THE EFFECT OF CREDIT RISK ON FINANCIAL
PERFORMANCE OF COMMERCIAL BANKS IN
ETHIOPIA
BY
TESFAYE MULUGETA
ADDIS ABABA UNIVERSITY
COLLEGE OF BUSINESS AND ECONOMICS
DEPARTMENT OF ACCOUNTING AND FINANCE
FEBRUARY, 2018
ADDIS ABABA, ETHIOPIA
THE EFFECT OF CREDIT RISK ON FINANCIAL
PERFORMANCE OF COMMERCIAL BANKS IN ETHIOPIA
A THESIS SUBMITTED TO ADDIS ABABA UNIVERSITY COLLEGE OF
BUSINESS AND ECONOMICS DEPARTMENT OF ACCOUNTING AND
FINANCE IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR
DEGREE OF MASTERS OF SCIENCE IN ACCOUNTING AND FINANCE
BY
TESFAYE MULUGETA
ADVISOR
ALEM HAGOS (PHD)
FEBRUARY, 2018
ADDIS ABABA, ETHIOPIA
I
Declaration
I, Tesfaye Mulugeta, hereby declare that the thesis work entitled “The effect of credit risk on
financial performance of commercial banks in Ethiopia” submitted by me for the award of the
degree of Masters of Science in Accounting and Finance of Addis Ababa University at Addis
Ababa, Ethiopia, is my original work and it has not been presented in any university. All
sources and materials used for this thesis have been duly acknowledged.
Name Tesfaye Mulugeta
Signature: _______________
Advisor’s Name Alem Hagos (PhD)
Signature: _______________
ii
Certification
Addis Ababa University
School of graduate studies
This is to certify that the thesis prepared by Tesfaye Mulugeta entitles: The effect of credit
risk on financial performance of commercial banks in Ethiopia and submitted in partial
fulfillment of the requirements for the degree of masters of Science in accounting and finance
compiles with the regulations of the university and meets the accepted standards with respect
to originality and quality.
Approved by:
Internal Examiner Habtamu B (PHD)
Signature_______________ Date___________
External Examiner Lamessa (PHD)
Signature_______________ Date___________
Advisor AlemHagos (PHD)
Signature_______________ Date___________
iii
Acknowledgment
First of all, my deepest and heartfelt thank goes to the Almighty God and his mother St.
Marry, who follow me in all aspect of my life. Next, I would like express my deepest
gratitude to my advisor, Alem Hagos (PhD), for his support, encouragement, invaluable
comments, advice and guidance at various stages of my study.
I am grateful to the staff of NBE, CBE, AB, DB, UB, CBO and LIB because of giving me the
required data voluntarily to conduct my research. Without their kind cooperation, this study
would not have been completed and became a reality
In addition, I am grateful to my friend Melaku Habte for spending his valuable time during
the research discussing ideas, also for sharing the experiences, model specification of the
study as well as his assistance during the analysis stage of the study.
Finally I would like to thank my entire friends for their immeasurable assistance throughout
my study.
iv
Table of Contents
Page
Declaration …………………………………………………………………………...…………….II
Certification .....................................................................................................................................III
Acknowledgment ............................................................................................................................IV
Table of contents ...............................................................................................................................V
List of acronyms and abbreviations .................................................................................................VI
List of Tables ..................................................................................................................................VII
List of Figures ...............................................................................................................................VIII
Abstract.............................................................................................................................................IX
CHAPTER ONE: INTRODUCTION ................................................................................................1
1.1 Background of the study ..............................................................................................................1
1.2 Statement of the problem .............................................................................................................3
1.3 Research question ........................................................................................................................4
1.4 Objectives of the study ................................................................................................................4
1.4.1 General Objective .....................................................................................................................4
1.4.2 Specific Objectives ...................................................................................................................4
1.5 Hypothesis ...................................................................................................................................4
1.6 Scope of the study........................................................................................................................5
1.7 Limitation of the study………………………………………………………………………….5
1.8 Significance of the study .............................................................................................................5
1.9 Organization of the study….........................................................................................................6
CHAPTER TWO: LITERATURE REVIEW ...................................................................................7
2.2 Theoretical Review…..................................................................................................................7
2.2.1 Meaning of risk ........................................................................................................................7
2.2.2 Meaning of Credit risk .............................................................................................................7
2.2.3 Types of Credit risk……….......................................................................................................8
2.2.4 Credit risk management…........................................................................................................9
2.2.5 Sources of Credit risk………………………………………..................................................10
2.2.6 Credit risk exposures in bank………......................................................................................11
2.2.7 General principles of sound credit risk management in banking............................................11
2.2.8 Profitability and risk of banks................................,................................................................20
2.2.9 Classification of loan and advances…………………………………………………………20
IV
2.2.10 Determinants of credit risk....................................................................................................22
2.3 Empirical Review........................................................................................................................25
2.4 Literature Gap.............................................................................................................................28
2.5 Conceptual Framework ..............................................................................................................28
CHAPTER THREE: RESEARCH METHODOLOGY...................................................................30
3.1 Research approach......................................................................................................................30
3.2 Research design..........................................................................................................................30
3.3 Data Type and Source.................................................................................................................30
3.4 Target commercial banks selection ...........................................................................................30
3.5 Method of data analysis .............................................................................................................31
3.6 Model specification.....................................................................................................................33
3.6.1 Variables description………………………………................................................................34
CHAPTER FOUR: DATA ANALYSIS AND DISCUSSION OF RESULTS................................37
4.1 Descriptive statistics ..................................................................................................................37
4.2 Correlation analysis ...................................................................................................................39
4.3 Regression model tests ...............................................................................................................40
4.3.1 Model selection .......................................................................................................................40
4.3.2 Tests for CLRM assumptions .................................................................................................41
4.4 Analysis of regression................................................................................................................46
4.4.1 Discussion of results ………..………………….....................................................................48
CHAPTER FIVE: SUMMARY, CONCULSION AND RECOMMENDATION….......................51
5.1 Summary of findings...................................................................................................................51
5.2 Conclusions ................................................................................................................................52
5.3 Recommendation .......................................................................................................................53
5.4 Future research considerations…................................................................................................54
References.........................................................................................................................................55
Appendixes.......................................................................................................................................59
V
List of acronyms and abbreviations
AB
Awash Bank
CBE
Commercial Bank of Ethiopia
CBO
Cooperative Bank of Oromia
CLRM
Classical Linear Regression Model
DB
Dashen Bank
DW
Durbin-Watson
IMF
International Monetary Fund
LIB
Lion International Bank
LLTR
Loan Loss to Total Loan Ratio
LPTLA
Loan Provision to Total Loan and Advance
NBE
National Bank of Ethiopia
NPL
Non-Performing Loan
NPLLP
Non-Performing Loan to Loan Provision
NPLTLA
Non-Performing Loan to Total Loan and Advance
OLS
Ordinary Least Square
ROA
Return on Asset
ROE
Return on Equity
TL
Total Loan
TLATD
Total Loan and Advance to Total Deposit
UB
United Bank
VI
List of Tables
Page
Table 4.1 Descriptive Statistics..............................................................................................................37
Table 4.2 Correlation Analysis of Variables..........................................................................................39
Table 4.3 Hausman Test ........................................................................................................................40
Table 4.4 Heteroskedasticity Test..........................................................................................................42
Table 4.5Breusch-Godfrey Serial Correlation LM Test........................................................................42
Table 4.6 Correlation Matrix between independent variables...............................................................45
Table 4.7 Regression result....................................................................................................................46
VII
List of figures
Page
Figure 2.1The conceptual framework or model of the study...................................................29
Figure 4.1 Normality test results..............................................................................................44
VIII
ABSTRACT
This study attempts to reveal the relationship between credit risk and financial performance
of commercial banks in Ethiopia. In order to investigate these study quantitative research
approach is employed based on documentary analysis. A panel data from six selected
commercial banks covering the ten-year period (2007-2016) is analyzed within the fixed
effects model on regression analysis and using E-view8 software. The study used one
dependent variable return on asset (ROA), four independent variables that are: nonperforming loan to total loan and advance ratio (NPLTLA), loan provision to total loan and
advance ratio (LPTLA), total loan and advance to total deposit ratio (TLATD) and the ratio
of non-performing loan to loan provision (NPLLP) as measures of credit risk. Both
descriptive statistics and regression analysis specifically fixed effects model were used to
analyze the relationships of the depended variable with explanatory variables. The regression
result show that non-performing loan to total loan and advance ratio, loan provision to total
loan and advance ratio and the ratio of non-performing loan to loan provision show negative
and significant effect at 1% and 5% significance level on financial performance of
commercial banks in Ethiopia. However, total loan and advance to total deposit ratio show
positive and significant effect at 1% significance level on financial performance of
commercial banks in Ethiopia. The research concluded that credit risk has significant effect
on financial performance of commercial banks in Ethiopia. Hence, the study recommend in
support of each variable for commercial banks of Ethiopia should enhance their capacity in
credit analysis and loan administration.
Keywords: Credit risk, Bank, Performance.
IX
CHAPTER ONE
1. Introduction
1.1 Background of the Study
Credit creation is the main income generating activity for the banks; however, this activity
involves huge risks to both the lender and the borrower (kargi, 2011). When financial
institutions issue loans, there is a risk of borrower default and when banks collect deposits
and on-lend them to other clients, they put client’s savings at risk (Bessis, 2003). The risk of
a trading partner not fulfilling his or her obligation as per the contract on due date or anytime
thereafter can greatly jeopardize the smooth functioning of a bank’s business. The default of
small number of borrowers may result to large losses for a financial institution which can
lead to massive financial distress affecting the whole economy (Bessis, 2003)
Credit risk is the potential that a credit borrower/counter party fails to meet the obligations on
agreed terms. There is always scope for the borrower to default from his commitments for
one or the other reason resulting in crystallization of credit risk by the financial institution
(Pandey, 2004). These losses could take the form of outright default or alternatively, losses
from changes in portfolio value arising from actual or perceived deterioration in credit
quality. Credit Risk management is necessary to minimize the risk and maximize financial
institution’s risk adjusted rate of return by assuming and maintaining credit exposure within
the acceptable parameters (Pandey, 2004).
Credit risk management is a structured approach to managing uncertainties through risk
assessment, developing strategies to manage it and mitigation of risk using managerial
resources (Kargi, 2011). The strategies include transferring to another party, avoiding the
risk, reducing the negative effects of the risk, and accepting some or all of the consequences
of a particular risk (Kargi, 2011).
Effective credit risk management is the process of managing an institution’s activities which
create credit risk exposures, in a manner that significantly reduces the likelihood that such
activities will impact negatively on a bank’s earnings and capital (NBE guideline issued in
2003). Credit risk is not confined to a bank’s loan portfolio, but can also exist in its other
assets and activities. Likewise, such risk can exist in both a bank’s on-balance sheet and its
off-balance sheet accounts (NBE guideline issued in 2003).
1
Credit or default risk is the risk that the promised cash flows from loans and securities held
by financial institutions may not be paid in full. Should a borrower default, both the principal
loaned and the interest payments expected are at risk (Saunders and Cornett, 2003). The
potential loss a financial institution can experience suggests that financial institutions need to
collect information about borrowers whose assets are in their portfolios and to monitor those
borrowers overtime (Saunders and Cornett, 2003). Credit risk is the uncertainty associated
with borrower’s loan repayments. In general when borrowers’ asset values exceed their
indebtedness they repay loans but when borrowers’ assets values are less than loan values,
they do not repay and they could therefore exercise their option to default (Sinkey, 2002).
To minimize credit risk, banks are encouraged to use the “know your customer” principle as
expounded by the Basel Committee on Banking Supervision (Kunt-Demirguc and
Detragiache, 1997; Parry, 1999; Kane and Rice, 1998). Knowledge of the customer means
that credit shall be granted only to those customers whom the commercial bank fully
understands their business operations.
The National Bank of Ethiopia (NBE) developed risk management guidelines for the purpose
of providing minimum direction to banks on risk management and create a working
framework consistent with international standards and best practices which require banks to
have a fully independent credit risk management responsible for capital adjustment and
provision for escalating non-performing loans (NBE revised the risk management framework,
2003).
Girma (2011) asserted that in a small country like Ethiopia, the financial sector is still in the
development phase and customer services are still in their infancy and banks revenue depends
heavily on lending activities and credit growth is central to any banking organizations profit.
Thus, appropriate credit risk management system is the requirement of all banks, adjusting all
complication of their credit portfolio. Origination of loan system has importance so there is a
need for suitable analysis of borrower’s credit worthiness (Girma, 2011). According to
national bank of Ethiopia currently commercial banking system in Ethiopia comprises 17
commercial banks; 16 privately owned commercial banks and 1 public owned commercial
bank. The loan portfolio is the largest asset and the main source of profitability for those
commercial banks. The findings of this study will give a good insight for commercial banks
of Ethiopia credit management experts and policy makers to increase its profitability and
2
market share by minimizing the credit risk factors. Therefore, the purpose of this study is to
investigate the effect of credit risk on financial performance of commercial banks in Ethiopia.
1.2 Statement of the problem
Bessis (2005) the aim of every banking institution is to operate profitably in order to maintain
its stability and improve in growth and expansion. For the commercial banking, lending
represents the heart of the industry. Loans are the dominant asset at most banks, generate the
largest share of operating income, and represents the bank’s greatest risk exposure (Bessis
2005).
Weak credit risk management is the primary cause of many commercial banks’ failures.
McMenamin (1999) and Hempeletetal (1994) carried out studies and found out that the
consistent element in the failures was the inadequacy of the bank’s credit risk management
system in the controlling of loan quality.
