A Comparative Study of Stability & Consumer Confidence between Islamic & Conventional Banks in Bangladesh Sisili Rahman, University of Asia Pacific Abstract- This study exhibits a comparative analysis of consumer confidence and bank-level stability factors between Islamic and conventional banks in Bangladesh. The study finds that despite having lower liquidity Islamic banks are able to provide higher consumer confidence levels than conventional banks. Islamic banks have reported very small Non performing asset (NPA), and shown a positive and significant relationship with liquidity implying that during the crisis Islamic banks tend to take rigid risk strategies compared to conventional banks. Cost income ratio (CIR) is inversely and insignificantly related in both types of banks. As increase in cost decreases the stability of the bank, profit before tax (PBT) gives expected positive and significant relationship in all cases. Increase in PBT increases the stability and consumer confidence level but the level of significance is higher in case of Islamic banks. CC represents consumers’ confidence and shows positive result in all cases but with an exception with conventional bank in TQ factor. Keywords: Islamic Bank, Conventional Bank, Stability, Consumer Confidence, Unit Root Test, Random Effect Regression I. INTRODUCTION The stability of the banking sector is instrumental in ensuring the steadiness of the entire financial system of an economy. Banks are the key players in payment system, money creation, savings, investment and ensuring overall economic goal. Bank’s stability is normally reflected by liquidity (LQ), return on asset (ROA), value (TQ), and consumers’ confidence (CC) measured by the percentage of deposit in the total liability. Islamic banks have a different and unique form of product mix which protect their stability in case of financial crisis. Generally, there is a distinctive characteristic of liquidity management of Islamic banks compared to conventional banks. Similarly, the unique product mix is complied with the requirement of Basel Accord. It often impacts different level of risk capital and credit risk. While being exposed to same market conditions of an economy, whether the product mix of Islamic banks has analogous consequence towards prevailing stability and confidence of the consumers like conventional banks is salient to determine. An econometric approach has been used for the comparison of stability and consumer confidence using actual data rather than any perceived or established outcome. It is done by examining the stability factors of bank performance (ROAV - volatility) and firm value (Tobin Q ) both in the Islamic and conventional commercial bank because the level of non-performing assets (NPA) as a result of bad loan screening lending to sub-prime borrowers (reflected in loan-loss provisioning). Finally, through sensitivity to deposit (consumer confidence) over total liabilities it is examined that http://www.ojs.unito.it/index.php/EJIF ISSN 2421-2172 1 EJIF – European Journal of Islamic Finance No13, August (2019) whether the results are different or same in the two banking sectors. II. LITERATURE REVIEW Several researches have been conducted on the stability factor of various types of banks and financial institutions. With the emergence of Islamic banks and its outstanding contribution to the financial market and country’s economy, authors have also conducted a number of studies to compare between these two major types of banks in many countries. This paper has been prepared based on the knowledge of previously conducted researches and similar result has been found in most cases. Rochet (1992) noted that capital regulations cannot control the risk taking behavior of the banks [14]. So effectiveness is largely dependent on whether the banks are maximizing their value. Banks often shift their product mix to riskier asset with higher leverage ratios. So to correct it they should use solvency ratio rather than leverage ratios. Many of the Islamic scholars believe that sphere of profit in Islamic banking is interrelated. Among other more recent studies, Alaro, Razzak and Hakeem (2011) found that in terms of risk management Islamic banks are more competent than conventional