean 3505.288 2399.507 148.2657
3
194.363
SD 3386.441 2355.421 126.0769
4
Min 789 41 596 11
Max 25808 1284 15145 658
CBO 10 Mean 7271 434 4651 282
430.743
SD 7673.826 4776.407 215.3985
7
Min 7,925 204 4,452 321
Max 35,987 1,854 23,058 1,345
DB 10
Mean 18,385 763 10,750 895
SD 8595.941 558.358 5911.898 333.9129
Min 500 44 266.65 40
Max 2153.322 198.198 1553.712 206.972
DGB 5 105.639
Mean 1155.19 705.4652 108.1944
6
SD 650.9672 59.9371 516.7448 64.94087
Min 929.44 70 506.74 72
Max 5090.526 222.592 3313.951 261.653
129.118
EB 5 Mean 2731.101 1802.157 148.9306
4
65.8528
SD 1672.677 1101.533 81.62977
5
Min 704 34 470 16
Max 11,640 603 7,374 368
LIB 10
Mean 4,075 224 2,555 179
SD 3715.691 209.494 2398.503 130.2484
 26

Min 3,296 132 2,220 172

Max 21,619 950 13,499 448
NIB 10 Mean 9,323 398 5,981 314
272.353
SD 5899.384 3740.155 74.96097
8
Min 184 6 113 2
Max 19927 987 9969 963
OIB 10 Mean 6268 338 3333 265
340.454
SD 6381.591 3320.855 293.9443
2
Min 3,616 124 2,152 135
Max 23,080 1,156 14,870 623
UB 10 Mean 10,256 478 6,417 347
351.175
SD 6006.888 4215.505 137.4002
2
Min 3,728 133 2,112 239
Max 20,506 1,285 14,785 972
WB 10 Mean 9,078 513 5,895 499
377.568
SD 5146.825 3999.464 223.172
8
Min 278 24 189 13
Max 10241 352 4995 503
ZB 10
Mean 3633 156 1941 244
SD 3193.892 114.221 1627.518 147.2299
Table 4.1 is about descriptive statistics of all DMU, input and output variables used in the study.
As we see in the above table there is high difference among input output variables in most of
DMU. This implies that there is high scale (size) difference between Ethiopia private commercial
banks. But this cannot affect DEA method from measuring efficiency of the banks.

4.2 Input Output correlation analysis

. Corr

(Obs=10)

| depo nonie tol nonii

-------------+------------------------------------------------

depo | 1.0000

nonie | 0.9944 1.0000

tol | 0.9990 0.9957 1.0000

 27

nonii | 0.9704 0.9775 0.9688 1.0000

Table 4.2 correlation between variables yearly mean of ten years period

The suitability of the variables (input and output) used for measuring efficiency should have
correlation between them, if the selected input and output have high correlation coefficient
between them they are suitable for measuring efficiency (Rhodes & Southwick, 1993; Charnes et
al., 1994) as cited by Marjanovic et al. 2018). As the correlation analysis result show there is high
correlation among all input and output. The correlation coefficient lies between, 99.44% and
96.88%. For instance a correlation coefficient of DEPO and TOL equals 99.9% and DEPO and
NON-IE have 99.4%, DEPO and NON-II have 97% correlation coefficient and TLO and NON-
IE have 99.6% and TOL and NON-II have 96.9% correlation coefficient. NON-IE and NON-II
also have high correlation coefficient of 97.8%.

4.3. Efficiency score of Ethiopian private commercial bank from 2009 to 2018
The analysis result of overall technical and pure technical efficiency of all PCBs for each of 10
years is presented and discussed below

Figure 4.1: efficiency of PCBs in CRS and VRS models for the year 2009

OTE AND PTE OF BANKS 2009

1.02
1
0.98
0.96
0.94
0.92
0.9
0.88
0.86
0.84
AB ADIB AIB BA BIB BUB CBO DB DGB EB LIB NIB OIB UB WB ZB MEAN

2009 TE CRS 2009 TE VRS

Source: author own computation from the data

 28

In 2009 from the sample of 10 banks, four banks (CBO, DB. NIB and WB) were 100% efficient
in both CRS and VRS method and OIB and ZB were efficient only in VRS method. In VRS
method only BA and in CRS method three banks (BA, OIB and ZB) were above 95% efficient
and their efficiency ratio is lies between 0.978 and 0.998. The other three banks (AIB, LIB and
UB) were above 90% efficient in both CRS and VRS methods. The mean efficiency for the year
was 0.973 and 0.976 for CRS and VRS respectively and 7 of the sample banks performed above
the mean and 3 of the banks fell below the mean efficiency.

Figure 4.2: efficiency of PCBs in CRS and VRS models for the year 2010

OTE AND PTE OF BANKS 2010

1.2

0.8

0.6

0.4

0.2

0
AB ADIB AIB BA BIB BUB CBO DB DGB EB LIB NIB OIB UB WB ZB MEAN

2010 TE CRS 2010 TE VRS

Source: author own computation from the data

In 2010 from the sample of 12 banks, six banks (AIB, BA, BUB, DB, WB and ZB) were 100%
efficient in both CRS and VRS method and BB were efficient only in VRS method. In VRS
method three banks (LIB, NIB and UB) and in CRS method also three banks (BB, NIB and UB)
were above 90% efficient and their efficiency ratio is lies between 0.906 and 0.962. The other
three banks (CBO, LIB and OIB) was below 90% efficient in CRS method and in VRS methods
two banks (CBO and OIB) was below 90% efficiency score . The mean efficiency for the year
was 0.93 and 0.945 for CRS and VRS respectively and 7 of the sample banks performed above
the mean and 5 of the banks fell below the mean efficiency in CRS method and in VRS method 8
of the sample banks performed above the mean and 4 of the banks fell below the mean
efficiency.
 29

Figure 4.3: efficiency of PCBs in CRS and VRS models for the year 2011

OTE AND PTE OF BANKS 2011

1.2

0.8

0.6

0.4

0.2

0
AB ADIB AIB BA BIB BUB CBO DB DGB EB LIB NIB OIB UB WB ZB MEAN

2011 TE CRS 2011 TE VRS

Source: author own computation from the data

In 2011 from the sample of 13 banks, in CRS method four banks (AIB, BUB, UB and ZB) were
100% efficient and three banks (BA, DB and NIB) was above 90% efficient and six banks (AB,
BIB, CBO, LIB, OIB and WB) were below 90%and their efficiency ratio is lies between 0.683
and 0.875. In VRS method most of the banks were efficient only three banks (CBO, LIB and
OIB) were inefficient and their efficiency score is below 90% and it was 0.684, 0.873 and 0.745
respectively .The mean efficiency for the year was 0.903 and 0.946 for CRS and VRS
respectively and 7 of the sample banks performed above the mean and 6 of the banks fell below
the mean efficiency in case of CRS and10 of the sample banks performed above the mean and 3
of the banks fell below the mean efficiency in VRS method.
 30

Figure 4.4: efficiency of PCBs in CRS and VRS models for the year 2012

OTE AND PTE OF BANKS 2012

1.2

0.8

0.6

0.4

0.2

0
AB ADIB AIB BA BIB BUB CBO DB DGB EB LIB NIB OIB UB WB ZB MEAN

2012 TE CRS 2012 TE VRS

Source: author own computation from the data

In 2012 from the sample of 14 banks, six banks (ADIB, AIB, BUB, DB, NIB and ZB) were
100% efficient in both CRS and VRS method and BIB and WB were efficient only in VRS
method. In VRS method two banks (BA and UB) and in CRS method three banks (BA, UB and
WB) were above 90% efficient and their efficiency ratio is lies between 0.948 and 0.996. The
other five banks (AB, BIB, CBO, LIB and OIB) was below 90% efficiency score in CRS method
and in VRS method except BIB all the other efficiency score were below 90%. The mean
efficiency for the year was 0.934 and 0.947 for CRS and VRS respectively and 9 of the sample
banks performed above the mean and 5 of the banks fell below the mean efficiency in CRS
method and in VRS method 10 of the sample banks performed above the mean and 4 of the
banks fell below the mean efficiency.
 31

