International Journal of Accounting and

Financial Management Research (IJAFMR)

ISSN(P): 2249-6882; ISSN(E): 2249-7994
Vol. 5, Issue 3, Jun 2015, 1-14
TJPRC Pvt. Ltd.

DETERMINANTS OF CREDIT RISK IN ETHIOPIAN PRIVATE

COMMERCIAL BANKS
ATAKELT HAILU ASFAW & P. VENI
1

Research Scholar, Department of Commerce and Management Studies, Andhra University, Visakhapatnam,
Andhra Pradesh, India
2

Professor, Department of Commerce and Management Studies, Andhra University, Visakhapatnam,

Andhra Pradesh, India

ABSTRACT
Optimal portfolio diversifications, establishing a comprehensive credit limit and loan pricing system as well as
Credit risk management strategy, policy and procedures without a clear picture of Credit risk drivers is just considered as
putting money on fire. Therefore, analyzing the link between the bank specific factors and credit risk indictor is
indispensably required using a panel data set over the period of 2006-2012. The three Panel data estimation method, pooled
OLS regression, fixed effect and random effect model, were used for extracting good result and F-test ascertained the
appropriateness of Pooled OLS regression model. Its result revealed that the credit growth and return on equity had
statistically significant negative impact on Credit risk indicator of the large Ethiopian private commercial banks. However,
inefficiency, and deposit rate had statistically insignificant positive influence on the Credit risk indicator. It means that
inefficient bank as well as those Banks that charge high deposit rate is likely to incur higher problem loan.

KEYWORDS: Loan Pricing System, OLS Regression, Credit Risk, F-Test

INTRODUCTION
There are many economic and financial reasons for justifying this study. First, Credit risk is the king of all risks.
Many authors ranked the types of risk in terms of importance for the banks and Credit risk got the first rank
((Atakelt & Veni, 2015) (Hussain & Al-Ajmi, 2012) (Alam & Masukujjaman, 2011)). Therefore, investigating the
determinants of Credit risk directly or indirectly means obtaining solution for a major problem of the banks.
Second, financial system of Ethiopia has shown progress since the financial reform of 1994. New and complex
financial and economic police, regulatory measure and directive have been evolving gradually. As a result, it is inevitable
to face multiple problems related to economic and financial policy measure that directly or indirectly influence the
Ethiopian banking business. Within this change, the issue of credit risk management did not receive its due attention in the
literature especially in developing country like Ethiopian. Therefore, this study serves as a benchmark for policy makers,
bankers and others on aspects of banking environments.
Third, the linkage between financial crisis or banking failure and Credit risk management system has also been
justified by many authors. For instance, Failure of practicing effective Credit risk management system was considered as
one of the main cause of financial crises in general and banking failure in particularly (( (Hussain & Al-Ajmi, 2012), ( AlTamimi & Al-Mazrooei, 2007) ,Van Gestel 2009), Saunders , 2008) , Fight A. (2003), Suresh P. (2010)). Practicing
www.tjprc.org

editor@tjprc.org

Atakelt Hailu Asfaw & P. Veni

effective Credit risk management system was a prescription given by several authors to minimizes banking problems and
failure ((Hull, 2007), (De Juan, 2004), ( Marrison, 2002) (Kolb, 1992)).
Understanding the impact or determinants of credit risk is tantamount with understanding the cause of diseases
under which treatment and prevention measures proceed easily. When determinants of credit risk are assessed properly, it
is possible to minimize the level of NPL, Credit loss, banking failure and crises.
Assessing the determinants of credit risk is the cornerstone for the effectiveness of risk management system and
practice. Optimal portfolio diversifications, establishing a comprehensive credit limit system and loan pricing, as well as
Credit risk management strategy, policy and procedures without a clear picture of credit risk drivers considered just like
driving a car without having a break and knowing final destination. Therefore, the success and survival of commercial
banks is greatly depending on effective Credit risk management system and practice (Atakelt & Veni, 2015).
Macroeconomic environment instability (specially increasing the level of unemployment, interest rate and
inflation) with bank specific factors (aggressive lending, liquidity problem, management quality, asset quality and capital
adequacy) may influence bank's performance and lead to insolvency and then finally bring the bank crisis
((Festic M. et al, 2011), ( Saunders & Cornett, 2003)). Banks are unable to meet their current financial obligation when
borrowers fail to repay their loan amount. As a result, it may face liquidity problems that in turn force the bank to sell its
asset less than what it worth in normal business condition and then finally may lead to insolvency.
Banks asset quality has been a major concern for a regulatory body. Problems related to Asset quality have a
multiplier effect that directly or indirectly flamed by both micro and macro-economic environment. Therefore, proper
assessment of banking environment means reducing adverse effect of banking business and solving the credit problem as
well as identifying the cause and consequence of whole banking and financial system.
Due to the fact mentioned above, Banks should give more emphasis on determinant of credit risk to survive while
doing their credit business. Study on the determinants of Credit risk is indispensably required to balance the contradicted
objective of the bank (liquidity, safety and profitability).
The safety, soundness, solvency and profitability of banking system greatly depend on macro and micro economic
condition. This study assumes that the level of asset quality depends on the fluctuation of bank specific economic
environments. Generally, this study focus on the bank level determinants of Credit risk, such as Credit growth,
profitability, bank size, inefficiency, diversification and deposit rate, in large private commercial banks of Ethiopia using
panel data model over the period of 2006- 2012.

