MACROECONOMIC DETERMINANTS OF NON-PERFORMING LOANS IN NIGERIA: AN EMPIRICAL ANALYSIS

Transcription

1 MACROECONOMIC DETERMINANTS OF NON-PERFORMING LOANS IN NIGERIA: AN EMPIRICAL ANALYSIS ABSTRACT Ogechi Adeola Fredrick Ikpesu Lagos Business School, Pan-Atlantic University, Lagos, Nigeria The intermediation role of banks which is vital for the development of any nation can be adversely affected by non-performing loans. This study investigated the macroeconomic determinants of non-performing loans in Nigeria, using time series data for the period 2005 to 2014 collated from Central Bank of Nigeria Statistical Bulletin, Nigeria Deposit Insurance Corporation annual report, World Bank Development Indicators and International Financial Statistics. The choice of the period 2005 to 2014 was because of its significance as the number of banks was reduced from 89 to 25 in 2005 due to the banking recapitalization exercise embarked upon by the regulatory authority which led to the consolidation of banks. The variables used in the study are nonperforming loan (NPL) as the dependent variable, while gross domestic product growth rate (GDPGR), inflation (INF), lending rate (LR), exchange rate (ER), money supply to gross domestic product (M2GDP) and unemployment rate (UR) as independent variables. The outcome of the regression result showed that GDPGR has a positive relationship with NPL. In addition, the result also revealed that INF and ER have a positive relationship with NPL while LR, M2GDP, and UR have a positive and significant relationship with NPL. Out of the six macroeconomic variables used in the study, it can be observed that only LR, M2GDP, and UR determines NPL in Nigeria while GDPGR, INF, and ER has a positive relationship with NPL but does not influence or determine NPL in Nigeria. Hence, the policy implication of this study is that the monetary authorities should ensure that the lending rate of interest charged on loans by deposit money banks is realistic to enable borrowers conveniently repay the borrowed fund. Finally, the government should direct its monetary and fiscal policies towards reducing unemployment by creating an enabling environment for business growth through the provision of social and infrastructural facilities. JEL classification: E44, E51, G21 Key words: Non-performing Loans, Macroeconomic factors, Credit, Banking System, Nigeria of the corresponding author: oadeola@lbs.edu.ng INTRODUCTION The intermediation role of banks which is vital for the development of any nation can be affected by nonperforming loans. Literature and practice have shown that non-performing loan is the leading cause of financial crises Brownbridge, 1998; Greenidge & Grosvenor, 2010). Fan and Shaffer, 2004; Giranome et al also documented in their research findings that bad management and inefficiency of banks have been the major cause of non-performing loans. The study conducted by Caprio & Klingebiel (2002) showed that in Indonesia, over 60 banks failed during the financial crisis caused by the increasing surge of non-performing loans. In another study conducted by Fofack (2005), it was showed that the banking crises caused by non-performing loans affected the majority of the Sub-Saharan African nations in the 1990 s. Similarly, Tracey and Leon (2011) are of the view that the decline in the quality of the bank assets is one of the signs of banks failure in both less developing and developed nations. They also noted that the decline in loan assets hinder the banks in performing its expected intermediation function. Westenmann (2003) indicated that the persistent increase in non-performing loan was one of the reasons for the credit crunch in Germany after the credit boom witnessed by the country. In another study conducted by Shingjergji (2013), the empirical result revealed that GDP growth, interest rate, and foreign exchange rate have a positive relationship with non-performing loan ratio while inflation has a negative relationship with non-performing loan ratio in the Albanian banking sector. The increasing volume of nonperforming loan has always been the major concern of banks management, regulators and policy makers because of its adverse effect on both the bank and the economy. For instance, non-performing loans reduce the earning capacity of the banks, leads to financial crises, bank distress, affects the level of domestic investment, put high pressure on government revenue in an attempt to bail the failed banks, and ultimately hinders the intermediation role of banks. Non-performing loans have been identified as the major threat to the profitability of banks (Sinkey, 1992; Aremu, Suberu & Oke, 2010) and in some cases, the reason for banks failure (Brownbridge, 1998). According to Stuart (2005), between 1999 and 2009, the level of bad loans was as high as 35% in commercial 481

