DETERMINANTS OF NON-PERFORMING LOAN OF COMMERCIAL

BANKS IN ETHIOPIA

BY

AZEB SHIMELIS

A THESIS SUBMITTED TO COLLEGE OF BUSINESS AND ECONOMICS OF ADDIS

ABABA UNIVERSITY IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR
AWARD OF THE DEGREE OF MASTER SCIENCE IN ACCOUNTING AND
FINANCE PROGRAM.

ADDIS ABABA UNIVERSITY

COLLEGE OF BUSINESS AND ECONOMICS

DEPARTMENT OF ACCOUNTING AND FINANCE

Feb 2022

Addis Ababa Ethiopia

1
 Determinants of Non-performing loan of 2022
commercial Banks in Ethiopia

DECLARATION

I, undersigned declare that this study entitled as “Determinant of non-performing loan of

commercial banks in Ethiopia.” is my own work. I have carried out the research work
independently with the guidance and support of the research advisor. This study had not been
submitted to any degree/diploma in this or any other institution where any of the content
presented is the result of input or data from a related collaborative research programme this is
duly acknowledged.

Name of student: Azeb Shimelis

Signature: _________________

Date:_____________________
 Determinants of Non-performing loan of commercial
Banks in Ethiopia
Certification

Addis Ababa University

School of Graduate Studies

This is to certify that the thesis “Determinant of non- performing loan of commercial banks in
Ethiopia.” was carried out by Azeb Shimelis under the supervision of Dr. Abebaw Kassie,
submitted in partial fulfillment of the requirements for the degree of Master of Science in
Accounting and Finance complies with the regulations of the University and meets the accepted
standards with respect to originality and quality.

Approved by:
Internal examiner: Abebe Yitayew (PHD) Signature __________ Date __________

External examiner: Dejene Mamo (PHD) Signature ________ __ Date _________

Advisor: Abebaw Kassie(pHD) Signature __________ Date _________

I
 Determinants of Non-performing loan of commercial
Banks in Ethiopia

ACKNOWLEDGEMENT

First and for most I would like to thank the Almighty God with his mother Virgin Marry who
gave me divine strength and the Most Merciful & endless support throughout my life.

I offer my sincerest gratitude to my Advisor Abebaw kassie (PHD), for his unreserved
assistance, suggestions and guidance.

I extend my deepest gratitude to my family, for their endless patience and support. In
particular, I thank my wonderful mother, W/ro Alganesh Mekonnen for all her support & her
silent praying for my success however she passed away before end of my paper (RIP), my dear
father Shimelis Egualle, and my beloved sisters and brothers especially my sister Hibret
Shimelis.

Finally, I want to thank all directly or indirectly participated in the accomplishment of the study.

i
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

Table of Contents

ACKNOWLEDGEMENT ..................................................................................................... i

Table of Contents ………………………………………………………………………....ii

List of Tables & Figure ......................................................................................................... v

List of Acronyms.................................................................................................................. vi

Abstract .............................................................................................................................. viii

CHAPTER ONE ................................................................................................................... 1

Introduction ........................................................................................................................... 1

1.1. Back ground of the study ............................................................................................. 1

1.2. Statement of the problem .......................................................................................... 3

1.3. Research Questions and Hypotheses......................................................................... 6

1.3.1 Research Question........................................................................................................ 6

1.3.2 Objectives of the Research ........................................................................................... 6

1.3.1. General Objective...................................................................................................... 6

1.3.2. Specific Objective ..................................................................................................... 6

1.4. Scope & Limitation of the research........................................................................... 6

1.5. Significance of the research ...................................................................................... 7

1.6. Organization of the paper .......................................................................................... 7

CHAPTER TWO................................................................................................................... 8

REVIEW OF RELATED LITERATURES .......................................................................... 8

2.1. Theoretical and Conceptual Review .............................................................................. 8

2.1.1. Definition of Non-performing loans ........................................................................... 9

2.1.2. Determinants of Non-performing loans .................................................................... 12

2.1.2.1. Bank specific factors .............................................................................................. 13

ii | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
2.1.2.2. Macroeconomic factors .......................................................................................... 15

2.3. Empirical Studies on the determinants of non-performing loans................................. 17

Research Hypothesis (RH) .................................................................................................. 20

2.4. Conclusions and knowledge Gap ................................................................................. 20

Chapter Three ...................................................................................................................... 22

3.1. Research Design & Methodology ................................................................................ 22

3.2. Sampling Design .......................................................................................................... 22

3.3. Data Analysis and Presentation.................................................................................... 23

3.4. Model Specification ................................................................................................... 24

3.5. Varianble of Study .................................................................................................. 25

CHAPTER FOUR ............................................................................................................... 27

4.1. Descriptive statistics..................................................................................................... 27

4.2. Test for the Classical Linear Regression Model (CLRM) Assumptions ..................... 29

4.2.1. Autocorrelation Test.................................................................................................. 30

4.2.2. Heteroscedasticity Test ............................................................................................. 31

4.2.3. Normality Test........................................................................................................... 31

4.2.4. Multicolinearity Test ................................................................................................. 32

4.3. Model Selection............................................................................................................ 34

4.3.1 Random Effect versus Fixed Effect Models .............................................................. 34

4.4. Result of Regression Analysis...................................................................................... 36

CHAPTER FIVE................................................................................................................. 41

SUMMARY, CONCLUSION AND RECOMMENDATION ........................................... 41

5.1. Summary ...................................................................................................................... 41

iii | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
5.3. Recommendations ........................................................................................................ 42

References ........................................................................................................................... 44

Appendix1: Model Selection and Basic Tests for CLRM Assumptions................................ i

Appendix2. Raw Data ......................................................................................................... v

iv | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
List of Tables & Figure

Table

Page

Table: 1.1.Minimum provision requirement…………………………………………………....11

Table:2.1. Summary of variables, their definition and expected sign. ………………..……….26

Table:3.1. Summary of descriptive statistics for dependent & independent Variable...............27

Table: 4.1.Resultsof random effect regression model…………………………………………..36

Table:5.1. Summary of actual and expected sign’s of explanatory variable..............................40

Figure1: Conceptual framwork…………………………………………………………………28

v|Page
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

List of Acronyms

AWB: Awash Bank

BUIB: Buna International Bank

BOA: Bank of Abyssinia

CAR: Capital Adequacy Ratio

CBB: Construction and Business Bank

CBE: Commercial Bank of Ethiopia

CLRM: Classical Linear Regression Model

CSA: Central Statistics Agency

DB: Dashen Bank

DR: Deposit rate

DW: Durbin-Watson

ECB: Ethiopian Commercial Banks

EXR: Exchange Rate

FEM: Fixed Effect Model

GDP: Gross Domestic Product

INF: General inflation rate

LTD: Loan to deposit ratio

NBE: National Bank of Ethiopia

NIB: Nib international Bank

NPL: Nonperforming loan

OLS: Ordinary Least Square

REM: Random Effect Model

vi | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
ROA: Return on Asset

ROE: Return on Equity

HB: Hibret Bank

UM: Un employment

WB: Wogagen Bank

vii | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

Abstract
This study Examine the bank Specific bank economic determinants of commercial banks in
Ethiopia .Non-Performing Loan (NPL) has been crucial factor these days in terms of Banking
sector sustainability and profitability. This paper intends to assess factors that may influence
the rising level of nonperforming loans in commercial banks of Ethiopia. The study uses,
Capital Adequacy Ratio and Loan to deposit ratio as specific determinants of nonperforming
loans and Inflation rate, Exchange rate and UN employment rate as macro economic factors.
Panel data from 2008-2018 was taken from published annual reports and NBE data base.
Descriptive Statistics, correlation analysis and random effect panel least square regression was
used to analyze data through STATA 14 application software. The analysis leads to the
conclusion that reduced level of nonperforming loans leads to increased banks performance. It
was also concluded that Exchange rate and un employment rate has got a negative significant
impact on non-performing loans. On the other hand the finding showed Insignificant
relationship loan to deposit, capital adequacy and inflation rate. The study suggests that,
commercial banks should consider specific and the macro economic factors before extending
loans

Key words: Non-performing loans; Macroeconomic determinants; Bank specific determinants

viii | P a g e
 Determinants of Non-performing loan of commercial
Banks in Ethiopia
CHAPTER ONE
Introduction

1.1. Back ground of the study

Banks can be defined as intermediaries between depositors and borrowers in an economy

(Heffernan, 1996). Banks are one of the most important financial intermediaries whose primary
function is to mobilize funds from surplus areas and pass them over (lend) to those who face
shortages (deficit) (Sontakke and Tiwari, 2013.By its nature banks face number of challenges
within internal and the external business environment, the nucleus of banks is known with risks
which include credit risk, market risk, interest rate risk, default risk, operational risk, exchange
rate risk (Aruwa& Musa, 2014). Basically, banks operate with three basic objectives which are
profitability, growth of assets and customer base. Asset quality is an aspect of bank
management which entails the evaluation of firm assets in order to facilitate the measurement of
the level and size of credit risk associated with its operation. Asset quality is micro prudential
determinants commercial banks soundness and profitability. It relates to the left-hand side of a
bank balance sheet and focused on the quality of loans which provides earnings for a bank
(Abata, 2014).

According to Felix A. et.al, (2008) Credit creation is the main income generating activity for the
banks. However, this activity involves huge risks to both the lender and the borrower. The risk
of a trading partner not fulfilling his or her obligation as per the contract on due date or anytime
thereafter can greatly expose the smooth functioning of a bank’s business International
Monetary Fund (IMF, 2008)defined non-performing loan (NPL) as any loan in which interest
and principal payments are more than 90 days overdue; or more than 90 days’ worth of interest
has been refinanced .On the other hand the (Basel Committee, 2011) puts non-performing loans
as loans left unpaid for a period of 90 days. Non-performing loans are those loans that are
ninety days or more past due or no longer accruing interest or non-performing loans are those
loans, which are not generating income. This is also articulated by (Caprio and Klingebiel ,
1996), as non-performing loans are those loans which are relatively long period of time do not
generate income that is, the principal and or interest on these loans have been left unpaid for at

1
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
least ninety days. In the other word, non-performing loans (NPL) exist when the borrower
default in making their payment according to the schedule agreed upon the acceptance of the
contract between lender and borrower. In general, an asset/loan becomes non-performing when
it ceases to generate income for the bank. The economic and financial costs of NPL’s are
significant. Graham and Humphrey (1978) suggested that, banks with larger amounts of NPL
have greater tendency to incur large amount of future losses, and hence, NPL should be
included as an indicator of the banking system stability. (Fofack and Hippolyte, (2005)Pointed
out that, these loans might negatively affect the level of private investment, increase deposit
liabilities and constrain the scope of bank credit. NPLs have an adverse effect on banking
sectors survival. Thus, since nonperforming loans had an adverse effect on the banking sectors’
survival, the cause for NPLs should be given due consideration. Its causes are different in
different countries that might be due to situational factors such as the level of economic
condition in which the banking sectors are operating and bank level factors. Consequently, this
issue attracted different researcher’s interest in different countries. Thus, many studies are
performed on the determinants of NPLs of financial sectors worldwide. For instance:-(Saba
et.al, 2012) Made study on the determinants of NPLs on US Banking sector and found as
lending rate had negative while inflation and Real GDP per capital had positive and significant
effect on NPLs. Besides, (Loizis et al. 2010) examined the determinants of NPLs in the Greek
financial sector using dynamic panel data model and found as real GDP growth rate, ROA and
ROE had negative whereas lending, unemployment and inflation rate had positive significant
while loan to deposit ratio and capital adequacy ratio had insignificant effect on NPLs. The
study of (Skarica, 2013) on the determinants of NPLs in Central and Eastern European countries
through fixed effect model was also found as GDP growth rate, unemployment rate and
inflation had negative and significant impact on NPLs. Similarly, (Carlos &Andres,(2012)
based on OLS model estimators found as NPLs have negative association with GDP growth rate
whereas a positive association with unemployment rate. This study is to examine the
determinants of NPLs by using panel data and a quantitative type of research in commercial
banks of Ethiopia.