banks [2]. Another study of Malaysia’s Financial Stability Report (2011) found that the countries where Islamic banks are major and key player of the economy are less instable than those which are managed by conventional banks. Hadeel Abu Loghod (2010) noted that the major reason behind the higher deposits and liquidity of Islamic banks are their specialized products such as Mudaraba, Musharakah, Murabahah, Ijarah, and profit and loss sharing mechanism [11]. In Bangladesh Islamic banks are adapting new techniques both quantitative and as well as qualitative to manage their credit risk and uphold the consumer confidence in their banking mechanisms. Hasan and Dridi (2010) found that the asset growth of Islamic banks was double than that of other conventional banks in 2007-2009 [18]. Again, higher loan to asset ratio negatively impacts the stability. Biancone & Radwan (2016) stated that variety is the financial instruments offered by Islamic finance has not only depicted a positive growth but also is recognized as a lucrative investment opportunity [26]. Size of the ban and net working capital have positive but insignificant relation with the liquidity risk in Islamic banks whereas negative relation s found in case of conventional bank for size of the bank [23]. Competitive condition could not define any significance in relationship between the weighted assets ratio and Islamic bank behavior. Cost income ratio is another measure of efficiency. The lower the ratio the higher the profitability will be. And it is negatively related to Z-scores. Banks often think that their poor financials maybe improved with higher loan disbursement [17]. The prohibition of predetermined income (interest income) with the commands of Quran is another prominent cause of lower non performing loans of Islamic banks compared to their counter parties [18]. Many studies [7,18 & 21] have found superior performance by Islamic banks and their larger contribution in keeping the economy stable. Iqbal and Molyneux (2005) have used frontier approach to conduct their study while others have used simple ratios, Z-scores and regression model [20]. (Berger, Hunter & Timme, 1993) By focusing on cost management by the two types of banks and it was found that the revenue sides held most inefficient forces [4]. The studies have proved that Islamic banks are more efficient and profitable than conventional banks in other countries as well. The http://www.ojs.unito.it/index.php/EJIF ISSN 2421-2172 2 EJIF – European Journal of Islamic Finance No13, August (2019) practice of Islamic banking in Bangladesh is in its mid 30s with the emergence of Islami Bank Bangladesh Limited in 1983. Several since then several studies have conducted to decide upon its performance and significance of the country’s financial sector. This paper is one to find out the recent stability factor and level of confidence of the consumers of the two major types of banks operated in Bangladesh. III. METHODOLOGY Based on the availability of data 7 Islamic Banks (listed) and 24 conventional banks are taken into consideration for conducting the study. The period of data set is 2007 to 2017. The data are collected from the annual reports of the banks listed in the country’s stock exchange for the above mentioned period. The variables are: Return on banks average asset (ROA), Non performing asset proxied by loan loss provision over total asset (NPA), Liquid asset over total asset (LIQ), Equity over earning (TQ), Consumer confidence proxied by deposit over total liabilities and equity (CC), Cost income Ratio (CIR), Profit before tax (PBT), Loan loss provision over total loan (LLP), Equity to asset ratio. (ETA), Net loan over total asset (NLTA). Here, ROA, NPA, LI and TQ are the dependent variables while rest others being the independent variables [15]. A. Unit Root Test- Augmented Dickey Fuller Accurate result cannot be derived from a non- stationary time series data set. So, the precondition of running any econometric analysis is to conduct a unit root test. Apart from that, according to Engle and Granger (1987), a long-run relationship exists only when there is a similar order of integration between the variables [29]. The Augmented Dickey-Fuller (ADF) test is extensively used