Figure 4.5: efficiency of PCBs in CRS and VRS models for the year 2013

OTE AND PTE OF BANKS 2013

1.2

0.8

0.6

0.4

0.2

0
AB ADIB AIB BA BIB BUB CBO DB DGB EB LIB NIB OIB UB WB ZB MEAN

2013 TE CRS 2013 TE VRS

Source: author own computation from the data

In 2013 from the sample of 14 banks, eight banks (ADIB, BA, BIB, CBO, DB, LIB, NIB and
ZB) were 100% efficient in both CRS and VRS method and AIB were efficient only in VRS
method. In VRS method two banks (BUB and WB) and in CRS method three banks (AIB and
WB) were above 90% efficient and their efficiency ratio is lies between 0.943 and 0.988. The
other four banks (AB, BUB, OIB and UB) was below 90% efficiency score in CRS method and
in VRS method except BUB all the other efficiency score were below 90%. The mean efficiency
for the year was 0.956 and 0.965 for CRS and VRS respectively and 9 of the sample banks
performed above the mean and 5 of the banks fell below the mean efficiency in CRS method and
in VRS method 10 of the sample banks performed above the mean and 4 of the banks fell below
the mean efficiency.
 32

Figure 4.6: efficiency of PCBs in CRS and VRS models for the year 2014

OTE AND PTE OF BANKS 2014

1.2

0.8

0.6

0.4

0.2

0
AB ADIB AIB BA BIB BUB CBO DB DGB EB LIB NIB OIB UB WB ZB MEAN

2014 TE CRS 2014 TE VRS

Source: author own computation from the data

In 2014 from the sample of 16 banks, five banks (ADIB, CBO, DB, NIB and ZB) were 100%
efficient in both CRS and VRS method and AIB and DGB were efficient only in VRS method. In
VRS method two banks (BIB and BUB) and in CRS method three banks (AIB, BIB and BUB)
were above 90% efficient and their efficiency ratio is lies between 0.909 and 0.963. The other
eight banks (AB, BIB, DGB, EB, LIB, OIB,UB and WB) was below 90% efficiency score in
CRS method and in VRS method except DGB all the other efficiency score were below 90%.
The mean efficiency for the year was 0.902and 0.92 for CRS and VRS respectively and 8 of the
sample banks performed above the mean and 8 of the banks fell below the mean efficiency in
both CRS and VRS methods
 33

Figure 4.7: efficiency of PCBs in CRS and VRS models for the year 2015

OTE AND PTE OF BANKS 2015

1.2

0.8

0.6

0.4

0.2

0
AB ADIB AIB BA BIB BUB CBO DB DGB EB LIB NIB OIB UB WB ZB MEAN

2015 TE CRS 2015 TE VRS

Source: author own computation from the data

In 2015 from the sample of 16 banks, only four banks (ADIB, CBO, EB and ZB) were 100%
efficient in both CRS and VRS method and the other four banks (AIB, DB, DGB and NIB) were
efficient only in VRS method. In CRS method three banks (AIB, DB and NIB) and in VRS
method only LIB, were their efficiency was above 90% and their efficiency ratio is lies between
0.946 and 0.995. The other nine banks (AB, BA, BIB, BUB, DGB, LIB, OIB, UB and WB) was
below 90% efficiency score in CRS method and in VRS method except DGB and LIB all the
other efficiency score were below 90%. The mean efficiency for the year was 0.897 and 0.936
for CRS and VRS respectively and 7 of the sample banks performed above the mean and 9 of the
banks fell below the mean efficiency in CRS method and in VRS methods 9 of the sample banks
performed above the mean and 7 them fell below the mean efficiency.
 34

Figure 4.8: efficiency of PCBs in CRS and VRS models for the year 2016

OTE AND PTE OF BANKS 2016

1.2

0.8

0.6

0.4

0.2

0
AB ADIB AIB BA BIB BUB CBO DB DGB EB LIB NIB OIB UB WB ZB MEAN

2016 TE CRS 2016 TE VRS

Source: author own computation from the data

In 2016 from the sample of 16 banks, only five banks (ADIB, BIB, DGB, EB and ZB) were
100% efficient in both CRS and VRS method and the other three banks (AIB, COB, and DB)
were efficient only in VRS method. In CRS method eight banks (AB, AIB, BUB, CBO, LIB,
NIB, UB and WB) and in VRS method (AB, BUB, LIB, NIB, UB and WB), were their
efficiency was above 90% and their efficiency ratio is lies between 0.91 and 0.999. The other
three banks (BA, DB, and OIB) was below 90% efficiency score in CRS method and in VRS
method two banks (BA and OIB) efficiency score were below 90%. The mean efficiency for the
year was 0.95 and 0.964 for CRS and VRS respectively and 10 of the sample banks performed
above the mean and 6 of the banks fell below the mean efficiency in CRS method and in VRS
methods 11 of the sample banks performed above the mean and 5 them fell below the mean
efficiency.
 35

Figure 4.9: efficiency of PCBs in CRS and VRS models for the year 2017

OTE AND PTE OF BANKS 2017

1.2

0.8

0.6

0.4

0.2

0
AB ADIB AIB BA BIB BUB CBO DB DGB EB LIB NIB OIB UB WB ZB MEAN

2017 TE CRS 2017 TE VRS

Source: author own computation from the data

In 2017 from the sample of 16 banks, seven banks (ADIB, AIB, BIB, DGB, EB, WB and ZB)
were 100% efficient in both CRS and VRS method and DB were efficient only in VRS method.
In CRS method six banks (BA, BUB, CBO, DB, NIB, and UB) and in VRS method (BA, BUB,
CBO, NIB, and UB) were their efficiency was above 90% and their efficiency ratio is lies
between 0.919 and 0.99. The other three banks (AB, LIB, and OIB) were with the efficiency
score of below 90% in both CRS and VRS method. The mean efficiency for the year was 0.945
and 0.955 for CRS and VRS respectively and 9 of the sample banks performed above the mean
and 7 of the banks fell below the mean efficiency in CRS method and in VRS methods 11 of the
sample banks performed above the mean and 5 them fell below the mean efficiency.
 36

Figure 4.10: efficiency of PCBs in CRS and VRS models for the year 2018

OTE AND PTE OF BANKS 2018

1.2

0.8

0.6

0.4

0.2

0
AB ADIB AIB BA BIB BUB CBO DB DGB EB LIB NIB OIB UB WB ZB MEAN

2018 TE CRS 2018 TE VRS

Source: author own computation from the data

In 2018 from the sample of 16 banks, five banks (AIB, BIB, DGB, EB and ZB) were 100%
efficient in both CRS and VRS method and DB, OIB and WB were efficient only in VRS
method. In CRS method only ADIB and in VRS method AB and ADIB were their efficiency was
above 90% and their efficiency ratio is lies between 0.911 and 0.948. The other ten banks (AB,
BA, BUB, CBO, DB, LIB, NIB, OIB, UB and WB) were with the efficiency score of below 90%
in CRS method and VRS method six banks (BA, BUB, CBO, LIB, NIB and UB). The mean
efficiency for the year was 0.893 and 0.933 for CRS and VRS respectively and 6 of the sample
banks performed above the mean and 10 of the banks fell below the mean efficiency in CRS
method and in VRS methods 9 of the sample banks performed above the mean and 7 them fell
below the mean efficiency.
 37

4.4 Scale efficiency of Ethiopian private commercial banks

Table 4.4 shows scale efficiency of Ethiopian private commercial banks for the period of ten
years.

YEAR
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
DMU SE SE SE SE SE SE SE SE SE SE
AB     0.875 0.972 0.999 0.993 0.983 0.999 0.999 0.921
ADIB       1 1 1 1 1 1 0.957
AIB 1 1 1 1 0.988 0.963 0.981 0.999 1 1
BA 1 1 0.964 0.998 1 0.989 0.966 0.99 1 0.922
BIB   0.906 0.844 0.876 1 0.984 0.984 1 1 1
BUB   1 1 1 0.934 0.991 0.993 1 0.979 0.994
CBO 1 0.966 0.999 0.986 1 1 1 0.997 0.993 0.988
DB 1 1 0.988 1 1 1 0.946 0.877 0.923 0.827
DGB           0.815 0.762 1 1 1
EB           0.987 1 1 1 1
LIB 0.981 0.974 0.988 0.994 1 0.993 0.881 0.976 0.97 0.987
NIB 1 0.98 0.929 1 1 1 0.956 0.989 0.993 0.988
OIB 0.998 0.982 0.99 0.98 0.997 0.997 0.969 0.99 0.991 0.851
UB 1 0.994 1 0.999 0.989 1 0.958 0.982 0.995 1
WB 1 1 0.851 0.993 0.969 0.99 0.977 0.979 1 0.892
ZB 0.993 1 1 1 1 1 1 1 1 1
MEAN 0.997 0.984 0.956 0.986 0.991 0.981 0.96 0.986 0.99 0.958
MAX 1 1 1 1 1 1 1 1 1 1
MIN 0.981 0.906 0.844 0.876 0.934 0.815 0.762 0.877 0.923 0.827
Note: some parts of the table are blank because the particulars bank were not start operating in
respect year and the data were not found for the blanked year.