2|Page
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

1.2. Statement of the problem

Banks exist to provide financial intermediation service while at the same time endeavor to
maximize profit and shareholders' value. Lending is considered the most important function for
fund utilization of Commercial Banks as major portion of their income is earned from loans and
advances (Radha, 1980).

NPLs affect the bank`s liquidity and profitability which are the main components for the overall
efficiency of the bank. An increase in NPLs leads high level of provision expenses, which
diminishes income. Again, mismatch of maturities between asset and liability creates liquidity
risk for the banks that deteriorate bank’s overall credit rating including its image (Badar and
Yasmin, 2013).Therefore; the determinants of NPLs should be given a due consideration
because of its adverse effect on survival of banks.

Thus, given the unique features of banking sector and environment in which they operate and
also rapid expansion of banking institutions in Ethiopia, there are strong wishes to conduct a
separate study on the determinants of NPLs of banking sector in Ethiopia. Besides, inconsistent
results in different studies among researchers are also another motive to conduct this study. To
this end, the main objective of this study was to examine the bank specific and macroeconomic
determinants of NPLs of commercial banks of Ethiopia.

The study of Saba et al. (2012) on the title of “Determinants of Nonperforming Loan on US
Banking sector” found negative significant effect of lending rate and positive significant effect
of real GDP per capital and inflation rate on NPL via OLS regression model. Similarly, the
study of Louzis et al.(2010) examined the determinants of NPLs in the Greek financial sector
using dynamic panel data model and found as real GDP growth rate, ROA and ROE had
negative where as lending, unemployment and inflation rate had positive significant while loan
to deposit ratio and capital adequacy ratio had insignificant effect on NPLs. However, Swamy
(2012) examined the determinants of NPLs in the Indian banking sector using panel data and
found as GDP growth rate, inflation, capital adequacy and bank lending rate have insignificant
effect on NPLs. Shingjergji (2013) who conducted study on “the impact of bank specific factors
on NPLs in Albanian banks system” utilized OLS estimation model and found as ROE have

3|Page
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
significant negative on NPLs. However, Ahmad and Bashir (2013) conducted a study on the
“Bank Specific Determinants of Nonperforming Loan” by static panel data model and found as
ROE has insignificant negative association with NPLs. Makri et al.(2014) identify the factors
affecting NPLs of Euro zone’s banking systems through difference Generalized Method of the
Moments (GMM) estimation. Accordingly, they found as ROA did not show any significant
impact on NPL ratio. However, Selma and Jouini (2013) conducted a study on Italy, Greece and
Spain for the period of 2004-2008 via panel data model and found a significant negative effect
of ROA on NPLs. similarly, Boudriga et al. (2009) conducted a study on the title “Problem
loans in the MENA countries via random-effects panel regression model and found as ROA has
significant negative effect on NPLs.

In addition to the above facts, there has been the study made by Wondimagegnehu (2012)
assessed the determinants of nonperforming loans in Ethiopian commercial banks from 2005-
2010. He used mixed research approach and Survey was conducted with professionals engaged
in both private and state owned Banks in Ethiopia holding different positions using self-
administered questionnaire, structured review of documents and records of banks. Result of the
study shows that poor credit assessment, failed loan monitoring, underdeveloped credit culture,
lenient credit terms and conditions, aggressive lending, compromised integrity, weak
institutional capacity, unfair competition among banks, willful default by borrowers and their
knowledge limitation, fund diversion for unintended purpose, over/under financing by banks
attribute to the causes of loan default was assessed the bank specific factors affecting NPLs via
OLS estimation model by the help of SPSS software.

Zelalem T. (2013) made an empirical study on Ethiopia commercial banks; the study examines
the bank-specific and macro-economic determinants of Non-performing loans (NPLs). He used
a mixed methods research approach by mixed documentary analysis (structured review of
documents) and in-depth interviews from 2000-2011 in eight commercial banks in Ethiopia to
assess determinants of Non-performing loans (NPLs). The result of the study shows that, loan
growth, financial performance, operational efficiency, effective exchange rate, inflation rate and
gross domestic product have negative and statistically significant relationship with banks’
NPLs. On the other hand, variables like bank size and state ownership have a positive and

4|Page
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
statistically significant relationship with banks’ NPLs. However, the relationship for average
lending rate and income diversification were found to be statistically insignificant.

Gadise (2014) Examined determinants of nonperforming loan empirical study in case of

commercial banks in Ethiopia and found that ROE,ROA, capital adequacy ratio, lending rate,
and effective tax rate had statistically significant effect on the level of NPLs from 2002-2013.
However, the results of fixed effect regression model revealed the insignificant effect of loan to
deposit ratio and inflation rate on the level of NPLs of commercial banks in Ethiopia for the
period under consideration.

Anisa (2015)the finding of the study failed to reject four research hypotheses that indicate the
relationship between bank’s nonperforming loan and cost efficiency, loan to deposit ratio,
deposit rate and interest rate whereas, three hypotheses indicating the relationship between
bank’s nonperforming loan and bank solvency, gross domestic product, and inflation rate
rejected (Solvency ratio and GDP growth rate had insignificant impact on bank’s NPL in
Ethiopia From 2004-2013.

Generally, NPLs are loans that are outstanding both in its principal and interest for a long period
of time contrary to the terms and conditions under the loan contract. Any loan facility that is not
up to date in terms of payment of principal and interest contrary to the terms of the loan
agreement is NPLs. Thus, the amount of nonperforming loan measures the quality of bank
assets (Tseganesh, 2012).

However, this study considers both macroeconomic factors such as Inflation rate, Exchange rate
and Unemployment rate, bank specific factors like loan to deposit ratio, ROE, ROA, and capital
adequacy ratio as determinant factors of NPLs. Besides, random effect model and version
14Stata software used in this study to examine the determinants of NPLs of commercial banks
in Ethiopia. as per the knowledge of the researcher, the Specific & macroeconomic determinants
of NPLs in Ethiopian banking sector, all the studies use the data before the year 2015 and some
variables are differ so, the current impact of the determinants on NPLs is not studied. The
purpose of this study is to examine the determinants of NPLs by using panel data and a
quantitative type of research in commercial banks of Ethiopia.

5|Page
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

1.3. Research Questions and Hypotheses

The general objective of this study is to investigate the determining factors of NPLs in
Ethiopian commercial banking sector. Having this general objective, the following sections
discuss the research questions and hypothesis that have been used for investigation purpose.

1.3.1 Research Question

R.Q. 1. What are the bank specific factors?

2. What are the non-bank specific factors?

1.3.2 Objectives of the Research

1.3.1. General Objective

The main objective of this study was to determine bank specific and macro economic factors
that could affect banks NPLs and to examine the relationship between these factors and the rate
of NPLs of commercial banks in Ethiopian.

1.3.2. Specific Objective

The paper has the following specific objective:

To measure the relationship between (CAR, LTD) With NPLs.

To measure the relationship between (INF,EXR, UM) With NPLs
To determine the Significance of each determinants with NPLs

1.4. Scope & Limitation of the research

This study is to examine the determinant of NPL of commercial banks in Ethiopia the study
include one state owned commercial banks namely Commercial Bank of Ethiopia and six
private banks namely AWB,DB,HB,WB,NIB and BOA have the data for the past ten
consecutive years, from 2008 to 2018. The research data was based on intensive secondary data
review.

The study relied on secondary data and limited to two bank specific (capital adequacy and loan
to deposit) and three macroeconomic factors (Inflation rate, Exchange rate and unemployment).

6|Page
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
This study only cover the financial aspect and limited only 11 years data from only seven
commercial banks, also it used only secondary data, researcher lack of experience to the study,
confidentiality of data because most financial institutions will not willingly disclose information
to researchers.

1.5. Significance of the research

Since the study is designed for investigate Determinants of NPLs in commercial banks first and
foremost the findings of the study will tremendously beneficial to researcher, academic,
regulatory body i.e. NBE policy makers and managements of commercial banks this study also
can assist the bankers in formulating different strategies and policies to reduce NPLs. As result,
this study will extend the existing literature by providing evidence on the determinants of NPLs
in Ethiopian commercial banking context by utilizing both macroeconomic and bank specific
variables. Apart from contributing to the literature, this study may also have important practical
implications for commercial banks mangers and bank regulators authorities in dealing with
NPLs management. Moreover, and as a stepping stone for those who want to make further study
on the issue of NPLs in the Ethiopian banking context afterwards. Finally, it may provide a
possible opportunity to all stake holders to gain deep knowledge about the leading cause of
NPLs in Ethiopian commercial banking sector.

This initiates the bank management and executives with applied knowledge on the management
of identified variables and provides them with understanding of activities that will enhance their
loan quality and play a vital role in filling gap in understanding the determinants of NPLs.

1.6. Organization of the paper

This thesis is organized into five chapters. The first chapter starts with background of the study,
statement of the problem, objective of the study, significance of the study, scope and limitation
of the study. The second chapter focuses on both theoretical and empirical review of related
literature. The third chapter deals with the research methodology. Chapter four deals with the
data analysis and presentation and the fifth chapter contain the conclusion and recommendation
of the study including the direction for further study.

7|Page
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

CHAPTER TWO
REVIEW OF RELATED LITERATURES

2.1. Theoretical and Conceptual Review

The literature on non-performing loans (NPLs) focus extensively on the macroeconomic and
bank-level determinants of non-performing loans with little or no focus on the role of financial
development for the persistence of non-performing loans. For instance, Nkusu (2011)
investigate the determinants of non-performing loans across 26 developed countries over the
1998 to 2009 period and find that deteriorating macroeconomic conditions such as: economic
growth and higher unemployment led to higher non-performing loans. Klein (2013), using
country-level data, investigate 16 CESEE countries over the1998 to 2011 period and find that
aggregate NPLs are negatively associated with credit growth, unemployment, gross domestic
product growth rate and inflation. Louzis et al (2012) investigate the determinants of non-
performing loans (NPLs) in the Greek banking sector for each loan category: consumer loans,
business loans and mortgages, and find that non-performing loans are significantly influenced
by management quality, GDP, unemployment, interest rates and public debt. Skarica (2014),
using country-level non-performing loans data, investigate the determinants of non-performing
loans among 7 countries in the Central and Eastern European (CEE) region during the third-
quarters of 2007 and 2012 and find that higher non-performing loans are significantly
associated with economic slowdown, unemployment and inflation. Beck et al (2015) examine
the macroeconomic determinants of non-performing loans (NPLs) across 91 countries and find
that non-performing loans are significantly affected by real GDP growth, share prices, exchange
rate and lending interest rate. Anastasiou et al (2016) focus on the Euro-area banking system
during the 1990 to 2015 period and find that income tax and output gap significantly influence
NPLs. In the literature, GDP growth rate is often associated with NPLs because NPLs are lower
during economic booms and are higher during recessionary periods (see, Skarica, 2014; Ozili,
2015; Beck et al, 2015; etc). Also, high unemployment levels are associated with high non-
performing loans because high unemployment can affect borrowers’ capacity to repay loans
(Klein, 2013; Nkusu, 2011; Ozili, 2018). The effect of inflation on non-performing loans is

8|Page
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
inconclusive in the literature, with mixed evidence (see, Klein, 2013; Beck et al, 2015, etc.).
Global risk-factors may also influence the persistence of non-performing loans.