for testing stationarity of the variables (Dickey, Fuller, 1979, 1981) [32]. Later on a modification was incorporated by Phillips and Perron (1988) to give it a more comprehensive look [33]. The test is conducted at individual variables in level log form and the first differenced log form. If the log forms or first differenced log forms reject the null hypothesis (H0: series has a unit root), the time series is stationary. The unit root test is run on the basis of the following model: ∆yt = ∂ + by(t-1) -1 + et (1) Here, ∆ = 1st difference operator ∂ = constant/intercept et = error term Sometimes variables have auto correlation. To deal with this problem Dickey Fuller had developed the following three models: 1. Type 0=No intercept, no trend 2. Type 1=Intercept but no trend 3. Type 2=Intercept and trend ∆yi = β1yi-1+ � yj p j=1 ∆yi-j+ εi (2) ∆yi =β0+ β1yi-1+ � yj p j=1 ∆yi-j+ εi (3) ∆yi =β0+ β1yi-1+β2i+ ∑ yj p j=1 ∆yi-j+ εi (4) To make the data set stationary differentiation is required. Ho= Variables is not stationary or got unit root. H1= Variables is stationary or does not have unit root. B. Granger Causality Being proposed in 1969 by Clive Granger, Granger causality is used to test the appropriateness of one time series data for another. It tests the “predictive causality”. Regression may reveal a mere causality only. Stationary data set s a prerequisite for testing Granger causality and VAR model. A series of t- http://www.ojs.unito.it/index.php/EJIF ISSN 2421-2172 3 EJIF – European Journal of Islamic Finance No13, August (2019) test and F-test on lagged values of X can determine whether X variable Granger cause Y variable (Y value is also lagged). If the time series is already stationary then level data is used for this test. If the data set is non-stationary then first differentiation is used and then 2nd differentiation f required. H0: Null Hypothesis: Variable X does not cause variable Y H1: Alternative Hypothesis: Variable X causes variable Y. IV. RESULTS AND INTERPRETATIONS A. Descriptive Statistics Here (Table 01) the mean of PBT is for Islamic bank (2480) is higher than that of conventional banks (2060). This difference is attributed to two main factors. First, to some extent the nature of accounting treatment of PBT in profit - loss sharing arrangement may be counted as a financial cost, and second but to a great extent Islamic banks earn high PBT owing to its prices. Another difference is shown in NPA because Islamic banks show NPA under profit loss sharing adjustment. The mean value of cost income ratio, CIR for Islamic bank is 0.32 and for nonislamic banks is 0.45. 1) Islamic Banks TABLE 01: DESCRIPTIVE STATISTICS OF ISLAMIC BANKS Particulars CC CI ETA LI NL NPA PB RO TQ Mean 0.8 0.3 0.02 0.1 0.7 0.03 24 0.02 5.92 Median 0.8 0.3 0.07 0.1 0.7 0.03 18 0.01 5.97 Maximu 1.0 1.2 0.13 0.2 8.5 0.08 12 0.24 37.0 Minimum 0.6 - - 0.0 0.0 0.02 - -0.10 -8.56 Std. Dev. 0.0 0.3 0.16 0.0 1.0 0.01 32 0.04 6.28 Skewness - 0.8 - 0.2 7.0 1.13 1E- 2.46 1.88 Kurtosis 4.0 3.2 8.90 2.5 53. 4.15 5E- 19.13 12.4 Sum 51. 32. 1.06 11. 48. 2.14 2E 1.06 372. Sum 0.2 7.1 1.67 0.1 64. 0.01 7E 0.10 2446 2) Conventional Banks TABLE 02: DESCRIPTIVE STATISTICS OF CONVENTIONAL BANKS Particula CC CI ETA LIQ NLT NP PBT ROA TQ Mean 0.74 0.4 0.075 0.26 0.680 0.05 2060 0.0114 8.320 Median 0.80 0.4 0.073 0.24 0.702 0.05 1.82E+ 0.0122 6.024 Maximu 0.90 0.8 0.154 0.6 1.044 0.17 7.11E+ 0.0323 37.49 Minimu 0.02 0.2 - 0.03 0.068 0.01 - - - Std. 0.21 0.1 0.033 0.08 0.129 0.03 1.86E+ 0.0162 6.759 Skewnes - 1.4 - 0.58 - 1.38 -3.69E- - 2.112 Kurtosis 9.88 6 18.09 5.14 9.288 4.66 3.539E- 98.211 9.513 Sum 106. 64. 