The table shows that the mean scale efficiency of Ethiopian private commercial banks is above
95%, it lies between 95.6% and 99.7%. The maximum score of SE is 100% and 76.2% is the lists
SE score of the ten years period. From the whole observed banks ZB, EB and ADIB were the
most scale efficient banks. In general the bank should increase or decreased there scale of
operation from 4.4% to 0.03%. For example if we see LIB it was inefficient for all period except
2013. When we see the inefficiency score of 2018, it was 1.3%, so that LIB should increase its
scale of operation in 1.3%.
 38

4.5. Mean efficiency score of overall technical, pure technical and scale for each
bank
Graph 4.1 a presentation of mean OTE, PTE and SE Efficiency score of Ethiopian private
commercial bank for the period of 2009 to 2018

MEAN OF OTE,PTE AND SE

1.2

0.8

0.6

0.4

0.2

0
AB ADIB AIB BA BIB BUB CBO DB DGB EB LIB NIB OIB UB WB ZB

TE PTE SE

This graph described in the table below in detail, it shows OTE, PTE and SE of all the observed
banks for the study period using mean of input and output. A DMU rich a line in point one is
100% efficient.

Table 4.5: mean efficiency table

 39

mean OTE, PTE and SE

DMU OTE PTE SE RTS
AB 0.8571 0.866337 0.989339 Decr
ADIB 1 1 1 Cons
AIB 1 1 1 cons
BA 0.906231 0.906962 0.999193 Incr
BIB 1 1 1 Cons
BUB 0.92155 0.933366 0.987341 Incr
CBO 0.858541 0.915594 0.937687 Decr
DB 0.973114 1 0.973114 Decr
DGB 1 1 1 Cons
EB 1 1 1 Cons
LIB 0.863809 0.864182 0.999569 Decr
NIB 0.960976 0.975522 0.985089 Incr
OIB 0.763015 0.787078 0.969427 Decr
UB 0.909286 0.916375 0.992264 Decr
WB 0.933956 1 0.933956 Decr
ZB 1 1 1 Cons
MEAN 0.910063 0.923394 0.985563 Decr
MXA 1 1 1  
MIN 0.763015 0.787078 0.933956  
SD 0.07104 0.066296 0.021703  

Table 4.5 shows the result of input oriented analysis of all DMU input output mean efficiency
score from the period of 2009 to 2018. The banks with the perfect score of 1 are efficient and
those less than 1 are inefficient.

AB

According to the result obtained from the mean efficiency table AB (ABAY BANK) were
inefficient with the efficiency score of 0.8571, 0.8663 and 0.9893 for OTE, PTE and SE
respectively.
 40

These implies that on average AB should have used 14.29% less resources to produce the same
level of output under CRS method and 13.37% less resources under VRS method this implies
that there were managerial inefficiency in the bank (inefficiency in resource utilization). In case
of scale efficiency, AB scale inefficiency was 1.07% and in decrease return to scale this means
that the bank was too large in its size and not operate at most productive scale.

ADIB

As of the mean efficiency table ADIB (ADDIS INTERNATIONAL BANK) was in the
efficiency frontier with the efficiency score of 100% for OTE, PTE and SE at all.

These implies that ADIB was fully efficient in using its resources (the bank used minimum input
to produce the current level of output) in both CRS and VRS method this implies that there were
good managerial performance in the bank (proper resource utilization). In case of scale
efficiency, ADIB scale efficiency was 100% and in constant return to scale this means that the
bank was at optimal scale size and operate at most productive scale.

AWB

The mean efficiency of AWB (AWASH BANK) was in the efficiency frontier with the
efficiency score of 100% for OTE, PTE and SE at all.

These implies that AWB was fully efficient in using its resources (the bank used minimum input
to produce the current level of output) in both CRS and VRS method this implies that there were
good managerial performance in the bank (proper resource utilization). In case of scale
efficiency, AWB scale efficiency was 100% and in constant return to scale this means that the
bank was at optimal scale size and operate at most productive scale.

BOA

According to the result obtained from the mean efficiency table BOA (BANK OF ABYSSINIA)
were inefficient with the efficiency score of 0.9062, 0.9069 and 0.9991 for OTE, PTE and SE
respectively.
 41

These implies that on average BOA should have used 9.38% fewer resources to produce the
same level of output under CRS method and 9.31% fewer resources under VRS method this
implies that there were managerial inefficiency in the bank (inefficiency in resource utilization).
In case of scale efficiency, BOA scale inefficiency was 0.09% and in increase return to scale this
means that the bank was small in its size. And in general the scale inefficiency is insignificant.

BB

The mean efficiency score of BB (BERHAN BANK) was 100% for OTE, PTE and SE at all and
the bank is in the efficiency frontier.

These implies that BB was fully efficient in using its resources (the bank used minimum input to
produce the current level of output) in both CRS and VRS method this implies that there were
good managerial performance in the bank (proper resource utilization). In case of scale
efficiency, BB scale efficiency was 100% and in constant return to scale this means that the bank
was at optimal scale size and operate at most productive scale.

BUIB

As showed in the mean efficiency table BUIB (BUNNA INTERNATIONAL BANK) were
inefficient with the efficiency score of 0.9215, 0.9333 and 0.9873 for OTE, PTE and SE
respectively.

These indicates that on average BUIB should have used 7.85% less resources to produce the
same level of output under CRS method and 6.67% less resources under VRS method this
implies that there were managerial inefficiency in the bank (inefficiency in resource utilization).
In case of scale efficiency, BUIB scale inefficiency was 1.27% and in increase return to scale
this means that the bank was too small in its size and not operate at most productive scale.

CBO
According to the result obtained from the mean efficiency table CBO (COOPERATIVE BANK
OF OROMIA) were inefficient with the efficiency score of 0.8585, 0.9155 and 0.9376 for OTE,
PTE and SE respectively.
 42

These shows that on average CBO should have used 14.15% fewer resources to produce the
same level of output under CRS method and 8.44% fewer resources under VRS method this
implies that there were managerial inefficiency in the bank (inefficiency in resource utilization).
In case of scale efficiency, CBO scale inefficiency was 6.23% and in decrease return to scale this
means that the bank was too large in its size and not operate at most productive scale.

DB

The mean efficiency of DB (DASHEN BANK) were inefficient with the efficiency score of
0.9731, 1 and 0.9731 for OTE, PTE and SE respectively. The CRS inefficiency of the bank is not
in case of technical inefficiency it was only scale inefficiency since DB PTE was 100%, this
shows there is no managerial inefficiency (the bank used minimum input to produce the current
level of output). In case of scale efficiency, DB scale inefficiency was 2.69% and in decrease
return to scale this means that the bank was too large in its size and not operate at most
productive scale.

DGB

The mean efficiency of DGB (DEBUB GLOBAL BANK) was in the efficiency frontier with the
efficiency score of 100% for OTE, PTE and SE at all.

These implies that DGB was fully efficient in using its resources (the bank used minimum input
to produce the current level of output) in both CRS and VRS method this implies that there were
good managerial performance in the bank (proper resource utilization). In case of scale
efficiency, DGB scale efficiency was 100% and in constant return to scale this means that the
bank was at optimal scale size and operate at most productive scale.

EB

As of the mean efficiency table EB (ENAT BANK) was in the efficiency frontier with the
efficiency score of 100% for OTE, PTE and SE at all.
 43

These implies that EB was fully efficient in using its resources (the bank used minimum input to
produce the current level of output) in both CRS and VRS method this implies that there were
good managerial performance in the bank (proper resource utilization). In case of scale
efficiency, EB scale efficiency was 100% and in constant return to scale this means that the bank
was at optimal scale size and operate at most productive scale.