Saundeers (2005) indicated that the very nature of the banking business is so sensitive
because more than 85% of their liability is deposits from depositors. Banks use these deposits
to generate credit for their borrowers, which in fact are a revenue generating activity for most
banks (Saundeers, 2005). This credit creation process exposes the banks to high default risk
which might led to financial distress including bankruptcy. All the same, beside other
services, banks must create credit for their clients to make some money, grow and survive
stiff competition at the market place (Saundeers, 2005).
Currently in Ethiopia seventeen commercial banks are in operation. According to NBE based
on their paid up capital commercial banks can be classified in to three categories that is big
size, medium size and small size.
Girma (2011) conducted a study on credit risk management and its impact on performance of
Ethiopian commercial Banks and Agegnehu (2013) conducted a study on credit risk
management and the performance of Ethiopian commercial banks and found that credit risk
and non-performing loan have been major challenges which affected the performance of
commercial banks in Ethiopia.
Girma (2011) and Agegnehu (2013) in their studies only selected that big size (CBE) and
medium sized commercial banks as a sample. Hence, it failed to disclose the literature gap by
incorporating those small sized commercial banks in their samples. Therefore, this study
3
incorporated those small sized commercial banks and tried to disclose the gap. The purpose
of this study is to investigate the effect of credit risk on financial performance of commercial
banks in Ethiopia.
1.3. Research questions
Based on the above statement of problem the researcher develops the following research
question.
Does credit risk influence the performance of Ethiopian commercial banks?
1.4 Objective of the study
1.4.1. General Objective
The main objective of the research is to investigate the effect of credit risk on financial
performance of commercial banks in Ethiopia.
1.4.2. Specific objectives
This study attempt to achieve the following specific objectives
1
To analyze the effect of non-performing loan to loan provision ratio on performance
of commercial banks in Ethiopia
2
To examine the effect of loan provision to total loan and advance ratio on
performance of commercial banks in Ethiopia.
3
To scrutinize the effect of non-performing loan to total loan and advance ratio on
performance of commercial banks in Ethiopia.
4
To investigate the effect of total loan and advance to total deposit ratio on
performance of commercial banks in Ethiopia.
1.5 Research hypothesis
Better credit risk management has high return on asset (ROA) and lower non-performing loan
and loan provision. Accordingly with the help of empirical studies on selected firms the study
was developed and tests the following hypothesis:
H1: Non-performing Loan to Loan provision ratio has negative and significant effect
on the performance of commercial banks in Ethiopia.
4
H2: Loan Provision to Total Loan and Advance ratio has negative and significant
effect on the performance of commercial banks in Ethiopia.
H3: Non-performing Loan to Total Loan and Advance ratio has negative and
significant effect on the performance of commercial banks in Ethiopia.
H4: Total Loan and Advance to Total Deposit ratio has a positive and significant
effect on the performance of commercial banks in Ethiopia.
1.6 Scope of the study
In terms of time sphere this study confined itself and only considered a time series data of 10
years (2007 - 2016) on the identified proxy performance indicators of Return on Asset and
proxy credit risk indicators of Non-Performing Loan to Loan Provision, Loan Provision to
Total Loan and Advances, Non-Performing Loan to Total Loan and Advances and Total
Loan and Advances to Total Deposit.
Considering the above mentioned circumstances, the results of the study are limited to
selected commercial banks as a sample and are generalized to all Ethiopian commercial
banks. Finally, the study used the quantitative approach and focus on the description of the
outputs from software and give the researcher own explanation.
1.7 Limitations the Study
This study is limited to detecting the relationship between credit risk management and
performance of Ethiopian commercial banks. Thus, other financial risks are not discussed. In
addition to this, the researcher only used return on asset as a performance measure and four
proxy credit risk indicators to measure the performance of commercial banks, as a result of
this it’s considered as a limitation for this particular study.
1.8 Significance of the study
The study has important for commercial bank credit risk managers and policy makers about
credit risk in the Ethiopian banking system and its impact on performance. Thus, the findings
have developed a framework for measuring and assessing credit risk which is an important
element for the financial stability unit at National Bank of Ethiopia. In general, the study will
have a great importance for commercial banking firms in order to make adequate control over
credit management system to make profitability sustainable. It can also act as a source of
5
literature for other scholars who intend to carry out further research on the effect of credit risk
on financial performance with specific reference to banking institutions
1.9 Organization of the Study
The study consists of five chapters. The first chapter deals with introductory part which
consists of background of the study, statement of the problem, objective(s) of the study,
significance of the study, and scope and limitation of the study. The second chapter deals
with review of theoretical and empirical literature. The third chapter deals with methodology
of the study. The fourth chapter presents the analysis and discussions of the results of the
study and finally, in the fifth chapter based on the analysis summary of major findings,
conclusions and recommendations has forward.
6
CHAPTER TWO
2. REVIEW OF RELATED LITERATURE
2.1
Introduction
This chapter summarizes the information from the available literature in the same field of
study. It reviews theories of credit management as well as empirical studies on credit risk
management, the gaps in existing literature described and developed conceptual framework.
2.2. Theoretical Review
2.2.2. Meaning of risk
As Ralph (2000) defined risk is the existence of uncertainty about future outcomes. Risk is a
key factor in economic life because people and firms make irrevocable investments in
research and product development, plant and equipment, inventory, and human capital,
without knowing whether the future cash flows from these investments will be sufficient to
compensate both debt and equity holders. If such real investments do not generate their
required returns, then the financial claims on these returns will decline in value (Ralph,
2000).
2.2.2. Meaning of Credit Risk
Financial institutions are exposed to a variety of risks among them; interest rate risk, foreign
exchange risk, market risk, liquidity risk, operational risk and financial risk (Yusuf, 2003;
Cooperman, Gardener and Mills, 2000). Credit or counterparty risk is defined as the chance
that a debtor or financial instrument issuer will not be able to pay interest or repay the
principal according to the terms specified in a credit agreement (Hennie van Greuning 2003).
Credit risk means that payments may be delayed or ultimately not paid at all, which can in
turn cause cash flow problems and affect liquidity (Hennie van Greuning 2003).
Credit risk is the risk of a loss resulting from the debtor's failure to meet its obligations to the
bank in full when due under the terms agreed (Raghavan, 2003). Credit risk is the potential
that a bank borrower or counterparty will fail to meet its obligations in accordance with
agreed terms. Generally the credit risk is associated with traditional lending activities of
banks and it is simply described as risk a loan not being repaid in part or in full (Hempel and
Simonson, 1999).
7
All banks have their own credit philosophy established in a formal written loan policy that
must be supported and communicated with an appropriate credit culture (Hempel and
Simonson, 1999). A credit culture is successful when all employees in the bank are aligned
with the management’s lending priorities (Hempel and Simonson, 1999).
2.2.3. Types of credit risk
Concerning the classifying of credit risk, different writers have expressed various criteria. For
example, Hennie (2003) list in his book that the three types of credit risk are personal or
consumer, company or corporate and sovereign or country risks, while Culp and Neves
(1998) pointed out realized default risk and resale risk being the two types of credit risk.
What is adopted here is part of the views from Horcher (2005), who defines more types of
credit risk, including default risk, counterparty pre-settlement risk, counterparty settlement
risk, legal risk, country or sovereign risk and concentration risk that are six. According to
Mckinley & Barrickman (1994), credit risk also contains transaction risk, intrinsic risk &
concentration risk.
1. Transaction Risk: It focuses on the volatility in credit quality and earnings resulting from
how the bank underwrites individual loan transactions. Transaction risk has three scopes:
selection, underwriting and operations (Mckinley & Barrickman, 1994).
2. Intrinsic Risk: It focuses on the risk inherent in certain lines of industry and loans to certain
industries. Commercial real estate construction loans are inherently more risky than consumer
loans. Intrinsic risk addresses the susceptibility to historic, predictive, and lending risk factors
that typify an industry or line of business. Historic elements address prior performance and
stability of the industry or line of business. Predictive elements focus on characteristics that
are subject to change and could positively or negatively affect future performance. Lending
fundamentals focus on how the collateral and terms offered in the industry or line of business
affect the intrinsic risk (Mckinley & Barrickman, 1994).
3. Concentration Risk: Concentration risk is the aggregation of transaction and intrinsic risk
within the portfolio and may result from loans to one borrower or one industry, geographic
area, or lines of business. Bank must define acceptable portfolio concentrations for each of
these aggregations. Portfolio diversify achieves an important objective. It allows a bank to
8
avoid disaster. Concentrations within a portfolio will determine the magnitude of difficulties
a bank will experience under adverse conditions (Mckinley & Barrickman, 1994).
2.2.4. Credit risk management
Experiences elsewhere in the world suggest that the key risk in a bank has been credit risk.
Indeed, failure to collect loans granted to customers has been the major factor behind the
collapse of many banks around the world (NBE guideline issued in 2003). Banks need to
manage credit risk inherent in the entire portfolio as well as the risk in individual credits or
transactions. Additionally, banks should be aware that credit risk does not exist in isolation
from other risks, but is closely intertwined with those risks (NBE guideline issued in 2003).
Effective credit risk management is the process of managing and institution’s activities which
create credit risk exposures, in a manner that significantly reduces the likelihood that such
activities will impact negatively on a bank’s earnings and capital. Credit risk is not confined
to a bank’s loan portfolio, but can also exist in its other assets and activities. Likewise, such
risk can exist in both a bank’s on-balance sheet and its off-balance sheet accounts (NBE
guideline issued in 2003).
Credit risk management is inherent in banking and is unavoidable. The basic function of bank
management is risk management. Banks assume credit risk when they act as intermediaries of
funds and credit risk management lies at the heart of commercial banking (Haim and Thierry,
2005). The business of banking is credit and credit is the primary basis on which a bank’s
quality and performance are adjusted. Credit risk is composed of default risk and credit
mitigation risk. Default risk is the risk that the counterparty will default on its obligations to
the investor (Haim and Thierry, 2005). In this risk, the credit quality deteriorates (or default
risk increases). Credit risk is more difficult to measure because data on both default and
recovery rates are not extensive, credit returns are highly skewed and fat tailed and longer
term time horizon and higher confidence levels are used in measuring credit risks. These are
problems in measuring credit risk that have inspired the development of several sophisticated
models and commercial software products for measuring portfolio credit risk (Haim and
Thierry, 2005).
9
Credit risk management maximizes bank’s risk adjusted rate of return by maintaining credit
risk exposure within acceptable limit in order to provide framework for understanding the
impact of credit risk management on banks’ profitability (Kargi, 2011).
Girma (2011) in a small country like Ethiopia, the financial sector is still in the development
phase and customer services are still in their infancy and banks revenue depends heavily on
lending activities and credit growth is central to any banking organizations profit (Kargi,
2011). Girma (2011) in his research paper supposed that nature of Ethiopian commercial
banking business is, so sensitive because more than 85% of their liability is deposits from
depositors, all banks aggressively use these deposits to generate credit for their borrowers to
make some money, grow and survive stiff competition at the market place and this credit
creation process exposes the banks to high default risk.
Basel (1999) credit risk management processes forces the banks to establish a clear process in
for approving new credit as well as for the extension to existing credit. These processes also
follow monitoring with particular care and other appropriate steps are taken to control or
mitigate the risk of connected lending (Basel, 1999). Credit risk management processes
enforce the banks to establish a clear process in for approving new credit as well as for the
extension to existing credit. These processes also follow monitoring with particular care and
other appropriate steps are taken to control or mitigate the risk of connected lending (Basel,
1999).
Banks have credit policies that guide them in the process of awarding credit. Credit control
policy is the general guideline governing the process of giving credit to bank customers
(Payle, 1997). The policy sets the rules on who should access credit, when and why one
should obtain the credit including repayment arrangements and necessary collaterals. The
methods of assessment and evaluation of risk of each prospective applicant are part of a
credit control policy (Payle, 1997).
2.2.5. Sources of Credit Risk
Coyle (2000) defines credit risk as losses from the refusal or inability of credit customers to
pay what is owed in full and on time. The main sources of credit risk include, limited
institutional capacity, inappropriate credit policies, volatile interest rates, poor management,
10
inappropriate laws, low capital and liquidity levels, directed lending, massive licensing of
banks, poor loan underwriting, reckless lending, poor credit assessment., no non-executive
directors, poor loan underwriting, laxity in credit assessment, poor lending practices,
government interference and inadequate supervision by the central bank (Coyle, 2000). To
minimize these risks, it is necessary for the financial system to have well-capitalized banks,
service to a wide range of customers, sharing of information about borrowers, stabilization of
interest rates, reduction in non-performing loans, increased bank deposits and increased credit
extended to borrowers. Loan defaults and nonperforming loans need to be reduced (Bank
Supervision Annual Report, 2006; Laker, 2007; Sandstorm, 2009).
Credit risk largely arises in assets shown on the balance sheet, but it can also show up on off
balance sheet in a variety of contingent obligations. Okorie (1998) identified poor project
supervision, evaluation and management; untimely loan disbursement; diversion of funds;
and dishonesty of loan beneficiaries as causes of loan default which ultimately leads to credit
risk.
Theoretically there are so many reasons as to why loans fail to perform. Some of these
include depressed economic conditions, high real interest rate, inflation, lenient terms of
credit, credit orientation, high credit growth and risk appetite, and poor monitoring among
others. Bercoff et al. (2002) categorizes causes of nonperforming loans to bank specific and
macroeconomic conditions.
2.2.6. Credit Risk Exposures in Banks
Generally, credit risk is related to the traditional bank lending activities, while it also comes
from holding bonds, interbank transactions, trade financing, foreign exchange transactions, in
the extension of commitments and guarantees, and the settlement of transactions. Various
financial instruments including acceptances, interbank transactions, financial futures,
guarantees, etc also increase banks’ credit risk (Bercoff et al. 2002).
2.2.7. General Principles of Sound Credit Risk Management in Banking
Credit risk is most simply defined as the potential that a bank borrower or counter party will
fail to meet his obligations in accordance with agreed terms (Adams & Mehran, 2004).