2.1.1. Definition of Non-performing loans

NPLs can be defined as defaulted loans, which banks are unable to profit from. Usually loans
falls due if no interest has been paid in 90 days, but this may vary between different countries
and actors. Defaulted loans force banks to take certain measures in order to recover and
securitize them in the best way. (Ernst & Young, 2004) The definition of impaired loans is loans
that have not expired, but it is uncertain whether the borrowers could repay their debts.
Generally banks and other credit institutes do not class them as potential non-performing loans
and therefore calculate with little or no loss provision from them. (Ernst & Young, 2004) Non-
performing loans are those loan facilities which borrowers often have difficulties repaying.

The term ‟bad loans‟ as described by Basu (2003), is used interchangeably with NPLs and
impaired loans as identified in Fofack (2005). Berger and De Young, (1997) also considers
these types of loans as “problem loans”. Thus these descriptions are used interchangeably
throughout the study. Theoretically, there is no global standard to define NPLs which could be
applied to all economies of the world (Hou 2006 and Bloem and Gorter 2002). Variations exist
in terms of the classification system, the scope, and contents. Such problem potentially adds to
disorder and uncertainty in the NPL issues (Hou 2006). Thus, the definition of NPLs varies
from one banking system to another according to banking laws and regulations (Issa 2009). In
practical terms, Quantitative and qualitative criteria are used individually or collectively by
credit institutions to identify the situation of the loan. A quantitative criterion uses numbers of
days or months to determine the weakness of borrowers to repay their debt, while a qualitative
criterion uses all the information about the future of loans and borrowers (Bloem and Gorter
2002).

In referring to the period of NPLs, Rose (2002 p.118) defined NPLs as "a loan is placed in the
NPLs category when any scheduled loan repayment is past due for more than 90 days". In
addition, Bloem and Freeman (2005 p.8) give the definition of NPLs as "a loan is NPLs when
payments of interest and/or principal are past due by 90 days or more, or interest payments

9|Page
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
equal to 90 days or more have been capitalized, refinanced, or delayed by agreement". Others
consider NPLs as a borrower stopping to repay the installments in a period of over six months.
For instance, Cho (2002 p.10) define NPLs as "a loan was considered NPLs only when it was
past due six months or more while provisioning requirements".

In light of the above discussion, a study for the International Monetary Fund (IMF), Cortawarria
et al. (2000) define NPLs according to region where they originate from. For instance, in
countries like France, Spain, Portugal, Switzerland and Norway, loans became NPLs when
principal and interest uncollected for more than 90 days. Others countries like Greece and Italy
used more than 90 days. In countries like U.K and Germany there is no explicit criteria to be
used in determining loans as good or bad.

As per NBE (2018), NPLs are defined as “loans or advances whose credit quality has
deteriorated such that full collection of principal and/or interest in accordance with the
contractual repayment terms of the loan or advances in question”. It further provides that:

“Short term loans are NPLs when principal and/or interest is due and uncollected for
90(ninety) consecutive days or more beyond the scheduled payment day or maturity. Medium
and long term loans are NPLs when principal and/or interest is due and uncollected for
12(twelve) consecutive months or more beyond the scheduled payment day or maturity”.

According to NBE (2018) directive, Ethiopian commercial banks are required to classify their
loans as pass, special mention, substandard, doubtful and loss.

 Pass: loans in this category are fully protected by the current financial and paying
capacity of the borrower and not subject to any criticism.

 Special mention: Short term loans past due for 30 days or more, but less than 90 days
and medium and long term loans past due for 6 month or more, but less than 12 months.

 Substandard: Short term loan past due for 90 days or more, but less than 180 days and
medium and long term loans past due for12 months or more, but less than 18 months

 Doubtful: Short term loan past due for 180 days or more, but less than 360 days and
medium and long term loans past due for18 months or more, but less than 3 years.

10 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
 Loss: Short term loan past due for 360 days or more, and Medium and long term loans
past due for3 years or more.

According to NBE (2018) directive, all Ethiopian commercial banks required holding
provisions for each loans mentioned above so as to absorb the potential losses in their loans
portfolio. The minimum provision requirements are mentioned below:

Table 2.1: Minimum provision requirement

Classification of loans Minimum percentage

Pass 1%

Special mention 3%

Substandard 20%

Doubtful 50%

Loss 100%

Source: NBE (2018)

Among the above classified loans, the last three loans such as, substandard, doubtful and loss
are considered as NPLs (NBE 2018). In fact, such types of classification under the NBE
directive banks are required to hold 50% provision.

Five Cs of Nonperforming/Bad loans

As noted by MacDonald (2006), there are five Cs of bad credits that represent the issues used to
guard against/prevent bad loans). These are:

11 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
 Complacency: refers the tendency to assume that because of the things were good in the
past, they will be good in the future. For instance, Assuming the past loan repayment
success since things have always worked out in the past.
 Carelessness: indicates the poor underwriting typically evidenced by inadequate loan
documentation, lack of current financial information or other pertinent information in
the credit files, and lack of protective covenants in the loan agreement. each of these
makes it difficult to monitor a borrower`s progress and identify problems before they are
unmanageable.
 Communication ineffectiveness: inability to clearly communicate the bank`s objectives
and policies. This is when loan problem can arise. Therefore, the bank management
must clearly and effectively with existing loans as soon as they appear. Communicate
and enforce the loan policies and loan officers should make the management aware of
specific problems.
 Contingencies: refers the lenders` tendency to play down/ignore circumstances in which
a loan might in default. It focuses on trying to make a deal work rather than identifying
down side risk.
 Competition: involves following the competitors` action rather than monitoring the
bank`s own credit standards. Banks, however, still have required expertise, experiences,
and customer focus to make them the preferred lender for many types of loan. Lending
is not just a matter of making loan and waiting for repayment. Loan must be monitored
and closely supervised to prevent loan losses (MacDonald, 2006).

2.1.2. Determinants of Non-performing loans

Researchers set different determinant factors for NPLs depending on multidimensional aspects
like economy and countries situation and give more emphasized for two grand factor bank
specific and macroeconomic factors. Macroeconomic factors on external events such as the
overall macroeconomic conditions which are likely to affect the borrowers’ capacity to repay
their loans while the Bank specific internal factor which attributes for the level of
nonperforming loan.

12 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

2.1.2.1. Bank specific factors

Bank-specific variables refer to those factors which characterized individual banks. Those
factors can be influenced by managerial decisions and usually associated with the specific
policy choices of a particular bank with regard to its efforts to maximize efficiency and improve
its risk management. Hence, bank specific variables that are usually theorized as determinates of
NPLs are include , loan growth, financial performance, bank size, ownership structure, the
quality of the loan portfolio and operational efficiency of bank’s. Hence, the following part of
this particular section clearly presents the bank-specific variables that are used in this study.

 Capital Adequacy Ratio

Capital adequacy is a measure of bank`s financial strength since it shows the ability to
Withstand/tolerate with operational and abnormal losses. It also represents the ability to
undertake additional business (Habtamu, 2012). As noted by Makriet al.(2014), CAR
determines risk behavior of banks. It is a measure of banks solvency and ability to absorb risk.
Thus, this ratio is used to protect depositors and promote stability and efficiency of financial
systems. According to Makriet al.(2014), there is negative relationship with NPLs indicating a
risky loan portfolio is marked by a high NPL (equivalent to high credit risk). However,
Djiogapand Ngomsi (2012) found positive association between NPLs and capital adequacy
ratio. It is measured by total Equity to total asset ratio. However, it is expected to have negative
association with NPLs in this study. This implies that well capitalized banks are less incentive
to take risk.

CAR= Total Equity

Total Asset

It is widely accepted in literature that low capital adequacy ratio is associated with higher
probability of loans default, because it induces the bank managers to involve in more risky
projects. The confirmation of this statement was found in (Berger and DeYoung, 1997) USA
banks analysis result of thinly capitalized banks take increased portfolio risk, which results in
Determinants of Non-Performing Loans in Ethiopian Commercial Banks higher levels of
problem loans in the future, which supports moral hazard theory developed by them. However,

13 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
Louzisetal. (2010) in his study examined the determinants of NPLs in the Greek financial sector
using dynamic panel data model and found as capital adequacy ratio had insignificant effect on
NPLs. According to Makri et al. (2014), there is negative relationship with NPLs indicating a
risky loan portfolio is marked by a high NPL (equivalent to high credit risk).

 Loan to Deposit Ratio

The LDR is not the only metric used to ascertain a bank's liquidity. Modern banks today have
multiple sources of finance beyond equities and deposits. The diversity of financing sources
reduces the importance of LDR in determining a bank's health. Basel III which is part of the
Basel Accords provides various complementary statistics to measure banking liquidity more
comprehensively.

a ratio between the banks total loans and total deposits. The ratio is generally expressed in
percentage terms If the ratio is lower than one, the bank relied on its own deposits to make loans
to its customers, without any outside borrowing. If on the other hand the ratio is greater than
one, the bank borrowed money which it re loaned at higher rates, rather than relying entirely on
its own deposits. Banks may not be earning an optimal return if the ratio is too low. If the ratio
is too high, the banks might not have enough liquidity to cover any unforeseen funding
requirements or economic crises. Banking analysts commonly used metric for assessing a bank's
liquidity.

The loan to deposit ratio is affected by the operational strategy of a bank’s management.
Excessive rapid loan growth declined bank’s capital levels and useful pointers the deterioration
of banks financial health and can be employed as early warning indicators of future problem
loans (Das and Ghosh 2007)As disclosed by Jimenez and Saurian (2006) loan growth is
considered as one of the most important causes of problem loans.

However, according to Sinkey and Green Walt (1991) a rapid expansion of loan may not be a
problem by itself, but such expansion leads to poor screening and lending to borrowers of
inferior quality.

14 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

2.1.2.2. Macroeconomic factors

The existing literature provides evidence that suggests a strong association between NPLs and
macroeconomic factors. Several macroeconomic factors which the literature proposes as
important determinants of NPLs are: real GDP growth, inflation rate, effective exchange rate,
real interest rate, unemployment rate, broad money supply (M2) and GDP per capital (Salas and
Suarina 2002, Fofack 2005 and Jimenez and Saurina 2005). This study only considers inflation
rate, exchange rate and Unemployment rate.

 Inflation rate
It is a situation in which the economies overall price level is rising. It represents sustained and
pervasive increment in aggregate price of goods and services resulting decline in purchasing
power of money. Accordingly, when inflation is high and unexpected, it can be very costly to an
economy. At the same time, inflation generally transfers resources from lender and savers to
borrowers since borrowers can repay their loans with birr that are worthless. It is determined as
the general consumer price index. This indicates that, as inflation increase, the cost of
borrowing gets more expensive and deteriorates the quality of loan portfolio.

There are ambiguous results regarding the relationship between NPLs and inflation rate.

According to Farhanet al.(2012), Skarica(2013), Klein(2013) and Tomak(2013) found as there

is a positive relationship between NPLs and Inflation rate. Theoretically, inflation should reduce
the real value of debt and hence make lending easier. However, high inflation may pass through
to nominal interest rates, reducing borrowers’ capacity to repay their debt. Through its attraction
with the tax system, it can increase tax burden by artificially increasing income and profits.
Besides, inflation cause firms to increase their costs of changing prices. Finally, it made
individuals to hold less cash and make more trips to banks since inflation lowers the real value
of money holdings. It can negatively affect the borrowers’ real income when wages are stick.

Keeping this information in mind, the relationship between NPLs and inflation is expected to be
negative for this study. In case, the figure amount of CPI was taken from CSA.

