Collision of NPLs
on the Financial Performance of Commercial Banks: A Case Study of Ethiopia
Hailu Megersa Tola
Dean
College
of Business and Economics
Ambo
University
Ethiopia
D. Guna Sankar
Department
of Accounting and Finance
Ambo
University
Ethiopia
^{ }
Abstract
Credit risk in banking relates
to the possibility that loans will not be paid or that investments will�� deteriorate in quality or go in to default
with resultant loss to the bank. This is the most obvious and most important
risk to the banking industry in terms of potential losses. Credit risk is not
confined to the risk that borrowers are unable to pay; it also includes the
risk of payments being delayed, which can also cause problems for the bank. In
order to protect their own interest and the wealth of bank depositors, banks
need to investigate and monitor the activities of the will be and existing
borrowers. Adequately managing of those risks related with credit is critical
for the survival and growth of any financial institution. The present case
study projects the effects of NonPerforming Assets on the Financial
Performance of Commercial Banks in Ethiopia.
Keywords: Commercial Banks, Effects, Ethiopia, Financial Performance, NPLs.
1. Introduction
A nonperforming
loan is a loan that is in default or close to being in default. Banks face different elements of risk that require to be
identified measured and managed. Managing these risks is a process by which one
identifies the risk, measures and quantifies the risk and develops strategies to
manage the risk. The banking industry is facing different types of risks
associated with its functions. But according to Van Gestel & Baesens
(2009), credit risk has been the most principal and perhaps the most important
risk type that has been present in finance, commerce and banks too. Credit risk
has been defined from different perspectives by different researchers and
organizations. Most researchers agreed with the definition given by Basel
(1999) who defines it as the potential that debtor or counterparty default in
satisfying contractually predetermined obligation according to the agreed up on
terms. Because failure of trading partner to repay its debts in full can
seriously damage the affair of the other partner, credit risk always has been
the vicinity of career throughout the world (Achoo & Tenguh, 2008).
According to Zewude (2011), for banks, the issue of
credit risk is of even of greater concern because of the higher level of
perceived risk resulting from the loan book which is the largest asset for any
commercial bank. Even though credit creation is the main income generating
activity for commercial banks, it involves a huge risk to both the banks and
the borrowers.
2. Statement of the Problem
Banks are exposed to risks like credit, market, operational,
interest rate and liquidity risk. The appropriate management of these risks is
a key issue to reduce the earnings risk of the bank, and to reduce the risk
that the bank becomes insolvent and depositors cannot be refunded. Banks use
deposits of their customers to generate credit for their borrowers, which in
fact is a revenue generating activity for the banks themselves. This credit
creation process exposes the banks to a high default risk which might lead to
financial distress including bankruptcy. The banks can either choose from the
proposed options or employ their own as long as it gives sound and fair
results.
The importance of the credit
risk management and its impact on performance has motivated researcher to
pursue this study. The research assumes that if the credit risk management is
sound, the performance (profit level) was satisfactory. The other way around,
if the credit risk management is poor, the performance (profit level) was
relatively lower. The central question is how significant is the impact of
credit risk management on performance (profitability).
3. Objectives of the Research
� The general objective of this study was to assess the
impact of NPLs on the performance of selected commercial banks in Ethiopia.
� To analyze the impact of credit risk management on the
performance of the bank.
� To determine the relationship between credit risk
management and performance in terms of profit for the commercial banks in
Ethiopia.
4. Hypotheses of the Research
Throughout the research, the following two hypotheses
were tested.
5. Significance of the Research
It was show the impact of NPLs on bank performance; it
was give a motivation to other researchers to conduct a research about the
NPL�s practices in the commercial banks and it was useful for financial
institutions by providing information about NPLs.
6. Scope of the Research
In Ethiopia, there were banks
which give service in number twenty one. This study was limited to a manageable
of five banks in commercial banks of Ethiopia and even if there were different
problems which need investigation, the aim of the study was to to assess the impact of NPLs on the performance of
selected commercial banks in Ethiopia.
These researches were limited on the measure of the
performance of commercial banks in terms of credit risk management under the
selected sample.� The study was employ
nonperforming loan ratio/NPLR/ and capital adequacy ratio/CAR/ as the
measuring instruments of credit risk management and return on asset /ROA/ as
indicator of performance in terms of profit.
7. Limitations of the Research
The researcher limited this study to only five commercial banks; the study
was limited to 10 years of bank data and the study was based on secondary data
only.
The researcher
decided to limit this study to Commercial Banks of Ethiopia namely, Awash
international bank, Bank of Abyssinia, Nib International Bank, Dashen Bank and
Commercial Bank of Ethiopia. These banks have been selected with criteria taken
as the five banks from other banks expected to have more than ten years of
experience on the lending activities.
8. Sampling Design
The researchers selected five major commercial banks in
Ethiopia and collected the necessary data from each bank. Those data was
collected from National Bank of Ethiopia annual report from 2007 to 2016, and
used for regression purpose. The reason why the researcher purposively selects
five banks is, to have more observations. For those banks with 10year life
span is selected. Therefore, there are 50 observations in the regression
analysis.
9. Source of Data and Data Collection Instrument
The main source of data for the study was found from the
audited balance sheet and income statement of five purposively selected banks.
From those banks, 10 consecutive years of balance sheet and income statement
report were used for the study. In our country it�s a must for banks to submit
its annual report to the NBE not only that they are supposed to submit their
off balance sheet too. So the researcher�s easily gets annual reports of all
selected banks from the NBE.
10. Data Analyzing Instrument
The data collected from the annual reports of the banks
were analyzed using multiple regression analysis: the relation of one dependent
variable to multiple independent variables. The regression output was obtained
using Statistical Package for Social Sciences (SPSS 20 version).
11. Model Specification
In this study, multiple regression models with two
independent variables were used. To measure for financial performance i.e. ROA
(Net Income/Total asset): for credit risk management are NPLR (Nonperforming
Loans/Total Loans) and CAR] respectively.
12. Inferential Analysis of Commercial Banks in Ethiopia
12.1 Diagnostic Tests
Here the researcher used regression command for handling the regression.
This is followed by the output of these SPSS commands.
Table 1
Variables Entered/Removedb 