Boateng (2004) asserts that the goal of credit risk management is to maximize a bank’s riskadjusted rate of return by maintaining credit risk exposure within acceptable parameters of
not more than five (5%) of default rate. Banks therefore need to manage the credit risk
11
inherent in the entire portfolio as well as the risk in individual credits or transactions. Banks
should also consider the relationships between credit risk and other risks. According to
Machiraju (2004) the effective management of credit risk is a critical component of a
comprehensive approach to risk management and essential to the long-term success of any
banking organization.
The Basel Committee on Banking Supervision (1999; 2000; 2001) having assessed the
challenges associated with banks management of credit globally, issued some guidelines that
have come to be regarded as benchmark credit risk management practices in order to promote
sound practices for managing credit risk. The report of the Basel Committee on Banking
Supervision (2000) on credit risk focused on some four main areas as critical in every credit
management process. These areas are establishing an appropriate credit environment,
operating a sound credit granting process, ensuring adequate controls over credit risk and
evaluation and enforcement of protective covenants.
A. Establishing an appropriate credit environment
According to Wheehem & Hunger (2008), the controlling and directing mind of every
organization is the board of directors. As with all other areas of a bank’s activities, the board
of directors has a critical role to play in overseeing the credit granting and credit risk
management functions of the bank. The board of directors, according to the report of the
Basel Committee (2000) should have responsibility for approving and periodically (at least
annually) reviewing the credit risk strategy and significant credit risk policies of the bank.
According to Saunders (2005), these strategies should reflect the bank’s tolerance for risk and
the level of profitability the bank expects to achieve for incurring various credit risks.
Each bank should develop a credit risk strategy or plan that establishes the objectives guiding
its credit-granting activities and adopt the necessary policies and procedures for conducting
such activities (Machiraju, 2004). The board needs to recognize that the strategy and policies
must cover the many activities of the bank in which credit exposure is a significant risk.
The strategy should include a statement of the bank’s willingness to grant credit based on
exposure type (for example, commercial, consumer, and real estate), economic sector,
geographical location, currency, maturity and anticipated profitability (Matyszak, 2007). This
might also include the identification of target markets and the overall characteristics that the
bank would want to achieve in its credit portfolio (including levels of diversification and
tolerances).
12
Again, Boateng’s (2004) study shows that the credit risk strategy of a bank should give
recognition to the goals of credit quality, earnings and growth. Every bank, regardless of size,
is in business to be profitable and, consequently, must determine the acceptable risk-return
trade-off for its activities, factoring in the cost of capital. A bank’s board of directors should
approve the bank’s strategy for selecting risks and maximizing profits. The board should
periodically review the financial results of the bank and, based on these results, determine if
changes need to be made to the strategy. The board must also determine the degree of the
bank’s capital adequacy (Boateng, 2004).
According to Wilson (1998), the credit risk strategy of any bank should provide continuity in
approach. Therefore, the strategy will need to take into account the cyclical aspects of the
economy and the resulting shifts in the composition and quality of the overall credit portfolio.
Although the strategy should be periodically assessed and amended, it should be viable in the
long-run and through various economic cycles (Machiraju, 2004).
Fotoh (2005), states that the credit risk strategies and policies should be effectively
communicated throughout the organization. All relevant personnel should clearly understand
the bank’s approach to granting and managing credit and should be held accountable for
complying with established policies and procedures. The board should ensure that senior
management is fully capable of managing the credit activities conducted by the bank and that
those activities are done within the risk strategy, policies and tolerances approved by the
board (Basel Committee, 2001). The board should also regularly (i.e. at least annually), either
within the credit risk strategy or within a statement of credit policy, approve the bank’s
overall credit-granting criteria (including general terms and conditions). In addition, it should
approve the manner in which the bank will organize its credit-granting functions, including
independent review of the credit granting and management function and the overall portfolio
(Wilson, 1998).
While members of the board of directors, particularly outside directors, can be important
sources of new business opportunities for the bank, once a potential credit is introduced, the
bank’s established processes should determine how much and at what terms credit is granted
(Machiraju, 2004). In order to avoid conflicts of interest, as asserted by Wilson (1998), it is
important that board members do not override the credit-granting and monitoring processes
of the bank.
13
Fotoh (2005) states that once the board of directors has come out with a sound credit
management environment, senior management, led by the Chief Executive Officer, should
have responsibility for implementing the credit risk strategy approved by the board of
directors and for developing policies and procedures for identifying, measuring, monitoring
and controlling credit risk. Such policies and procedures should address credit risk in all of
the bank’s activities and at both the individual credit and portfolio levels. Senior management
of a bank is responsible for implementing the credit risk strategy approved by the board of
directors. The responsibility for implementing the strategy includes ensuring that the bank’s
credit-granting activities conform to the established strategy, that written procedures are
developed and implemented, and that loan approval and review responsibilities are clearly
and properly assigned Fotoh (2005. Senior management must also ensure that there is a
periodic independent internal assessment of the banks credit-granting and management
functions.
According to Boateng (2004), a cornerstone of safe and sound banking is the design and
implementation of written policies and procedures related to identifying, measuring,
monitoring and controlling credit risk. Credit policies establish the framework for lending
and guide the credit-granting activities of the bank. Credit policies should address such topics
as target markets, portfolio mix, price and non-price terms, the structure of limits, and
approval authorities (Basel committee, 2001). Such policies, according to Harper (2008),
should be clearly defined, consistent with prudent banking practices and relevant regulatory
requirements, and adequate for the nature of the bank and this may be difficult for very small
banks.
However, there should be adequate checks and balances in place to promote sound credit
decisions. The policies should be designed and implemented within the context of internal
and external factors such as the bank’s market position, trade area, staff capabilities and
technology. Policies and procedures that are properly developed and implemented enable the
bank to: (i) maintain sound credit-granting standards, (ii) monitor and control credit risk, (iii)
properly evaluate new business opportunities; and (iv) identify and administer problem
credits (Machiraju, 2004).
According to Sinkey (1998), banks consider the involvement of the Chief Executive Officer
(CEO), information generation and processing, and supervision as key elements of their risk
management and reporting systems. The components of a bank’s overall risk management
14
and reporting system focuses on such factors as: corporate organization structure,
organization of risk management, organization of lending, approval process, credit
administration, risk management function and loan quality reporting.
The Chief Executive Officer must be involved in the formulation and implementation of
credit policies that should incorporate the overall risk management and reporting system of
the bank. According to Sinkey (1998), every bank must have a credit policy that will guide
the credit activities and thereby reduce credit risk and improve profitability. Generally, a loan
policy consists of five major components:
General Policies:- Management, Trade area, Balance loan portfolio, Portfolio administration,
Loan-to-deposit ratio, Legal loan limit, Lending authority, Loan responsibility, Interest Rates,
Loan repayment, Collateral, Credit information and documentation, Delinquency ratio, Loan
loss Reserves, Charge-offs, Extensions of renewals of past due loans, Consumer laws and
regulations(Sinkey, 1998).
Specific Loan Categories:- Commercial loans, Agricultural loans, Mortgage loans,
Installment and branch bank loans, VISA and revolving credits, Mortgage banking
subsidiary, Personal loans (Sinkey, 1998).
Miscellaneous Loan Policies:- Loan to Executive Officers, directors and shareholders,
Employee loans, Mortgage- Banking subsidiary, Conflict of interest (Sinkey, 1998).
Quality Control:- Credit Department, Loan Review Department, Recovery Department
(Sinkey, 1998).
Committees:- Directors loan committee, Officers loan committee, Loan Review Committee
(Sinkey, 1998).
B. Operating a sound credit granting process
The Basel Committee (2000; 2001) asserts that in order to maintain a sound credit portfolio, a
bank must have an established formal transaction evaluation and approval process for the
granting of credits. Approvals should be made in accordance with the bank’s written
guidelines and granted by the appropriate level of management. There should be a clear audit
trail documenting that the approval process was complied with and identifying the
individual(s) and/or committee(s) providing input as well as making the credit decision
(Boateng, 2004). According to Wilson (1998), banks often benefit from the establishment of
15
specialist credit groups to analyze and approve credits related to significant product lines,
types of credit facilities and industrial and geographic sectors.
Banks should invest in adequate credit decision-making resources so that they are able to
make sound credit decisions consistent with their credit strategy and meet competitive time,
pricing and structuring pressures (Khambata, 1996).
Each credit proposal should be subjected to careful analysis by a qualified credit analyst with
expertise commensurate with the size and complexity of the transaction. According to
Boateng (2004), an effective evaluation process establishes minimum requirements for the
information on which the analysis is to be based. There should be policies in place regarding
the information and documentation needed to approve new credits, renew existing credits
and/or change the terms and conditions of previously approved credits.
According to Machiraju (2004), one of the management principles that banks have employed
in their customer information gathering process is screening. Screening involves the process
of identifying only reliable and creditworthy customers from a pool of numerous applicants
for financial assistance. Banks screen “good” credit risk from “bad” ones so as to make
profitable loans. Screening is usually carried out before a loan is granted.
Effective screening requires banks to collect accurate and reliable information from potential
borrowers. The aim is to evaluate the default risk of their customers. The potential borrower
is normally required to supply the loan officer with information about their background,
income and net worth. Different credit risk models ranging from qualitative to quantitative
ones may be used to facilitate the screening process to arrive at an informed decision.
Schonbucher (2000) and Machiraju (2004) assert that banks have traditionally focused on the
principles of five Cs to estimate borrowers’ creditworthiness. These five C’s are:
i.
Character. This refers to the borrower’s personal characteristics such as honesty,
willingness and commitment to pay debt. Borrowers who demonstrate high level of
integrity and commitment to repay their debts are considered favorable for credit
(Schonbucher, 2000 and Machiraju, 2004).
ii.
Capacity. This also refers to borrowers’ ability to contain and service debt judging
from the success or otherwise of the venture into which the credit facility is
employed. Borrowers who exhibit successful business performance over a reasonable
16
past period are also considered favorable for credit facility (Schonbucher, 2000 and
Machiraju, 2004).
iii.
Capital. This refers to the financial condition of the borrower. Where the borrower
has a reasonable amount of financial assets in excess of his financial liabilities, such a
borrower is considered favorable for credit facility (Schonbucher, 2000 and
Machiraju, 2004).
iv.
Collateral. These are assets, normally movable or unmovable property, pledged
against the performance of an obligation. Examples of collateral are buildings,
inventory and account receivables. Borrowers with a lot more assets to pledge as
collateral are considered favorable for credit facility (Schonbucher, 2000 and
Machiraju, 2004).
v.
Condition. This refers to the economic situation or condition prevailing at the time of
the loan application. In periods of recession borrowers find it quite difficult to obtain
credit facility (Schonbucher, 2000 and Machiraju, 2004).
In addition to the five Cs, Machiraju (2004) asserts that bankers and analysts have employed
many different models to assess the default risk on loans and bonds. These vary from
relatively qualitative to highly quantitative models. Further, these models are not mutually
exclusive, in that a financial institutions manager may use more than one to reach a credit
pricing or loan quantity rationing decision.
The information received will be the basis for any internal evaluation or rating assigned to
the credit and the accuracy and adequacy of the information are critical to management for
making appropriate judgments about the acceptability of the credit. Banks must develop a
corps of credit risk officers who have the experience, knowledge and background to exercise
prudent judgment in assessing, approving and managing credit risks. A bank’s credit-granting
and approval process should establish accountability for decisions taken and designate who
has the absolute authority to approve credits or changes in credit terms (Machiraju, 2004).
C. Ensuring adequate controls over credit risk
In order to ensure adequate controls over credit, Ganesan (2000) asserts that there must be
credit limits set for each officer whose duties have something to do with credit granting.
17
Material transactions with related parties should be subject to the approval of the board of
directors (excluding board members with conflicts of interest), and in certain circumstances
(e.g. a large loan to a major shareholder) reported to the banking supervisory authorities.
Banks must also consider the time frame for granting credit since time is of particular
importance to borrowers. Borrowers usually require credit within a given time, and for such
credits to be worthwhile they must be granted within the period the facility is required.
According to Hubbard (2000), if a borrower requires a credit within, say, one month, the
lending bank must meet such time period without undue delays. This implies that lending
institutions must make known in unequivocal terms to the borrowers the terms and conditions
to granting the credit. Having granted credit there is the need for maintaining an appropriate
credit administration, measurement and monitoring process. Again, banks must establish a
system of independent, continuous assessment of clients’ operational results, looking out for
early warning signs of operational difficulties (Mueller, 1998; Rachev, Schwartz &
Khindanova, 2000).
D. Maintaining an Appropriate Credit Administration, Measurement and Monitoring
Process
Credit administration is a critical element in maintaining the safety and soundness of a bank.
Once a credit is granted, it is the responsibility of the bank to ensure that credit is properly
maintained. This includes keeping the credit file up to date, obtaining current financial
information, sending out notices and preparing various documents such as loan agreements,
and follow-up and inspection reports (Wesley, 1993).
Credit administration, as emphasized by Wesley (1993), can play a vital role in the success of
a bank, since it is influential in building and maintaining a safe credit environment and
usually saves the institution from lending sins. Therefore, banks should never neglect the
effectiveness of their credit administration operations. Then talking about credit risk
measurement in banks, it is required that banks should adopt effective methodologies for
assessing the credit risk inherent both in the exposures to individual borrowers and credit
portfolios, and this will be explained in details later (Wesley, 1993). The last focus in this
area of principles is related to credit risk monitoring, which is definitely a must in banks’ risk
management procedure. Banks should keep track on the borrowers’ current financial
conditions and ensure their compliance with the covenants. Both cash flows and collateral
18
adequacy should be ensured and the potential problem credits should be considered. In this
way, banks are well in control of their credit qualities as well as all the related situations, and
can react to any future changes timely and readily (Wesley, 1993). A proper credit
monitoring system will provide the basis for taking prompt corrective actions when warning
signs point to deterioration in the financial health of the borrower. The bank has to assess the
credit worthiness of the borrower and even after the loan is granted, interim monitoring is
required until when the borrower has finished repaying the loan. This monitoring is very
important because with the uncertainty in the future, any potential event that can cause a
borrower to default payment can be fast identified or, a mechanism can be put in place on
time to reduce the frequency and/or intensify of a loss should it occur (Wesley, 1993).