15 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
 Exchange rate

like inflation a change in effective exchange rate can also affects borrowers’ debt servicing
capacity through different channels and its impact on NPL can be positive or negative (Nkusu
2011). As noted in Pasha and Khemraj (2009), depreciation of the exchange rate can have
mixed implications on borrowers’ debt servicing capacity. On the one hand, it can improve the
competitiveness of export- oriented firms. As long as the value of domestic currency
depreciated (lower), export- oriented firms can dominate the international

market at lower price since their production cost is covered in domestic currency which has
lower value than foreign currency and their revenue is collected in foreign currency which has
higher value as compared to the domestic currency. Hence, depreciation of exchange rate can
improve the debt-servicing capacity of export-oriented borrowers. On the other hand, it can
adversely affect the debt-servicing capacity of borrowers who borrow in foreign currency
(import-oriented firms).

One of the main determinants of economic instability is the exchange rate volatility (Zameer&
Siddiqi, 2010). Its impact on NPL can be positive or negative (Nkusu 2011). Depreciation of the
exchange rate can have mixed implications on borrowers‟ debt servicing capacity. A negative
correlation of real effective exchange rate and NPLs was founded by Castro
(2013),Zribi&Boujelbene(2011),Gunsel (2012), Vogiazas&Nikolaidou (2011) and Fofack
(2005).Other studies such as: Kalirai&Scheicher (2002) and Aver (2008) have not identified
any relationship between foreign exchange fluctuation and NLPs ratio. like inflation a change in
effective exchange rate can also affects borrower debt servicing capacity through different
channels and its impact on NPL can be positive or negative (Nkusu 2011).

 Unemployment rate

Is a phenomenon that occurs when a person who is actively searching for employment is unable
to find work? An increase in the unemployment rate could influence negatively the cash flow
streams of households and increase the debt burden. Skarica (2013) revealed that
unemployment rate has statistically significant negative association with NPLs with justification
of rising recession and falling during expansions and growth has an impact on the levels of

16 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
NPLs. With regards to firms, increases in unemployment may signal a decrease production as a
consequence of a drop in effective demand. Measures of unemployment appear to be a good
predictor of problem loans in all countries (Gambera, 2000). Iuga and Lazea (2012) examined
the influence of unemployment rate on non-performing loans in Romania during 2008-2011.
The results reveal that there is a strong correlation between unemployment rate and
nonperforming loans. Nkusu (2011), Farhan et al. (2012), Selma and Jouini (2013), Akinlo and
Emmanuel (2014), Vogiazes and Nikolaidu (2011), and Dragomirescu (2014) and Bofondi and
Ropele (2011) also found positive and significant relationship between NPLs and
unemployment. The reason behind the result is explained by the literatures as increase in
unemployment negatively affects income of individuals thereby increasing their debt burden. In
addition, increased unemployment in the economy negatively affected the demand for products
of firms which ultimately affected the production/sales of the firms, which led to a decline in
revenues of the firms and a fragile debt conditions.

2.3. Empirical Studies on the determinants of non-performing loans

There are many studies that have been conducted so far in determinants of bank’s NPLs. Their
results have shown that, banks NPLs are determined by either internal or external factors.
Hence, the following section presents the empirical evidence on the determinants of bank’s
NPLs with a particular focus on those studies that have been conducted more recently, as far as
they are the best indicators of the current situation.

Keeton and Morris (1987) investigated the causes of loan losses for a sample of nearly 2,500
US commercial banks for the period 1979–1985. Using simple linear regressions, they found
out local economic conditions along with the poor performance of certain sectors like
agriculture and energy explain the variation in loan losses recorded by the banks. The study also
stated that commercial banks with greater risk desire tend to record higher losses.

Salas and Saurian (2002) analyze problem loans of the Spanish commercial and savings banks
and find that credit risk is determined by microeconomic individual bank level variables, such
as bank size net interest margin, capital ratio and market power, in addition to real GDP growth.
Hu et al (2006) examined the relationship of ownership structure, size of banks and income

17 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
diversification with NPLs of commercial banks in Taiwan with a panel dataset covering the
period 1996-1999. The study shows that banks with higher government ownership recorded
lower NPLs. Hu et al (2006) also show that bank size is negatively related to NPLs while
diversification has not found a significant association with banks NPLs in Taiwan commercial
banking sector.

Louzis et al., (2010) examined the determinants of NPLs in the Greek financial sector using
dynamic panel data model and found as real GDP growth rate, ROA and ROE had negative
whereas lending, unemployment and inflation rate had positive significant while loan to deposit
ratio and capital adequacy ratio had insignificant effect on NPLs.

Vogiazas and Nikolaidou (2011) investigated the credit risk determinants of the Bulgarian
banking sector by means of time series modeling approach covering the time period from
January 2001 to December 2010. The results indicate that, the macro economic and financial
markets‟ variables, specifically the unemployment rate, the construction index, the industrial
production index and the real effective exchange rate jointly with the credit growth and the
global financial crisis influence the NPLs of Bulgarian banks.

Farhan et al. (2012) studied the economic determinants of Non-Performing Loans in 10

Pakistani banks by using a primary data collected via a structured questionnaire from 201
bankers who are involved in the lending decisions or analyze the credit risk or handling NPLs
portfolio. Correlation and regression analysis was carried out to analyze the impact of six
independent variables on NPLs. According to the results Pakistani bankers perceive that Interest
Rate, Energy Crisis, Unemployment, Inflation, and Exchange Rate has a significant positive
relationship with the non-performing loans of Pakistani banking sector while GDP growth has
significant negative relationship with the non-performing loans of Pakistani banking sector.

Louzis et al. (2012) identified factors that result in nonperforming loans in the banking sector of
Greece by adopting a dynamic panel data method using both bank-specific and macroeconomic
variables. The empirical research revealed that non-performing loans can be explained by
macroeconomic variables, including GDP, interest rate, public debt, and unemployment, as well
as the bank-specific variable of management quality.

18 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
Janvisloo and Muhammad (2013) analyze relationship between Non-Performing Loans (NPL)
and macroeconomic variables by using a dynamic panel data model in Malaysian commercial
banking system for period 1997-2012. The results show that there is a strong evidence cyclical
sensitivity of loan quality in Malaysia commercial banking system. Lending interest rate and
foreign direct investment outflow are the most effective factors on NPL ratio with simultaneous
positive effects and a reverse effect with one-year delay. The results also reveal that the impact
of external shocks on the domestic banking system is more than internal shocks.

Bucur and Dragomirescu (2014) studied the influence of macroeconomic conditions on credit
risk on the Romanian banking sector for the period 2008-2013. They revealed that the credit
risk is significantly and negatively affected by the exchange rate fluctuation and significantly
and positively affected by the unemployment rate. A study by Rahman et al (2017) on the
impact of financial ratios on non-performing loans of Bangladesh commercial banks applied an
econometric model to find out correlations among financial ratios and a sample of 96
observations has been analyzed from 20 banks out of 30 listed commercial banks during 2010-
2015. The study mostly agrees with the existing literature that, credit-deposit ratio, net interest
margin have a positive influence on the non-performing loans and capital adequacy ratio, return
on assets have a negative influence on the non-performing loans. It also reveals that, sensitive
sector’s loan, priority sector’s loan have significant positive influence on the non-performing
loans and unsecured loans, profit per employee, investment deposit ratio have significant
negative impact on gross non-performing loan.

Warue (2013) investigated the effects of bank specific and macroeconomic factors on
nonperforming loans in commercial banks of Kenya. Real GDP, GDP per capita, lending
interest rates, inflation, government expenditure, export and imports and exchange rate are the
variables included as macroeconomic factors. Whereas, credit risk management techniques,
bank structures, and quality management are factors included as bank specific factors. The
period covered under this study was 1995 to 2009.Secondary and primary data of 44
commercial banks in Kenya was taken. And a causal-comparative research design based on
bank structures was adopted. The study found evidence that per capita income was negative and
significantly related to NPL levels across bank ownership categories further, return on assets

19 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
(ROA) was negative and significantly related to NPLs levels in large banks and small banks but
insignificant in medium banks. In addition the study found that return on asset (ROA) was
negative and significant in local banks and government banks but not in foreign banks.
However the study found no evidence that banks asset size was related to NPLs levels across all
bank categories in Kenya. Akinlo and Emmanuel (2014), studied determinants of non-
performing loans in Nigeria over the period 1981-2011. The study provides a macroeconomic
model for non-performing loans and the analysis confirms that in the long run, economic
growth is negatively related to non-performing loan. On the other hand, unemployment, credit
to the private sector and exchange rate exerts positive influence on nonperforming loans in
Nigeria. In addition, the study reveals that in the short run, credits to the private sector,
exchange rate, lending rate and stock market index are the main determinants of non-performing
loans.

Research Hypothesis (RH)

Hypothesis is developed after supporting theoretical framework or comes from prior literature
and studies on the topic, so as to answer specific research question and to achieve the general
objective, the current study proposed the following research hypothesis.

 H1: There is a negative and significant relationship between CAR and NPLs.
 H2: There is a positive & significant relationship b/n loans to deposit & NPLs.
 H3: There is negative & significant relationship between Inflation rate & NPLs.
 H4: There is negative and significant relationship b/n Exchange rate and NPLs.
 H5: There is a positive & significant relationship b/n Unemployment & NPLs.

2.4. Conclusions and knowledge Gap

Theoretical and empirical reviews mentioned in the above concluded that NPLs have a great
impact on the operation of banks. The empirical studies show that banks NPLs can be
determined by many variables which can be grouped as macroeconomic and bank specific
factors. However, Most of the literatures that are discussed so far appeared to have focused on
studies that were conducted in the banking sector of different countries outside Ethiopia.
Consequently, the Banking sector of Ethiopia has so far received inadequate attention in the

20 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
literature. Conclusions cannot be made on the causes of NPLs of the banking sector of a given
country based on the finding of other countries because the studies were carried out based on
the data from diverse countries. Apart from the data originated from, those literatures by
themselves provided contradictory conclusions because of different models and methodologies
they used.

In the Ethiopia context, there was different study regarding Determinants of nonperforming
loan, Wondimagegnehu (2012) the related studies conducted by assessed the determinants of
NPLs in Ethiopian commercial banks focusing on bank-specific variables. The study by
Zelalem (2013) Gadise (2014), Anisa (2015), Mekdes A.(2017)Sitina A. (2018)study focused
on specific variables & macroeconomic variables. Accordingly, as per the knowledge of the
researcher, the Specific & macroeconomic determinants of NPLs in Ethiopian banking sector,
all the studies use the data before the year 2016 and some variables are differ so, the current
impact of the determinants on NPLs is not studied.

2.5. Conceptual frame work

The purpose of this study is to examine the determinants of NPLs of commercial banks in
Ethiopia. Based on the objective of the study, the following conceptual model is framed.

Figure: 1. Conceptual frame work

 Inflation
Non -performing Loan

Rate
 Capital
 Exchang
Adequacy
e Rate
Ratio
Specific Macro  Unemplo
 Loan To
Factors Economic yment
Deposit
Factors Rate

Source: Develop for the study

21 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

CHAPTER THREE
Research Methodology

This chapter discussed in detail the research methodology that has been followed when
conducting the research study. It detailed the research methodology that was used in the study
which includes the research design, data collection and analysis methods. Variables description,
model specification.

3.1. Research Design & Methodology

Research design is a master plan specifying the methods and procedures for collecting and
analyzing the required data. The choice of research design depends on objectives that the
researchers want to achieve (John, 2007). Since this study was designed to examine the
relationships between NPLs and its determinants, a logical reasoning either deductive or
inductive is required. Deductive reasoning starts from laws or principles and generalizes to
particular instance whereas inductive reasoning starts from observed data and develops a
generalization from facts to theory.