Model 
Variables Entered 
Variables Removed 
Method 
1 
CAR,
NPLR^{a} 
. 
Enter 
a.
All requested variables entered. 


b.
Dependent Variable: ROA 

Source: SPSS regression out put
Table one shows the variables entered or variables
removed from the study at any point of time from the beginning till the end of
the work. As it is explained the variables entered in column�� two are independent variables of the study
i.e., capital adequacy ratio and nonperforming loan ratio. Since there was no
variable removed from the study, variable removed column is free. The last
column shows the method that was used by the researcher, enter method was used
to remove or enter the variables. All variables are entered on the above table.
The dependent variable which is return on asset explained in the bottom of the
table.
Table 2: Linearity of the
Variables
Model Summary^{b} 

Model 
R 
R Square 
Adjusted R Square 
Std. Error of the Estimate 
Change Statistics 

R Square Change 
F Change 
df1 
df2 
Sig. F Change 

1 
.422^{a} 
.178 
.143 
.00571 
.178 
5.080 
2 
47 
.010 
a.
Predictors: (Constant), CAR, NPLR 







b.
Dependent Variable: ROA 






Source: SPSS regression out put
Table 2; demonstrates about large R, which shows the
multiple correlation coefficients and the correlation between the observed and
predicted values of the dependent variables. And the value of R for models
produced by the regression procedure range from 0 to 1. The larger the value of
R display that there is strong relationship among observed and predicted value.
In our case R is 0.422.
R square tells us, how much of variance in the dependent
variable is explained by our independent variable. So, in our case we were
known how much NPLR and CAR explained by ROA.��
As of R and the value of R square ranges between 0 and 1, beside to that
small value indicates that the model does not fit the data well. As the table
indicates, the independent variable explained the dependent variable by 17.8%.
This means that our model using two predicted variables (NPLR & CAR)
explain about 17.8% variance of our dependent variable (ROA). Right next to R
square we get adjusted R square. If we had small sample size the R square value
in the sample tend to be a little over estimated and little optimistic over
estimation of what probably really happening in the population. So, Adjusted R
square corrects this value to provide a better estimation of what actually
happening in the population. In our case Adjusted R square is .143. Standard
error of the estimate this is basically gives an idea of how much our
prediction might be off. If the number is large the more variability it
indicates from the table, we have Standard error of the estimate (0.006) which
is very small and good.
R square is significant at 5 % level of significance as
the SE <.05.
Table 3: ANOVA Table
ANOVA^{b} 

Model 
Sum of Squares 
df 
Mean Square 
F 
Sig. 