E. The role of supervisors
Although the board of directors and senior management bear the ultimate responsibility for
an effective system of credit risk management, supervisors should, as part of their ongoing
supervisory activities, assess the system in place at individual banks to identify, measure,
monitor and control credit risk. This should include an assessment of any measurement tools
(such as internal risk ratings and credit risk models) used by the bank. In addition, they
should determine that the board of directors effectively oversees the credit risk management
process of the bank and that management monitors risk positions, and compliance with and
appropriateness of policies (Wheehem & Hunger, 2008).
To evaluate the quality of credit risk management systems, supervisors can take a number of
approaches. A key element in such an evaluation is the determination by supervisors that the
bank is utilizing sound asset valuation procedures. Most typically, supervisors, or the external
auditors on whose work they partially rely, conduct a review of the quality of a sample of
individual credits. In those instances where the supervisory analysis agrees with the internal
analysis conducted by the bank, a higher degree of dependence can be placed on the use of
such internal reviews for assessing the overall quality of the credit portfolio and the adequacy
of provisions and reserves. Supervisors or external auditors should also assess the quality of a
bank’s own validation process where internal risk ratings and/or credit risk models are used.
Supervisors should also review the results of any independent internal reviews of the creditgranting and credit administration functions. Supervisors should also make use of any
reviews conducted by the bank’s external auditors, where available (Wheehem & Hunger,
2008).
19
2.2.8 Profitability and risk of banks
Berger and DeYoung (1997) banking profitability may also reflect the risk taking behavior
of managers. Banks with high profitability are less pressured to revenue creation and thus less
constrained to engage in risk credit offerings. At the same time, inefficient banks are more
likely to experience high level of problem loans. Poor management can imply weak
monitoring for both operating costs and credit quality of customers, which will include high
levels of capital losses. Under this bad management hypothesis advances by Berger and
DeYoung (1997), mangers lack competencies to effectively assess and control risks incurred
when lending to new customers. Garciya-Marco and Robels-Fernendz (2007) found that
profit maximizing policies will be accompanied by higher level of risk. The acceptance and
management of financial risk is inherent to the business of banking and banks roles as
financial intermediaries. Risk management as commonly perceived does not mean
minimizing risk; rather the goal of risk management is to optimize risk-reward trade-off.
Notwithstanding the fact that banks are in the business of taking risk, it should be recognized
that an institution need not engage in business in a manner that unnecessarily imposes risk
upon it nor should it absorb risk that can be transferred to other participants. Rather it should
accept those risks that are uniquely part of the array of bank’s services. An important aspect
regarding various risk categories is their correlation Garciya-Marco and Robels-Fernendz
(2007).
2.2.9. Classification of loan and advances
National Bank of Ethiopia asserted that banks’ advances can be classified into five categories,
depending on their performance in terms of possibility of repayment. These categorizations
are:
1. Pass
Advances in these categories are those for which the borrower is up-to-date with repayment
of both principal and interest. Indications that an overdraft or loan is current indicate regular
activity on the account with no sign that a hardcore of debt is building up. NBE asserts that a
minimum of 1% of the aggregate outstanding balance of current advances should be provided
against possible future defaults. This enables the bank to set aside part of current profits or
earnings to offset any future default by beneficiaries of current advances.
2. Sub Mention
20
The loans and advances with pre-established repayment programs past due 30(thirty) days or
more but less than 90(ninety) days, or overdraft and loan or advances that do not have preestablished re payment program, NBE asserts that minimum of 3% of aggregate outstanding
of the current advances should be provided against possible future defaults.
3. Substandard advances
The loans and advances with pre-established repayment programs past due 90(ninety) days or
more but less than 180(one hundred eighty) days, or overdraft and loan or advances that do
not have pre-established re payment program, Substandard advances display well-defined
credit weaknesses that jeopardize the liquidation of the debt. These categories of advances are
advances for which the borrower’s cash flow is not significant enough to meet current
maturing debt, as well as the borrower lacking sufficient working capital to meet his
operating needs. For substandard advances, a provision of about 20% of the aggregate net
unsecured outstanding amount should be provided against current profit NBE (SBB/43/2008)
4. Doubtful advances
The loans and advances with pre-established repayment programs past due 180(one hundred
eighty) days or more but less than 360(three hundred sixty) days, or overdraft and loan or
advances that do not have pre-established re payment program, These categories of advances
exhibit all the weaknesses inherent in advances classified as substandard with the added
characteristics that the advances are not well secured and the weaknesses make collection or
liquidation in full on the basis of current facts, conditions and values highly questionable and
improbable. The possibility of loss is extremely high, but because of certain important and
reasonably specific pending factors, which may work to the advantage and strengthening of
the advance, its classification as an estimated loss is deferred until its exact status is
determined. NBE (SBB/43/2008) states that as much as 50% of the aggregate net unsecured
outstanding balance of doubtful advances should be provided against current profits to off-set
future defaults.
5. Loss advances
Non-performing loans or advances with pre-established repayment program past due
360(three hundred sixty) days or more. Advances classified as loss advances are classified
uncollectible and of such little value that their continuation as recoverable advances is not
worthwhile. Non-performing advances which are overdue for 360 or more days are classified
21
as loss advances. According to the NBE (SBB/43/2008) as much as 100% of the aggregate
net unsecured outstanding balance is provided for against current profits.
2.2.10. Determinants of credit risk
1. Proxy credit risk variables
I.
Nonperforming loans
Nonperforming loans (NPLs) refer to those financial assets from which banks no longer
receive interest and/or installment payments as scheduled. It is a loan that is in default or
close to being in default. Many loans become non-performing later being in default for
3months, but this can depend on the contract terms. A loan is nonperforming when payments
of interest and principal are past due by 3 months or more, or at least 3 months of interest
payments have been capitalized, refinanced or delayed by agreement, or payments are less
than 3months overdue, but there are other good reasons to doubt that payments will be made
in full (International Monetary Fund (IMF), 2001)
They are known as non-performing because the loan comes to an end to perform or generate
income for the bank. Choudhury et al. (2002) state that NPL is not a uniclass but it is a
multiclass concept, which means that NPLs can be classified into different varieties. Usually
based on the length of overdue of the said loans, NPLs are viewed as a typical byproduct of
financial crisis: they are not a main product of the lending function but rather an accidental
occurrence of the lending process, one that has enormous potential to deepen the severity and
duration of financial crisis and to complicate macroeconomic management (Woo, 2000). This
is because NPLs can bring down investors confidence in the banking system, piling up
unproductive economic resources even though depreciations are taken care of, and impeding
the resource allocation process.
In a bank-centered financial system, NPLs can further thwart economic recovery by shrinking
operating margin and eroding the capital base of the banks to advance new loans. This is
sometimes referred to as credit crunch (Bernanke et al., 1991).
In addition, NPLs, if created by the borrowers willingly and left unresolved, might act as a
contagious financial malaise by driving good borrowers out of the financial market. Further,
Muniappan (2002) argues that a bank with high level of NPLs is forced to incur carrying
costs on non-income yielding assets that not only strike at profitability but also at the capital
adequacy of a bank, and in consequence, the bank faces difficulties in augmenting capital
resources. Bonin and Huang (2001) also state that the probability of banking crises increases
22
if financial risk is not eliminated quickly. Such crises not only lower living standards but can
also eliminate many of the achievements of economic reform overnight.
II.
Loan provision
Loan loss provision is an expense set aside as an allowance for uncollected loans loan
payments. This provision is used to cover a number of factors associated with potential loan
losses including bad loans, customer defaults and renegotiated terms of a loan that incur
lower than previous estimated payment. https://www.investopedia.com
Loan Loss Provision' is a non-cash expense for banks to account for future losses on loan
defaults. Banks assume that a certain percentage of loans will default or become slow-paying.
Banks enter a percentage as an expense when calculating their pre-tax incomes. This
guarantees a bank's solvency and capitalization if and when the defaults occur. The ratio of
Loan Provision to Total loan and Advance (LP/TLA) is a measure for credit risk. It indicates
banks’ ability to generate income before the expectation of default occurs (Agegnehu, 2013)
III.
Loan to deposit
The loan to deposit ratio is used to calculate a lending institutions ability to cover
withdrawals made by customers. A lending institution that accepts deposit must have a
certain measure of liquidity to maintain its normal daily operations. Loans given to its
customer are mostly not considered liquid meaning that they are investments over long period
of time. Although a bank will keep a certain level of mandatory reserves, they may also
choose to keep a percentage of their non-lending investing in short term securities to ensure
that any monies needed can be accessed in the short term (Agegnehu, 2013).
Loan to deposit = total loan and advance/ total deposit
IV.
Non-Performing Loan to Loan Provision
Is a loan that is in default or close to being in default out of loan provision in future or
become slow-paying. The ratio of non-performing loan to loan provision (NPLLP) used as a
measure for credit risk. The ratio of non-performing loan to loan provision, which measures
how much a bank is suffer default in year relative to its reserved loan provision, is used to
23
measure the effect of non-performing loans to loan provision on performance of commercial
banks in Ethiopia (Girma, 201).
Non-performing Loan to Loan provision = Non-performing Loan/ Loan provision
2. Performance measure
Return on Asset (ROA)
Prior works on credit risk management and bank performance indicated that Return on assets
(ROA) was an important measurement used in comparing the financial performance of banks
(Girma, 2011; Agegnehu, 2013). It is defined as Profit after Tax divided by Total Asset;
reflect how well bank managers are using the banks real investment resource to generate
profit. It shows the profit earned per dollar of assets and most importantly, reflects the
management ability to utilize the bank’s financial and real investment resources to generate
profits. For any bank, ROA depends on the bank’s policy decisions as well as uncontrollable
factors relating to the economy and government regulations. Many regulators believe return
on assets is the best measure of bank efficiency from the two major alternative measures of
profitability, namely ROA and ROE. As highlighted by Athanasoglou et al. (2008) and
Sufian (2011), many scholars suggest that ROA is the key ratio for the evaluation of bank
performance given that ROA is not distorted by high equity multipliers, while ROE
disregards the risks associated with high leverage and financial leverage.
24
2.3 Empirical Review
Credit Risk Management is a serious threat to the performance of banks; therefore various
researchers have examined the effect of credit risk management on banks in varying
dimensions.
Benson Mwai Karugu (2015) the study sought to analyze the effect of credit risk management
practices on profitability of listed commercial banks at Nairobi Kenya. A descriptive research
design was adopted. The population comprised of listed commercial banks where a sample of
55 employees was purposively sampled. The results indicated that credit risk governance had
a positive and significant effect on profitability. Based on the study findings the study
concluded that credit appraisal, debt collection and credit risk governance have a significant
positive effect on profitability. It is thus recommended that commercial banks should have
stringent credit appraisal and debt collection policies, credit personnel at all levels must work
in co-ordination in order to ensure that credit is collected in a timely manner and banks
should also adopt credit risk governance frameworks which can be attained by making the
process of interaction between senior management and the Board more effective.
Ugirase Josiane Magnifique (2013) examined the effect of credit risk management on the
financial performance of commercial banks in Rwanda. The findings of the study conclude
that credit risk identification, credit scoring mechanism and credit analysis and assessment
are good predictors of the model consequently those three indicators used of credit risk
management have shown a positive relationship with the financial performance of
commercial banks in Rwanda.
The prevailing relationship between profitability and credit risk is further complicated by the
finding of Kithinji (2010).Employing a regression analysis on data collected from financial
reports of commercial banks in Kenya for the period of 2004 to 2008 concluded that
profitability of commercial banks measured by ROA did not show significant relationship
with credit risk measures.
Joyce Wangari Githaiga (2015) the study was undertaken to analyze the effects of credit risk
management practices on the performance of Financial Banking Institutions. The study
attempted to establish if there exists any relationship between the credit risk management
25
determinants by use of CAMEL indicators and financial performance of commercial banks in
Kenya. The finding of the study concludes that Credit Risk has a strong negative relationship
with financial performance. This indicates that poor credit risk or high non-performing loans
to total assets related to poor bank performance. Thus, Commercial banks with high credit
risk and low non-performing loans are more profitable than the others.
Kargi (2011) estimated the effect of credit risk on the profitability on Nigerian banks. Data on
credit risk and profitability ratios were collected from 2004 to 2008.The analysis of this data
involved descriptive, correlation and regression techniques. The result was that the credit risk
management has the substantial influence on the profitability of Nigerian banks. Other
findings on the study showed
that a rise in nonperforming loans negatively affected
profitability and liquidity.
The ratio of loan loss to total loans (LLTR) is also significant determinant of banks profit
(Sufian et al (2008). The rise in LLTR represents a rise in the credit risk the banks are
exposed to. Hence higher credit risk affects profitability of a bank adversely. A study carried
out by Vong et al (2009) revealed that, a loan loss provision is inversely related to the
performance of banks in Macao.
Epure and Lafuente (2012) studied the impact of risk on the performance of banks in costRican Banking industry during 1998-2007.The results showed the performance has an inverse
relationship with non-performing loans and capital adequacy related positively with
performance.
As short (1979) argues, size is closely related to the capital adequacy of a bank since
relatively large banks tends to rise less expensive capital and, hence, appear more profitable
the finding of also shows that return on equity ROE and return on asset ROA all indicating
profitability where negatively related to the ratio of non-performing loan to total loan
NPL/TL, of financial institutions therefore decrease profitability. According to Hempel and
Simomson (1999), non-performing loans is a major credit risk indicator, depleting
profitability and therefore a bank can lower its credit risk exposure by reducing it.