LITERATURE REVIEW
There are several studies related to the determinants of credit risk at macroeconomic and micro level of banks1.
Researcher tried to summarize the main finding of some selected studies mainly on the area of macroeconomic and banks
specific determinants of credit risk.
Awojobi & Amel (2011) employed panel data for analysis the determinants of Credit risk efficiency of Nigerian
banking industry. Capital adequacy, proxy for Credit risk efficiency, was the independent variable while bank specific
determinants: Credit risk (total loan over the asset), insolvency risk (current asset over current liability), Interest sensitivity

Impact Factor (JCC): 4.8567

Index Copernicus Value (ICV): 3.0

Determinants of Credit Risk in

Ethiopian Private Commercial Banks

ratio, market risk, management quality, ROA and bank size and macroeconomic determinants: growth and inflation, were
used as explanatory varibles. Researchers found that Credit risk, insolvency risk, market risk, bank size and economic
growth had a positive influence on credit risk efficiency. However, management quality and inflation had a negative
impact on credit risk efficiency.
Abdullah, A. et al (2012) conducted research using Johansens co-integration test to assess the long-term relationship
between Credit risk and bank specific factors. Researchers found that Bank size had a positive and significant relationship
with credit risk in domestic banks. Liquid assets and credit risk had negative and significant in foreign banks.
Ganic, M. (2012) conducted research on Bank Specific Determinants of Credit Risk in the Banking Sector of
Bosnia and Herzegovina using the panel regression model and found that inefficiency and credit growth had a significant
negative influence on credit risk while ROE and deposit rate had significant positive impact on credit risk. However,
capital adequacy, liquidity, market power, ROA and reserve ratio had an insignificant impact on credit risk.
The impact of an economic condition on borrowers credibility or credit quality was widely evidenced in the
literature. Several authors found that a favorable economic condition reduce the level of Nonperforming loan. Thiagarajan,
S. et al (2011), Das and Ghosh (2007), Zribi N.& Boujelbene Y. (2011), Fainstein G. (2011), Salas and Saurina (2002),
Castro V. 2013) found that a significant negative relationship between GDP growth and the level of nonperforming loan.
Prakash & Poudel (2013) conducted research on Macroeconomic Determinants of Credit Risk in Nepalese
Banking Industry and found that inflation and foreign exchange rate influence credit risk negatively while GDP growth,
growth of Broad Money Supply and Market Interest Rate failed to influence credit risk in the Nepalese banking industry.
Many authors also strongly link the loan problem with macroeconomic variables2.
Several studies conducted in the area of macro-economic and bank specific determinants of credit risk. For
instance, Abdullah, A., et al (2012), Awojobi & Amel (2011) , Ahmad & Bashir (2013) , Aman & Zaman (2010) , Castro
(2013) , Bucur & Dramgoirescu (2014) , Fainstein G. (2011) , Zribi & Boujelbene, (2011) were some of the studies.
For instance, Thiagarajan, S. at el (2011), Das and Ghosh (2007), Zribi N.& Boujelbene Y. (2011), Fainstein G.
(2011) , Salas and Saurina (2002), Castro V. 2013) found that a significant negative relationship between GDP growth and
the level of nonperforming loan.
It manifested from the above literature that the level of asset quality is influenced by several macro and micro
economic factors. Generally, Bank size, deposit rate, inefficiency, diversification, profitability, credit growth and capital
adequacy indicators are important bank specific factors that were mostly employed in the study related to Credit risk
determinant while GDP, inflation, exchange rate, interest rate, money supply, and unemployment rate are some of widely
employed macroeconomic determinants of credit risk.
The main objective of this study is to find out bank specific Credit risk determinants, Credit growth, profitability,
bank size, inefficiency, diversification and deposit rate, in large private commercial banks of Ethiopia using panel data
model over the period of 2006 to 2012.

www.tjprc.org

editor@tjprc.org

Atakelt Hailu Asfaw & P. Veni

RESEARCH QUESTION AND HYPOTHESES

The aim of This study is to verify the hypothesis whether the bank specific factors such as ROE, Credit growth,
bank size, deposit rate, diversification and inefficiency have an influence on the level of asset quality (credit risk) in
Ethiopian private commercial banks using panel data model over the period of 2006 to 2012.
Main research question is that how the bank specific factors affect credit risk in Ethiopian private commercial
banks. Based on this objective and research questions, researcher tried to test whether bank specific factors have significant
influence on the credit risk of Ethiopian commercial banks.
H1: There is a significant positive correlation between each Bank specific determinant and Credit risk.