1 
Regression 
.000 
2 
.000 
5.080 
.010^{a} 

Residual 
.002 
47 
.000 



Total 
.002 
49 




a.
Predictors: (Constant), CAR, NPLR 




b.
Dependent Variable: ROA 





Source: SPSS Regression Out put
ANOVA, table summarizes the output of the analysis of
variance. In regression row, the output for regression displays information
about the variation accounted for by the existing model. Residual displays
information about the variation that is not accounted for by the model. And
total in the table shows the sum of regression and residual. Mean square is the
sum of squares divided by the degrees of freedom. And F statistics is the
regression mean square divided by the residual mean square. If the significance
value of the F statistics is small, then the independent variable does a good
job in explaining the variation in the dependent variables.
Table 4:�
Collinearity Diagnostics Test Table
Collinearity Diagnostics^{a} 

Model 
Dimension 
Eigenvalue 
Condition Index 
Variance Proportions 

(Constant) 
NPLR 
CAR 

1 
1 
2.561 
1.000 
.01 
.06 
.01 
2 
.387 
2.574 
.03 
.92 
.05 

3 
.052 
7.005 
.96 
.02 
.94 

a.
Dependent Variable: ROA 



Source: SPSS Regression Out put
Table 5 is a table which displays statistics that help to
determine whether there are any problems with collinearity or not. Collinearity
(multicollinearity) is the undesirable situation where the correlations among
the independent variables are strong.
Eigenvalues proved an indication of how many different
dimensions are there among the independent variables. When several Eigen values
are close to zero, the variables are highly interring correlated and small
changes in the data values may lead to large changes in the estimates of the
coefficients.
Condition index are the square roots of the ratios of the
largest eigenvalue to each successive eigenvalue. A condition index greater
than 15 indicates a possible problem and an index greater than 30 suggests a
serious problem with collinearity (SPSS output).
Even if eigenvalues are used for checking the existence
of collinearity, the best way is conditional index. So in this research case,
since conditional index value scored around 1, 2 and 7, from this ground the
researcher can say that there is no multicollinearity among independent
variables.
Table 5 Residuals Statistics
Residuals Statistics^{a} 


Minimum 
Maximum 
Mean 
Std. Deviation 
N 
Predicted
Value 
.0186 
.0331 
.0280 
.00260 
50 
Residual 
.01952 
.01035 
.00000 
.00559 
50 
Std.
Predicted Value 
3.609 
1.973 
.000 
1.000 
50 
Std.
Residual 
3.416 
1.812 
.000 
.979 
50 
a.
Dependent Variable: ROA 