In Ethiopian case, studies on the relationship between credit risk and performance of
commercial banks in Ethiopia are few through many studies documented that credit risk is
26
among the major challenges of bank in Ethiopia. Of these studies, Tibebu (2011) studied the
effect of credit risk management on the performance of commercial banks in Ethiopia. Used
secondary date from annual reports of the commercial banks showed that there is a negative
relationship between credit risk and performance of commercial banks in Ethiopia
Girma (2011) examined Credit Risk Management and Its Impact on Performance on
Ethiopian commercial Banks. Quantitative research design is employed under the quantitative
research design survey method is used. The data were collected by cross sectional survey
method. The estimation results showed that result of return on asset (ROA) on the regression
shows that non-performing loan and loan provision of the financial institution is significantly
negatively related to performance.
On the other hand, Agegnehu (2011) describes credit risk management and performance of
Ethiopian commercial banks. This study attempts to reveal the relationship between credit
risk and performance of commercial banks in Ethiopia. In order to investigate these issues
quantitative research approach is utilized based on documentary analysis. A panel data from
seven selected commercial banks covering the eleven-year period (2001-2011) is analyzed
within the fixed effects framework. The findings of the study showed that Non-performing
loan had statistically significant negative relationship with ROA.
In addition, Hailu (2016) in their study investigated the impact of credit risk on profitability
performance of selected public and private commercial banks of Ethiopia. Used secondary
date from annual reports of the selected public and private commercial banks of Ethiopia
showed that there is a negative relationship between credit risk and performance of
commercial banks in Ethiopia.
27
2.4. Literature gap
Empirical studies were conducted on effect of credit risk in performance of commercial
banks in Ethiopia by Girma (2011), Tibebu (2011) and Agegnehu (2013). However, in their
studies only selected sample that big size (CBE) and medium sized commercial banks as a
sample. Hence, it failed to disclose the literature gap by incorporating those small sized
commercial banks in their samples. Therefore, this study incorporated those small sized
commercial banks and tried to disclose the gap. The purpose of this study is to investigate
credit risk management practice and its effect on performance of Ethiopian commercial
banks. The study will contribute a lot for both public and private commercial banks credit
management department and policy makers in order to identified and mitigated the bank
credit risk factors and in turn to improve its performance.
2.5. Conceptual Framework
The main objective of this study is to examine the effect of Credit risk on financial
performance of commercial banks in Ethiopia. Based on the objective of the study, the
following conceptual model is framed. As it described previously in the related literature
review parts, bank performance measured by Return on asset can be affected by credit risk
proxy variables of Non-Performing Loan to Total Loan and Advance (NPLTLA), NonPerforming Loan to Loan Provision (NPLLP), Loan Provision to Total Loan and Advance
(LPTLA) and Total Loan and Advance to Total Deposit (TLATD).
Thus, the following conceptual model is framed to summarize the main focus and scope of
this study in terms of variables included.
28
Non-performing
loan to loan
provision
Loan provision to
total loan and
advance
Credit
Risk
Non-performing
loan to total loan
and advance
Total loan and
advance to total
deposit
29
Return on
Asset (ROA)
CHAPTER THREE
3. RESEARCH METHODOLOGY
This chapter highlights about the research methodology of the study and comprises research
approach, research design, target population and sampling unit, sampling design and sample
size, data analysis, model specification, validity and Ethical consideration sections.
3.1. Research Approach
The aim of this research is to investigate the effect of credit risk on financial performance of
commercial banks in Ethiopia. Therefore, in order to achieve the aforementioned objective
the researcher employed quantitative research approach.
3.2. Research design
According Durrheim (2004) research design is a strategic framework for action that serves as
a bridge between research questions and the execution, or implementation of the research
strategy. Thus, in order to answer the research questions casual research design used. The
purpose of this study is to investigate the effect of credit risk on financial performance of
commercial banks in Ethiopia in so doing, casual research design has adopt in collecting,
analyzing and interpreting data to address the objective(s) of the study.
3.3 Data Type and Source
In this study the researcher has employed quantitative research approach. Only secondary
data are used for the study. The secondary data was collected from the financial statements of
the selected banks. From those banks, the study obtained data by considering the proxy credit
risk indicators of Non-Performing Loan to Loan Provision, Loan Provision to Total Loan and
Advances, Non-Performing Loan to Total Loan and Advance, and Total Loan and Advance
to Total Deposit and performance proxy indicators of ROA of the period covered from 2007
to 2016 G.C.
3.4 Target commercial banks selection criteria and technique
Currently in Ethiopia seventeen commercial banks are in operation. According to NBE based
on their paid up capital commercial banks can be classified in to three categories that is big
size, medium size and small size. Big size consists of Commercial bank of Ethiopia, Medium
30
size consists of Awash Bank, Dashen Bank, United Bank, Bank of Abyssinia, Wogagen Bank
and Nib International Bank and Small size consists of Cooperative Bank of Oromia, Lion
International Bank, Buna International bank, Oromia International Bank, Abay Bank, Enat
Bank, Birhan Bank, Zemen Bank, Addis International Bank and Debub Global Bank.
Therefore, the researcher classified in to three stratums, from those the researcher used six
commercial banks as a sample namely Commercial bank of Ethiopia, Awash Bank, Dashen
Bank, United Bank, Corporate Bank of Oromia and Lion International Bank which is one big
size, three medium sizes and two small sizes by using purposive sampling technique in terms
of their capital from each stratum and based on their loan portfolio. Purposive sampling
targets a particular group of people.
3.5. Method of data Analysis
To achieve objective of the study, the study mainly concentrated on quantitative analysis.
Hence, the researcher used econometric model to identify and measure the effect of credit
risk on financial performance of Ethiopian commercial banks and used Ordinary Least
Square (OLS) method using Eviews-8 econometric software package for the study.
According to Brooks (2008) regression is concerned with describing and evaluating the
relationship between a given variable (usually called the dependent variable) and one or more
other variables (usually known as the independent variables. Thus, the study adopted panel
data regression model to examine the effect of credit risk on financial performance of
commercial banks in Ethiopia.
As stated by Brooks (2008) panel data is favoured for situation often arises in financial
modelling where we have data comprising both time series and cross-sectional elements. In
addition, we can address a broader range of issues and tackle more complex problems with
panel data than would be possible with pure time-series or pure cross-sectional data alone.
Accordingly, the study model focused on panel data technique that comprises both crosssectional elements and time-series elements; the cross-sectional element is reflected by the
different Ethiopian commercial banks (six) and the time-series element is revealed by the
period of study (2007-2016). Therefore, the collected panel data is analyzed using descriptive
statistics, correlations and multiple linear regression analysis. The rational for choosing
Ordinary Least Square (OLS) is that, if the Classical Linear Regression Model (CLRM)
assumptions hold true, then the estimators determined by OLS will have a number of
desirable properties, and are known as Best Linear Unbiased Estimators (Brooks, 2008).
31
Diagnostic checking is done to test whether the sample is consistent with the following
assumptions. According to Brooks (2008), the assumptions of ordinary least squares are:
I.
The errors have zero mean (E(ut ) = 0)
II.
Variance of the errors is constant (Var(ut) = σ2 <∞)
III.
Covariance between the error terms over time is zero (cov(ui, uj ) = 0 for i ≠ j
IV.
Test for Normality (ut ∼N(0, σ2)
V.
Multicollinearity Test
If all the above assumptions are consistent with the sample, E-view result will be accurate
and reliable. The following tests are done in this research to test the above assumptions.
I.
The errors have zero mean (E(ut ) = 0)
Relay on Brooks (2008), the first assumption required is that the average value of the errors is
zero. In fact, if a constant term is included in the regression equation, this assumption will
never be violated.
II.
Variance of the errors is constant (Var(ut) = σ2 <∞) (heteroscedasticity)
According to Brooks (2008), the variance of the errors is constant this is known as the
assumption of homoscedasticity. If the errors do not have a constant variance, they are said to
be heteroscedastic. If heteroscedasticity occur, the estimators of the ordinary least square
method are inefficient and hypothesis testing is no longer reliable or valid as it will
underestimate the variances and standard errors. There are several tests to detect the
Heteroscedasticity problem, which are Park Test, Glesjer Test, Breusch-Pagan-Goldfrey Test,
White’s Test and Autoregressive Conditional Heteroscedasticity (ARCH) test. In this study,
the popular white test was employed to test for the presence of heteroscedasticity.
III.
Covariance between the error terms over time is zero (cov(ui, uj ) = 0 for i ≠ j
(Autocorrelation)
According to Brooks (2008), when the error term for any observation is related to the error
term of other observation, it indicates that autocorrelation problem exist in this model. In the
case of autocorrelation problem, the estimated parameters can still remain unbiased and
consistent, but it is inefficient. The result of T-test, F-test or the confidence interval will
become invalid due to the variances of estimators tend to be underestimated or overestimated.
Due to the invalid hypothesis testing, it may lead to misleading results on the significance of
32
parameters in the model. Therefore, the study test for the existence of autocorrelation, the
popular Durbin–Watson test and Breusch-Godfrey test were employed.
VI.
Normality (ut ∼N(0, σ2)
As per Brooks (2008) normality tests are used to determine if a data set is well-modeled by a
normal distribution. With the normality assumption, ordinary least square estimation can be
easily derived and would be much more valid and straight forward. This study used Jarque
Bera Test (JB test) to find out whether the error term is normally distributed or not.
IV.
Multicollinearity
According to Brooks (2008), Multicollinearity will occur when some or all of the
independent variables are highly correlated with one another. If the multicollinearity occurs,
the regression model is unable to tell which independent variables are influencing the
dependent variable. This study used high pair-wise correlation coefficients method to test the
presence of multicollinearity problem in a regression model, because it shows the correlation
of independent variables between each other one by one. Malhotra (2007) stated that
multicollinearity problems exists when the correlation coefficient among explanatory
variables should be greater than 0.75. However, Brooks (2008) mentioned that if the
correlation coefficient along with the independent variables is 0.8 and above,
multicollinearity problems will be existed.
3.6. Model Specification
To investigate the effect of credit risk on financial performance of commercial banks in
Ethiopia following general multiple linear regression models used:
Yi = β0 + βXi+ µi…………………………………………… (1)
Where:
Yi –Dependent Variable
β0 - Constant term,
Xi - Explanatory or independent Variables
µi
- Disturbance term
33
Hence, based on the above general multiple linear regression models and on the basis of
selected variables for the study the specific empirical model presented as follows:
ROA =f (NPLTLA, LPTLA, NPLLP, TLATD)
Y= α + β1X1 + β2X2 + β3X3+ β4X4+ µ
Where: Y= Return on Asset (dependent variable) - performance measure
Proxy credit risk indicators
X1= Non-Performing Loan to Total Loan and Advances
X2= Loan Provision to Total Loan and Advance
X3= Non-Performing Loan to Loan Provision
X4= Total Loan and Advance to Total Deposit
µ= disturbance term
Also α is an intercept and β is the parameter of explanatory variable and it measures by what
amount the dependent variable (ROA) increases or decreases when the specific explanatory
variables increases or decreases by a unit.
3.6.1. Variables description
In this section explained the variables used as dependent and independent (explanatory)
variables in the study. The definitions and measurements that are used for these variables are
described and accordingly with the help of empirical data hypothesis developed as follows:
Dependent Variable
Return on Asset (ROA)
Prior works on credit risk management and bank performance indicated that Return on assets
(ROA) was an important measurement used in comparing the financial performance of banks
(Girma, 2011; Agegnehu, 2013). It is defined as Profit after Tax divided by Total Asset;
reflect how well bank managers are using the banks real investment resource to generate
profit. It shows the profit earned per dollar of assets and most importantly, reflects the
management ability to utilize the bank’s financial and real investment resources to generate
profits. For any bank, ROA depends on the bank’s policy decisions as well as uncontrollable
factors relating to the economy and government regulations. Many regulators believe return
on assets is the best measure of bank efficiency from the two major alternative measures of
profitability, namely ROA and ROE. As highlighted by Athanasoglou et al. (2008) and
34
Sufian (2011), many scholars suggest that ROA is the key ratio for the evaluation of bank
performance given that ROA is not distorted by high equity multipliers, while ROE
disregards the risks associated with high leverage and financial leverage.
Independent Variables
Non-Performing Loan to Loan Provision
NPLLP: is a loan that is in default or close to being in default out of loan provision in future
or become slow-paying. The ratio of non-performing loan to loan provision (NPLLP) used as
a measure for credit risk. The ratio of non-performing loan to loan provision, which measures
how much a bank is suffer default in year relative to its reserved loan provision, is used to
measure the effect of non-performing loans to loan provision on performance of commercial
banks in Ethiopia.
Loan provision to Total loan and advance
LPTLA; first 'Loan Loss Provision' is a non-cash expense for banks to account for future
losses on loan defaults. Banks assume that a certain percentage of loans will default or
become slow-paying. Banks enter a percentage as an expense when calculating their pre-tax
incomes. This guarantees a bank's solvency and capitalization if and when the defaults occur.
The ratio of Loan Provision to Total loan and Advance (LP/TLA) used as a measure for
credit risk. It indicates banks’ ability to generate income before the expectation of default
occurs (Agegnehu, 2013).
Non-performing loan to Total loan and advances
NPLTLA is a loan that is in default or close to being in default. Many loans become nonperforming after being in default for 90 days, but this can depend on the contract terms. NPL
amount has been presented using different names, such as, impaired loans, problem loans,
doubtful claims and loan allowances. However, the definitions of those are similar to the
definition of NPLs. NPL amount is provided in the Notes to financial statements under Loans
section. Total Loan (TL) amount, is the denominator of the ratio, has been gathered by adding
two types of loans: loans to institutions and loans to the public. The researcher has collected
the non-performing loan amount provided in off balance sheet of the banks that is submitted
to NBE from 2007 until 2016.