METHODOLOGY
The mean intention of the researcher is to find out the link between bank specific determinants of credit risk and
credit risk indicator of the private commercial bank in Ethiopia. Panel data model was used in order to achieve this
objective. It is obvious that research should be done mainly based on previous literature. The majority of reviewed
previous researches used panel data model to analyze the determinants of credit risk (Awojobi & Amel , 2011) , (Ganic,
M., 2012), (Prakash & Poudel (2013), (Zribi & Boujelbene , 2011), (Misman, F., 2012 ) ((Das & Ghosh, 2007),
(Thiagarajan, S. et al., 2011), (Jimenez & Saurina, 2006), (Misman, F. , 2012), (Castro V., 2013)).
Totally, 19 commercial banks, where 17 private commercial banks and two public commercial banks, are
currently operating in the Ethiopian banking system. Based on the classification of NBE, There were six big private
commercial banks out of 17 private commercial banks operating in the country during the study period. Population of the
study was the six large private commercial banks.
In ordered to take balanced secondary data, from six large private commercial banks, the six-year periods of panel
data set starting from 2007 to 2012 were taken. Bank of Abyssinia, Wegagan bank, United bank, Awash international bank,
Dashin bank and Nib international bank were the six large private commercial banks based on their asset size3.
Based on asset size, NBE classified commercial banks in to three : large commercial banks for those commercial
banks with total asset greater than Birr 9 billion and medium commercial banks with total asset size between birr 3 billion
and birr 9 billion and finally small commercial banks with asset size less than birr 3 billion. Therefore, this study excludes
public commercial bank as well as small and medium private commercial banks.
Specification of Model and Variables
A panel data model where both feature of cross sectional and time series panel data set are taken in to
consideration in the estimation of the parameters of the study. It deals with multiple observations of the same entities over
multiple period of time. The general function of the Model of the study is written as follows:
Credit risk indicator= (bank specific determinants of Credit risk)

Symbolically:
Yt= + 1x1t + 2 x2t + . + kxkt +Uit...

Impact Factor (JCC): 4.8567

...........2

Index Copernicus Value (ICV): 3.0

Determinants of Credit Risk in

Ethiopian Private Commercial Banks

Where:
i= 1, 2, 3..

....6: six large private commercial banks

t= 1, 2, 3.

..6: six years: 2007-2012

yit is stands for credit risk indicator for the bank i at timet period. Here i represent for the ith cross-sectional
unit while t stand for the tth time period.

and are vector of the parameters

Xkt is the values of the kth independent variables for bank i at a time t:

Ut is the error term

With the variables of the study:
CRt = + 1CGt + 2ROEt + 3DRt + 4INFt + 5DVt + 6BSt +Ut..3
Where
CR- stands for the Credit risk ratio, DR: Deposit rate, CG: Credit growth rate ROE: Return on equity, INE:

Inefficiency, DV: Diversification and BS: Bank size

Variable of the Study
There are multiple factors that affect Credit risk management activities of the banks. For instance, profitability,
loan growth, Bank size, deposit and lending rate, branch expansion, efficiency, level of diversification, capital adequacy,
level of liquidity and ownership structure were some of the bank specific Credit risk determinants mostly reported in
literatures4.
Is no single unique variable that indicate the level of credit risk and being considered as a proxy for the credit risk
indicator (dependent variable). Different authors used different asset quality measure as an indicator of credit risk5. it is
depand on the the level of disclosuring asset quality data for external users. It is difficult to get data related to NPL from
most commercial banks and regulatory body due to its confidentiality. Therefore, the choice of the variable, credit risk
indicator greatly depends on the availability of data related to asset quality.
Different financial ratios were used to indicate the level of credit risk (asset quality) in several studies and books
as well as banks report. Financial ratio such as the ratio of NPL to total loan, loan to total asset, risk-weighted assets to
total assets, loan loss reserve to total loans, loan losses to total loan, loan loss provison to total loans and Provision for loan
losses to total assets as well as total loan to total deposit and total loan to equity were mostly used as a proxy for credit risk
in several credit risk determinants related literature. Generally, the ratio of nonperforming loan to total loan was considered

See Ganic, M. (2012), Zribi & Boujelbene (2011). Thiagarajan S. et al (2011), Das & Ghosh (2007), Swamy V.( 2012),
Misman, F. (2012), Meyer and Yeager (2001), Zribi & Boujelbene (2011)
5
In most books and studies, ratio of Provision for loan losses over total operating income, Provision for loan losses over
total loans, Provision for loan losses over total assets and Non-Performing loans over total loans are an indicator of asset
quality (Casu B. et al. ,2006, P. 214). Almost all banks are using such parameters as standard to determine their asset
quality.
www.tjprc.org

editor@tjprc.org

Atakelt Hailu Asfaw & P. Veni

as the Crisk indicator in most literatures6.