2 banks in Nigeria. The rising non-performing loans ratio in the banks has been attributed to poor and ineffective credit administration which stems from poor loan processing, undue interference in the process of loan granting, inadequate or lack of collateral amongst others (Umoh, 1994). In Nigeria, like most developing countries, nonperforming loans pose challenges for credit risk management. Graphical Representation of the Trends of Non-Performing Loans in Nigeria NPL NPL Source: Author s computation. Notes: This depict the graphical representation of the trends of Non-Performing Loans in Nigeria The graph above depicts the trends of non-performing loans ratio (NPL) in the banking industry in Nigeria within the periods of 2005 to The periods from 2005 to 2008 showed that the rate of NPL fell from 18.1% to 6.3%. The reason for the fall was because banks in Nigeria had just recapitalized from 2 billion naira to 25 billion naira in 2005 thus making banks stronger, well equipped in areas of personnel and technology. However, within the periods 2008 to 2009, NPL rose from 6.3% to 37.3%. This upsurge of NPL in the banking sector was due to the global financial crises in 2007, which impact was felt in Nigeria between the periods 2008 to Also, the period between 2009 to 2010, witnessed a drastic fall in NPL from 37.3% to 20.1%. This decline was due to the measures put in place by the regulatory authority (CBN) such as the establishment of Asset Management Corporation of Nigeria (AMCON) which major objective is to purchase non-performing loans of banks. The periods of 2011 to 2014 also witnessed a fall in NPL in the bank industry because of the institution of strong corporate governance and risk management framework within the banks and the role played by AMCON. This study is unique because it examines the macroeconomic impact of non-performing loans in Nigeria, focusing on the significant period of 2005 when the numbers of banks have reduced to 25 due to the banking recapitalization exercise embarked upon by the regulatory authority which led to the consolidation of banks to the period Also, the study is significant because it employed the industry data on non-performing loan obtained from World Development Indicator (WDI) and Nigeria Deposit Insurance Corporation (NDIC) in the analysis. The choice of the industry data of non-performing loan is because it captures the whole data of banks NPL in the Nigeria. LITERATURE There are various empirical studies on the macroeconomic determinant of non-performing loan in the literature. In a study conducted by Amuakwa & Boakye (2014) in Ghana, it was showed that both bank specific variables (previous year s NPL, bank size, net interest margin (NIM), and current year s loan growth) and macroeconomic variables (previous year s inflation, real gross domestic product (GDP) per capital growth and real effective exchange rate) significantly affect NPL. The study also revealed that for large banks, both bank specific variables and macroeconomic variables affect NPL whereas, in small banks, only bank specific variables (previous year s NPLs and current year s loan growth) are vital in explaining NPLs. Waweru & Kalani (2009) using a non-parametric study, examined the causes of non-performing loan in Kenya with a sample of 30 managers from ten big banks. In the study, economic recession was adjudged the most important external factor 482