35
Total loan and advance to Total Deposit
TLATD; the formula for the loan to deposit ratio is exactly as its name implies, total loans
and advance divided by total deposits. A lending institution that accepts deposits must have a
certain measure of liquidity to maintain its normal daily operations. Loans given to its
customers are mostly not considered liquid meaning that they are investments over a longer
period of time. Although a bank will keep a certain level of mandatory reserves, they may
also choose to keep a percentage of their non-lending investing in short term securities to
ensure that any money needed can be accessed in the short term. The ratio of Total loan &
Advances to Total deposit (TLA/TD) it used as indicators of credit risk. For banks, it is how
much they have coming in (deposits) vs how much they have going out (loans). The more
money the bank has loaned out generates more interest income provided the loans are to
secure borrowers. Deposits are an obligation (debts) of the bank has to the depositors. So, a
healthy bank has lots of secure loans generating lots of income (interest) to cover depositor's
accounts.
36
CHAPTER FOUR
4. DATA ANALYSIS AND DISCUSSION OF RESULT
As the researcher discussed in the previous chapters the major objective of this study is to
investigate the effect of credit risk on financial performance of Ethiopian commercial banks.
Therefore, this chapter deals with the results and analysis of the findings and it contains three
sections. The first section presented descriptive and correlation analysis on variables of the
study; the second section presented fulfillment of the classical linear regression model
(CLRM) assumptions; the third section laid down the results of regression.
The data analysis procedures used for ratio scale measurement and the ratio of the specified
dependent and independent variables were calculated for further statistical analysis. The
collected data was analyzed by the aid of the statistical software Eview8.
4.1 Descriptive statistics
This section presents the descriptive statistics of dependent and explanatory variables used in
this study. The dependent variable used in this study was ROA, while the explanatory
variables are NPLTLA, NPLLP, LPTLA and TLATD. Table 4.1 Summary of Descriptive
Statistics of dependent and independent variables for six commercial banks over the sample
period of 2007 - 2016 with a total of 60 observations. The table shows the mean, minimum,
maximum, standard deviation and number of observations for the dependent variable and
independent variables of firms’ performance.
Table 4.1 Descriptive statistics
ROA
NPLLP
LPTLA
NPLTLA
TLATD
Mean
0.025790
0.646007
0.030677
0.026548
0.577008
Median
0.027515
0.644215
0.021800
0.020400
0.563130
Maximum
0.044700
0.975500
0.315350
0.198540
0.901820
Minimum
-0.018800
0.121490
0.002440
0.002742
0.297430
Std. Dev.
0.010317
0.208977
0.042687
0.029111
0.133128
Source: - annual report of sample commercial banks computed using E-views 8
37
As indicated in the table 4.1, the firms’ performance measured by return on asset shows that
sampled of Ethiopian commercial banks achieved 2.579% on average after tax profit over the
last ten years from 2007 to 2016. From the total sample, return on asset had a maximum of
4.47% and minimum of -1.88%. It means that the most profitable commercial banks among
the sampled earned 4.47 cents of profit after tax for a single birr (1.00) invested in the assets
of the firm. On the other hand, not profitable commercial banks of the sampled lost -1.88
cents of profit after tax for each birr (1.00) invested in the assets of the firm and the value of
return on asset deviate from its mean by 1.03%.
The average value for non-performing loan to loan provision as measured by ratio of
commercial banks was 64.60% with a maximum of 97.55% and a minimum of 12.15%. It
implies that there is large amount of non-performing which tends to have default risk. On the
other hand, sampled of Ethiopian commercial banks who have excess uncollected amount has
average 0.64 cents from the reserved loan provision birr (1.00) and the value of nonperforming loan to loan provision deviate from its mean by 20.89%.
On the other hand, the loan provision to total loan and advance ratio indicated by the range
between 31.53% and 0.24%. The mean value is equals 3.06 %. The relatively high range
between the minimum and maximum value implies that the most efficient bank has a
profitable capability compared to the least efficient bank. The standard deviation statistics for
loan provision to total loan and advance ratio was 4.26% which indicates that the written
amount of loan loss variation between the selected banks was medium.
The average value of Non-performing loan to total loan advance rate equals 2.54% with a
maximum of 19.85% and its minimum value was 0.27%. The standard deviation statistics for
NPLR is 2.91%.
This means despite the inverse relationship that exists between non-
performing loan and performance and the value of non-performing loan to total loan and
advance deviate from its mean by 2.91%.
Finally, the average value of the total loan to deposit was 57.70% with a minimum 29.74%
and maximum of 90.18%. The standard deviation statics for this was 13.31%. This shows the
existence relatively high variation of loan to deposit ratio between the selected banks
compared with the variation in ROA.
38
4.2 Correlation Analysis
Correlation measures the degree of linear association between variables. Values of the
correlation coefficient are always ranged between +1 and -1. A correlation coefficient of +1
indicates that the existence of a perfect positive association between the two variables, while
a correlation coefficient of -1 indicates perfect negative association. A correlation coefficient
of zero, on the other hand, indicates the absence of relationship (association) between two
variables (Brooks, 2008).The table below shows the correlation matrix among dependent and
independent variables.
Table 4.2 Correlation Analysis of Variables
ROA
NPLLP
LPTLA
NPLTLA
TLATD
ROA
1.000000
-0.227633
-0.330013
-0.283095
0.267086
NPLLP
-0.227633
1.000000
0.187480
-0.338694
0.091506
LPTLA
-0.330013
0.187480
1.000000
-0.106438
0.029683
NPLTLA
-0.283095
-0.338694
-0.106438
1.000000
-0.047537
TLATD
0.267086
0.091506
0.029683
-0.047537
1.000000
Source: - annual report of sample commercial banks computed using E-views 8
The results in table 4.2 revealed that a correlation between dependent variables with
independent variables. According to the result non-performing loan to loan provision, loan
provision to total loan and advance and non-performing loan to total loan and advance have
negative correlation with return on asset and total loan and advance to total deposit have
positive correlation with return on asset for measurement of Ethiopian commercial banks’
performance. It refers that when non-performing loan to loan provision, loan provision to
total loan and advance and non-performing loan to total loan and advance variables increases,
performance of Ethiopian commercial banks will be go down and when total loan and
advance to total deposit increases, performance of Ethiopian commercial banks will be go up.
The coefficient estimates of correlation in the above table shows -0.227633, -0.330013, 0.283095 and 0.267086 for loan provision to non-performing loan, loan provision to total
loan and advance, non-performing loan to total loan and advance and total loan and advance
to total deposit respectively. This implies that non-performing loan to loan provision, loan
39
provision to total loan and advance and non-performing loan to total loan and advance are
negatively correlated with return on asset and total loan and advance to total deposit are
positively correlated with return on asset.
4.3 Regression model tests
For valid hypothesis testing and to make data available for reliable results, the test of
assumption of regression model is required. Accordingly, the study has gone through the
most critical regression diagnostic tests consisting of model specification tests,
heteroskedasticity, autocorrelation, normality and multicollinearity test accordingly.
4.3.1 Model Selection (Random Effect versus Fixed Effect Models)
As Brooks (2008) referring on his book, there are broadly two classes of panel estimator
approaches that can be employed in financial research: fixed effects models and random
effects models. The choice between both approaches is done by running a Hausman test. To
conduct a Hausman test the number of cross section should be greater than the number of
coefficients to be estimated. In this study the numbers of cross section are greater than the
number of coefficients to be estimated so it is possible to conduct a Hausman test. Therefore
a fixed cross-sectional effect is specified in the estimation to capture the effects of different
commercial banks.
Table 4.3 Hausman Test
Correlated Random Effects - Hausman Test
Equation: Untitled
Test cross-section random effects
Test Summary
Cross-section random
Chi-Sq. statistic
13.831540
Chi-Sq. d. f.
Prob.
4
0.0079
Source: - annual report of sample commercial banks computed using E-views 8.
The Hausman model selection test for this study has a p-value of 0.0079 for the regression
models. On this score, fixed effect model is preferable. Therefore, the study has 10 time
series and 6 cross sectional which is relevant to fixed effect model.
According to Brooks (2008) it is often said that the random effects model is more appropriate
when the entities in the sample can be thought of as having been randomly selected from the
40
population, but a fixed effect model is more reasonable when the entities in the sample
effectively represent the entire population. Thus, the sample for this study was not selected
randomly instead it selected rationally that can effectively represent the total number of
population, due to this it is appropriate for fixed effect model selection.
4.3.2. Tests for the Classical Linear Regression Model (CLRM) assumptions
To maintain the data validity and robustness of the regressed result of the research, the basic
classical linear regression model (CLRM) assumptions must be tested for identifying any
misspecification and correcting them so as to augment the research quality (Brooks,2008).
There are different CLRM assumptions that need to be satisfied and that are tested in this
study, which are: errors equal zero mean test, model specification, heteroskedasticity,
autocorrelation, normality and multicollinearity test.
I.
The errors have zero mean (E(ut ) = 0)
This part shows the test for the assumptions of classical linear regression model (CLRM)
namely the error have zero mean, heteroscedasticity, autocorrelation, normality and
multicollinearity. Relay on Brooks (2008), the first assumption required is that the average
value of the errors is zero. In fact, if a constant term is included in the regression equation,
this assumption will never be violated. Hence, study’s regression model has included a
constant term, so that this assumption was not violated.
II.
Test for heteroskedasticity assumption (var(ut ) = σ2 <∞)
As indicated by Brooks (2008), this assumption requires that the variance of the errors to be
constant. If the errors do not have a constant variance, it is said that the assumption of
homoscedasticity has been violated. This violation is termed as heteroscedasticity. In this
study test was used to test for existence of heteroscedasticity across the range of explanatory
variables.
41
Table 4.4 Heteroskedasticity Test
Heteroskedasticity Test: White
F-statistic
1.240639
Prob. F(5,54)
0.3031
Obs*R-squared
6.182260
Prob. Chi-Square(5)
0.2889
Scaled explained SS
5.569338
Prob. Chi-Square(5)
0.3504
Source: - annual report of sample commercial banks computed using E-views 8
In this case, both the F- statistic and R-squared versions of the test statistic give the same
conclusion that there is no evidence for the presence of heteroscedasticity, since the p-values
are considerably in excess of 0.05 and also the third version of the test statistic, ‘Scaled
explained SS’, which as the name suggests is based on a normalized version of the explained
sum of squares from the auxiliary regression, suggests also that there is no evidence of
heteroscedasticity. Thus, there is no evidence of heteroscedasticity. Thus, the conclusion of
the test has shown that no evidence of heteroscedasticity.
III.
Test for autocorrelation assumption (cov(ui, uj ) = 0 for i ≠ j
This assumption stated that the covariance between the error terms over time (or cross
sectionals, for that type of data) is zero. In other words, it is assumed that the errors are
uncorrelated with one another. If the errors are not uncorrelated with one another, it would be
stated that they are auto correlated or that they are serially correlated (Brooks, 2008).
The Durbin-Watson test statistic of 1.617 is close to two, so that there is no evidence for the
presence of autocorrelation.
Another test for the existence of autocorrelation is by using Breusch-Godfrey test.
Table 4.5 Breusch-Godfrey Serial Correlation LM Test
Breusch-Godfrey Serial Correlation LM Test:
F-statistic
2.341503
Prob. F(2,52)
0.1062
Obs*R-squared
4.957047
Prob. Chi-Square(2)
0.0839
Source: - annual report of sample commercial banks computed using E-views 8
Both versions of the test; F- statistic and R-squared version of the test indicate that the null
hypothesis of no autocorrelation should not be rejected, since the p-values are considerably in
42
excess of 0.05. The conclusion from both versions of the test described that there is no
autocorrelation.
IV. Test of normality (ut∼
∼N(0, σ2)
As stated by Brooks (2008), if the residuals are normally distributed, the histogram should be
bell-shaped and the Bera-Jarque statistic would be significant. This means that JarqueBera
formalizes this by testing the residuals for normality and testing whether the coefficient of
skeweness and kurtosis are ≈ 0 and ≈ 3 respectively. Normality assumption of the regression
model can be tested with the Jarque- Bera measure. Skewness measures the extent to which a
distribution is not symmetric about its mean value and kurtosis measures how it is fat the tails
of the distribution. If the JarqueBera value is greater than 0.05, it’s an indicator for the
presence of normality (Brooks, 2008).
In addition, it is quite often the case that one very extreme residuals cause a rejection of the
normality assumption. Such observations would appear in the tails of the distribution, which
enters into the definition of kurtosis, to be very large. Such observations that do not fit in with
the pattern of the remainder of the data are known as outliers. If this is the case, one way to
improve the chances of error normality is to use dummy variables (Brooks, 2008). In line
with this, the study included one dummy variable (DUM608) to adjust the normality
distribution. Thus, the figure 4.1 shows the result of normality by including one dummy
variable.
43
Figure 4.1 Normality Test Result
12
Series: Standardized Residuals
Sample 2007 2016
Observations 60
10
8
6
4
2
Mean
Median
Maximum
Minimum
Std. Dev.
Skewness
Kurtosis
7.18e-18
-0.000108
0.015564
-0.016122
0.006418
-0.061458
3.224340
Jarque-Bera
Probability
0.163592
0.921460
0
-0.015
-0.010
-0.005
0.000
0.005
0.010
0.015
Source: - annual report of sample commercial banks computed using E-views 8
The above diagram witnesses that normality assumption holds, i.e., the coefficient of kurtosis
was close to 3, skewness was close to zero and the Bera-Jarque statistic has a P-value of
0.921460 implying that the data were consistent with a normal distribution assumption.
V. Test for multicollinearity
As referred by Brooks (2008), an implicit assumption that is made when using the OLS
estimation method is that the explanatory variables are not correlated with one another. If
there is no relationship between the explanatory variables, they would be said to be
orthogonal to one another. However, a problem occurs when the explanatory variables are
very highly correlated with each other, and this problem is known as multicollinearity.