Fainstein, G. (2011) , Thiagarajan S. et al (2011), Prakash & Poudel (2013) , Ganic, M. (2012), Castro V.(2013)
Das & Ghosh (2007), Swamy V.( 2012) Misman, F. (2012), Meyer and Yeager (2001) used the Ratio of non-performing
loans to total loan as a proxy for credit risk. Zribi & Boujelbene (2011) used the ratio of risk-weighted assets to total assets
as a measure of bank credit risk. Rama mohana rao K. and Tekeste B. (2013) used loan loss reserve to total loans was used
as a proxy for credit risk. In this paper the dependent variable (credit risk indicator) is defined as the allocation of loan loss
provision relative to total loans and advance.
Credit growth, profitability, Bank size, efficiency, diversification and deposit rate were used as the banks specific
Credit risk determinants in this study while the ratio of loan loss provision to total loans and advance was used as a proxy
for the credit risk indicator (dependent variable). The variable of the study are clearly described belew.
Deposit Rate
Deposit rate is measured as the ratio of interest expense to total deposit. It is one of the determinants of
nonperforming loan due to its influentiality in pricing loans. Increment of cost of fund enforces the bank to charge high
lending rate that in turn influences the repayment capacity of borrowers.
Today in complex banking environment, banks make competition through different ways. Charging attractive
interest rate is one among others. Charging high deposit rate has multiple impacts on credit risk. First, offering high deposit
rate directly influence the lending rate. Landing rate in turn, especially charging high-risk premium for risky borrowers,
have multiple effect on borrowers loan repayment capacity. Therefore, it has positive impact on loan quality since
charging high interest rate means raising borrowing costs, which increase loan burden.
Second, offering high deposit rate encourage depositors or improve the level of deposit mobilization and create an
excess loanable fund. The Deposited money should not be kept idle since the deposit is not free of cost. As a result, it
enforces the bank to increase its lending volume and thus credit risk. Growth in deposit creates excessive loanable fund
that motivated the bank to take credit risk or extend credit to riskier borrowers.
Most authors consider NPL ratio as Credit risk indicators. NPL ratio = amount of NPLs at the end of the
year/gross loan portfolio amount at the end of the year. ( Casu B et al, 2006), (Cousin, 2007), (Fainstein G., 2011)
National Bank of Ethiopia [NBE] ,2009).
Many authors also reported the bond between deposit rate and non-performing loan. For instance, deposit rate and
level of NPL positively correlated ((Ganic, M. (2012), (Ahmad & Bashir, 2013),). Ahmad & Bashir (2003) argued that the
effect of deposit rate on Non-performing loan depending upon the level of risk management system being practiced.
Coefficient of deposit rate is expected to be positive.
Annual Credit Growth Rate
It is obvious that the probability of non-repayment of the loan will increase with the level of credit growth.
Theoretically, the increment of any unit of credit is not without bearing the risk. Creation of an additional unit of credit is
only possible through taking risks. Therefore, there is default risk whenever the banks take risk to extend credit.

Impact Factor (JCC): 4.8567

Index Copernicus Value (ICV): 3.0

Determinants of Credit Risk in

Ethiopian Private Commercial Banks

The impact of Credit growth on nonperforming loan was extensively reported in several literatures. The finding of
Das & Ghosh, (2007), Jimenez & Saurina (2006), Thiagarajan, S., et al (2011), Ahmad & Bashir (2013) ascertained the
positive impact of credit growth on credit risk. However, Ganic, M. (2012) found a negative relationship between credit
risk and credit growth.
Bank Profitability
Bank profitability is usually measured by return on equity, return on assets and net interest margin. Return on
equity is Ratio of Net income to average stockholders' equity. It indicates what the shareholders of the bank earning from
their equity investment. Most of the studies were analyzed using ROE as banks profitability indicator (Abdullah, A. et al.
(2012). Ganic, M. (2012), (Shingjergji, 2013)). MACDoNALD S. and Koch T (2007) note as ROE is vital for performance
analysis specially for indicating long-term sustainability and survival of the bank. Therefore, ROE is one of the vital
measures of bank performance (profitability indicators) and negative sign will be expected on this ratio.
Asset Size
Bank size has become quite fashionable determinant of credit risk recently and measured by log of total asset.
There is no clear-cut evidence that shows the exact relationship between bank size and problem loan. The negative impact
of bank size on credit risk was reported in some literature7 . It is justified that big banks have ability to deal with credit risk
by formulating sound and effective Credit risk management system, introducing modern risk management instrument and
technology as well as a better portfolio diversification opportunity and having Competitive advantage on economies of
scale. On the other hand, there is a big to fail finding stating that bank size and credit risk positively correlated. In this
case, there is a massive mobilization of fund through branch expansion and paying attractive deposit rate. As a result, bank
can extend credit, which exposed the bank to credit risk ( (Das & Ghosh, 2007), (Zribi & Boujelbene , 2011), (Abdullah A.
et al, 2012) and (Misman, F., 2012)). Thirdly, Awojobi & Amel (2011) found that bank size is irrelevant factor on credit
risk. This is due to the common supervision and regulatory measure taken by the apex bank of the country under which all
banks either big or small obey such measure in a similar way.
Inefficiency
In this study, inefficiency is measured by the ratio of operating expenses to total income. Inefficient banks tend to
face high operating expense that flamed by nonperforming loan related costs. Inefficient banks Credit risk management
system and practice characterized with poor credit analysis and credit monitoring system lead to poor credit quality.
Efficient banks have sound and effective Credit strategy, policy and procedure with a strong credit culture that enable to
undertake Credit risk management function properly and reduce operating expense while improving operating income. As
a result, positive coefficient will be expected on this explanatory variable.
Thiagarajan, S. et al (2011), Ganic, M. (2012), Rashid K. et al (2014), Das and Ghosh (2007) used inefficiency in
their study and found that inefficiency was positively related with problem loans. Diversification is also one of the major
determinants of credit risk and measured by ratio of non interest income to total income8.