3 that causes NPL and also the unwillingness of customer in disclosing key information during loan application process was noted to be the main customer specific factor. Furthermore, the study highlighted that the main bank specific factor contributing to non-performing debt problem in Kenya is the absence of aggressive debt collection policy. Louzis, Vouldis, & Metaxas (2012) examine what determines non-performing loan in the banking sector of Greece by adopting a dynamic panel data method using both bank-specific and macroeconomic variables. The empirical research revealed that non-performing loan can be explained by the following macroeconomic variables (GDP, Interest rate, Public Debt and Unemployment) and management quality. Beck, Jakubik, & Piloiu, (2013) examine the macroeconomic determinant of non-performing loan across 75 countries using panel data. The result of the study revealed that real GDP growth, share prices, exchange rate, and the lending interest rate are the variables found to significantly affect non-performing loan ratios. In another study conducted by Somoye (2010), the researcher documented in his research outcomes that earning risk is one of the key factors that determines non-performing loan, while other factors include interest rate and monetary policy rate. Fofack (2005) carried out an empirical analysis on the determinant of nonperforming loan in Sub-Saharan Africa. Empirical evidence from the study revealed that economic growth, real exchange rate appreciation, real interest rate, NIMs, and inter-bank loans are the significant determinants of NPLs in Sub-Saharan Africa. Another study carried out by Rajan & Dhal (2003) highlighted that banks size, growth rate of GDP, credit orientation and credit terms are the major determinants of non-performing loan in India. Bonilla (2011) investigated the determinants of non-performing loan indices in both Spain and Italy within the period January 2004 to March 2012 using used credit growth, wage, inflation, unemployment and GDP as macroeconomic variables. The outcome of the study revealed that out of the five macroeconomic variables, unemployment, wage, and GDP significantly affect non-performing loan indices in both Spain and Italy. Saba, Kouser & Azeem (2012) investigated the determinants of non-performing loan in US banking sector within the period The variables used in the study were Real GDP per Capita, Inflation, and Total Loans as independent variables, and non-performing loan ratio as dependent variable. The outcome of the study revealed that Real GDP per Capita, Inflation, and Total Loans have a significant impact on non-performing loan ratio. Anisa (2015) investigated what determines non-performing loan in Ethiopia for the period 2004 to 2013 using both bank-specific and macroeconomic variables. The empirical study revealed that deposit rate; loan to deposit ratio and lending interest rate has a positive and significant impact on banks non-performing loan while Bank solvency ratio and gross national product (GDP) growth rate and inflation rate has a negative as well as statistically insignificant impact on banks non-performing loan. Jimenez and Saurina (2006) in their empirical study also provided evidence that the NPL ratio can be explained by GDP growth, real interest rates, and credit conditions. In another study conducted by Khemraj & Pasha (2009) on the determinant of non-performing loan in Guyana, the researchers in their findings revealed that real effective exchange rate has a positive and significant impact on non-performing loan. The result further revealed that GDP growth is negatively related to non-performing loan. Adebola, Sulaiman & Dahalan (2011) used ARDL approach to examine the determinant of non-performing loans in Islamic banking in Malaysia within the period 2007:1 to 2009:12. The empirical result revealed the existence of cointegration among the variables. The result further showed that interest rate has a positive and a significant impact on non-performing loan, while the relationship between industrial production index and non-performing loan is positive but insignificant. It was also indicated in their findings that producer price index has a negative and significant impact. Using time series data, Akinlo and Emmanuel (2014) examines the determinant of non-performing loan in Nigeria within the periods The result of their findings revealed that both economic growth and non-performing loan have a negative relationship, while exchange rate, unemployment, credit to the private sector have a positive relationship with non-performing loan. Their result further revealed that in the short-run, exchange rate, stock market index, credit to the private sector and lending rate is what determines non-performing loan. Messai & Jouini (2013) carried out a study by investigating what determines non-performing loan within the period , using a sample of 85 banks in Italy, Greece and Spain. The