LIB

According to the result obtained from the mean efficiency table LIB (LION INTERNATIONAL
BANK) were inefficient with the efficiency score of 0.8638, 0.8641 and 0.9995 for OTE, PTE
and SE respectively.

These implies that on average LIB should have used 13.62% fewer resources to produce the
same level of output under CRS method and 13.59% less resources under VRS method this
implies that there were managerial inefficiency in the bank (inefficiency in resource utilization).
In case of scale efficiency, LIB scale inefficiency was 0.05% and in decrease return to scale this
means that the bank was large in its size. And in general the scale inefficiency is insignificant.

NIB

As showed in the mean efficiency table NIB (NIB INTERNATIONAL BANK) were inefficient
with the efficiency score of 0.9609, 0.9755 and 0.9850 for OTE, PTE and SE respectively.

These indicates that on average NIB should have used 3.91% less resources to produce the same
level of output under CRS method and 2.45% less resources under VRS method this implies that
there were some managerial inefficiency in the bank (inefficiency in resource utilization). In case
of scale efficiency, NIB scale inefficiency was 1.5% and in increase return to scale this means
that the bank was too small in its size and not operate at most productive scale.

OIB

According to the result obtained from the mean efficiency table OIB (OROMIA
INTERNATIONAL BANK) were inefficient with the efficiency score of 0.7630, 0.7870 and
0.9694 for OTE, PTE and SE respectively.
 44

These shows that on average OIB should have used 23.7% less resources to produce the same
level of output under CRS method and 21.3% less resources under VRS method this implies that
there were high managerial inefficiency in the bank (inefficient in resource utilization) relative to
the observed banks . In case of scale efficiency, OIB scale inefficiency was 3.06% and in
decrease return to scale this means that the bank was too large in its size and not operate at most
productive scale.

UB

As showed in the mean efficiency table UB (UNITED BANK) were inefficient with the
efficiency score of 0.9092, 0.9163 and 0.9922 for OTE, PTE and SE respectively.

These indicates that on average UB should have used 9.08% fewer resources to produce the same
level of output under CRS method and 8.37% fewer resources under VRS method this implies
that there were managerial inefficiency in the bank (inefficiency in resource utilization). In case
of scale efficiency, UB scale inefficiency was 0.78% and in decrease return to scale this means
that the bank was large in its size. And in general the scale inefficiency is insignificant.

WB

As of the mean efficiency result WB (WEGAGEN BANK) were inefficient with the efficiency
score of 0.9339, 1 and 0.9339 for OTE, PTE and SE respectively. The CRS inefficiency of the
bank is not in case of technical inefficiency it was only scale inefficiency as WB PTE was 100%,
this shows there was no managerial inefficiency (the bank used minimum input to produce the
current level of output).

In case of scale efficiency, WB is the most scale inefficient bank among the observed banks
with the inefficiency score of 6.61% and it was in decrease return to scale this means that the
bank was too large in its size and not operate at most productive scale.

ZB

The mean efficiency of ZB (ZEMEN BANK) was in the efficiency frontier with the efficiency
score of 100% for OTE, PTE and SE at all.
 45

These implies that ZB was fully efficient in using its resources (the bank used minimum input to
produce the current level of output) in both CRS and VRS method this implies that there were
good managerial performance in the bank (proper resource utilization). In case of scale
efficiency, ZB scale efficiency was 100% and in constant return to scale this means that the bank
was at optimal scale size and operate at most productive scale.

As the shows from the evaluated 16 banks Only 6 banks were technically efficient in CCR model
and the rest 10 banks are technically inefficient with the score of 0.763 to 0.973 and in VRS
model 8 banks were technically efficient and half of the banks were technically inefficient with
the score of 0.787 to 0.975. DB and WB are technically efficient in VRS model but technically
inefficient in CCR model it’s because of scale inefficiency. In case of DB this result is in line
with Yedersal (2018) in his study also DB inefficiency is because of scale inefficiency.
OIB is the least efficient bank in both CCR and VRS with the score of 0.763015 and 0.787078
respectively. In case of scale efficiency only 6 banks were efficient and the other 10 banks are
inefficient with the score of 0.933 to 0.999. The least efficient bank in scale efficiency is WB
with the score of 0.933956.
The mean efficiency of the Ethiopian private commercial banks during the study period is
0.910063, 0.923394 and 0.985563 for OTE, PTE and SE respectively. This implies that on
average the banks could have saved 9% and 7.66% of input from the present consumption of
input to produce the present level of output for CRS and VRS method respectively. And in case
of SE the banks were inefficient and in decreasing return to scale, these implies that banks are
not in proper size (they are too large) and not operate at most productive scale (operate over
productive scale).
In general Only 6 banks were 100% efficient in the evaluation period and the result shows that
most of the banks are inefficient this means the banks were inefficient in resource utilization and
not operate at most productive scale. According to the result the inefficiency of Ethiopian private
commercial banks is mostly managerial inefficiency (inefficient utilization of resource) rather
than scale inefficiency. And this result is unlike to Gamachis (2016) and inline to Fasika (2016)
and Yedersal (2018) the result of this two study show that mean SE of the banks in Ethiopia is
better than PTE under their study period.
 46

4.6. Slack value

Slack
Slack Slack Slack Reduction and addition in input output
DMU Score (NON- NON- NON-
(DEPO) (NON-IE) (TLAA)
II) DEPO IE TLAA II
AB 0.8571 0 0 0 0 14.29% 14.29% 0 0
ADI
1 0 0 0 0
B 0 0 0 0
AWB 1 0 0 0 0 0 0 0 0
16.85
BOA 0.906231 0 0 0 62.63372
9.38% 9.38% 0 %
BB 1 0 0 0 0 0 0 0 0
BUIB 0.92155 0 0 0 13.6889 7.85% 7.85% 0 9.23%
11.53
CBO 0.858541 0 -13.7978 0 32.52364
14.15% 17.33% 0 %
DB 0.973114 -432.868 0 0 0 5.05% 2.70% 0 0
DGB 1 0 0 0 0 0 0 0 0
EB 1 0 0 0 0 0 0 0 0
LIB 0.863809 0 0 0 0 13.62% 13.62% 0 0
19.31
NIB 0.960976 0 0 0 60.63848
3.90% 3.90% 0 %
OIB 0.763015 0 0 0 0 23.70% 23.70% 0 0
18.48
UB 0.909286 0 0 0 64.12547
9.07% 9.07% 0 %
WB 0.933956 0 0 0 0 6.60% 6.60% 0 0
ZB 1 0 0 0 0 0 0 0 0
Table 4.6 shows the slack value for inefficient banks and the required improvements in each
input and output variables

Kumar. S and Gulati. R (2008) as stated in their study, Slack occurs for inefficient DMU only in a
particular DEA run. These slack give information about the parts which an incompetent DMU
needs to improve.
For these study model which is input oriented, the input slack indicate the excess amount of
input used and the output slack shows the amount of under produced output.
The above table indicating that from the observed 16 banks all inefficient banks need
improvement in both input variables, for DEPO the banks slacks (the deposit money need to
decreased) is lies between 3.90% and 23.70%, in case of NON-IE (the non-interest expenses
need to decreased) the slack is lies between 2.70% and 23.70%.in case of output variables all
DMU have zero slack of TLAA so no need of enhancement in this output variable and for NON-
 47

II (the other output variable) only five banks need improvement which are BOA, BUIB, CBO,
NIB and UB with a percentage of 16.85%, 9.23%, 11.53%, 19.31% and 18.48% respectively.
Unlike to the others CBO and DB need to decrease additional amount from their input in
addition to inefficient percentage. CBO need to decrease 13,797,800 birr from non-interest
expenses and DB need to decrease 432,868,000 from its deposit.

CHAPTER FIVE
 48

5. SUMMARY OF FINDING CONCLUSION AND RECOMMENDATION

5.1 Introduction
This chapter presents the main finding of the analysis in the summary of the finding section, and
conclusion, recommendation and suggestion for future research is presents in the other sections.

5.2 Summary of findings

This part of the chapter summarized the major findings of the analysis that obtained from
MAXDEA 7, which is overall technical, pure technical and scale efficiency of PCBs and input
output slack of DMU.

The finding from the analysis reveal that Form the period of 10 years under the study only in two
years in 2012 and in 2013 from the entire banks 50% of the sample banks were fully efficient.

AB and OIB did not have any full efficiency in overall technical and scale efficiency score in
any of the ten years period and ZB and ADIB were inefficient only in one year from the entire
period of the study.