7
8

For instance, see the finding of Thiagarajan, S. et al (2011). Zribi, N., & Boujelbene, Y. (2011). Das & Ghosh (2007) Misman F. (2012).

Non-interest income includes service charges on deposits, trust fees, advisory fees, servicing fees, net trading profits from
trading books, and commissions and fees from off balance sheet items.

www.tjprc.org

editor@tjprc.org

Atakelt Hailu Asfaw & P. Veni

PROVISION FOR DOUBTFUL LOAN TO TOTAL LOANS AND ADVANCE

Changes in the level of asset quality directly influence the volume of loan loss provision. It is obvious that when
the banks loan loss provision is high, it means high risk associated with credit portfolio and expecting high credit loss. In
other word, the level of asset quality and loan loss provision moves together positively. Therefore, Loan loss provision
clearly shows the level of credit risk since it is determined depends upon the level of asset quality. As a result, the
researcher used ratio of loan provision to a total loan as a proxy for credit risk.

RESULT AND DISCUSSION OF THE STUDY

The description of this study started by giving insight about the credit risk indicators of the Ethiopian commercial
banks and then assessing assumption of the model and analysis of the study finally.
Specific loan provisions indicates the level of credit quality of the bank since it is an expense that the bank set
aside as an allowance against deteriorated loans9. It is one of the vital Credit risk management instruments that build the
banks ability to absorb the credit loss arises from bad loans by allocating capital under Loan loss provision account (
(Atakelt & Veni ,2015), (Van Greuning H. ,2010)). Therefore, it indicates the effectiveness of the Credit risk management
system and practice due to the fact that loss loan provision is determined based on the level of risk factors associated with
borrowers.
Figure 1 shows the trend of Loan loss provision of Ethiopian commercial banks. High capital was allocated on
provision in 2003 and 2012, which means that banks incurred high credit risk during to withstand the adverse credit effect.
Banks loan loss provision increased from 2000 to 2003, then came down until 2009, and finally showed upward
movement contentiously.

Source: National Bank of Ethiopia

Figure 1: Trend of Loan Loss Provision of Commercial Banks
It is necessary to assess the basic assumption of the method before undertaking further analysis. The result of
normality test, the p. value of Kolmoprou-Smimay test for each variable is greater than 0.071, ascertain the normality of
the distribution of the data for each variable.
Several authors have suggested that Multicollinearity problem exists if the correlation between each pair of
independent variable is greater than 0.8 ((Gujarati, N. ,2003,P. 341-375) , (Pallant, J. (2007, P.149) , Yfield, A. (2009, P.
9

In order to make adequate provision for loans, some regulatory bodies classify provision in to two. 1. Bank to allocate
certain percent of their capital on performing loans (standard asset). 2. Specific Loan loss provision where bank allocate
certain percentage of its capital as a provision based on the classification of nonperforming asset (sub standard asset,
doubtful asset and loss asset).

Impact Factor (JCC): 4.8567

Index Copernicus Value (ICV): 3.0

Determinants of Credit Risk in

Ethiopian Private Commercial Banks

233, 242). It can be clearly seen from table 2 that the correlation between each pair of independent variable is less than
0.515. Since the coefficients of all independent variables are less than 0.515, there is no Multicollinearity problem in the
study. Furthermore, Durbin watesn test, with value of 0.9, also shows that the residuals are independents.
Table 2: Correlations Coefficient of Variables
Variables
Correlations Coefficient
CG ROE INF
BS
DR
CR
CG
1
.066 -.023 -.119 .039 -.51**
ROE
1
.346* .515** .014 -.131
INF
1
.368* -.172 -.419*
Pearson Correlation
BS
1
.236 -.116
DR
1
.501
CR
1
Note: * Correlation is significant at the 0.05 level
**. Correlation is significant at the 0.01 level
Source: Own Computation
The result of correlation coefficient shows that all variables are statistically significant and negatively correlated
with the credit risk ratio except deposit rate. Deposit rate has significant positive correlation with credit risk.
Estimation of Panel Data Regression Model
Table 2 shows the estimation result of the three models of the retained variable using Stata 10 (diversification was
excluded due to Mulicollinearity problem). Researcher tried to explore good result by employing the three methods of
estimation10.
The three methods of estimation of linear panel data models (pooled OLS, fixed and random effect model) show
that Credit growth, bank profitability and bank size had a negative influence on credit risk while inefficiency and deposit
rate had a positive influence on credit risk. The two variables, inefficiency and deposit rate, had statistically insignificant
influence on credit risk in any method of estimation. In previous literature, Credit growth, deposit rate, ROE and
inefficiency were positively correlated with Credit risk.
Table 2: Estimation Result of Parameters using Panel Data Model
Variable
CR
CG
ROE
BS
INE
DR
Cons
Model

10

Pooled Regression Model

Coef.
P. value
-.165403
0.003
-.0020505
0.040
-.0655315
0.059
.4195253
0.514
.0248789
0.989
.5114174
0.000
=
+