empirical research findings showed that GDP growth rate is negatively related to non-performing loan while unemployment and interest rate are positively related to non-performing loan. Oluyemi (2005) highlighted that many of the banks owners and directors abuse their position by participating in activities that promote their self-interest, like granting unsecured loan and lending above the statutory lending limit which has resulted in the high non-performing loan in Nigeria. In another study conducted by Inekwe (2013), the research explored the link between GDP and NPL in Nigeria between 1995 to The research outcome indicated that there is a significant relationship between GDP and non-performing loan in Nigeria. From the available literature, it can be inferred that the macroeconomic determinant of non-performing loans is country specific. Hence, this study aims to investigate the macroeconomic determinants of nonperforming loan in Nigeria using industry data of non-performing loan which captures the whole data of banks NPL in the industry. 483

4 METHODOLOGY Sources of Data In assessing the macroeconomic determinants of non-performing loan in Nigeria, time series data for the period 2005 to 2014 was collated from Central Bank of Nigeria Statistical Bulletin, Nigeria Deposit Insurance Corporation annual report, World Bank Development Indicators and International Financial Statistics. The choice of the period 2005 to 2014 was because of its significance as the number of banks was reduced from 89 to 25 in 2005 due to the banking recapitalization exercise initiated by the CBN which led to the consolidation of banks. The variables used in the study are non-performing loan (NPL) as the dependent variable, while gross domestic product growth rate(gdpgr), inflation (INF), lending rate (LR), exchange rate (ER), money supply to gross domestic product (M2GDP) and unemployment rate (UR) as independent variables. TABLE 1. DATA DESCRIPTION AND SOURCES VARIABLES DEFINITIONS DATA SOURCES EXPECTED RELATIONSHIP Non-Performing Loan A loan is said to be nonperforming when interest and principal are left unpaid over the period of 90 days World Development Indicator(WDI), Nigeria Deposit Insurance Corporation(NDIC) annual Dependent Variables Gross Domestic Product Growth Rate(GDPGR) Inflation(INF) Lending Rate (LR) Exchange Rate (ER) M2GDP Unemployment Rate (UNPR) This is the growth rate of Gross Domestic Product. It refers to how a country gross domestic product grows from one year to another. This refers to the increase in the price of goods and services due to large volume of money in circulation This refers to the interest rate charged by banks on loan and advances This refers to the price at which a country s currency is exchanged for another country s currency. This refers to Percentage Ratio of Broad Money to Gross Domestic Product This refers to the percentage of the total labour force who are unemployed. Source: Authors, 2016 Notes: This table shows the data abbreviation and description. Analytical Techniques report Central Bank of Nigeria (CBN) Statistical Bulletin World Development Indicator(WDI) World Development Indicator(WDI) World Development Indicator(WDI) Central Bank of Nigeria (CBN) Statistical Bulletin International Financial Statistics(IFS) - Indeterminate (+/-) + Indeterminate (+/-) The study employed the multiple regression techniques of Ordinary Least Square to establish the macroeconomic determinant of non-performing loans in Nigeria. Analytic test such as autocorrelation test, heteroscedasticity test, and stability was also carried out. Normality Test This test was conducted to determine if the data are normally distributed. A distribution is normal if the residuals are normally distributed, the histogram is bell-shaped, the Kurtosis approaches 3 and the Jarque- Bera statistic p> 5%