3.2. Sampling Design

Sample design deals with sample frame, sample size and sampling technique. Sampling is a
technique of selecting a suitable sample for the purpose determining parameters of the whole
population. Population is the list of elements from which the sample may be drawn (John,
2007). A sample is drawn to overcome the constraints of covering the entire population with the
intent of generalizing the findings to the entire population. There are Twenty Commercial banks
in Ethiopia. These are commercial bank of Ethiopia, Awash Bank , Dashen Bank, Abyssinia
Bank, Wegagen Bank, Hibret Bank, Nib International Bank, Cooperative Bank of Oromiya,
Lion International Bank, Oromia International Bank, Zemen Bank, Buna International Bank,
Berhan International Bank, Abay Bank, Addis International Bank, Debub Global Bank, Enat
Bank, Shebele Bank, Zemzem Bank, Sinqe Bank and Amhara Bank.

22 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
As noted by Kothari (2004), good sample design must be viable in the context of time and
funds available for the research study. Besides, judgmental sampling offers the researcher to
deliberately select items for the sample concerning the choice of items as supreme based on the
selection criteria set by the researcher.

Therefore, the data for this study collect from seven commercial banks in the country. This is
due to the fact that since the primary aim of this study is to examine the determinants of
nonperforming loans of commercial banks in Ethiopia, it is better to make generalization for the
banking sector of the country based on data drawn from sample bank which is much more
experienced in the industry. Further, lending is not a one night process rather it comes by
making operation for some consecutive years since the bank should have to accept deposit to
grant loans to reach stage of suffering from poor asset quality. Thus, as one can understand from
objective of the study, the researcher aimed to examine the determinants of nonperforming
loans of commercial banks in Ethiopia. Thus, the researcher used 11 years data of selected
commercial banks that provide financial statements consecutively from 2008-2018 periods.

To this end, the sample size of this study is not less than specified sample size required for
ones’ study since the accuracy and validity of the works never guaranteed by increasing the
sample size beyond specified limit. This is due to the fact that increasing the number of sample
size beyond the specified sample size required for ones’ study never add value to the accuracy
of the study rather it made information unmanageable due to redundancy(Ayalew, 2011).That is
why this study used Seven experienced commercial bank in Ethiopia from eighteen banks in the
country.

3.3. Data Analysis and Presentation

As noted by Kothari (2004), data has to be analyzed in line with the purpose of the research
plan after data collection. Accordingly, secondary data collected from NBE, CSA and head
office of each respective bank were analyzed to determine its suitability, reliability, adequacy
and accuracy. Thus, this study utilized both descriptive and econometric analysis based on a
panel data from 2008-2018 to examine the relationship between the NPLs and its determinant

23 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
factors in commercial banks of Ethiopia. The data collected from different sources were coded,
checked and entered to simple excel program to make the data ready for analysis.

3.4. Model Specification

Data collected from audited financial statement of commercial banks included in the sample and
publications of NBE and CSA. Thus, collected data analyzed using descriptive statistics and
multiple linear regression analysis. The descriptive statistics (Mean values and standard
deviations) used to analyzed the general trends of the data from 2008 to 2018. A multiple linear
regression model was used to determine the relative importance of each independent variable in
explaining the variation of NPLs in ECBs. According to Hausman and Siekpe (2008) emphasize
that regression methods have become an integral component of any data analysis concerned
with describing the relationship between a response variable and one or more explanatory
variables. Multiple regressions used in statistics to find the best fitting and most parsimonious,
yet reasonable model to describe the relationship between an outcome (dependent) variable and
a set of independent (explanatory) variables. To explain the relationship between dependent and
independent variable the following general multiple regression equations adopted.

Yit= βo + βXit + εit

Where: -

Yit :- is the dependent variable for firm ‘i’ in year‘t’

β0:- is the constant term

β :- is the coefficient of the independent variables of the study,

Xit :- is the independent variable for firm ‘i’ in year‘t’ and

Εit:-the normal error term.

The study uses the above general model to examine the determinant of NPL based on selected
variables as follows:

NPLit= β0+β1 (CAR)it+β2(LTD)it+β3(INF)+β4(EXR)it +β5(UNE)it+εit

Where;

24 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
 β0 is an intercept
 β1, β2, β3, β4, and β5 represent estimated coefficient for specific bank i at time t,
 CAR, LTD, INF, EXR and UNE represent capital adequacy ratio, Loan to deposit ratio,
Inflation rate, Exchange rate and unemployment respectively.
 εit represents error terms for intentionally/unintentionally omitted or added variables.
The multiple linear regressions model was conducted by the OLS method using stata14
econometric software package. According to Petra (2007) OLS outperforms the other
estimators when the cross section is small and the time dimension is short. According to
Brooks (2008) OLS or linear least squares is a method to estimate the slope and
intercept in a linear regression model. Therefore, as far as the above facts true in the
study used OLS method. The rational for choosing OLS is that, if the Classical Linear
Regression Model (CLRM) assumptions hold true, then the estimation determined by
OLS have a number of desirable properties, and are known as Best Linear Unbiased
Estimators (Brooks 2008). The following section discussed CLRM assumptions and
their diagnostics test result.

3.5. Varianble of Study

For the analyses purpose, nonperforming loan treated as the dependant variable; NPL rate
calculated by using the ratio of impaired loan to total loan, whereas, return on asset, return on
equity, capital adequacy and loan to deposit treated as independent variable from bank specific
factors and inflation rate, exchange rate and unemployment rate independent variables of
determinants of NPL from macroeconomic factors.

25 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
Table: 2.1 Summary of variables, their definition and expected sign.

Name of Type of Measurement Expected Sign

variable variables

NPL Dependent the ratio of non-performing loans to total loans

CAR Independent Total equity/Total asset -

LTD Independent Loan to deposit ratio Loan/Deposit +

INF Independent Customer price index -

EXR Independent the annual exchange rate b/n birr and USD -/+

UNE Independent The annual unemployment rate +

26 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

CHAPTER FOUR
DATA ANALYSIS AND INTERPRETATION

4. Results and Discussion

The purpose of this chapter is to present results and analysis of data involved in this study.
Accordingly, the descriptive statistics of all the variables used in this study and the results of
hypothesis testing i.e. the estimated parameters of the regression equation, their significance, the
connection between the independent variables and dependent variable according to the sign and
the value of the parameters for the regression model are presented and discussed in detail.

4.1. Descriptive statistics

The summary of descriptive statistics that was intended to give general descriptions about the
data (both dependent and independent variables) is presented in Table 2. The total number of
observation for each variable was 77 (i.e., 11 years data for 7 banks adjusted for some outlier
values). Accordingly, mean, standard deviation, minimum and maximum values of each
variable were used so as to show the overall trend of the data over the period under
consideration.

Table: 3.1. Summary of descriptive statistics for dependent and independent variables

NPL CAR LTD INFR EXR UM

Mean 3.519058 11.75297 .5453718 .4401299 15.18736 1.953709

Std.Dev 2.200386 5.456087 .1858303 1.690874 6.269562 .6381607

Min. .013 .11 .1146 .028 .261 .0208

Max 9.01 21 1.033 10.7 22.41 2.39

Obs 77 77 77 77 77 77

Source: Stata 14

27 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
The descriptive statistics for dependent and independent variables are presented and their results
are discussed in the table 3.1 above. For both dependent and independent variables value of
minimum, maximum, mean and standard deviation are presented. Nonperforming loan is treated
as the dependant variable which is calculated by using the ratio of impaired loan to total loan,
whereas return on asset, return on equity capital adequacy and loan to deposit treated as
independent variable from bank specific factors and inflation rate, exchange rate& un
employment rate as independent variables of determinants of NPL from macroeconomic factors
The mean of NPL of the seven banks under the study was 3.51 percent and the standard
deviation was 2.2 %. The mean of the NPL was less than the minimum value of the NPL ratio
requirement of national Bank of Ethiopia (5%). This indicate that there is a good performance
towards minimizing NPL in ECBs on the study period. The standard deviation of 2.2 %
indicated that NPLs of commercial banks are far from the mean value of 3.51 percent on
average by 2.2%. But when the NPL ratio of each banks are compared, there is great variation
between them. i.e., it is ranging from 0.013 percent to 9.01 percent indicated the margin that
NPLs ratio of Ethiopian commercial banks ranged over the sample period a low standard
deviation indicates that the data point tend to be very close to the mean, whereas high standard
deviation indicates that the data point are spread out over a large range of values. Brooks,(2008)

Capital adequacy ratio shows the proportion of owner’s equity to total asset. The mean value for
capital adequacy ratio was 11.75% whereas the maximum level was 21% and minimum value
was 0.11% with a standard deviation of 5.46% .The average amount of capital adequacy is
greater than the minimum capital requirement 8% of the NBE showing that ECBs has ability to
bear loss results from loan default.

The mean value of loan to deposit ratio was 0.55 percent which shows that the average value of
banks loan to deposit indicate Banks used the depositors fund for credit which exposed to
default risk, again it tells us on average loans are the most important asset for commercial banks
in Ethiopia. The standard deviation 0.18 percent reveals that there was high variation towards
the mean among banks in Ethiopia. The minimum and maximum was 0.11 percent and 1.03
percent respectively, suggesting that the ECBs concentrate on lending business which is low

28 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
exposure to risk by using depositors’ money. The maximum value also raises a surprise on how
banks lend excess of their total loan and engaged in high risk taking activity.

Inflation rate has 44.01 percent mean value, minimum value of 2.8 percent and maximum value
of 10.7 percent. Standard deviation 1.69 percent this indicate This that inflation rate in Ethiopia
during the study period remains unstable. Exchange rate has a mean value of 15.19 percent,
minimum value 0.26 percent and maximum value of 22.41 percent. This result tells us that the
depreciation of Ethiopian birr against USD dollar was very high during the period between 2008
and 2018. I.e. the USD dollar against birr was appreciated from birr 15.19 in 2008 to birr 22.41
in2018 which is about 47.53% inflated.

Unemployment rate has a mean value 1.95 percent, minimum value 0.02 percent and maximum
value of 2.39 percent. Standard deviation 0.64 percent this implies that unemployment is stable
as compared to the others macro economic factors but UM increasing highly in Ethiopia during
the study period.

4.2. Test for the Classical Linear Regression Model (CLRM) Assumptions
Descriptive statistics part of the study shows the mean, standard deviation, minimum and
maximum values of the dependent and explanatory variables including the number of
observation for each variable during the period under consideration, that is from 2008-
2018.However, this section provide test for the classical linear regression model (CLRM)
assumptions such as normality, heteroscedasticity, autocorrelation and multi-co linearity tests.
The linearity of the parameter is assumed since the model applies linear ordinary least
square(OLS). The objective of the model is to predict the strength and direction of association
among the dependent and independent variables. Thus, in order to maintain the validity and
robustness of the regression result of the research in CLRM, it is better to satisfy basic
assumption CLRM.As noted by Brooks (2008), when these assumptions are satisfied, it is
considered as all available information is used in the model. However, if these assumptions are
violated, there will be data that left out of the model. Accordingly, before applying the model
for testing the significance of the slopes and analyzing the regressed result, normality, multi-co

29 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
linearity, autocorrelation and heteroscedasticity tests are made for identifying misspecification
of data if any so as to fulfill research quality.

4.2.1. Autocorrelation Test

The researcher tested the autocorrelation assumptions that imply zero covariance of error terms
over time. That means errors associated with one observation are uncorrelated with the errors of
any other observation. As noted by Gujarati (2004), the best renowned test for detecting serial
correlation is Durbin Watson test. Accordingly, if the “d” computed nearest to 2 in application,
it is assumed that there is no autocorrelation problem. Thus, shown as follows

Heteroscedasticity and autocorrelations test

Breusch-Pagan / Cook-Weisberg test 0.0507

Durbin-Watson stat 1.15323

Note: Breusch-Pagan / Cook-Weisberg test for heteroscedasticity whereas Durbin-Watson test

for autocorrelation

Breusch-Pagan / Cook-Weisberg test for heteroskedasticity

Ho: Constant variance

Variables: fitted values of npl

chi2 (1) = 3.82

Prob>chi2 = 0.0507

Computed “d” in this study was 1.15 which is implying the absence of autocorrelation problem.
Thus, this implies that error terms are not correlated with one another for different observation
in this study.