+
2

1
..+

Fixed Effect
Coef.
P. Value
-.1024854
0.038
-.0014856
0.138
-.072284
-.072284
.9492798
0.285
-.984171
0.539
.5163931
0.000
= + 1
+ 2
..+

it

Random Effect Model

Coef.
P. Value
-.1654
0.001
-0.0020
0.032
-0.0655315
0.050
.4195253
0.509
.0248789
0.989
.5114174
0.000
= + 1
+
2
..+
+
( +
)

Generally, there are three method of estimating panel data model such as pooled model, fixed effect and random effect
(Asteriou & Hall, 2007, P. 345)

www.tjprc.org

editor@tjprc.org

10

Atakelt Hailu Asfaw & P. Veni

The Pooled OLS Regression OF Credit Risk Ratio Function

Even though the pooled OLS model uses data that composed of both time series and cross-section data, it has
some strength and weakness11. One of the basic advantages of the pooled OLS model is that it increases the accuracy of the
estimation due to its possibility of increasing sample size. In other side, it assumes that there are no differences among the
sample banks or all sample banks are assumed to be homogenous, which is an unrealistic assumption (Asteriou & Hall,
2007, P. 345).
In the pooled OLS method of estimation, credit growth and ROE had statistically significant negative impact on
credit risk. However, inefficiency, and deposit rate had statistically insignificant positive influence on the Credit risk ratio.
Furthermore, Bank size had positive but statistically insignificant impact on credit risk.
Statistically significant and negative coefficient of return on equity (ROE) shows an inverse relationship between
profitability and credit risk ratio. This result is not as expected and also inconsistence with previous research. This is not an
exceptional output of this study even other, like Shingjergji, A. (2013), Swamy V.2012) found similar results of
inconsistency. It may be due to the reason that profitable banks have the capacity to implement modern Credit risk
management system, technology and hiring trained man power/experts as well as having credit information access so that it
contributes for minimizing problem loan.
Credit growth had statistically significant negative influence on credit risk. It is inconsistence with the result of
Das & Ghosh, (2007), Jimenez & Saurina (2006), Thiagarajan, S., et al (2011), Ahmad & Bashir (2013) who found a
positive influence of Credit growth on credit risk. This is due to the reason that the banks may develop the best experience
of dealing with borrowers (build the capacity of solving the borrowers problem by giving consultant and other service to
improve their loan repayment), developing strong credit risk culture as well as develop sound Credit risk management
system whenever a problem loan arise due to credit growth.
In a similar manner, Bank size and credit risk are also correlated negatively but not statistically significant. It is
due to the fact that large banks have ability to deal with credit risk by formulating sound and effective Credit risk
management system, introducing modern risk management instruments and adopt new technology as well as a better
portfolio diversification opportunity and gaining competitive advantage on economies of scale so that contribute for
minimizing problem loan.
Operating inefficiency had insignificant positive influence on credit risk. It implies that increasing the level of
inefficiency tends to increase problem loan. There is statistically insignificant positive relationship between deposit rate
and credit risk. This means that increasing the deposit rate directly influences landing rate that create repayment Burdon on
borrowers and influence loan quality.

11

It is just running the OLS regressions by pooling all individual observations together in to one data set and imposing a
common set of parameters across them. Therefore, it confronts the basic strength and weakness of OLS regression model.
increasing accuracy of the estimation due to increment of sample size is one of the advantage of pooled OLS model, and
also Gujarati, N. ,(2003,P. 307) noted that pooled OLS model may improve the relative precision of the estimated
parameters since it include all observation in a regression.

Impact Factor (JCC): 4.8567

Index Copernicus Value (ICV): 3.0

11

Determinants of Credit Risk in

Ethiopian Private Commercial Banks

Fixed Effect Method of Estimation

In the fixed effect method of estimation12, ROE and bank size had statistically insignificant negative impact on
credit risk while inefficiency and deposit rate had positive impact but statistically insignificant. However, Credit growth
had statistically significant negative impact on the credit risk ratio. Finally, credit growth, bank profitability and bank size
had statistically significant negative impact on credit risk while inefficiency and deposit rate had positive impact but
statistically insignificant.
So here, one of the vital questions to be raised is that, which model is an appropriate method of estimation. In
order to choose one, either pooled OLS or fixed effect model, standard F-test can be used13. If the F- statistic value is less
than the critical value, accept th null hypothesis and indicate the appropriateness of the pooled OLS model while fixed
effect model if null hypothesis is rejected. The F value is obtained through the following formula (Gujarati N. 2003, P.643
)
=(

)(

)/( )

The result of F(5,24) value Clearly indicates that the F- value , 0.419 (for 5 numerator df and 24 denominator df),
is insignificant and, therefore, the pooled OLS model seems to be appropriate. In addition to this, the fixed effect model
may not produce such much better estimation result, then pooled OLS model since the degree of heterogeneity of the
sample banks in somewhat reduced (all sample banks are large commercial banks and no difference in ownership
structure). Therefore, they are the same in terms of structure and size. However, there is some degree of heterogeneity, in
managerial skill, number of branch and branch location as well as credit culture/philosophy etc, among sample banks.
The final test, Hausman test, is conducted to check the appropriateness of the random effect model the pooled
OLS method of estimation. If the result of such test is significant, p-value is less than 0.05, pooled OLS model will be
appropriate while random effect model will be preferred if the null hypothesis is rejected (Gujarati, N., 2003, P. 650). The
result of Hausman test attested the superiority of pooled OLS model.(see the table 5).