10.83 37.5 98.02 8.46 2.97E+ 1.6474 1198. Sum Sq. 6.32 1.6 0.160 0.94 2.404 0.21 4.92E+ 0.0377 6532. ADF test is conducted to test the stationary of the data set. If there is a trend in data set then 1st differentiation is needed to be conducted to remove the nonstationary property. If not fully removed then 2nd differentiation is applied. In this study all the variable are already stationary. The probability for ROA is 0.00, TQ is 0.01, CIR is 0.00, CC is 0.002, PBT is 0.003, NLTA is 0.03, LIQ is 0.004. That means variables do not have any unit root. TABLE 03: GRANGER CAUSALITY TEST Null Hypothesis: F-Statistic Probability CC does not Granger Cause LIQ 11.3918 2.00E-05 CC does not Granger Cause NPA 0.12935 0.0788 CC does not Granger Cause ROA 0.07598 0.0269 CC does not Granger Cause TQ 8.41658 0.0003 CIR does not Granger Cause LIQ 0.11311 0.02931 CIR does not Granger Cause NPA 0.26278 0.07693 CIR does not Granger Cause ROA 21.1024 8.00E-09 CIR does not Granger Cause TQ 0.24811 0.03806 ETA does not Granger Cause LIQ 0.77133 0.0342 ETA does not Granger Cause NPA 0.94309 0.00916 ETA does not Granger Cause ROA 4.005 0.0201 ETA does not Granger Cause TQ 1.93172 0.0483 NLTA does not Granger Cause LIQ 0.44216 0.0434 NLTA does not Granger Cause NPA 1.40924 0.02474 NLTA does not Granger Cause ROA 20.5199 1.00E-08 http://www.ojs.unito.it/index.php/EJIF ISSN 2421-2172 4 EJIF – European Journal of Islamic Finance No13, August (2019) NLTA does not Granger Cause TQ 2.7111 0.0696 PBT does not Granger Cause NPA 0.64837 0.0243 PBT does not Granger Cause ROA 0.02564 0.05247 PBT does not Granger Cause TQ 3.08221 0.0487 PBT does not Granger Cause LIQ 0.44744 0.0401 As the value of the almost all of the probability is less than 0.05 so there is a causality relation between them. But few variables show the opposite outcome. B. Fixed Vs. Random Effect- Hausman Test The probability of the Hausman Test is derived as 1.00 and it indicates that Random Effect Model is best suitable for this data set to conduct regression model. 90% confidence level is considered and 10% level of significance is allowed in conducting the regression model. C. Random Effect Regression TABLE 04: COEFFICIENTS OF THE INDEPENDENT VARIABLES WITH ROA Islamic Bank Conventional Bank All Bank Cons 0.141295 0.004964 0.026824 CC 0.145595 0.003866 0.003138 CIR -0.024832 -0.025941 -0.053455 PBT 2.80E-12 7.01E-12 2.98E-12 ETA 0.127956 -0.007491 0.002977 NLTA 0.014979 0.016668 -0.010159 TABLE 05: COEFFICIENTS OF THE INDEPENDENT VARIABLES WITH TQ Islamic Bank Conventional Bank All Bank Cons 17.56869 0.80658 10.41641 CC -28.93446 -9.349169 -9.584313 CIR -4.688297 -19.01269 -9.9664 PBT -5.43E-10 1.12E-09 -6.86E-10 ETA -11.75549 112.2729 -29.33747 NLTA -1.708535 1.995513 -0.578676 TABLE 06: COEFFICIENTS OF THE INDEPENDENT VARIABLES WITH LIQ Islamic Bank Conventional Bank All Bank Cons 0.036708 0.227645 0.20999 CC 0.144223 0.002817 -0.017617 CIR 0.039501 0.082031 0.05965 PBT 2.41E-12 -1.24E-12 1.22E-12 ETA 0.157697 -0.217199 0.204524 NLTA 0.004615 -0.025824 0.007305 TABLE 07: COEFFICIENTS OF THE INDEPENDENT VARIABLES WITH NPA Islamic Bank Conventional Bank All Bank CC 0.046574 0.088268 0.083362 CC 0.019791 0.027896 -0.058131 CIR 0.000376 0.044299 0.011664 PBT 7.22E-13 2.66E-12 1.71E-12 ETA 0.001104 -0.168077 0.039833 NLTA 0.001963 -0.068857 0.000984 These tables (4,5,6,7) give the comparative results of the four regressions, which shows how the two types of banks are impacted by changes in financial conditions. ROA which is the return on average asset is one stability factor. CIR is inversely and insignificantly related in both types of banks. A rise in cost decreases the stability of the bank. ETA is significantly and positively related in case of Islamic bank. But it shows inverse and insignificant result for conventional banks. PBT gives expected positive and significant relationship in all cases. Increase in PBT increases the stability and consumer confidence level. Furthermore, Islamic banks reported very small NPA, and have shown a positive and significant relationship with liquidity. NLTA shows expected inverse relation with liquidity factor but it is insignificant in case of Islamic banks. CC represents consumers’ confidence. And positive result in all the cases but with an exception with conventional bank in TQ factor has been found. http://www.ojs.unito.it/index.php/EJIF ISSN 2421-2172 5 EJIF – European Journal of Islamic Finance No13, August (2019) And the result is more significant with the Islamic banks than conventional banks. The effect of instability is measured by Tobin’s Q model. It also assesses whether there is similar impact on conventional and Islamic banks. Tobin Q for all banks is inversely, but significantly, related to bank specific factors of PBT, ETA, NLTA, CC, CIR for all banks. This indicates that the bank value increases with the decrease in stability