Malhotra (2007) stated that multicollinearity problems exists when the correlation coefficient
among explanatory variables should be greater than 0.75. However, Brooks (2008) mentioned
that if the correlation coefficient along with the independent variables is 0.8 and above,
multicollinearity problems will be existed.
44
Table 4.6 Correlation Matrix between independent variables
NPLLP
LPTLA
NPLTLA
NPLLP
1
LPTLA
0.187480
1
NPLTLA
-0.338694
-0.106438
1
TLATD
0.091506
0.029683
-0.047537
TLATD
1
Source: - annual report of sample commercial banks computed using E-views 8
The method used in this study to test the existence of multicollinearity was by checking the
Pearson correlation between the independent variables. The correlations between the
independent variables are shown in table 4.6 above. All correlation results are below 0.75,
which indicates that multicollinearity is not a problem for this study.
45
4.4 Analysis of regression
This section presents the empirical findings from the econometric output results on effect of
credit risk on commercial banks performance in Ethiopia. Table 4.7 below shows regression
results between the dependent variable (ROA) and explanatory variables. Under the
following regression outputs the beta coefficient may be negative or positive; beta indicates
that each variable’s level of influence on the dependent variable
Regression result
Empirical model: The empirical model used in the study in order to indentify the effect of
credit risk on commercial banks financial performance:ROA= β0 + β1NPLLP+ β2LPTLA + β3NPLTLA + β4TLATD + ε
Table 4.9 Regression result
Dependent Variable: ROA
Method: Panel Least Squares
Date: 01/05/18 Time: 01:06
Sample: 2007 2016
Periods included: 10
Cross-sections included: 6
Total panel (balanced) observations: 60
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
NPLLP
LPTLA
NPLTLA
TLATD
DUM607
0.029195
-0.012455
-0.100671
-0.071777
0.017800
-0.038167
0.005033
0.004674
0.020471
0.031487
0.006634
0.006881
5.800643
-2.664748
-4.917701
-2.279591
2.683401
-5.546418
0.0000
0.0104
0.0000
0.0270
0.0099
0.0000
Effects Specification
Cross-section fixed (dummy variables)
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
0.700112
0.638910
0.006199
0.001883
225.9373
11.43942
0.000000
Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat
0.025790
0.010317
-7.164576
-6.780613
-7.014387
1.616724
Source: - annual report of sample commercial banks computed using E-views 8
46
This section discusses in detail the analysis of the results for each explanatory variable and
their effect on Ethiopian commercial banks performance. Furthermore, the discussion
analyzed the statistical findings of the study in relation to the previous empirical evidences.
Hence, the following discussions present the interpretation on the fixed effects model
regression results.
P-value indicates at what percentage or precession level of each variable is significant. The
R-squared value measures how well the regression model explains the actual variations in the
dependent variable (Brooks, 2008). R-squared statistics and the adjusted- R squared statistics
of the model was 70.01% and 63.89% respectively. The adjusted R-squared value 63.89%
indicates that the dependent variable of return on asset (ROA) of Ethiopian commercial banks
is well explained by the independent variables that are listed in the model. Thus, these
variables collectively are good explanatory variables to identify the effect of credit risks on
commercial bank financial performance in Ethiopia. The regression F-statistic (11.43942) and
the p-value of zero attached to the test statistic reveal that the null hypothesis that all of the
coefficients are jointly zero should be rejected. Thus, it implies that the independent variables
in the model were able to explain variations in the dependent variable. The coefficient for
NPLLP is -0.012455 on ROA which indicates that the non-performing loan to loan provision
of the commercial banks had negative relationship with ROA and also the relationship is
significant at 5% level of significant. And also, the coefficient for LPTLA is -0.100671on
ROA which refers that non-performing loan to total loan and advance had negative and
significant relation with ROA at 1% level of significant. Next, the coefficient for NPLTLA is
of -0.071777 had negative and significant relation with ROA at 5% level of significant
respectively.
However, TLATD is 0.017800 on ROA, which implies that TLATD had positive relation
with ROA but it is significant at 1% level of significant. The negative relationships indicate
that there is an inverse relationship between the explanatory variables and ROA. Thus,
increasing of those variables will lead to a decrease in ROA of Ethiopian commercial banks.
On the other hand the positive relationships indicate that there is a direct relationship between
the remaining one independent variables and ROA. Decreasing of this variable will led to a
decline in ROA Ethiopian commercial banks.
47
4.4.1 Discussion of results
Non-performing loan to Loan provision
H1: non-performing loan to loan provision has negative and statistically significant effect on
Ethiopia commercial banks performance.
According to the regression result of non-performing loan to loan provision (NPLLP) has a
negative relationship with Ethiopian commercial banks performance by a coefficient estimate
of -0.012455. This means that holding other independent variables constant and when one
percent increases in non-performing loan to loan provision, consequently it reduces return on
asset (ROA) of Ethiopian commercial banks by 1.24% and the p value of non-performing
loan to loan provision (NPLLP) is 0.0104 reveals that it is statistically significant at 5% level
of significance. Accordingly, the result supports the working hypothesis that non-performing
loan to loan provision has negative and statistically significant effect on performance of
commercial banks in Ethiopia for the period of 2007 to 2016. Thus, this outcome is consistent
with prior study of Girma (2011) that point out the negative significant effect of nonperforming loan to loan provision on performance.
As implied from the above finding, non-performing loan to loan provision is adversely
affecting the performance of the commercial banks during the study period in Ethiopia.
Hence, the possible reason is associated with the non-performing loan, customers who is
likely fail to pay their debt to the commercial banks on due date.
Loan Provision to Total Loan and Advance
H2: loan provision to total loan and advance has negative and statistically significant effect
on Ethiopia commercial banks performance
The results of fixed effect model table 4.9 indicated that the relationship between loan
provision to total loan and advance has a negative relationship with Ethiopian commercial
banks performance by a coefficient estimate of -0.100671. This means that holding other
independent variables constant and when one percent increases in total loan and advance to
total deposit, as a result it decreases return on asset (ROA) of Ethiopian commercial banks by
10.06% and the p value of loan provision to total loan and advance is 0.0000 discloses that it
is statistically significant at 1% level of significance
Hence, this significant negative relationship between loan provision and ROA similar to with
the fact which suggest that firms income before excluding provision also affect performance
48
of commercial banks. This finding is not consistent with the literatures Girma (2011). In the
case of Girma (2011) its negative relationship with ROA and is not statically significant at
5% level of significance.
Non-Performing Loan to Total Loan and Advance
H3: Non-performing loan to total loan and advance has negative and statistically significant
effect on Ethiopia commercial banks performance.
The ratio of non-performing loan to total loan and advance, which measures how much a
bank is suffer default in year relative to its gross loan disbursed, is used to measure the effect
of non-performing loans on performance of commercial banks in Ethiopia. According to the
regression result of non-performing loan to total loan and Advance (NPLTLA) has a negative
relationship with Ethiopian commercial banks performance by a coefficient estimate of 0.071777. This means that holding other independent variables constant and when one
percent increases in non-performing loan to total loan and Advance, consequently it
decreases return on asset (ROA) of Ethiopian commercial banks by 7.17% and the p value of
NPLTLA is 0.0270 shows that it is statistically significant at 5% level of significance.
Accordingly, the result supported the working hypothesis that non-performing loan to total
loan and Advance has negative and statistically significant effect on performance of
commercial banks in Ethiopia for the period of 2007 to 2016. This result shows that, high
NPL increase the level of default rate and reduces the performance of commercial banks in
Ethiopia. Hence, the possible reason is that the borrowers fail to pay the loan. This finding is
consistent with the literatures Girma (2011) and Agegnehu (2013) both concluded that there
is a negative relationship between non-performing loan to total loan and advance and
performance of commercial banks in Ethiopia
Total Loan and Advance to Total Deposit
H4: Total loan and advance to total deposit has positive and statistically significant effect on
Ethiopia commercial banks performance.
In accordance with the regression result of total loan and advance to total deposit (TLATD)
has a positive relationship with Ethiopian commercial banks performance by a coefficient
estimate of 0.017800. This means that holding other independent variables constant and when
one percent increases in total loan and advance to total deposit, as a result it increase return
49
on asset (ROA) of Ethiopian commercial banks by 1.798 and the p value of TLATD is 0.0099
discloses that it is statistically significant at 1% level of significance and the result supported
the working hypothesis that Total loan and advance to total deposit has positive and
statistically significant effect on Ethiopia commercial banks performance for the period of
2007 to 2016. Hence, this is the result of the interest rate difference between what the banks
charging on loan and what they actually paying on the deposits. Thus, this outcome is
consistent with prior study of Agegnehu (2013), concluded that there is a positive relationship
between total loan and advance to total deposit and performance of commercial banks in
Ethiopia
Generally this chapter discussed the results of the analysis and then presented the discussions
of these results using the appropriate method. Accordingly, the chapter discussed the
descriptive analysis, correlations between the variables and through the regressions analyses;
it illustrates how the independent variables influence the dependent variable. Thus, a
discussion of the result indicates that non-performing loan, total loan and advance and
provision profit were statistically significant credit risk factors that determine the
performance of banks in Ethiopia. The next chapter presents conclusions and
recommendations of the study.
50
CHAPTER FIVE
1. SUMMARY, CONCULSION AND RECOMMENDATION
The preceding chapter presented the results and discussion, while this chapter deals with
summary, conclusion and recommendations based on the findings of the study. Accordingly
this, chapter is organized into three subsections.
5.1 Summary of findings
The research general objective was to examine the effect of credit risk on financial
performance of commercial banks in Ethiopia. The study used ten (10) years period of time
from 2007-2016 data from six (6) selected commercial banks in Ethiopia. It carried out by
constructing a balanced panel regression model using OLS and as per the Hausman test, fixed
effect model was adopted for secondary data obtained from audited annual report.
The overall result obtained from the regression model indicates that credit risk has an effect
on performance of commercial banks in Ethiopia. The dependent variable used to measure
commercial banks performance was return on asset and independent variables these are nonperforming loan to total loan and advance, loan provision to total loan and advance, nonperforming loan to loan provision and total loan and advance to total deposit in order to attain
the objective of the study.
From the regression result, non-performing loan to total loan and advance, loan provision to
total loan and advance and non-performing loan to loan provision had negative and
significant effect on Ethiopian commercial banks performance. However, total loan and
advance to total deposit had positive and significant effect on Ethiopian commercial banks
performance.
51
5.2 Conclusions
This study aimed to identify the relationship between credit risk and financial performance of
Ethiopian banks. In doing so, previous studies on credit risk have been reviewed and it is
summarized that the performance of bank is usually affected by non-performing and total
loan. Both are originated from bank accounts (balance sheets and/or profit and loss accounts)
and therefore could be termed micro or bank-specific factors for performance included in
asset quality.
To achieve the intended objectives the study has employed quantitative research approach.
The data were collected purposively from a sample of six banks over the time period from
2007-2016. The collected data were analyzed by employing fixed effect model using
econometrics software E-view8.
In order to conduct the empirical analysis, one dependent variable and four independent
variables were selected from prominent previous research works namely non-performing loan
to total loan and advance ratio, total loan and advance to total deposit ratio, non-performing
loan to loan provision ratio and loan provision to total loan and advance ratio. The results of
the fixed effect estimation model showed the existence of the relationship between
performance and four independent variables.
The findings of the study on the effect of credit on financial performance of commercial
banks in Ethiopia for the sample suggest the following conclusions.
According to the regression results, the findings indicated that bank credit risk measured in terms
of Non-performing loan to total loan and advance had statistically significant negative
relationship with ROA, which was in line with previous studies. A negative sign suggests that
banks with high default rate affect performance negatively.
On the other side, the findings indicated that bank credit risk measured in terms of loan to
deposit ratio of banks in this study, the result shows that as there was positive and statistically
significant relationship with ROA. This implies that bank charge more than what the bank
incurring as interest expense for the depositors and the more loan the bank give will have a
significant effect on banks profitability
Next, the findings indicated that bank credit risk measured in terms of non-performing loan to
loan provision shows that negative and statistically significant relationship with ROA. This
implies that the borrower failed to pay.
52
Finally, the result showed a negative relationship between loan provision to total loan and
advance and performance with statistical significance. This implies that commercial banks
income before excluding provision also affects performance of commercial banks.
Generally, the study has developed four hypotheses. Accordingly, all hypotheses of the credit
risk variables were significantly related with bank performance. Therefore, all variables are
accepted and this particular study concludes that most bank financial performances are affect
by credit risk factors.
5.3 Recommendation
In light of the major finding obtained from the results, the following recommendations are
made.
Banks non-performing loan to total loan and advance ratio, ratio of loan Provision to total
loan and advance, ratio of non-performing loan to loan provision and ratio of total loan &
advances to total deposit are significant key credit risk drivers of performance of
commercials banks in Ethiopia.
Therefore, banks should pay greater attention to these significant variables in
determining their credit risk management policy. By establishing standards and
overall objectives to reduce the level of credit exposures
The study also shows that, performance of banks in Ethiopia mainly negatively
influenced by non-performing loan and loan provision. Therefore, it is
recommendable for banks need for strong credit risk and loan service process
management must be adopted to keep the level of non-performing loan and loan
provision as low as possible which will enable to maintain high performance of
commercial banks in Ethiopia.
Finally, by improving the loan to deposit ratio the bank should maximize the profit. In
addition the loan also should be given up to some stage where the liquidity risk does not
exist.
.
53
5.4 Further Research Consideration
This study only considered the effect of credit risk on financial performance of commercial
banks in Ethiopia. However, it is recommendable for potential researchers to further asses
other factors of credit risk that can affect firm’s performance by incorporating additional
bank specific, macro-economic factors and qualitative data.