Generally the efficiency score of each bank during the study period is inconsistent except ADIB,
DGB, EB and ZB.

The outcome of the mean efficiency of the banks shows that from the sample of 16 banks only 6
banks were efficient in overall technical, pure technical and scale efficiency. From the inefficient
banks eight of them (AB, BOA, BUIB, CBO, LIB, NIB, OIB and UB) inefficiency were mostly
caused in technical inefficiency.

DB and WB were efficient in pure technical efficiency and the inefficiency of these two banks is
because of scale inefficiency. OIB were the least efficient bank in overall technical and pure
technical efficiency score and WB is least efficient bank in scale efficiency score.

The mean efficiency of PCBs during the study period is 91%, 92.33% and 98.56% for OTE, PTE
and SE respectively. This implies that on average the banks could have to save 9% and 7.66% of
input from the present consumption of input to produce the present level of output for CRS and
VRS method respectively. In scale efficiency also the banks were inefficient and in decreasing
return to scale, these implies that banks are not in proper size (they are too large).
 49

According to the result the inefficiency of Ethiopian private commercial banks is mostly
managerial inefficiency (inefficient utilization of resource) rather than scale inefficiency.

Based on the slack value result from the observed 16 banks all inefficient banks need
improvement in both input variables, for DEPO the banks slacks (the deposit money need to
decreased) is lies between 3.90% and 23.70%, in case of NON-IE (the non-interest expenses
need to decreased) the slack is lies between 2.70% and 23.70%.in case of output variables all
DMU have zero slack of TOL so no need of enhancement in this output variable and for NON-II
(the other output variable) only five banks need improvement which are BOA, BUIB, CBO, NIB
and UB with a percentage of 16.85%, 9.23%, 11.53%, 19.31% and 18.48% respectively. And
specifically CBO need to decrease 13,797,800 birr from non-interest expenses and DB need to
decrease 432,868,000 from its deposit in addition to the inefficiency ratio of decreasing.

5.3 Conclusion

This study is about measuring of technical and scale efficiency of Ethiopian private commercial
banks. The study used DEA method to analysis to measure the efficiency of the banks for a
period of 10 years from 2009 to 2018. Both CRS and VRS models of DEA were used to analysis
OTE, PTE and SE of 16 PCBs and to select the input and output variables the researcher
followed the intermediation approach based on this the input variables are deposit and non-
interest expenses and the output variables are total loan and non-interest income.

The result indicated that the level of OTE, PTE, and SE of PCBs is 91%, 92.33% and 98.56%
respectively

From the sampled 16 banks 6 banks (ADIB, AIB, BB, DGB, EB and ZB) were in the efficiency
frontier and OIB is the least efficient bank in resource utilization and WB were scale inefficient
from the entire sample banks.

According to the outcome of the study most of private commercial banks in Ethiopia during the
study period were inefficient and this inefficiency is caused on both pure technical inefficiency
(managerial inefficiency) and scale inefficiency (not operate at most productive scale). Still in
most of the inefficient banks (8 banks) the source of overall technical inefficiency is pure
technical inefficiency.
 50

In general According to the result the inefficiency of Ethiopian private commercial banks is
mostly managerial inefficiency (inefficient utilization of resource) rather than scale inefficiency.

In case of input output improvements the inefficient banks should decreased deposit money from
a minimum of 3.9% to a maximum of 23.7%, non-interest expenses need to decreased a
minimum of 2.7% to a maximum of 23.7% and need to enhancement non-interest income. In
addition to that CBO need to decrease 13,797,800 birr from non-interest expenses and DB need
to decrease 432,868,000 from its deposit.

5.4 Recommendation

To be OT efficient most of the inefficient banks need to improve their resources utilization. For
this the inefficient banks should learn from the efficient banks especially from ZB and ADIB.

Most of the inefficient bank should revise their business strategy and/or used properly their
business strategy. Since most of the banks were in decrease return to scale rather than expansion
better to offer services like agency banking and electronic banking system.

As a result of the analysis the study conclude that the inefficiency of Ethiopian private
commercial banks is mostly managerial inefficiency so that, banks need to develop their
manager’s capacity. Through experience sharing and giving updated training.

BOA, BUIB, CBO, NIB and UB need to focus on income diversifying from fee based and
commotion based incomes.

The recommendation for national bank of Ethiopia is that to adopt DEA method of analysis to
evaluate the banking sector, this will help NBE in order to get new insight about the sector and
its efficiency this will support for regulation and policy making.

5.5 Suggestion for feature research

 51

Feature researchers can apply DEA method of analysis using production approach to evaluate
branch level efficiency of each bank this will help the bank in strategic planning

Feature researchers can apply DEA method using different input and outputs for evaluating the
financial sectors, such as banks and insurances.

References
 52

Adamu, T. K. & Kenenisa, L. D. (2017). Financial performance of commercial banks in Ethiopia: an application of
camel model. Zenith international journal of multidisciplinary research, vol.7 (12), pp. 208-221
Aderaw, G., & Singh. M., (2016) Development of Financial Sector in Ethiopia: Literature Review, Development of
Financial Sector in Ethiopia: Literature Review, Vol.7, No.7, 2016
Agu c.c., (2004), Efficiency of Commercial Banking in the Gambia: African Review of Money Finance and
Banking pp31-50.
Akhtar, M.H., & Nishat, M., (2002), X-efficiency analysis of commercial banks in Pakistan: A preliminary
investigation, The Pakistan Development Review, 567-580.
Alvi, m. (2016) A Manual for Selecting Sampling Techniques in Research. Munich Personal RePEc Archive,
University of Karachi, Iqra University, MPRA Paper No. 70218,
Aly, H.Y., Grabowski, R., Pasurka, C. and Rangan, N. (1990). Technical, Scale, and Allocative Efficiencies in U.S.
banking: An Empirical Investigation, Review of Economics and Statistics, DOI: 10.2307/2109710,
https://www.researchgate.net/publication/24094832.
Avikaran N. K., (2006), Productivity Analysis in the service sector with Data Envelopment Analysis, Third Edition,
University of Queensland QLD 4072 Australia.
Avkiran, (2000) Avkiran, N.K. (2000) ‘Rising productivity of Australian trading banks under deregulation 1986–
1995’, Journal of Economic & Finance, Vol. 24,No.2,pp.122–140

Baidya, M. K. & Mitra, D. (2012), An analysis of the technical efficiency of Indian public sector banks through
DEA approach. Int. J. Business Performance Management, Vol. 13, Nos. 3/4, 2012
Balcerzak, A. P., Kliestik, T., Streimikiene, D., & Smrcka, L. (2017). Non-Parametric Approach to Measuring the
Efficiency of Banking Sectors in European Union Countries, Acta Polytechnic Hungarica, Vol. 14, No. 7,
2017
Banker, R. D., Charnes, A., Cooper, W.W. (1984). Some models for estimating technical and scale inefficiencies in
data envelopment analysis. Management Science, Vol. 30, No.9
Bauer, P. W., Berger, A. N., Ferrier, G. D., & Humphrey, D. B. (1998). Consistency conditions for regulatory
analysis of financial institutions: a comparison of frontier efficiency methods. Journal of Economics and
business, 50(2), 85-114.
Camp, W. G. (2001). Formulating and Evaluating Theoretical Frameworks for Career and Technical Education
Research. Journal of Vocational Educational Research, 26 (1), 27-39.
Chandrasekar, Sendhil, R., & Ramasundaram, P. (2017). e Compendium of Training-cum-Workshop on Data
Analysis Tools and Approaches (DATA) in Agricultural Sciences, March 2017
Charnes, A., Cooper, W., E. (1978). Measuring efficiency of decision making units. European Journal of operational
Research, 429-444.
Coelli, T.J et al (1998) An Introduction to Efficiency and Productivity Analysis,Boston, Dodrecht, London, Kluwer
Academic Publishers.
Creswell, J.W., (2014). Research design: qualitative, quantitative, and mixed methods approaches / 4th Ed.
Thousand Oaks, California: Sage Publications
 53