30 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

4.2.2. Heteroscedasticity Test

In the classical linear regression model, one of the basic assumptions is Homoskedasticity
assumption that states as the probability distribution of the disturbance term remains same for
all observations. That is the variance of each uiis the same for all values of the explanatory
variable. However, if the disturbance terms do not have the same variance, this condition of non
constant variance or non-homogeneity of variance is known as heteroscedasticity (Bedru and
Seid, 2005).Accordingly, in order to detect the heteroscedasticity problems, Breusch-Pagan or
Cook-Weisberg test was utilized in this study. This test states that if the p-value is significant at
95confidence interval, the data has heteroscedasticity problem, whereas if the value is
insignificant (greater than 0.05), the data has no heteroscedasticity problem. Thus, as shown in
appendix 1A, there is no heteroscedasticity problem for this study hence the p value is 5%
showing insignificant value.

4.2.3. Normality Test

One assumption of classical linear regression model (CLRM) is the normal distribution of the
residual part of the model. As noted by Gujarati (2004), OLS estimators are BLUE regardless of
whether the ui are normally distributed or not. If the disturbances (ui) are independently and
identically distributed with zero mean and constant variance and if the explanatory variables are
constant in repeated samples, the OLS coefficient estimators are asymptotically normally
distributed with means equal to the corresponding β’s. However, as per the central limit
theorem, if the disturbances are not normally distributed, the OLS estimators are still normally
distributed approximately if there are large-sample data. Thus, since the sample size for this
study is large enough, it is approximately considered as normally distributed. This implies that
residuals are asymptotically normal in this study.

31 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
Skewness/Kurtosis tests for Normality

Probability(P-value) Decision rule p<0.05 Prob>chi2

Kurtosis 0.5928 Do not reject 0.7035

Skewness 0.5260

Sktest’ shows the number of observations (which is 77 here) and the probability of skewness
which is 0.5260 implying that skewness is asymptotically normally distributed (p-value of
skewness > 0.05). Similarly, Pr(Kurtosis) which is 0.5928 indicates that kurtosis is also
asymptotically distributed (p-value of kurtosis > 0.05). Finally, chi (2) is 0.7035 which is
greater than 0.05 implying its significance at a 5% level. Consequently, the null hypothesis
cannot be rejected. Therefore, according to the Skewness test for normality, residuals show
normal distribution

4.2.4. Multicolinearity Test

The term Multi- co linearity indicates the existence of exact linear association among some or
all explanatory variables in the regression model. When independent variables are multi
collinear, there is overlapping or sharing of predictive power. Thus, if multi- co linearity is
perfect, the regression coefficients of the independent variables are undetermined and their
standard errors are immeasurable (Gujarati, 2004). The multi-colinearity makes significant
variables insignificant by increasing p-value since increased p-value lowers the t-statistics
value. Thus, the panel regression results with multi-co linearity will shows significant variables
as insignificant variables.

The multi-co linearity problem is solved by dropping highly correlated variables (Ahmad and
Bashir, 2013).Then, the result provide more significant variables than before.

This is due to the fact that when explanatory variables are highly correlated with one another
,they share the same information. Thus, the multi-co linearity problem reduces the individual
explanatory variables’ predictive power. That is none of the predictor variables may contribute

32 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
uniquely and significantly to the prediction model after the other independent variables is
included (Theodros, 2011).

In case, Pearson correlation matrix and Variance inflation factor (VIF) are used for testing
multi-co linearity in this study. Pearson correlation matrix is a technique used for testing multi-
colinearity of explanatory variables by investigating their relationship and also useful to
measure the propensity of how much the independent variables influence the dependent variable
Wooldridge (2005) cited in Theodros (2005).

As noted by Gujarati (2004), the correlation analysis is made to describe the strength of
relationship or degree of linear association between two or more variables. In Pearson
correlation matrix, the values of the correlation coefficient range between -1 and +1. A
correlation coefficient of +1 indicates that the two variables have perfect positive relation; while
a correlation coefficient of -1 indicates as two or more variables have perfect negative relation.
A correlation coefficient of 0, on the other hand indicates that there is no linear relationship
between two variables (Bedru and Seid, 2005).Besides, as noted by Brooks (2008), zero
correlation among explanatory variables is not occurring in any practical work. Thus, even if
there is some indication for the existence of zero correlation among the explanatory variables,

It does not have a great effect on the accuracy. Accordingly, Pearson correlation matrix is
applied to examine the association between NPLs ratio, loan to deposit, capital adequacy ratio,
return on equity, and return on asset, exchange rate and unemployment rate where
nonperforming loans are considered as dependent variable whereas loan to deposit ratio, return
on equity, return on asset, capital adequacy ratio, exchange rate and unemployment rate are
explanatory variables used in this study.

Thus, as it can be seen from appendix 1B, the result of Pearson correlation matrix indicates that
NPLs has positive correlation with Exchange rate, while the correlation between NPLs ratio
with capital adequacy ratio, loan to deposit ratio, return on equity, and exchange rate is
negative. Besides, the result of correlation analysis made in the above table clearly indicates that
there is no significant multi-co linearity problems among explanatory variables since each of
them are not above 0.8 thresholds. As noted by in Gujarati (2004), a serious problem for Multi-
co linearity is occurred if the correlation is about 0.8 or larger. However, multi-co linearity
33 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
between explanatory variables may result wrong sign in the estimated coefficients and bias the
standard errors of coefficients (Theodros, 2011). To overcome this problem, VIF test was
conducted. That means, the larger the value of VIF indicates the more co linearity of the
variables with each other. According to the rule of thumb, if VIF of a variable exceeds 10, the
variable is said to be highly collinear (BedruandSeid, 2005). Based on the result indicated in
appendix C, there is no multi-co linearity problem in this study. This is due to the fact that the
mean of VIF of variables is 1.89 which is much lower than the threshold of 10.The VIF for each
variable also very low. This indicates that the explanatory variables included in the model were
not correlated with each other. To sum up, beside the descriptive statistics, correlation analysis
is made for explanatory variable to detect the multi-co linearity problem in the regression
model. In case, there is no multi- co linearity problem between variables. Thus, the explanatory
variables are the basic determinants of NPLs of commercial banks in Ethiopia.

4.3. Model Selection

4.3.1 Random Effect versus Fixed Effect Models

Econometrics model used to examine the impact of, capital adequacy ratio, loan to deposit ratio,
exchange rate, inflation rate and unemployment rate on nonperforming loans of commercial
banks in Ethiopia was panel data regression model which is either fixed-effects or random-
effect model.

For Hausman test, the null and alternative hypotheses are as follows:

Ho: uiis not correlated with Xi (random- effects model appropriate)

H1: uiis correlated with Xi (fixed-effects model appropriate)

Thus, to test the null hypothesis, it requires comparing the estimates from the random-effects
and the fixed-effects estimator. Random-effect estimator is consistent under the null hypothesis,
but inconsistent under the alternative hypothesis whereas fixed-effect estimator is consistent
under both the null and alternative hypothesis. If the estimates for the random-effects estimators
are not significantly different from the estimates for the fixed-effects estimator, then the null
hypotheses is accepted and conclude that uiis not correlated with Xi, and therefore the random-

34 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
effect model is the appropriate model. If the estimates for the random effect estimator are
significantly differ from the estimates for the fixed-effect estimator, the null is rejected and
conclude that uiis correlated with Xi, and therefore the Random-effect model is the appropriate
model for the study.

Accordingly, appendix1D demonstrates the Hausman Specification Test that used to decide the
best model for this study. The decision rule, for Hausman Specification test is rejecting the null
hypothesis when the p-value is significant. Thus, shown as follows

Hausman specification Test

Coefficients

| (b) (B) (b-B) sqrt (diag (V_b-V_B))

| fixed random Difference

car -.0773113 -.0390364 -.038275 .0326953

ltd .5548807 .9368932 -.3820125 .2180878

infr -.0267443 -.0196363 -.0073079 .0182502

exr -.1536101 -.1646722 .0110621 .0137784

um 3.283308 3.193562 .0897455 .0848435

b = consistent under Ho and Ha; obtained from xtreg

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

Test: Ho: difference in coefficients not systematic

chi2(7) = (b-B)'[(V_b-V_B)^(-1)](b-B)

= 3.48

Prob>chi2 = 0.6266

(V_b-V_B is not positive definite)

35 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
Hausman specification test for this study has a p-value of 0.6266 for the regression models. This
indicates that p-value is not significant and then the null hypothesis is accepted justifying as
Random effect models appropriate for the given data set in this study.

4.4. Result of Regression Analysis

The empirical evidence on the determinants of Ethiopian commercial banks’ non-performing

loan is studied based on strongly balanced panel data, where all the variables are observed for
each cross-section and each time period. The study has a time series segment spanning from the
period 2008 up to 2018 and a cross section segment which considered Seven Ethiopian
commercial banks, namely Commercial Bank of Ethiopia, Dashen Bank, Wegagen Bank,
Awash Bank, Bank of Abyssinia, Hibret Bank and Nib International Bank. The study used
multiple liner regression equation to analyze the relationship between Ethiopian commercial
banks non-performing loan and determinant variables. All the proposed independent variables
(i.e. CAR, LTD, ROA, INFR, EXR and UN) were regressed with respect to the dependent
variable (NPLs). The following linear regression model is developed.

Table: 4.2. Results of Random effect regression model

Explanatory variables Coefficient Std.err z p>/t/

Capital Adequacy ratio -.0390364 .0562641 -0.69 0.488***

Loan to Deposit .9368932 1.077799 0.87 0.385***

Inflation Rate -.0194363 .1182269 -0.16 0.869***

Exchange Rate -.1646722 .0.446796 -3.69 0.000*

Un employment 3.193562 .366659 8.71 0.000*

Constant -.2629038 1.136211 -0.23 0.817

36 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

R2 = 0.53

Rho=0.211

Prob>F=0.0000

Source: Developed for the research through Stata 14

Notes:*significant at 1%,and ***insignificant

Thus, based on the above table 4.2, the following model was developed to examine the
determinants of NPLs in this study.

NPL=-.2629038-.0390364CAR+.9368932LTD-.0194363NFR-.1646722EXR+3.193562+ε

As shown in the above table 4.2.coefficient of determination was 53% revealing that 53% of
variation in NPLs ratio is explained by the selected explanatory variables (capital adequacy
ratio, loan to deposit ratio, ratio, Inflation rate, exchange rate and unemployment rate. Besides,
Rho displays that 21.1% variation in NPLs is due to entity specific characteristics of the
selected cross sectional entities i.e. commercial banks in Ethiopia. Furthermore, Since F-
statistics is designed to jointly test the impact of explanatory variables on dependent variables;
F-statistics of this model has a p-value of 0.0000 indicating rejecting of the null hypothesis.
This implies that all selected explanatory variables can affect the level of NPLs in common.