5 Autocorrelation Test The study adopts the Breusch-Godfrey serial correlation LM test to determine whether the time series data are serially correlated. Heteroskedasticity Test The study employed the Breusch-Pagan-Godfrey test also known as the white test to determine if there is the presence of heteroskedasticity in our model. Stability Test This test was conducted to determine if the model used in the study is stable. In achieving this, the CUSUM test was employed. Model Specification In examining the macroeconomic determinants of non-performing loan in Nigeria, the model of the study is expressed as follows: NPL= F (GDPGR, INF, LR, log(er), M2GDP, UNPR) (1) The econometric model is expressed as follows: NPL= β O +β 1 GDPGR + β 2 INF + β 3 LR + β 4 log(er) + β 5 M2GDP + β 6 UNPR + U t (2) Where NPL= Non-performing loan which is non-performing loan to total loan GDPGR= Gross domestic product growth rate INF= Inflation LR= Lending rate ER= Exchange rate M2GDP= Percentage Ratio of Broad Money to Gross Domestic Product UNPR= Unemployment rate. U t = Error terms β O, β 1, β 2, β 3, β 4, β5, and β 6 are the respective coefficient terms. RESULTS AND DISCUSSION TABLE 2. OLS REGRESSION RESULT Dependent Variable D(NPL) Variable Coefficient Std. Error t-statistic Prob. D(GDPGR) D(INF) D(LR) D(LOG(ER)) D(M2GDP) D(UR) R-squared Adjusted R-squared Durbin-Watson stat Notes: These figures show the estimate of OLS Regression Result 485

6 Table 2 depicts the OLS regression result. The outcome of the regression result showed that GDPGR has a positive relationship with NPL. This result is consistent with similar findings by (Inekwe, 2013; Shingjergji, 2013). Although previous empirical studies suggest a negative relationship between GDPGR and NPL. In addition, the result also revealed that INF and ER have a positive relationship with NPL while LR, M2GDP and UR have a positive and significant relationship with NPL. Furthermore, the result also showed that the R- squared is (99%) which implies that 99% of the total variation in NPL is explained by the explanatory variables. The Adjusted R-squared of (98%) implies that the explanatory variables were robust in explaining the variation in NPL thus showing a good fit. Out of the six macroeconomic variables used in the study, it can be observed that only LR, M2GDP, and UR determines NPL in Nigeria while GDPGR, INF, and LR has a positive relationship with NPL but does not influence or determine NPL in Nigeria. Normality Test Figure 1, depicts the graph of the normality test. As shown in the graph, the residuals are normally distributed as the histogram is bell-shaped, the Kurtosis approaches 3. Also, the Jarque- Bera test further indicates that the residual is normally distributed as the p-value is more than 0.05 percent. Figure 1: Graph of Normality Test Series: Residuals Sample Observations 9 Mean Median Maximum Minimum Std. Dev Skewness Kurtosis Jarque-Bera Probability Notes: This depicts the graph of the Normality test Autocorrelation Test Table 3 depicts the outcome of the autocorrelation test. The results revealed that the variables are not serially correlated as the p-value of the Prob. Chi-Square (2) of the Obs*R-squared of the Breusch-Godfrey serial correlation LM test is more than 0.05 percent. Hence, we accept the null hypothesis that there is no serial correlation. Table 3. Breusch-Godfrey Serial Correlation LM Test F-statistic Prob. F(2,1) Obs*R-squared Prob. Chi-Square(2) Notes: These figures show the estimate of Serial Correlation LM Test Heteroskedasticity Test 486

7 Table 4 depicts the result Heteroskedasticity test. The outcomes of the result suggest that there is no presence of heteroskedasticity in the model as the Prob. Chi-Square (2) of the Obs*R-squared of the Breusch- Pagan- Godfrey is more than 0.05 percent. Hence, we accept the null hypothesis that there is no heteroscedasticity (i.e. presence of homoscedasticity which is desirable. TABLE4. HETEROSKEDASTICITY TEST: BREUSCH-PAGAN-GODFREY F-statistic Prob. F(6,2) Obs*R-squared Prob. Chi-Square(6) Scaled explained SS Prob. Chi-Square(6) Notes: These figures show the estimate of Heteroskedasticity test Stability Test Figure 1 depicts the graph of the CUSUM test. This test shows whether the model is stable or not. The model is stable if the whole sum of recursive error is between the two critical lines. As shown below, the blue line is between the two critical (red) lines, an indication that the model is stable. 6 FIGURE 2. GRAPH OF CUSUM TEST CUSUM Notes: This depicts the graph of CUSUM test 5% Significance Correlation Table 5 depicts the result of the correlation matrix between the dependent variable and the independent variables. As seen in the table below, the growth rate of the gross domestic product is positively correlated with non-performing indicated by the correlation of This correlation shows that as the growth rate in gross domestic product increase, non-performing loan increase. Also, inflation, lending rate, exchange rate, money supply to gross domestic product and unemployment rate are all positively correlated with non-performing loan as seen in the table. This indicates that an increase in any of this independent variables leads to an increase in non-performing loan. 487