Dakito, A. (2015). Assessment of Banking Performance Using Capital Adequacy in Ethiopia. Economics. Vol. 4,
No. 6, 2015, pp. 106-111.
Debasish, S.S. (2006). Efficiency Performance in Indian Banking, Use of Data Envelopment Analysis. Global
business review, 7: 2, Sage Publications New Delhi/Thousand Oaks/London, DOI:
10.1177/097215090600700209
Emrouznejad, A., & Cabanda, E. (2015). Introduction to Data Envelopment Analysis and its Applications,
Handbook of Research on Strategic Performance Management and Measurement Using Data Envelopment
Analysis: 235 255. IGI Global, USA.
Erasmus, C., & Makina, D. (2014). An Empirical Study of Bank Efficiency in South Africa Using the Standard and
Alternative Approaches to Data Envelopment Analysis (DEA), Journal of Economics and Behavioral
Studies Vol. 6, No. 4, pp. 310-317, Apr 2014 (ISSN: 2220-6140)
Fasika, Z. A. (2016). Evaluating the Technical Efficiency of Commercial Banks in Ethiopia: A Data Envelopment
Analysis. European Journal of Business and Management, Vol.8, No.28, 2016.
Fentaw, L. & Sharma, D. (2017). Determinants of Banks Profitability: Review and Assessment. International
Journal of Management & Business Studies, Vol. 7. Iss ue 1, Jan - March 2017
Gamachis, G. (2016). Technical Efficiency and Productivity of Ethiopian Commercial Banks: Data Envelopment
Analysis (DEA). International Journal of Scientific and Research Publications, Volume 6, Issue 9,
September 2016 860 ISSN 2250-3153
Grover, V. K. (2015) Research Approach: An overview. Golden Research Thoughts, Volume-4 (8) ISSN 2231-5063
Gudata, A. (2015). Financial performance analysis in banking sector (in selected commercial banks in Ethiopia).
International Journal of Current Research Vol. 7, Issue, 10, pp.21883-21886
Gulati, R. (2011). Evaluation of technical, pure technical and scale efficiencies of Indian banks: An analysis from
cross-sectional perspective. presentation in The 13th Annual Conference on Money and Finance in the
Indian Economy on 25-26th February, 2011Indira Gandhi Institute of Development Research, Mumbai
Habtamu, B. A. (2015). Financial Performance of the Ethiopian Banking Sector. International Journal of Science
and Research (IJSR) ISSN (Online): 2319-7064 Index Copernicus Value (2013): 6.14
Horvatova, E. (2018). Technical Efficiency of Banks in Central and Eastern Europe. International Journal of
Financial Stud, 6, 66; doi: 10.3390/ijfs6030066
Islamia, J. M. (2016). Research Design, Research in Social Science: Interdisciplinary Perspectives, New Delhi
Kablan, S. (2007). Measuring bank efficiency in developing countries: the case of WAEMU (West African
economic monetary union). African Economic Research Consortium, may 2007
Kablan, S. (2010). Banking Efficiency and Financial Development in Sub-Saharan Africa. IMF Working Paper
African Department June 2010, JEL Classification Numbers: G21, O16, O55
Kamau, A. W. (2011). Intermediation efficiency and productivity of the banking sector in Kenya. Interdisciplinary
Journal of Research in Business, Vol 1, Issue 9 pp12 26.
Kapur,D. and Abebaw, K. G. (2012). Financial performance and ownership structure of Ethiopian commercial
banks. Journal of Economics and International Finance, Vol. 4 (1), pp. 1–8, 7, DOI: 10.5897/JEIF11.036
 54

Karimzadeh, M. (2012). Efficiency Analysis by using Data Envelop Analysis Model: Evidence from Indian Banks.
International Journal of Latest Trends in Finance & Economic Sciences, Vol-2 No. 3 September, 2012
Kizito, E.U. (2012).The Place of Financial Markets in the Development Process: Evidence from Nigeria. Journal of
Economics and Behavioral Studies, Vol. 4, No. 11, pp. 649-659, Nov 2012 (ISSN: 2220-6140)

Kocisova, K. (2013). Technical efficiency of top 50 world banks, Journal of Applied Economic Sciences, Volume
VIII, Issue 3(25), Fall 2013 Technical University of Kosice Faculty of Economics, Slovak Republic
Kokobe, S. A. & Birhanu, D. N. (2015). Determinants of Financial Performance of Commercial Banks in Ethiopia.
Journal of Business Management and Economics, 3: 11 November (2015).
Kothari, C.R. (2004). Research Methodology; methods and Techniques. University of Rajasthan Jaipur (India);
New Age International (P) Ltd,

Kounetas, K., & Tsekouras, K.(2007). Measuring Scale Efficiency Change Using a Translog Distance Function.
International Journal of Business and Economics, 6(1), 63-69.
Kumar, N. & Singh, A. (2015), Measuring Technical and Scale Efficiency of Banks in India Using DEA. IOSR
Journal of Business and Management, Volume 17, Issue 1.Ver. II (Jan. 2015), PP 66-71

Kumar. S and Gulati. R (2008). An Examination of Technical, Pure Technical, and Scale Efficiencies in Indian
Public Sector Banks using Data Envelopment Analysis. Eurasian Journal of Business and Economics, 1 (2),
33-69.
Luccheti, R., L. Papi, and A. Zazzaro (2000) Banks’ Efficiency and Economic Growth: A Micro-macro Approach.
Italy: Universita di Ancona
Marjanovic. I, Stankovic. J. J and Popovic. Z (2018). Efficiency estimation of commercial banks based on
financial performance: input oriented DEA CRS/VRS models. Economic Themes, 56(2): 239-252
Melaku, A. (2017). Financial performance analysis of private commercial banks of Ethiopia: camel ratings.
International journal of scientific and research publications, volume 7, issue 10, Issn 2250-3153
Mokhtar, H.S.A., Abdullah, N. & Alhabshi, S.M. (2008). Efficiency and competition of Islamic banking in
Malaysia. Humanomics, Vol. 24 No. 1, 2008 pp. 28-48, 0828-8666, DOI 10.1108/08288660810851450
Mongid, A., & Tahir, I. M. (2010). Technical and scale efficiency of Indonesian rural banks. Journal of Banks and
Bank Systems, Volume 5, Issue 3, 2010
Mousa, G. A. (2015). Financial Ratios versus Data Envelopment Analysis: The Efficiency Assessment of Banking
Sector in Bahrain Bourse. International Journal of Business and Statistical Analysis, vol 2.(2), ISSN (2384-
4663). July-2015
Muhabie, M. M. (2015). Evaluation of the Financial Performance of Banking Sectors in Ethiopia: The Case of
Zemen Bank. Global Journal of Management and Business Research: C Finance, Volume 15 Issue 9
Version 1.0 Year 2015
Nand, 2014 Review. International Journal of Advance Research and Innovation, ISSN 2347 – 3258, Volume 1
(2014) 120-126
 55

Ncube, M. (2009), Efficiency of banking sector in South Africa. “African Economic conference” 11 – 13 November
2009 • United Nations Conference Centre, Addis Ababa-Ethiopia.
Nguyen, Viet H. (2007), Measuring Efficiency of Vietnamese Commercial Banks: An Application of Data
Envelopment Analysis (DEA), in Khac Minh Nguyen and Thanh Long Giang (ed.), Technical Efficiency
and Productivity Growth in Vietnam, Publishing House of Social Labor, Tokyo
Nigmonov, A. (2010). Bank Performance and Efficiency in Uzbekistan Eurasian. Journal of Business and
Economics, 3 (5), 1-25
Othman, F. M., Mohd-Zamil, N. A., Abdul Rasid, S. Z., Vakilbashi, A. & Mokhber, M.(2016). Data Envelopment
Analysis: A Tool of Measuring Efficiency in Banking Sector, International Journal of Economics and
Financial Issues, 2016, 6(3), 911-916.
Pasiouras, F. (2006) Estimating the Technical and Scale Efficiency of the Greek Commercial Banks, 2006, School
of Management, University of Bath. Performance of conventional banks in Singapore”, Journal of Banking
and Finance, 26, 79-98.
Rahel, T.& Maru S. (2015).Determinants of the Financial Performance of a Private Commercial Bank in Ethiopia.
Journal of Business and Administrative Studies (JBAS) Vol. 7 No. 2
Rahi, S. (2017). Research Design and Methods: A Systematic Review of Research Paradigms, Sampling Issues and
Instruments Development. International Journal of Economics & Management Sciences, Volume 6 • Issue 2
• 1000403, ISSN: 2162-6359
Rajha K. S. (2016). Determinants of Non-Performing Loans: Evidence from the Jordanian Banking Sector. Journal
of Finance and Bank Management, Vol. 4, No. 1, pp. 125-136
Rao, M. and Tekeste, B. (2012). Cost Efficiency and Ownership Structure of Commercial Banks in Ethiopia: An
application of non-parametric approach, European Journal of Business and Management Vol. 4, No.10
Raphael, G. (2012). Commercial banks efficiency in Tanzania: A non-parametric approach. European Journal of
Business and Management, Vol. 4. (21), 2012.
San, O. T. & Heng, T. B. (2013). Factor affecting the profitability of Malaysian commercial bank, African Journal
of Business Management, 7(8). 649-660
Sanjeev, G.M. (2006). Data Envelopment Analysis (DEA): For Measuring Technical Efficiency of Banks. Journal of
Business Perspective, Vol 10 (1)
Sufian, F. and Chong, R.R., 2008. Determinants of bank profitability in a developing economy: Empirical evidence
from the Philippines. Asian academy of management journal of accounting and finance, 4(2), pp.91-112.
Tahir, I.M., Abu Bakar, N.M. & Haron, S. (2009). Estimating Technical and Scale Efficiency of Malaysian
Commercial Banks: A Non-Parametric Approach. International Review of Business Research Papers, Vol.5
No. 1 January 2009 Pp. 113- 123
Takbiri, O., Mohammadi, M., & Naderi, B. (2015). The efficiency of bank branches: Management Science Letters,
5 (2015) 1111–1116 , Department of Industrial Engineering, Faculty of Engineering, Kharazmi University,
Tehran, Iran
 56