54
REFERENCE
Adams, R., & Mehran, H.(2004). Is corporate governance different for bank holding
companies? Federal Reserve Bank of New York Economic Policy review pp 123-142
Agegnehu (2013) Credit risk management and the performance of Ethiopian commercial
banks. (Master thesis) Mekele University
Analyzing and managing banking risk second edition by Hennie van Greuning Sonja
Brajovic Bratanovic, April 2003. World Bank Publications
Athanasoglou, P., Brissimis, S N., and Delis, M D. (2005), ‘‘Bank-specific, industry specific
and macroeconomic determinants of bank profitability’’ MPRA Paper No.153
Basel committee on banking supervision (1999) “principles for the management of credit
risk” Basel committee publication
Basel committee on banking supervision (2000) “best practice for credit risk” Basel
committee publication
Basel committee on banking supervision (2001) “sound practices for the management of
operation risk” Basel committee publication
Basel committee on banking supervision “bank of international settlements” revised
framework 2004.
Bessis, J. (2003), Risk Management in Banking, John Wiley & Sons, Chichester, New York
2nd edition, 272-278.
Berger, A. N., & DeYoung, R.(1997)., Problem loans and cost efficiency in commercial
banks. Journal of Banking and Finance, 21(6), 849-870
Bikker, J.A., Hu, H (2002) cyclical patterns in profit, provisioning and lending of banks
procyclicality of the new basel capital requirements.
Boateng, G (2004). Credit risk management in banks: the case of scandinaviska enskilida
banking university of Skovde, Stockholm
Brooks, C. (2008). Introductory Econometrics for Finance, 2nd edition, Cambridge
University Press, New York
55
Charlittesville (2002) V.A. Association for investment Management Research,
Coyle B. (2000), frame work for credit risk management, mc Graw-hill companies
Culp, C. L. & Neves, A. M. P. (1998) Credit and Interest Rate Risk in the Business of
Banking Derivatives Quarterly, 4(4), 19-35.
Demirguc-Kunt, A., Huizinga, H, (1998) ‘‘Determinants of commercial bank interest margins
and profitability: some international evidence’’ World Bank Economic Review 13
Epure and Lafuente (2012) studied the impact of risk on the performance of banks in costRican Banking industry during 1998-2007
Fotoh, D.M (2005) commercial bank management and profitability, the case of Cameroon
bank, university of Skovde, Stockholm
Girma(2011) Credit Risk Management and Its Impact on Performance on Ethiopian
commercial Banks. (Master thesis) Addis Ababa University
Hailu (2016) the Impact of Credit Risk on Profitability Performance in Selected Public and
Private Commercial Banks of Ethiopia. (Master’s thesis) rift valley university
Haim, Levy, & Post,T. (2005). Investments, P 554, book on investment offers a topical and
thorough introduction to securities, securities market and investment strategies.
Hempel, G., Coleman, A. & Simon, D. (1986), Bank management text and cases, Fourth
Edition, John Wiley & Son (Asia) Pte Ltd, New York, 150-175.
Hennie, V. G., (2003).Analyzing and Managing Banking Risk: A Framework for Assessing
Corporate Governance and Financial Risk, 2ND edition. Washington DC: World
Bank Publications.
Horcher, K. A., (2005). Essentials of Financial Risk Management Hoboken: John Wiley &
Sons, Incorporated.
Kargi, h.s (2011) Credit risk and the performance of nigerian banks, Ahmadubello
University, zaria.
Kithinji (2010) the prevailing relationship between profitability and credit risk in Kenya
banks from 2004 – 2008
56
Kosmidou (2007), factors influencing the profitability of domestic and foreign commercial
banks in European Union
Kunt-Demirguc, D., Detragiache, E. &Ouarada, M. (2013) vol. 45, Bank Capital Lessons
from the Financial Crisis, Journal of Money, Credit and Banking,
Laker A, (2007), Bank Supervision Annual Report, Los Angeles Lakers community
relationship department, Los Angeles
Machiraju (2004) assert that banks have traditionally focused on the principles of five Cs to
estimate borrowers and Modern commercial banking. Vikas publication House PVT
Matyszak, P. (2007). Ancient Rome on five denarri a day, New York Thames & Hudson
publication
McMenamin, J., (1999), An Introduction to Credit Risk Management, John Wiley & Son
(Asia) Pte Ltd, New York, 100-112.
Mueller, H. P. (1998). Credit risk policy: the anchor of credit culture. Journal of Commercial
Lending, Vol. 78
MwaiKarugu (2015) the study sought to analyze the effect of credit risk management
practices on profitability of listed commercial banks at Nairobi Kenya.
National bank of Ethiopia “Bank Risk Management Guidelines (Revised)” issued in 2003.
Okorie, A. (1998) management of risk and default in agricultural lending in ijere,
Pandey, I.M (2004), Financial Management, Vikas Publishers House, New Delhi, 2 (7), 108150.
Pasiouras et al. (2007) factors influencing the profitability of domestic and foreign
commercial banks in European Union
Payle D (1997), “Bank Risk Management“, Working Paper NO.272, University of RBA
Rachev S. and Schwartz E. (2000), modeling of credit risk, research paper, university of
Karlsruhe.
Ravi Prakash Sharma Poudel(2012) article on The impact of credit risk management on
financial performance of commercial banks In Nepal
57
Richard S. (2010) assessment of the credit risk management practice of kakum rural bank
limited, university of cape coast
Rosemary (2011) the effect of credit management on the financial performance of
microfinance institutions in Kenya (MBA) School of business, university of Nairobi
R.S. Raghavan (2003), Article of Risk Management in Banks. The chartered accountant
Sandstorm A. (2009); “political risk in credit evaluation” World Bank group
Saunders, D. & Cornett, A. (2007), Financial Markets and Institutions: An Introduction to the
Risk management Approach, McGraw Hill Pub, 1(2), 36-49
Schonbucher (2000 assert that banks have traditionally focused on the principles of five Cs to
estimate borrowers’
Sinkey, J. F. (1998) Commercial bank financial management, 6th edition, prentice hall USA
Smirlock, M (1985), evidence on non relationship concentration and profitability in banking.
Journal of money, credit, and banking 17(1): 69-83
Tibebu (2011), the relationship between credit risk and performance of commercial banks in
Ethiopia, Addis Ababa University, College of business and economics
Ugirase Josiane Magnifique (2013) examined the effect of credit risk management on the
financial performance of commercial banks in Rwanda (MBA) University of Nairobi
Vong et al (2009) determinants of banks performance in Macao
Wilson, J. S. G. (1998). Bank asset and liability management, Magazine of Bank
Administration, p.20-22
Wheehem, T. L & Hunger, D. J (2008) Strategic management and business policy, New
Jersey; Pearson education Inc
Yuqi Li, Determinants of Banks. Profitability and its Implication on Risk Management
Practices: Panel Evidence from the UK in the Period 1999-2006
58
59
Appendix 1:-Descriptive Analysis
ROA
NPLLP
LPTLA
NPLTLA
TLATD
Mean
0.025790
0.646007
0.030677
0.026548
0.577008
Median
0.027515
0.644215
0.021800
0.020400
0.563130
Maximum
0.044700
0.975500
0.315350
0.198540
0.901820
Minimum
-0.0188
0.121490
0.002440
0.002742
0.297430
Std. Dev.
0.010317
0.208977
0.042687
0.029111
0.133128
Skewness
-1.773881
-0.304644
5.379252
4.081153
0.575580
Kurtosis
7.897890
2.454215
34.94681
22.84512
3.031872
Jarque-Bera
91.43987
1.672780
2840.861
1151.130
3.315459
Probability
0.000000
0.433272
0.000000
0.000000
0.190571
Sum
1.547414
38.76039
1.840592
1.592891
34.62045
Sum Sq. Dev.
0.006280
2.576616
0.107510
0.050001
1.045667
Observations
60
60
60
60
60
Appendix 2:- Correlation Analysis
ROA
NPLLP
LPTLA
NPLTLA
TLATD
ROA
1.000000
-0.22763
-0.33001
-0.2831
0.267086
NPLLP
-0.227633
1.000000
0.187480
-0.33869
0.091506
LPTLA
-0.330013
0.187480
1.000000
-0.10644
0.029683
NPLTLA
-0.283095
-0.33869
-0.10644
1.000000
-0.04754
TLATD
0.267086
0.091506
0.029683
-0.04754
1.000000
60
Appendix 3:-Hausman Test
Correlated Random Effects - Hausman Test
Equation: Untitled
Test cross-section random effects
Test Summary
Chi-Sq. Statistic
Chi-Sq. d.f.
Prob.
13.831540
4
0.0079
Cross-section random
** WARNING: estimated cross-section random effects variance is zero.
Cross-section random effects test comparisons:
Variable
LPTLA
NPLLP
NPLTLA
TLATD
Fixed
-0.096164
-0.019725
-0.108454
0.018013
Random
Var(Diff.)
Prob.
-0.077361
-0.016546
-0.147788
0.022275
0.000075
0.000004
0.000123
0.000011
0.0300
0.1273
0.0004
0.1996
Cross-section random effects test equation:
Dependent Variable: ROA
Method: Panel Least Squares
Date: 01/05/18 Time: 22:16
Sample: 2007 2016
Periods included: 10
Cross-sections included: 6
Total panel (balanced) observations: 60
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
LPTLA
NPLLP
NPLTLA
TLATD
0.033969
-0.096164
-0.019725
-0.108454
0.018013
0.006263
0.025835
0.005667
0.038882
0.008378
5.423328
-3.722193
-3.480826
-2.789296
2.149923
0.0000
0.0005
0.0010
0.0075
0.0364
Effects Specification
Cross-section fixed (dummy variables)
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
0.511838
0.423969
0.007830
0.003066
211.3202
5.825013
0.000016
Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat
61
0.025790
0.010317
-6.710673
-6.361616
-6.574138
1.384258
Appendix 4:-Test of Heteroskedasticity
Heteroskedasticity Test: White
F-statistic
Obs*R-squared
Scaled explained SS
1.240639
6.182260
5.569338
Prob. F(5,54)
Prob. Chi-Square(5)
Prob. Chi-Square(5)
0.3031
0.2889
0.3504
Test Equation:
Dependent Variable: RESID^2
Method: Least Squares
Date: 01/05/18 Time: 00:53
Sample: 1 60
Included observations: 60
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
NPLLP^2
LPTLA^2
NPLTLA^2
TLATD^2
DUM607^2
6.46E-05
3.28E-05
-0.000276
0.000531
-0.000110
-4.97E-05
2.22E-05
3.24E-05
0.000617
0.001541
4.83E-05
6.18E-05
2.910658
1.013273
-0.447255
0.344371
-2.275512
-0.804241
0.0052
0.3154
0.6565
0.7319
0.0269
0.4248
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
0.103038
0.019986
6.03E-05
1.96E-07
500.9838
1.240639
0.303122
Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat
62
4.05E-05
6.09E-05
-16.49946
-16.29003
-16.41754
2.051263
Appendix 5: - Test of autocorrelation
Breusch-Godfrey Serial Correlation LM Test:
F-statistic
Obs*R-squared
2.341503
4.957047
Prob. F(2,52)
Prob. Chi-Square(2)
0.1062
0.0839
Test Equation:
Dependent Variable: RESID
Method: Least Squares
Date: 01/05/18 Time: 00:55
Sample: 1 60
Included observations: 60
Presample missing value lagged residuals set to zero.
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
NPLLP
LPTLA
NPLTLA
TLATD
DUM607
RESID(-1)
RESID(-2)
0.000363
-0.000150
-0.004544
0.011194
-0.000854
0.004085
0.283478
0.071083
0.004879
0.004506
0.020588
0.032945
0.006492
0.007283
0.148717
0.140998
0.074313
-0.033184
-0.220709
0.339773
-0.131544
0.560888
1.906153
0.504140
0.9410
0.9737
0.8262
0.7354
0.8959
0.5773
0.0622
0.6163
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
0.082617
-0.040876
0.006548
0.002230
220.8715
0.669001
0.697120
Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat
63
6.14E-18
0.006418
-7.095715
-6.816469
-6.986487
1.873766
Appendix 6:-Normality
12
Series: Standardized Residuals
Sample 2007 2016
Observations 60
10
8
6
4
2
Mean
Median
Maximum
Minimum
Std. Dev.
Skewness
Kurtosis
7.18e-18
-0.000108
0.015564
-0.016122
0.006418
-0.061458
3.224340
Jarque-Bera
Probability
0.163592
0.921460
0
-0.015
-0.010
-0.005
0.000
0.005
0.010
0.015
Appendix 7:- Multicollinearity
NPLLP
LPTLA
NPLTLA
TLATD
NPLLP
1.000000
0.187480
-0.338694
0.091506
LPTLA
0.187480
1.000000
-0.106438
0.029683
NPLTLA
-0.338694
-0.106438
1.000000
-0.047537
TLATD
0.091506
0.029683
-0.047537
1.000000
64
Appendix 7:- Regression result
Dependent Variable: ROA
Method: Panel Least Squares
Date: 01/05/18 Time: 01:06
Sample: 2007 2016
Periods included: 10
Cross-sections included: 6
Total panel (balanced) observations: 60
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
NPLLP
LPTLA
NPLTLA
TLATD
DUM607
0.029195
-0.012455
-0.100671
-0.071777
0.017800
-0.038167
0.005033
0.004674
0.020471
0.031487
0.006634
0.006881
5.800643
-2.664748
-4.917701
-2.279591
2.683401
-5.546418
0.0000
0.0104
0.0000
0.0270
0.0099
0.0000
Effects Specification
Cross-section fixed (dummy variables)
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
0.700112
0.638910
0.006199
0.001883
225.9373
11.43942
0.000000
Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat
65
0.025790
0.010317
-7.164576
-6.780613
-7.014387
1.616724