8 TABLE 5: CORRELATION MATRIX BETWEEN DEPENDENT AND INDEPENDENT VARIABLES D(NPL) D(GDPGR) D(INF) D(LR) D(LOG(ER)) D(M2GDP) D(UR) D(NPL) Notes: These figures show the estimate of the correlation matrix between the dependent and independent variables CONCLUSION This study investigated the macroeconomic determinants of non-performing loans in Nigeria, using time series data for the period 2005 to 2014 collated from Central Bank of Nigeria Statistical Bulletin, Nigeria Deposit Insurance Corporation annual report, World Bank Development Indicators and International Financial Statistics. The choice of the period 2005 to 2014 was because of its significance as the number of banks reduced from 89 to 25 in 2005 due to the banking recapitalization exercise embarked upon by the regulatory authority which led to the consolidation of banks. The variables used in the study are non-performing loan (NPL) as the dependent variable, while gross domestic product growth rate(gdpgr), inflation (INF), lending rate (LR), exchange rate (ER), money supply to gross domestic product (M2GDP) and unemployment rate (UR) as independent variables. The outcome of the regression result showed that GDPGR has a positive relationship with NPL. This result is consistent with similar findings by (Inekwe, 2013; Shingjergji, 2013). Although previous empirical studies suggest a negative relationship between GDPGR and NPL In addition, the result further revealed that INF and ER have a positive relationship with NPL while LR, M2GDP, and UR have a positive and significant relationship with NPL. Furthermore, the result also showed that the R-squared is (99%) which implies that 99% of the total variation in NPL is explained by the explanatory variables. The Adjusted R-squared of (98%) implies that the explanatory variables were robust in explaining the variation in NPL thus showing a good fit. Out of the six macroeconomic variables used in the study, it can be observed that only LR, M2GDP, and UR determines NPL in Nigeria while GDPGR, INF, and LR has a positive relationship with NPL but does not influence or determine NPL in Nigeria. The policy implication of this study is that the monetary authorities should ensure that the lending rate of interest charged on loans by deposit money banks is realistic to enable borrowers conveniently repay the borrowed fund. Finally, the government should direct their monetary and fiscal policies towards reducing unemployment by creating an enabling environment for business growth through the provision of social and infrastructural facilities. FURTHER STUDY This research paper focused on a single country study. Future research effort should be directed towards multi - country study for a comparative purpose like investigating the macroeconomic determinant of non-performing loans in West African countries. REFERENCES Adebola, SS, Sulaiman, W, Yusoff, W, & Dahalan, J, 2011, An ARDL approach to the determinants of nonperforming loans in Islamic banking system in Malaysia, Kuwait Chapter of Arabian Journal of Business and Management Review, vol,1, no 1, pp Akinlo, O, and Emmanuel, M, 2014, Determinants of Non-Performing Loans in Nigeria, Accounting & Taxation, vol. 6, no. 2, pp.21. Alton, RG, and Hazen, JH, 2001, As Economy Flounders, Do we see A Rise in Problem Loans? Federal Reserve Bank of St Louis. Amuakwa Mensah, F, and Boakye Adjei, A, 2014, Determinants of non performing loans in Ghana banking industry. International Journal of Computational Economics and Econometrics, vol. 5 no.1, pp