Tesfaye, B. L. (2014). Efficiency in the Ethiopian Banking System: An Application of Data Envelopment Analysis.
European Journal of Business and Management, Vol.6, No.23, 2014
Thu Huong, T. T. & Firoz, B. (2016). Evaluating the Efficiency of Vietnamese Commercial Banks using Data
Envelopment Analysis, In: 6th Annual International Conference on Accounting and Finance, Singapore,
30th to 31st May 2016.
Tony, A., & Alemayehu, G. (2001). Ethiopia’s new financial sector and its regulation, Finance and Development
Research Program. Working paper series paper no 39 August 2001
Wozniewska, G.(2008). Methods of measuring the efficiency of commercial banks: an example of Polish banks.
Banking Department Wroclaw University of Economics, 53–345 Wroclaw, ul. Komandorska,118/120,
Poland

Yadav, R. (2015). Technical efficiency, pure technical efficiency and scale efficiency of Russian commercial banks:
an empirical analysis (2007-2014), Int.J.Eco.Res. ISSN: 2229-6158 2015, v 6 i6, 52 - 59
Yannick, G. S., Hongzhong, Z. & Thierry, B. (2016). 12th International Strategic Management Conference, ISMC
2016, 28-30 October 2016, Antalya, Turkey Wuhan University of Technology, Wuhan, China

Yidersal, D. D. (2018). Measurement of Commercial Bank Efficiency in Ethiopia: an Application to Data

Envelopment Analysis (DEA). European Journal of Business and Management, ISSN 2222-2839 (Online)
Vol.10, No.13, 201
 1

YEAR
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

TE TE TE TE TE TE TE TE TE TE TE TE TE TE TE TE TE TE TE TE
DMU CRS VRS CRS VRS CRS VRS CRS VRS CRS VRS CRS VRS CRS VRS CRS VRS CRS VRS CRS VRS
0.875 1 0.855 0.88 0.887 0.888 0.867 0.873 0.869 0.884 0.91 0.911 0.864 0.865 0.839 0.911
AB        
ADIB             1 1 1 1 1 1 1 1 1 1 1 1 0.948 0.992

AIB 0.904 0.904 1 1 1 1 1 1 0.988 1 0.963 1 0.981 1 0.999 1 1 1 1 1

BA 0.978 0.978 1 1 0.964 1 0.948 0.949 1 1 0.909 0.92 0.789 0.817 0.851 0.86 0.915 0.915 0.816 0.885

BIB     0.906 1 0.844 1 0.876 1 1 1 0.868 0.881 0.873 0.887 1 1 1 1 1 1

BUB     1 1 1 1 1 1 0.898 0.962 0.923 0.931 0.875 0.881 0.965 0.965 0.944 0.965 0.824 0.829

CBO 1 1 0.774 0.8 0.683 0.684 0.791 0.802 1 1 1 1 1 1 0.997 1 0.919 0.925 0.765 0.774

DB 1 1 1 1 0.988 1 1 1 1 1 1 1 0.946 1 0.877 1 0.923 1 0.827 1

DGB                     0.815 1 0.762 1 1 1 1 1 1 1

EB                     0.836 0.847 1 1 1 1 1 1 1 1

LIB 0.913 0.93 0.896 0.92 0.863 0.873 0.885 0.891 1 1 0.848 0.854 0.876 0.995 0.972 0.996 0.847 0.873 0.805 0.815

NIB 1 1 0.962 0.982 0.929 1 1 1 1 1 1 1 0.956 1 0.927 0.937 0.966 0.973 0.885 0.896

OIB 0.998 1 0.692 0.704 0.738 0.745 0.726 0.741 0.809 0.812 0.748 0.75 0.845 0.873 0.791 0.798 0.76 0.767 0.851 1

UB 0.947 0.947 0.925 0.93 1 1 0.994 0.996 0.86 0.869 0.834 0.834 0.778 0.812 0.94 0.958 0.985 0.99 0.833 0.833

WB 1 1 1 1 0.851 1 0.993 1 0.943 0.974 0.813 0.822 0.806 0.825 0.969 0.99 1 1 0.892 1

ZB 0.993 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

MEAN 0.973 0.976 0.93 0.945 0.903 0.946 0.934 0.947 0.956 0.965 0.902 0.92 0.897 0.936 0.95 0.964 0.945 0.955 0.893 0.933

MAX 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
MIN 0.904 0.904 0.692 0.704 0.683 0.684 0.726 0.741 0.809 0.812 0.748 0.75 0.762 0.812 0.791 0.798 0.76 0.767 0.765 0.774
NO.EF.BA 4 6 6 7 4 10 6 8 8 9 5 7 4 8 5 8 7 8 5 8
PERC 0.4 0.6 0.5 0.583 0.308 0.769 0.429 0.571 0.571 0.643 0.313 0.438 0.25 0.5 0.313 0.5 0.438 0.5 0.313 0.5
 1

Appendix Two
All banks data for each year
COMPAN NON- NON-
Y DEPO IE TLO II
AB 263 15 158 7
AB 779 50 452 54
AB 1476 90 843 82
1,474.8
AB 2,518.22 154.53 3 130.43
2,311.3
AB 3,623.81 222.4 4 213.2
3,073.6
AB 4,832.58 293.36 9 215.69
4,209.5
AB 6,792.20 380.19 7 298.7
539.81 5898.45 443.06
9566.089
AB 7 5 2
3205.828 179.44 2000.03 143.19
SD 7 98 7 39
min 263 15 158 7
539.81 5898.45 443.06
9566.089
max 7 5 2
3731.362 218.16 2302.61 180.51
mean 4 21 1 03
ADIB 211 21 154.488 23
34.100 52.627
ADIB 561 43 328 66
ADIB 792.41 60 511.04 85
ADIB 1109.6 88 762.35 112
1049.78
ADIB 1562.54 119 9 142
ADIB 2271.7 171 1561.56 178
207.38 2034.99 201.46
2970.377
ADIB 4 6 9
983.9333 69.834 683.133 65.091
SD 9 38 8 23
min 211 21 154.488 23
207.38 2034.99 201.46
2970.377
max 4 6 9
1354.089 100.06 914.603 113.44
mean 6 92 3 24