CONCLUSIONS
Optimal portfolio diversifications, establishing a comprehensive credit limit system and loan pricing, as well as
Credit risk management strategy, policy and procedures without a clear picture of credit risk drivers considered just like
driving a car without having a break and knowing final destination. Therefore, analyzing the link between the bank specific
factors and credit risk indictor is indispensably required in today complex banking environment.
The main objective of this study is to analyze the Credit risk determinants of large Ethiopian private commercial
banks using a panel data set over the period of 2006-2012. Researcher tried to explore good result by employing the three
methods of panel data estimation (pooled OLS, fixed and random effect model) and different test, F-statistical and
Hausman test , were carried out to chose the best result and finally, pooled OLS regression model found to be appropriate.
Credit growth had significant negative impact on problem loans due to the strong and unified credit risk culture,
12

In this model, each individual firm has either its own intercept or slop coefficient, heterogeneity among individual bank
is considered (there is a difference in skill manpower, number of branch, branch location, credit culture/philosophy
among sample banks). Furthermore, it is time-invariant variables and should not be changed with time t.
13
F-statistics is used to check whether pooled OlS or fixed effect model estimation is appropriate (Gujarati, N. D.
(2003)).

www.tjprc.org

editor@tjprc.org

12

Atakelt Hailu Asfaw & P. Veni

developing sound Credit risk management system as well as gaining best experience of dealing with borrowers and
building the capacity of solving the repayment problem.
Bank Profitability indicator had statistically significant negative relationship with credit problem. It is
inconsistence with previous studies and it may be due to the reason that profitable banks have the capacity to implement
modern Credit risk management system, technology and hiring trained man power/experts as well as having credit
information access so that it contributes for minimizing problem loan.
In a similar manner, Bank size and credit risk was also correlated negatively but not statistically significant. It is
due to the fact that large banks have ability to deal with credit risk by formulating sound and effective Credit risk
management system, introducing modern risk management instruments and adopt new technology as well as a better
portfolio diversification opportunity and gaining competitive advantage on economies of scale so that contribute for
minimizing problem loan. Finding further revealed that inefficient bank as well as those Banks that charge high deposit
rate is likely to incur higher problem loan. Finally, it can be concluded that Credit growth and bank profitability had
statistically significant negative impact on problem loan while inefficiency and deposit rate had statistically insignificant
positive impact on problem loans.

REFERENCES
1.

Abdelrahim, K. E. (2013). Effectiveness of Credit Risk Management of Saudi Banks in the Light of Global
Financial Crisis: A Qualitative Study. Asian Transactions on Basic and Applied Sciences , 03 (02), 71-91

2.

Abdullah, A., Khan, A. Q., & Nazir, N. (2012). A comparative study of credit risk management: A case study of
domestic and foreign banks in Pakistan. Academic Research International , 371-377.

3.

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

4.

Alam, M. Z., & Masukujjaman, M. (2011). Risk Management Practices: A Critical Diagnosis of Some Selected
Commercial in Bangladesh. Journal 16 of Business and Technology , 06 (01), 16-35.

5.

Al-Tamim H.A.H., A.-M. F. (2007). Banks risk management: a comparison study of i national and foreign banks.
The journal of risk finance, 394-409.

6.

Aman, Q., & Zaman, K. u. (2010). Credit Risk Performance of Private, State Owned and Foreign Banks on the
Economy of Pakistan. International Research Journal of Finance and Economics (57).

7.

Asteriou, D., & Hall, S. G. (2007). Applied econometrics:A modern approach (2nd edition ed.). Palgrave.

8.

Atakelt, H. A., & Veni, p. (2015). CREDIT RISK MANAGEMENT TOOLS PRACTICED IN ETHIOPIAN
COMMERCIAL BANKS. International Journal of Social Science & Interdisciplinary Research , 4 (5).

9.

Atakelt, H. A., & Veni, P. (2015). Empharical Study on Credit Risk Management Practice of Commercial banks.
Research Journal of Finance and Accounting , 6.

10. Awojobi, O., & Amel, R. (2011). Analysing Risk Management in Banks: Evidence of Bank Efficiency and
Macroeconomic Impact. Journal of Money, Investment and Banking (22), 148-161.