Source: SPSS Residual Out put
Table 6, tells about the residual and predicted value.
For each case, the predicted value is the value predicted by the regression
model and for each case; the residual is the difference between the observed
value of the dependent variable and the value predicted by the model. Residuals
are estimate of the true errors in the model, if the model is appropriate for the
data, the residuals should follow a normal distribution. Standardized predicted
values are predicted values standardize to have mean 0 and standard deviation
of 1. In short standardize residuals are ordinary residuals divided by the
sample standard deviation of the residual and have mean of 0 and standard
deviation of 1.
12.2 Test of Normality of Residuals
One of the assumptions of linear regression analysis is
that the residual is normally distributed, at the mean of zero and standard
deviation of one. All of the results from the examiner command suggest that the
residual or the error terms are normally distributed. The skewness and kurtosis
are near to 0. As one can observe from the histogram and pp plot it looks
normal. Based on these results, the residuals from this regression appear to
conform to the assumption of being normally distributed.
Figure: 1. Histogram (Test of Normality)
Source: SPSS regression out put
Figure: 2 Normal
pp plot of RegressiS on tandardized Residual
Source: SPSS Regression Out put
The above Figures show whether the data are normally
distributed or not. The error term should be normally distributed at the mean
of 0 and standard devotion 1, here in this model the mean is approximately 0
and the standard devotion is 0.979 approximately 1, so the model is normally
distributed. The researcher watched from the histogram and from the p p plot
too.
12.3 Relationships between ROA and NPLR
Figure 3: Scatter Diagram for ROA vs. NPLR
Source:� SPSS Regression Out put
The scatter diagram above shows the negative relationship
between the two variables hence a negative gradient. This confirms the
coefficient of the NPLR of the variable in the regression equation and hence
nonperforming loans is good indicator of return on asset from the above
results.
The points are closely clustered at one point. This may
indicate the nature of performance of the bank and the level of shareholding
and if the institution is public listed or private companies. The nature of
relationship does not give a positive relationship to the whole banking sector
in this research analysis.
12.4 Relationship between ROA and CAR
Figure 4: Scatter Diagram for
ROA vs. CAR
From the scatter diagram above the points along the line
of the best fit are observed to have a big dispersion in regard to the line.
The Positive points make line look like horizontal reducing the gradient/slope
between the two variables.
12.5 Descriptive Analysis of Commercial Banks in Ethiopia
The Relationship between Credit Risk Management
and Profitability at CB of Ethiopia
Graph 5
Scatter Diagram for ROA, CAR and NPLR of
Commercial Bank of Ethiopia
From the above graph it�s that observed the relations between ROA, NPLR
& CAR of commercial bank of Ethiopia.
When NPLR reaches its maximum at (15), ROA reaches its
minimum at (2). This means that 15% from the total loan are nonperforming or
default to be paid by the bank customer. So an increase trend of ROA indicates
that the profitability of the company is improving. Conversely, a decreasing
trend means that profitability is deteriorating. So, 2% indicates a
deteriorating profitability of CB of Ethiopia.
�During (20072008)
CBE have low performance regarding credit risk management in terms of (NPLR)
and from (20092016) dramatically decrease and ROA is above NPLR this shows
that CB of Ethiopia manages its default loan properly and the profit of the
bank also increase dramatically.
CAR of commercial bank of Ethiopia indicate normal trend
throughout the years.
�12.6 The Relationship between
Credit Risk Management and Profitability at BOA
Graph 6The Relationship between Credit Risk
Management and Profitability at BOA
Scatter Diagram for ROA, CAR & NPLR OF Bank of Abyssinia
From the above graph were observed the trend of ROA, NPLR & CAR for
Bank of Abyssinia.
During 2008 NPLR reaches its maximum with 9.8 % this
shows that from the total loan 9.8 % default to be paid by the customer.
NPLR shows a zigzag trend from 20082013 and it has
decrease initially. Finally from 20142016 it decreases at an increasing rate.
In recent year the trend shows BOA has managed its NPLR.
From this trend were observed that when the number of
NPLR reaches its maximum ROA reaches its minimum from our sample of 10 years of
data.
ROA is greater than NPLR from year 2012 this shows BOA
was managed its default loans properly.
CAR shows high trend which means that BOA kept high
capital for risk weighted sum for bank assets.
The Relationship between Credit
Risk Management and Profitability at AIB
Graph 7
Scatter Diagram for ROA, CAR &NPLR of Awash International Bank
From the above graph were observed the relations between ROA, NPLR &
CAR of Awash International Bank
Like other banks, AIB have bad NPLR from (20072010)
years but it decreases continuously from (20112016) as the above figure
indicated.
ROA is below NPLR From (20072010) and above NPLR from
(20112016) this indicates that Awash International Bank have managed its NPLR
properly during years.
CAR of Awash International Bank is little to high
relative from the above mentioned banks.
�The
Relationship between Credit Risk Management and Profitability at NIB
Graph 8
Scatter Diagram for ROA, CAR &NPLR of Nib International Bank
The above graph shows the relation between ROA, CAR & NPLR of Nib
International Bank.
NIB has relatively lower NPLR when compared to that of