AIB 4,962 155 2,713 201

AIB 6106 181 3146 383
3986.46
AIB 7744 213 4 533
AIB 9204 295 5504.61 442
 2

AIB 12,545 474 7,710 598

9,176.3
AIB 15,040 617.05 6 832.07
18,520. 12,482.
AIB 42 800.91 04 839.49
22,832. 1056.2 15450.7
AIB 03 3 8 901.42
30,590. 1499.1 22576.3 1176.1
AIB 90 3 4 8
1933.9 31049.0 1203.1
AIB 43451.38 75 6 16
607.25 9312.12 335.07
SD 12267.1 67 4 81
min 4,962 155 2,713 201
max 43,451 1,934 31,049 1,203
mean 17,100 723 11,379 711
BA 4494.19 147.25 2708.96 128.92
BA 5,139 145 3153 207
3315.68
BA 6,075 196 7 246
3897.40
BA 6,771 226 6 226
BA 8,496 226 4,702 226
BA 9,096 345 5,061 451
11,118. 5,905.2
BA 17 462 2 354
13,634. 8011.60
BA 96 713 9 535
20,700. 13927.2
BA 81 1075 4 789
1452.1 17780.9 552.70
BA 25794.54 47 6 9
7059.109 447.34 5073.40 206.69
SD 1 48 2 29
min 4494.19 145.00 2708.96 128.92
1452.1 17780.9
max 25794.54 5 6 789.00
mean 11131.97 498.74 6846.31 371.56
BIB 238 12.441 153 2
BIB 694 23 331.818 33
BIB 932 34.57 499.55 51.5
BIB 1,593 50.19 979 55.29
1184.69
BIB 2,012 116.16 5 86.37
BIB 3067.9 161.97 1875.49 162.38
3701.65
BIB 5296.52 298.9 1 325.83
 3

5254.24
BIB 7592.4 465.89 5 476.83
694.29 10097.0 434.97
BIB 10889 8 4 1
3633.957 236.96 3261.77 183.30
SD 6 94 5 72
min 238 12.441 153 2
694.29 10097.0
max 10889 8 4 476.83
3590.535 206.37 2675.16 180.90
mean 6 99 5 79
BUB 240 18 192 11
BUB 491 34 366.261 39
BUB 903 48 651.94 47
BUB 1,548 65 949 51
1,343.3
BUB 2,152 133.12 0 115.58
2,417.9
BUB 3,501.04 210.28 4 173.81
3631.84
BUB 5,384.60 309.78 4 243.8
5201.67
BUB 7,479.58 405.82 3 293.34
575.78 6841.60 359.86
9848.374
BUB 7 3 1
3386.441 194.36 2355.42 126.07
SD 2 34 1 69
min 240 18 192 11
575.78 6841.60 359.86
9848.374
nax 7 3 1
3505.288 199.97 2399.50 148.26
mean 2 63 7 57
CBO 789 40.7 596 11
CBO 1372 65 722 54
CBO 1,980 81 802 97
CBO 2798 104.97 1384 131.94
CBO 4465 189 2,116 301
3644.11
CBO 5450 306 5 460
CBO 7368 608 6566.04 524
5851.65
CBO 8402 808 7 250
9679.60
CBO 14277 855 2 331
1283.9 15144.9 657.98
25807.59
CBO 08 3 8
7673.826 430.74 4776.40 215.39
SD 3 37 7 85
 4

min 789 41 596 11

max 25808 1284 15145 658
mean 7271 434 4651 282
DB 7,925 204 4,452 321
10,144. 5,048.8
DB 55 257.89 4 481.67
11,841. 6,217.5
DB 24 327.04 4 678.51
14,065. 8,123.8
DB 60 421.86 1 827.63
15,851. 8,862.3
DB 26 513.98 2 796.05
17,681. 9,429.6 1004.1
DB 34 614.25 3 7
19,814. 11,526. 1101.0
DB 11 884.17 99 5
22,758. 1041.8 12695.1 1211.6
DB 50 5 2 4
27,782. 1514.0 18082.8 1344.7
DB 52 8 9 2
1854.3 23057.5 1186.6
35986.8
DB 85 3 89
8595.941 558.35 5911.89 333.91
SD 1 8 8 29
min 7,925 204 4,452 321
max 35,987 1,854 23,058 1,345
mean 18,385 763 10,750 895
DGB 500 44 266.65 40
DGB 819.34 67 334.91 63
DGB 871.76 94 591.284 103
DGB 1,431.53 125 780.77 128
198.19 1553.71 206.97
2153.322
DGB 8 2 2
650.9671 59.937 516.744 64.940
SD 7 1 8 87
min 500 44 266.65 40
198.19 1553.71 206.97
2153.322
max 8 2 2
1155.190 105.63 705.465 108.19
mean 4 96 2 44
EB 929.44 74 506.74 78
1133.60
EB 1,565.22 70 7 72
1615.51
EB 2,389.27 109 5 132
EB 3,681.05 170 2440.97 201
EB 5090.526 222.59 3313.95 261.65
 5

2 1 3
1672.677 65.852 1101.53 81.629
SD 4 85 3 77
min 929.44 70 506.74 72
222.59 3313.95 261.65
max 5090.526 2 1 3
2731.101 129.11 1802.15 148.93
mean 2 84 7 06
LIB 704 34 470 16
LIB 1018 41 584 55
LIB 1,297 52 676 65
LIB 1,737 75 970.663 104
LIB 2106 91 1318 128
1,541.1
LIB 2687 141 7 133
LIB 4457 307 2,831 312
LIB 6333.56 440.91 4303.39 367.6
LIB 8774.85 451.04 5485.68 279.07
602.88 7374.04
LIB 11639.59 4 1 334.33
3715.691 209.49 2398.50 130.24
SD 2 4 3 84
min 704 34 470 16
max 11,640 603 7,374 368
mean 4,075 224 2,555 179
NIB 3,296 132 2220 172
NIB 4,127 182 2,546 290
NIB 5,157 193 2,767 324
NIB 5,838 218 3709 326
NIB 6,655 275 4543 281
5,407.7
NIB 7,923.29 276 4 281
6,894.0
NIB 9,774.11 467 4 321
12,423.0 7511.98
NIB 2 547 4 290
16,416.4
NIB 4 736 10711.3 448
21619.23 949.84 13498.6 407.96
NIB 6 7 5 5
5899.384 272.35 3740.15 74.960
SD 1 38 5 97
min 3,296 132 2,220 172
max 21,619 950 13,499 448
mean 9,323 398 5,981 314
 6

OIB 184 6 113 2

OIB 821 44 369 50
OIB 1526 64 662 91
OIB 2117 109 1020 107
OIB 3050 178 1,621 138
2,531.6
OIB 5004 239 1 219
4,706.5
OIB 7290 391 7 247
5165.74
OIB 9348 579 7 247
OIB 13414.13 787 7175.54 583
986.93 9968.57 962.83
19927.02
OIB 3 3 9
340.45 3320.85 293.94
6381.591
SD 42 5 43
min 184 6 113 2
max 19927 987 9969 963
mean 6268 338 3333 265
UB 3,616 124 2,152 135
UB 4725 159 2614 260
UB 6,066 163 3,277 292
UB 6758 226 4085 313
UB 8,063 353 4,711 305
5,069.6
UB 8,904.98 400.76 2 243.98
11,804.3 6,860.0
UB 6 589.84 8 385.85
UB 13037.64 719.26 8534.36 447.64
11996.3
UB 16505.15 887.31 1 461.76
1156.4 14869.8 622.73
23079.65
UB 32 9 8
6006.887 351.17 4215.50 137.40
SD 6 52 5 02
min 3,616 124 2,152 135
max 23,080 1,156 14,870 623
mean 10,256 478 6,417 347
WB 3,728 133 2,112 239
WB 3,923 172 2474 318
WB 5957 257 2910 500
3565.67
WB 5758 252 4 408
WB 7,550.66 326 4,690 366
WB 8,384.48 438 4,604.4 409
 7

2
6071.91
WB 9,870.94 582 5 473
11,078.5 7506.21
WB 5 712 5 509
14,018.2 10235.0
WB 3 968 7 798
1285.4 14785.0 971.90
20506.13
WB 68 4 9
5146.824 377.56 3999.46 223.17
SD 5 88 4 2
min 3,728 133 2,112 239
max 20,506 1,285 14,785 972
mean 9,078 513 5,895 499
ZB 278 24.2 189 13
ZB 688 42 384 103
ZB 1,163 55 645.225 157
ZB 1,793 77 1012.69 164
ZB 2,506 181 1,370 255
1,429.9
ZB 3,031 120 6 250
2,156.6
ZB 3,823 165 9 239
3253.94
ZB 5,487 227 2 336
ZB 7,323 315 3970.61 503
351.66 423.31
10241.32 4995.01
ZB 6 2
3193.892 114.22 1627.51 147.22
SD 1 1 8 99
min 278 24 189 13
max 10241 352 4995 503
mean 3633 156 1941 244