 Capital adequacy ratio

Regarding capital adequacy ratio that determines the risk taking behavior of banks, this study
identifies statistically at 5% insignificant level and Negative impact of capital adequacy ratio on
NPLs. Thus, regression result of random effect model in the above table 4.2 is inconsistent with
the hypothesis developed in this study. The study hypothesized that there is a Negative
association between CAR and NPLs of banks. This Negative sign indicates indirect relationship
between capital adequacy ratio and NPLs. Thus, it implies that for one unit change in the banks’
capital adequacy ratio, keeping other thing constant had resulted 0.488 unit changes on the
levels of NPLs in opposite direction. The coefficient of CAR on the regression is -.0390 and p-

37 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
value is 0.488. Holding other variables constant, an increase in the ratio of total capital equity to
total assets results a -0.0390% decrease in NPLs of ECBs. But since the P-value is greater than
0.05 it is statistically insignificant. Which means in the study period, unlike the result of most
prior studies CAR does not have a significant impact on explaining the variability of NPLs in
ECBs. This result is consistent with the finding of Louzis et al. (2012). The justification behind
this result is high capital adequacy ratio leads banks to involve in high risky activities, create
risky loan portfolios but since the loan collection techniques are developed the risky loan
portfolios of ECBs are controlled and as a result NPLs does not increase significantly

 Loan to deposit (LTD)

Table 4.2 showed that the coefficient of loan to deposit is 0.9368 and positively insignificant at
5% (p-value=0.385) statistically insignificant level. This means, holding other factors constant,
an increase/decrease in Loan to deposit by one unit will result a 0.9368 increase/decrease of
Ethiopian commercial banks NPL. Generally, reject the null hypothesis since there is a positive
insignificant relationship between Loan to deposit and non-performing loans. Thus, regression
result of random effect model in the above table 4.2 is consistent with the hypothesis developed
in this study. The loan to deposit ratio measures the liquidity and reflects the risk attitude of
banks.

A positive in significant effect of loan to deposit on Ethiopian commercial banks of NPL

suggest that borrower wants loan and gives priority to banks provide loan with low interest rate
rather than not provide loan. Ethiopian commercial banks to spread their base and to increase
their deposit use loan as an enticement to attract a new customer. The above mutual interest of
lender and borrower contribute for the growth of lone to deposit ratio.

 Inflation rate

In this study the coefficient estimate of inflation was negative but statistically insignificant at
(P- value of 0.869). The negative coefficient estimate of inflation (-.01943) indicates a negative
association with NPLs. This implies that when inflation increases, the level of NPLs decreases
and vice versa. Theoretically, high inflation should reduce the real value of debt and hence
make debt servicing easier. However, high inflation may pass through to nominal interest rates

38 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
and weaken some borrowers’ ability to service debt by reducing real income when wages are
sticky (Skarica, 2014)

 Exchange rate

Exchange rate had negative impact on the level of NPLs having a coefficient -0.1646 and it is
statistically significant at 1% (p=0.0000). Thus, regression result of random effect model in the
above table 4.2 is consistent with the hypothesis developed in this study. Therefore, the null
hypothesis is not rejected. Holding other variables constant, a 1% increase on exchange rate
leads 16 % decrease in NPLs of ECBs. This result indicates that exchange rate is the most
important determining factor of NPLs in ECBs within the study period. A depreciation of the
exchange rate can have mixed implications on NPLs of banks. On the other hand, it can
improve the competitiveness of export-oriented firms and ultimately increase their ability to
service debt (Fofack, 2005). On the other, it can negatively affect the debt-servicing capacity of
borrowers who borrow in foreign currency (import-oriented firms). Since the number of import
oriented firms is higher in Ethiopia, the coefficient estimate of exchange rate (EXR) in this
particular study is negative association with NPLs of ECBs. Thus, the null hypothesis that there
is negative relationship between exchange rate and non performing loan in commercial banks in
Ethiopia should be accepted. This result consistent with the findings of sitina (2018).

 Un employment

The coefficient estimate of real unemployment rate is 3.1935 and it is significant at 1% (p-
value= 0.000) so the null hypothesis is accepted. Thus, regression result of random effect model
in the above table 4.2 is consistent with the hypothesis developed in this study. An increase in
unemployment results a decrease in NPLs of ECBs. This result is attributable to the chronic
shortage of hard currency that creates unutilized labor due to the dalliance of row materials and
products. The longer the queue of accessing hard currency the unproduced output that should
have been rendered by labor force of the companies is exacerbated and the organization baring
unnecessary wage expenses preference of the organization is to pay the Bank than retaining it is
employees in fear of future business. This result consistent with the studies of Valahzaghard et
al (2012) and Louzis,Vouldis& Metaxas (2012) , Farhan et al.,(2012), Akinlo and Emmanuel
(2014), Abid, Ouertani, Zouari-Ghorbel (2014) .
39 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
Table: 5.1Summary of actual and expected sign’s of explanatory variable

Name of Type the Measurement Expeced Actual Decision

variable variables Sign Sign

NPL Dependent the ratio of non-performing

loans to total loans

CAR Independent Total equity/Total asset - - -ve& in significant

Not accepted

LTD Independent Loan to deposit ratio + + +ve &

Loan/Deposit insignificant

Not accepted

INFR Independent Customer Price Index - - -ve& insignificant

Not accepted

EXR Independent the annual exchange rate b/n -/+ - -ve& significant
birr and USD accepted

UNE Independent The annual unemployment + + +ve & significant

rate accepted

Source: Developed by the researcher

40 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

CHAPTER FIVE
SUMMARY, CONCLUSION AND RECOMMENDATION

This chapter discussed the conclusions and recommendations of the study. The chapter
organized in to three sections, the first section 5.1 presents Summary 5.2. Presents the
conclusions of the study and section 5.3 presents the recommendations provide depend on the
findings of the study.

5.1. Summary
The broad objective of the study was to investigate bank specific and macroeconomic
determinants of NPLs in Ethiopian commercial banks. The study used quantitative research
approach. The study applied descriptive statistics and multiple linear regression analysis to
analyzed and identify the influences of bank specific and macroeconomic factor on non-
performing loans of seven sampled Ethiopia commercial banks. A sample of 77observations has
been analyzed over the period from 2008-2018 and used strongly balance panel data.
Regression analysis and descriptive statistics were employed on secondary data collected from
the sample financial statement of banks, NBE & CSA. The multiple linear regressions model
was conducted by the ordinary listing square and CLRM assumptions test of the models no
evidence for the presence of normality, heteroscedasticity, multi-colinearity and autocorrelation
problem. The study shows the cause-effect relationship between the bank specific,
macroeconomic factor and non-performing loans of Ethiopian commercial banks. The study
uses two bank specific variables and three macroeconomic factors such as Capital adequacy
ratio, Loan to deposit ratio, Inflation rate, Exchange rate and Unemployment. The findings of
the study suggested the following conclusions.

The study suggested that EXR& UM statistically significant effect on the level of NPLs.
However, the regression model revealed the insignificant effect of CAR, LTD& INFR on the
level of NPLs of commercial banks in Ethiopia for the period under consideration.

41 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
The regression analysis of the empirical determinants of NPLs presented in this study suggested
the relationship of LTD with the volume of Ethiopian commercial banks nonperforming loan is
positive. In this general setting, Therefore Ethiopian commercial Banks that anticipate un
collective loan may create provisions to reduce or absorb their risk this motivate Ethiopian
commercial banks to provide more loan and engaged in risky activities and ultimately the
probability that loans became NPLs will increased with the same manner. At the same time,
NPL of Ethiopian commercial banks negatively affected by CAR with NPLs of Ethiopian
commercial bank. The result indicates that higher capital requirement protects Ethiopian
commercial banks from providing loans to riskier projects. High ratio of non-performing loans
in banking system or rising tendency leads to increase in allowance to be allocated for
aforementioned loans and thus, to a decrease in the profitability and capital adequacy ratio of
the banks. Considered from the point of economics, increase in non-performing loans,
negatively effects economic growth by causing to a decrease in loan-able funds (MEHMET
İSLAMOĞLU, 2015) Non-performing loan can enhance the insolvency of banks leading to
bank failure.

5.2. Conclusion
Ethiopian Commercial Banks are showing a good performance on avoiding NPLs. NPLs of
Ethiopian commercial Banks are highly affected by macroeconomic factors unemployment, and
Exchange rates are found to be the most significant variable which explains the variation of
NPLs in ECBs. This is because the country’s economy is import oriented. From the bank
specific variables CAR and LTD are founded to be insignificant in explaining the variation of
NPLs of ECBs.

5.3. Recommendations

Based on the finding of regression analysis and conclusion the following recommendation is
forward.

42 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
 Bank management give attention for bank specific factor such as deposit, loan,
capital and their return, also give more emphasis on the asset management decision
and give priority for current asset specially loan in order to reduce the level of
nonperforming loans. Frequently the study recommends that Ethiopian Commercial
Banks to balance their loan in proportion with customers’ deposit.
 Bank managers thoroughly inspect client data and information during the credit
analysis stage to reduce information asymmetry also Bank management know their
customers (KYC) before granting loans to them, after granting focus on timely
collection of loan by strictly follow up
 Credit analysts and loan officers provide financial counseling to borrowers.
Furthermore, banks are encouraged to diversify more by adopting more non-interest
income activities.
 Within the bank, stress tests of banks’ loan quality and liquidity risk must be
conducted on a regular basis. Regulators should also focus on a continual managerial
performance evaluation in order to improve the stability of the financial system.
Lastly, Government can also assist in reducing the level of NPLs by stimulating the
economy through the improvement of infrastructure, foreign investments, and
economic growth so as to control unemployment.

Finally, the study sought to investigate the determinants of NPLs in Ethiopian commercial
banks. However, the variables used in the econometrics model did not include all factors that
can affect NPLs of Ethiopian commercial banks. Thus, future research could incorporate
external factors such as GDP, growth rate, size, ownership, money supply and corporate
governance. It is also recommended for future researchers to study the determinants of NPLs in
different economic sectors and year of study.

43 | P a g e
 Determinants of Non-performing loan of commercial
Banks in Ethiopia
References
Abid, L., M.N. Ouertani and S. Zouari-Ghorbel, 2014, Macroeconomic and Bank- Specific
Determinants of Household‘s Non-Performing Loans in Tunisia: A Dynamic Panel
Data, Procedia Economics and Finance, 13(2014): 58-68.

Ahmad, F., & Bashir, T. (2013). Explanatory Power of Bank Specific variables as Determinants of
Non-Performing Loans: Evidence from Pakistan Banking Sector, World Applied
Sciences Journal, 9, 22.

Anisa U. (2015).Determinants of nonperforming loan: an empirical study on commercial banks of

Ethiopia.

Armstrong, J. (2008). Perdue signs film tax-credit boost. Retrieved July 31, 2014,
www.jacksonville.com

Arpa, M., Giulini, I., Ittner, A., & Pauer, F. (2001). The Influence of Macroeconomic
developments on Austrian banks: implications for banking supervision. BIS Paper No.1,
pp/ 91-116.

Ayalew, T 2009, ‘Legal Problems in Realizing Non-Performing Loans of Banks in Ethiopia’, A

thesis, Addis Ababa University.

Basel Committee.2001. Sound Practices for the Management and Supervision of Operational Risk.
Basel Committee on Banking Supervision.

Bedru&Seid, (2005) Econometrics: A Teaching Material For Distance Students Majoring In

Economics (Mekelle University).

Bercoff, J, Julian, G & Franque, G 2002, ‘Argentinean Banks, Credit Growth and the Tequila

Berger, A & De Young,R 1997. ‘Problem loans and cost efficiency in commercial banks‘Journal

Bester, H 1999‘Collateral Default Risk and Relationship Lending’an empirical study on

Bloem,M.A.& Gorter N.C 2001, ‘Treatment of Non-Performing Loans in Macroeconomic

Bofondi, M. and Ropele, T 2011 ‘Macroeconomic Determinants of Bad Loans: Evidence

fromItalianBanks’, Occasional Papers, 89

44
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
Boudriga, A. Boulila, T., Neila, & Jellouli, S. (2009). Bank Specific, Business and Institutional,
Environment Determinants of Nonperforming Loans: Evidence from MENA Countries,
Journal of Financial Economic Policy, Vol. 1, pp. 286-318

Boudriga, A., Taktak, N. B., & Jellouli, S. (2009). Banking supervision and nonperforming loans:
a cross-country analysis. Journal of Financial Economic Policy , 286-318.