Impact Factor (JCC): 4.8567

Index Copernicus Value (ICV): 3.0

13

Determinants of Credit Risk in

Ethiopian Private Commercial Banks

11. Casu, B., Girardone, C., & Molyneux, P. (2006). INTRODUCTION TO BANKING. British: Pearson Education
Limited.
12. Castro, V. (2013). Macroeconomic determinants of the credit risk in the banking system: The case of the GIPSI.
Economic Modelling , 31.
13. Das, A., & Ghosh, S. (2007). Determinants of Credit Risk in Indian Stateowned Banks: An Empirical
Investigation. MRP , paper no. 17301.
14. Fainstein, G. (2011). The Comparative Analysis of Credit Risk Determinants In the Banking Sector of the Baltic
States. Review of Economics & Finance , 20-45.
15. Festic, M., Kavkler, A., & Repina, S. (2011). The macroeconomic sources of systemic risk in the banking sectors
of five new EU member states. Journal of Banking & Finance .
16. Fight A. (2003). Understanding International Bank Risk, John Wiley & Sons Ltd
17. Ganic, M. (2012). Bank Specific Determinants of Credit Risk An Empirical Study on the Banking Sector of
Bosnia and Herzegovina. International Journal of Economic Practices and Theories , 4 (4).
18. Greuning, H. v., & Bratanovic, S. B. (2009). Analyzing Banking Risk: Framework for Assessing Corporate
Governance and Risk Management. Washington, D.C.: The world bank.
19. Gestel T. and Baesens B., T. A. (2009). Credit Risk Management: Basic Concepts: financial risk
components,rating analysis, models, economic and regulatory capital. New york: Oxford University Press
20. Gujarati, N. D. (2003). Basic Econometrics (4rth Edion ed.). USA: McGraw-Hill.
21. Hussain, H. A., & Al-Ajmi, J. (2012). Risk management practices of conventional and Islamic banks in Bahrain.
The Journal of Risk Finance , 13, 215-239.
22. Hussein A. Hassan Al-Tamimi, F. M.-M. (2007). Banks' risk management: a comparison study of UAE. Emerald ,
8 (4 pp. 394 - 409), 396.
23. Jimenez, G., & Saurina, J. (2006). Credit Cycles, Credit Risk, and Prudential Regulation. International Journal of
Central Banking , 2 (2).
24. Joseph, C. (1988). Advanced Credit Risk Analysis and Management. John Wiley & Sons.
25. Joseph, C. (2013). Advanced Credit Risk Analysis and Management. Great Britain: John Wiley & Sons.
26. Macdonald, S. S., & Koch, T. W. (2006). Management of banking. In S. S. Macdonald, & T. W.
27. Misman, F. N. (2012 ). Financing Structures, Bank Specific Variables and Credit Risk: Malaysian Islamic
28. Banks. Journal of Business and Policy Research , 7 (1), 102 - 114.
29. Pallant, J. (2007). SPSS Survival Manual: A Step by Step Guide to Data Analysis using SPSS for Windows third
edition (3rd ed.). New York: McGraw-Hill Education.
30. Prakash, R., & Poudel, S. (2013). Macroeconomic Determinants of Credit Risk in Nepalese Banking Industry.
Proceedings of 21st International Business Research Conference. Toronto: Ryerson University.
www.tjprc.org

editor@tjprc.org

14

Atakelt Hailu Asfaw & P. Veni

31. Saunders, A., & Cornett, M. M. (2003). Financial institutions management: a risk management approach.
In financial institutions management: A risk management approach (4rth ed., pp. 138-152). New York: McGrawHill/Irwin
32. Saunders, A., & Allen, L. (2002). New Approaches to Value at Risk and Other Paradigms (2nd ed.). New York.:
John Wiley & Sons, Inc.
33. Shingjergji, A. (2013). The Impact of Bank Specific Variables on the Non Performing Loans Ratio in the
Albanian Banking System. Research Journal of Finance and Accounting, 4 (7).
34. SINGH, A. (2013). Credit risk management in Indian commercial banks. International Journal of Marketing,
Financial Services & Management Research, 2 (7), 147-151.
35. Swamy, V. (2012). Impact of Macroeconomic and endogenous factors on non performing Bank assets.
The International Journal of Banking and Finance, 9 (1), 27-47.
36. Suresh P.(2010). Management of Banking and Financial Service ,Second Edition ,dorling Kindersley India pvt.ltd
37. Rama mohana rao K. and Tekeste B. (2013) Cost Efficiency and Ownership Structure of Commercial Banks in
Ethiopia: An application of non-parametric approach. European Journal of Business and Management, 4.
38. Rashid, R. N., Azid, T., & Malik, S. (2014). Microeconomic Determinants of Credit Risk Management in
Pakistan: A Case Study of Banking Sector. Pakistan Journal of Social Sciences, 34 (1), 177-192.
39. Thiagarajan, S., & Ramachandran, A. (2011). An Empirical Analysis and Comparative Study of Credit Risk
Ratios between Public and Private Sector Commercial Banks in India. International Research Journal of Finance
and Economics (65).
40. Thiagarajan, S., Ayyappan, S., & Ramachandran, A. (2011). Credit Risk Determinants of Public and Private
Sector Banks in India. European Journal of Economics, Finance and Administrative Sciences (34).
41. Yfield, A. (2009). Discovering Statistics using SPSS (3rd ed.). New Delhi: SAGE Publications India Pvt Ltd.
42. Zribi, N., & Boujelbene, Y. (2011). The factors influencing bank credit risk: The case of Tunisia. Journal of
Accounting and Taxation, 3 (4), 70-78.

Impact Factor (JCC): 4.8567

Index Copernicus Value (ICV): 3.0