9 Anisa, U, 2015, Determinants of nonperforming loan: An empirical study on commercial banks of Ethiopia, (Doctoral dissertation, AAU). Beck, R, Jakubik, P, and Piloiu, A, 2013, Non-performing loans: What matters in addition to the economic cycle?. Bonilla, CA, O, 2011, Macroeconomic determinants of the Non-Performing Loans in Spain and Italy, (Doctoral dissertation, University of Leicester). Bossone, B, 2001, Do banks have a future?: A study on banking and finance as we move into the third millennium Journal of banking & finance, vol. 25, no. 12, Brownbridge, M. 1998, The causes of financial distress in local banks in Africa and implications for prudential policy, United Nations Conference on Trade and Development. Caprio, G, and Klingebiel, D, 2002, Episodes of systemic and borderline banking crises, World Bank, Washington. DC, mimeo. Fan, L, and Shaffer, S, 2004, Efficiency versus risk in large domestic US banks, Managerial Finance, vol. 30, no. 9, pp Fofack, H, 2005, Non-performing loans in Sub-Saharan Africa: causal analysis and macroeconomic implications, World Bank Policy Research Working Paper, pp Girardone, C, Molyneux, P, and Gardener, EP, 2004, Analyzing the determinants of bank efficiency: the case of Italian banks, Applied Economics, vol. 36, no. 3 36(3), Greenidge, K., and Grosvenor, T. 2010, Forecasting non-performing loans in Barbados, Journal of Business, Finance and Economics in Emerging Economies, vol. 5, no. 1, Guy, K 2011, Non-performing loans, The Central Bank of Barbados Economic Review, Vol. XXXVII, No. 1, pp Inekwe, M, 2013, The Relationship between Real GDP and Non-Performing Loans: Evidence from Nigeria, International Journal of Capacity Building in Education and Management (IJCBEM), Vol. 2, no. 1, pp 1-7 Jéminez, G, and Saurina, J. 2006, Credit Cycles, Credit Risk, and Prudential Regulation, International Journal of Central Banking, 2 Khemraj, T, and Pasha, S, 2009, The determinants of non-performing loans: An econometric case study of Guyana, In Caribbean Centre for Banking and Finance Bi-annual Conference on Banking and Finance, St. Augustine, Trinidad. Kroszner, RS, 2002 in Waweru, NM. and Kalani, VM 2009, Commercial banking crises in Kenya: causes and remedies, African Journal of Accounting, Economics, Finance and Banking Research, Vol. 4, no. 4, pp Louzis, D P, Vouldis, AT, and Metaxas, VL 2012, Macroeconomic and bank-specific determinants of nonperforming loans in Greece: A comparative study of mortgage, business and consumer loan portfolios,. Journal of Banking & Finance, vol. 36, no. 4, pp Messai, AS, and Jouini, F, 2013, Micro and macro determinants of non-performing loans, International journal of economics and financial issues, vol. 3, no. 4, pp.852. Oluyemi, SA, 2005, Banking Sector Reforms and the Imperatives of Good Corporate Governance in the Nigerian Banking System, NDIC Quarterly, vol.16, no.1 pp Rajan, R, and Dhal, SC, 2003, Non-performing loans and terms of credit of public sector banks in India: An empirical assessment Occasional Papers, vol. 24, no.3, pp Reinhart, C, and Rogoff, K. 2010, From Financial Crash to Debt Crisis, NBER Working Paper Saba, I, Kouser, R, and Azeem, M, 2012, Determinants of Non Performing Loans: Case of US Banking Sector, The Romanian Economic Journal, Year, vol.ixv no. 44. Sheefeni, JP, 2015, The Impact of Macroeconomic Determinants on Non-Performing Loans in Namibia, International Review of Research in Emerging Markets and the Global Economy (IRREM), vol.1, no.4. Shingjergji, A, 2013, The Impact of Macroeconomic Variables on the Non Performing Loans in the Albanian Banking System During , Academic Journal of Interdisciplinary Studies, vol. 2, no.9, pp Somoye, RO, C, 2010, The Variation of Risks on Non-Performing Loans on Bank Performances in Nigeria, Indian Journal of Economics & Business, vol. 9, no.1. Tracey, M, and Leon, H, 2011, The impact of non-performing loans on loan growth,imf Working Papers. Waweru, N, and Kalani, VM, 2009, Commercial banking crises in Kenya: Causes and remedies. African, Journal of Accounting, Economics, Finance and Banking Research, vol. 4, no

10 Westermann, F, 2003, Special. In CESifo forum (Vol. 4, no. 1, p. 36). Institut für Wirtschaftsforschung (Ifo). APPENDIX Dependent Variable: D(NPL) Method: Least Squares Date: 02/24/16 Time: 16:16 Sample (adjusted): Included observations: 9 after adjustments Variable Coefficient Std. Error t-statistic Prob. D(GDPGR) D(INF) D(LR) D(LOG(ER)) D(M2GDP) D(UR) R-squared Mean dependent var Adjusted R-squared S.D. dependent var S.E. of regression Akaike info criterion Sum squared resid Schwarz criterion Log likelihood Hannan-Quinn criterion Durbin-Watson stat Source: Authors' Computation and EView 9 Output. Table 3. Breusch-Godfrey Serial Correlation LM Test F-statistic Prob. F(2,1) Obs*R-squared Prob. Chi-Square(2) Table 4. Heteroskedasticity Test: Breusch-Pagan-Godfrey F-statistic Prob. F(6,2) Obs*R-squared Prob. Chi-Square(6) Scaled explained SS Prob. Chi-Square(6)