Breuer, JB2006 ‘Problem bank loans, conflicts of interest and institutions’, Journal of Financial

Brooks, C 2008,’Introductory Econometrics for Finance’, 2nd edn, Cambridge University

Press,New York.

Bucur, I. A. (2013), On the Economic Development of Romania as a Member of E.U. in the Frame
of the Global Economic Crisis, Ovidius University Annals – Economic Sciences Series,
13 (1), 152-157.

Caprio, G Jr and Klinggebiel D(1996): Bank Insolvency-Bad Luck, Bad Policy or Bad Banking,
Annual World Bank Conference on Development Economics.

Carbo V.S & Rodriguez, F 2007, ‘The determinants of bank margins in European banking’Journal
of Bankingand Finance, 31(7), 2043-2063.

Creswell, J W (2003). Research design: qualitative, quantitative and mixed methods approaches,
Sage Publications, California, 2nd ed.

Crisis: A Duration Analysis’(Unpublished).

Daniel T. (2010). Issues of non-performing loan: Privately owned commercial banks in Ethiopia.
Addis Ababa University.

Demirguc-Kunt, Asli, "Deposit-Institution Failures: A Review of the Empirical Literature,"

Espinoza, R. and A. Prasad, 2010, Nonperforming Loans in the GCC Banking Systems and their
Macroeconomic Effects, IMF Working Paper No. 10/224.

Fama, E.F., 1980, Agency Problems and the Theory of the Firm, The Journal of Political
Economy, 88(2): 288-307.

45 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
Farhan Muhammad, SattarAmmara, C. HussainAbrar, Khalil Fareeha. (2012). Economic
Determinants of Non-Performing Loans: Perception of Pakistani Bankers European;.
Journal of Business and Management, Vol4, No.19, .

Federal Reserve Bank of Cleveland, Economic Review, Quarter 4, 1989.

Financial Contracting.

Fofack, H2005, „Nonperforming loans in Sub-Saharan Africa: Causal Analysis and

Macroeconomic implication‟ World Bank Policy Research Working Paper 3769,
//econ.worldbank.org.

fromTurkey:Journalof Empirical Research, 3(8):933-943.

GadiseGezu (2014),determinants of nonperforming loans: empirical study in case of commercial

banks in Ethiopia

Godlewski, C., 2005, Capital Regulation and Credit Risk Taking: Empirical Evidence from Banks
in Emerging Market Economies, Journal of Banking Regulation, 6(2): 128-14

Habtamu G (2015). Assessment of factors affecting nonperforming loans. The case of Ethiopian
private banks. Addis Ababa University, Addis Ababa, Ethiopia.

Hoque, M.Z. and Hossain, M.Z., 2008. Flawed Interest rate policy and loan default: experience
from developing country. International review of business research papers, Col.5, No.5,
235-246.

International Monetary Fund. 2006. Spain: Financial Sector Assessment Program–Technical Note–
Stress Testing Methodology and Results, IMF Country Report No.06/216, (Washington
DC: International Monetary Fund).

Jensen, M.C. (1976). Ownership structure theory rather than capital structure in the crucial to be
determined. Norges Bank Economic Bulletin 78, 65–76.

Jimenez G, Saurina J., 2006. Credit cycles, credit risk, and prudential regulation, International
Journal of Central Banking, 2 (2), pp. 65-98.

46 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
Karim, M.Z., Chan, S.G., & Hassan, S. (2010). Bank Efficiency and Non-Performing Loans:
Evidence from Malaysia and Singapore.

Keeton, W. and Morris, C.S. (1987)“Why do Banks‟ Loan Loses Differ?”, Federal Reserve Bank
of Kynsas City, Economic Review, pp 3-21.

Khemraj Tarron, and Pasha S. (2009). The determinants of non-performing loans: an econometric
case study of Guyana.‖ Presented at the Caribbean Centre for Banking and Finance Bi-
annual Conference on Banking and Finance, St. Augustine, Trinidad. 61

Klein, N. (2013). Non-Performing Loans in CESEE: Determinants and Macro-economic

Performance, IMF working paper 13/72.

Kwambai, K. D., & Wandera, M. (2013), Effects Of Credit Information Sharing On

Nonperforming Loans: The Case of Kenya Commercial Bank Kenya, European
Scientific Journal, May 2013 Edition, (9)13.

Louzis, D. P., Vouldis, A. T., & Metaxas, V. L. (2011). Macroeconomic and bank-specific
determinants of non-performing loans in Greece: A comparative study of mortgage,
business and consumer loan portfolios. Journal of Banking & Finance

Mesele Beju 2016 Determinants of nonperforming loans of ethopian commercial banks.

ST.MARY’S UNIVERSITY

Messai A.S, and Jouini, F., 2013.Micro and macro determinants of non-performing Loans,
International Journal of Economics and Financial Issues. 3 (4), pp. 852-860.

NBE (2018): Asset Classification and Provisioning Directive No. SBB/69/2018. National Bank of
Ethiopia, Addis Ababa.

NBE, (2012). National Bank of Ethiopia Directive No. SBB/52/2012, 2012

Nkusu, M. (2011). Nonperforming Loans and Macrofinancial Vulnerabilities in Advanced

Economies. IMF Working Paper 11/161.

of Banking and Finance, 21: 8498,70.

47 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
Pasha,S and Khemraj, T 2009,” The determinants of non-performing loans: an econometric case
study of Guyana”, The Guyana Bank(Unpublished).

Saba, I., Kouser ,R. and Azeem ‘M. 2012: Determinants of nonperforming Loans: Case of
USBanking Sector. International Journal of Banking and Finance; No. 44 Pp 479-88

Salas, V and Saurina, J2002, „Credit risk in two institutional regimes: Spanish Commercial and
Savings banks‟, Journal of Financial Services Research, Vol. 22, No 3, PP. 203-224.

Sapienza, P., 2004, The effects of government ownership on bank lending,Journal of Financial
Economics, 72(2): 357-384.

Selma M.A. and Jouini F., 2013. Macro and Micro determinants of Nonperforming loans
inTunisia, International Journal of Economics and Financial Issues 13 (3), pp. 852-860.

Sitina Amino (2018)Determinants of Non-performing Loans in Ethiopian Commercial Banks

Skarica, B 2013, Determinants of Non-Performing Loans in Central And Eastern

EuropeanCountries: Financial Theory and Practice. 38 (1) 37-59

Stability, Vol.2, No.3, pp. 226–285.

Statistics’, IMF Working Paper, WP/01/209

Swamy, V 2012 “Impact of Macroeconomic and Endogenous Factors on Nonperforming

BankAssets” International Journal of Banking and Finance, Vol. 9.

Theodros K.(2011). Determinants of dividend payout: an Empirical study on bank industry in

Ethiopia (Master Thesis, Addis Ababa University, Addis Ababa, Ethiopia).

Tomak, S.2013, Determinants of Commercial Banks ‘lending Behavior: Evidence

TseganeshTesyaye (2012) Determinants of Banks Liquidity and Their Impact on Financial

Performance: Published thesis (MSc), University Addis Ababa, Ethiopia

Vogiazas, D &Nikolaidou, E 2011, “Credit risk determinants in the Bulgarian‟, Vogiazas-

Nikolaidou, PP. 177-189.

Wanjirumuriithi 2013the causes of non-performing loans in commercial banks in Kenya by

monicah.

48 | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia
Warue, B.N. (2010). Credit Growth, Problem Loans and Credit Risk Provisioning in Spain.
Financial Crisis. IMF working paper 00/33, March: 2-5. 63

Wondimagegnehu N (2012). Determinants of Non Performing Loans: The case of Ethiopian

Banks, University of South Africa.

YaredA(2018) determinants of nonperforming loans: empirical study of commercial banks in

Ethiopia.

YonasA(2017) Assessment of the determinants of non-performing loans the case of commercial

banks in Ethiopia.

Yosef Fekadu(2018),Determinants of non- Performing Loans: Evidence from Commercial Banks

in Ethiopia

Zelalem T. (2013)Determinants of Non-performing Loans: Empirical Study on Ethiopian

Commercial Banks.

49 | P a g e
 Determinants of Non-performing loan of commercial
Banks in Ethiopia

APPENDICES

Appendix1: Model Selection and Basic Tests for CLRM Assumptions

Appendix.1A:- Heteroscedasticity and autocorrelations test

Breusch-Pagan / Cook-Weisberg test 0.0507

Durbin-Watson stat 1.15323

Note: Breusch-Pagan / Cook-Weisberg test for heteroscedasticity whereas Durbin-Watson

statfor autocorrelation

Breusch-Pagan / Cook-Weisberg test for heteroskedasticity

Ho: Constant variance

Variables: fitted values of npl

chi2(1) = 3.82

Prob>chi2 = 0.0507

i
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

Appendix 1B

Skewness/Kurtosis tests for Normality

------ joint ------

Variable | Obs Pr(Skewness) Pr(Kurtosis) adj chi2(2) Prob>chi2

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

res | 77 0.5260 0.5928 0.70 0.7035

Appendix.1C:- Pearson Correlation Matrix

NPL CAR LTD INFR EXR UM

NPL 1.0000

CAR 0.1706 1.0000

LTD 0.1384 -0.2933 1.0000

INFR -0.2390 -0.3430 -0.0062 1.0000

EXR -0.0053 0.6842 -0.4761 -0.4035 1.0000

UM 0.5703 0.5945 -0.2651 -0.4846 0.6065 1.0000

Note: NPL stands for Nonperforming loan ratio, ROA for Return on asset,ROE for Return on
equity ,CAR for Capital Adequacy ratio ,LTD for Loan to deposit, , INFR for Inflation rate,
EXR for Exchange Rate, UM for Un employment Rate.

Source: stata version 14

ii | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

Appendix.1D:- summary of VIF

Variable VIF 1/VIF

EXR 2.62 0.381953

CAR 2.08 0.479993

UM 1.95 0.512547

INFR 1.43 0.699591

LTD 1.38 0.722948

Mean VIF 1.89

Source: stata version 14

iii | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

Appendix.1E Summary of descriptive statistics for dependent and independent variable

Variable Obs Mean Std. Dev. Min Max

NPL 77 3.519058 2.200386 0.013 9.01

CAR 77 11.75297 5.456087 0.11 21

LTD 77 0.5453718 0.1858303 0.1146 1.033

INFR 77 0.4401299 1.690874 0.28 10.7

EXR 77 15.18736 6.269562 0.261 22.41

UM 77 1.953709 0.6381607 0.0208 2.39

iv | P a g e
 Determinants of Non-performing loan of commercial 2022
Banks in Ethiopia

Appendix 1F: Hausman specification Test

--------------------------------------------------------------------------------------------------

Coefficients –

-----------------------------------------------------------------------------------------------

| (b) (B) (b-B) sqrt (diag (V_b-V_B))

| fixed random Difference sqrt(diag(V_b-V_B) S.E

car -.0773113 .0390364 -.038275 .0326953

ltd .5548807 .9368932 -.3820125 .2180878

infr -.0267443 -.0194363 .0073079 .0182502

exr -.1536101 -.1646722 .0110621 .0137748

um 3.283308 3.193562 .0897455 .0848435

b = consistent under Ho and Ha; obtained from xtreg

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

Test: Ho: difference in coefficients not systematic

chi2(7) = (b-B)'[(V_b-V_B)^(-1)](b-B)

= 3.48

Prob>chi2 = 0.6266

(V_b-V_B is not positive definite)

Source: stata version 14

v|Page