the other banks.
From (20072010), NPLR is above ROA and from (20112016)
ROA is greater than NPLR. NIB has managed its default loan properly.
CAR of NIB are very high like other private banks
The Relationship between Credit Risk Management
and Profitability at DB
Graph 9
Scatter Diagram for ROA, CAR &NPLR of Dashen Bank
The above graph shows the relation between ROA, CAR & NPLR of Dashen
bank
Dashen bank NPLR is low from year (20072016). This shows
that the bank managed loan properly. It should be a good example for other
banks regarding Loan management.
CAR of Dashen is high like other private banks.
13. Findings of the Research
14. Conclusion and Recommendation
The purpose of this chapter is to review the whole thesis
and highlight future researcher directions. The next section displays
recommendation made by the researcher for all concerning issues.
Conclusion
Recommendation
Based on the findings and conclusions of the study the
following recommendations are given.
References
Achou, F. Takang and Tenguh C.
Ntui (2008), �Bank performance and credit riskmanagement�, University essay from
H�gskolaniSk�vde/Institutionenf�rteknikochsamh�lle;
H�gskolaniSk�vde/Institutionenf�rteknikochsamh�lle
Afriyie & Akotey, 2012
Credit risk management & profitability of selected rural banks.
Altunbas,
Y., 2005. Mergers and acquisitions and bank performance in Europe � The role of
strategic similarities.European central bank, Working Paper Size Series, No.
398.
Anthony,
M.C., 1997. Commercial bank risk management; an analysis of the process.The Wharton
school.University of Pennsylvania.Financial Institutions Centers.
Asari
(2011) Determinants of Nonperformin loans evidence from commercial banks of
Ethiopia.
Athanasoglou,
P., S.N. Brissimis and M.D. Delis, 2005.Bankspecific, industry specific and macroeconomic
determinants of bank profitability. MPRA Paper No. 153.
A. V. Vedpuriswar, 2009 Credit
risk management & profitability of Commercial Banks in Ethiopia.
Awoyemi
Samuel Olausi,Banks ( 2014 ) the impact of credit risk management on the banks
of Nigeria.
Basel
Committee on Banking Supervision, 1982.Management of banks� international lending, country risk
analysis and country exposure measurement and control.Available from www.bis.org.
B.
Ravi Kumar, �Financial Performance Evaluation � A Case Study of Bharat Heavy
Electricals Limited,� EPRA International Journal of Economic and Business
Review, EISSN: 2347 � 9671, pp. 207 � 211, Vol. 3, Issue. 11, November  2015,
published by EPRA Trust � Tiruchirappalli.
Basel
Committee on Banking Supervision, 1999. Principles for the management of credit Risk, CH � 4002
basel, Switzerland Bank for International Settlements.
Basel
Committee on Banking Supervision, 2001. Risk management practices and regulatory capital:
Crosssectional comparison. Basel Committee on Banking Supervision.Available
from www.bis.org.
B. Ravi Kumar et. al. �Empirical
Analysis on Financial Performance through Cash Flow Statements,� International
Journal of Accounting & Finance Review, PISSN:�� 2576 � 1285, EISSN: 2576 � 1293, pp. 1 � 12,
Vol. 3, No. 1, September � 2018, published by Centre for Research on Islamic
Banking & Finance and Business � USA.
Basel
Committee on Banking Supervision, 2006. Studies on credit risk concentration, an overview of
the issues and a synopsis of the results from the research task force project.
Available from www.bis.org.
B. Ravi Kumar et. al.�Genesis for
Increase of NPAs in Indian Banks,� Journal of Banking and Finance Management,
pp. 1 � 8, Volume 1, Issue 1, July � 2018, published by Sryahwa Publications �
USA.
Bessis,
J., 2002. Risk management in banking.John Wiley & Sons.
Bobakovia,
I.V., 2003. Raising the profitability of commercial banks, BIATEC, Retrieved on
April, 2005, 11. Available from http/www/nbs.SK/BIATEC/.
Bourke,
P., 1989.Concentration and other determinants of bank profitability in Europe, North
America and Australia. Journal of Banking and Finance, 13(1): 65�79.
Brownbrigde, M, 1998. The causes of
financial distress in local banks in Africa and implications for prudential
policy, UNCTAD OSG/DP/132.
Boudriga, Taktak & Jellouli
(2009), Determinants of nonperforming loans in licensed Commercial Banks
evidence from srilanka.
B. Ravi Kumar et.al. �Non
Performing Assets in Public Sector Banks: A Cause Analysis,� American Finance
& Banking Review, PISSN: 2576 � 1226, EISSN: 2576 � 1234, pp. 14 � 19, Vol.
2, No. 2, September � 2018, published by Centre for Research on Islamic Banking
& Finance and Business � USA.
Choudhry, 2011, Relation
between collateralizing loans &occurance on nonperforming loans.
De Young,
R. and G. Whalen, 1994. Banking industry consolidation: Efficiency issues,
Working Paper No. 110. A Conference of the Jerome Levy Economics Institute,
Office of the Comptroller of the Currency, Washington, DC.
Eastern
Caribbean Central Bank, 2009 Factors affecting Nonperforming loans in
Ethiopian banks.
Felix,
A.T. and T.N. Claudine, 2008. Bank performance and credit risk management, Available
from http//www.essays.se/essay/55d5c0bd4/.
Gestel & Baesens, 2008 Factors
affecting Nonperforming loans in Taiwan
Heffernan,
S., 1996.Modern banking in theory and practice.Chichester: John Wiley and Sons.
International Journal of Management & Sustainability,2014,3(5): 295306
Ismaila,
2011 Determinants of nonperforming loans
Kauko, 2012, p.196 External
deficits & nonperforming loans in the recent financial crisis������
Copyrights
Copyright
for this article is retained by the author(s), with first publication rights
granted to the journal. This is an openaccess article distributed under the
terms and conditions of the Creative Commons Attribution license
(http://creativecommons.org/licenses/by/4.0/).