factor. On the other hand, a positive relationship for PBT is found for conventional banks indicating that profitability factors posit significantly for bank values. D. Significance of Model TABLE 8: SIGNIFICANCE TEST Islamic Bank ROA TQ LIQ NPA F-Statistics 4.368033 3.71891 2.02392 0.53405 Prob (F-Stat) 0.001957 0.00555 0.08889 0.047 Conventional Bank ROA TQ LIQ NPA F-Statistics 57.08982 14.0647 0.47745 4.64342 Prob (F-Stat) 0.05 0.01 0.09 0.0006 All Bank ROA TQ LIQ NPA F-Statistics 11.11861 10.6274 1.04802 3.13205 Prob (F-Stat) 0.005 0.002 0.039 0.00957 As it is known that if the P value of F statistics is less than 0.10 then the model is significant. The lesser the value the more significant the model become. It is seen that in most of the cases P value of F statistics is less than 0.10 so the regression models are significant enough to describe the relationship among dependent and independent variables. V. FINDINGS AND CONCLUSION Conventional banks charge average fixed interest rate regardless of the profitability of the project. Islamic banks charge proportional cost of financing in according to the profitability of the project which is totally contrary to the system of conventional. Conventional banks do not always utilize their full investment opportunities. Some portion is remained unused. On the other hand, Islamic banks make the best use of their investment opportunities which were untapped in the economy. This leads to a direct linkage between the success of the project and the income of Islamic Banks. In other words, the rise and fall in the projects’ return financed by a bank effects its net income. Thus the Islamic banking system is very much concerned about the performance of the project for which financing has been provided. Productivity is hampered in two folds in conventional banks. First of all, the capacity and resources are not used to recover non performing loans rather these are being involved in choosing new projects. Loan loss provision also hampers profitability. On the other hand, Islamic bank practice profit-loss-sharing mechanism which is a fruitful way to recover loans. The borrower faces a fixed or nonflexible loan payment schedule in conventional banks. It burdens them with an uncertainty in cash flows. Consequently, the economy with more of conventional banks faces a cyclical volatility in its performance. This instability continues because of the fixed payment schedule. But the Islamic banks receive a certain portion of the income derived by its investments in addition to the principal payment. And it is not obligatory to make payment when there is no earning. Loss is also shared by the Islamic banks in proportionate basis. This mechanism encourages the entrepreneurs to a good extent. They need not http://www.ojs.unito.it/index.php/EJIF ISSN 2421-2172 6 EJIF – European Journal of Islamic Finance No13, August (2019) to pay a higher (fixed) amount when lower profit is earned. In this way the gap or spread between profits and payment commitment is lessened by large margin. All this issues have uphold the consumers’ faith in Islamic banks more than conventional banks. And thus deposit is also higher in case of Islamic banks. As it is said before that stability and performance comes hand in hand with consumer confidence, Islamic Banks have managed to outperform in stability factor as well. REFERENCES [1] AI-Abdullatif, S. Abdullah, “The application of the AAOIFI accounting standards by the Islamic banking sector in Saudi Arabia”, Doctoral dissertation, Durham University, 2007. [2] Alaro, A. Razzak., & M. Hakeem, “Financial engineering and financial stability: the role of Islamic financial system”, Journal of Islamic Economics, Banking and Finance, vol. 7(1), 25-38, 2011. [3] A. Levin, C. F. Lin, and C. S. J Chu, “Unit root tests in panel data: asymptotic and finite-sample properties”, Journal of econometrics, vol. 108(1), pp. 1-24, 2002. [4] A. N. Berger, W. C. Hunter, and S. G. Timme, “The efficiency of financial institutions: A review and preview of research past, present and future”, Journal of Banking & Finance, vol. 17(2-3), pp. 221-249, 1993. [5] A.Q.M. S. Arif, Secretary General, Central Shari‘a Board for Islamic Banks of Bangladesh, An Interview, 16 April, 2014. [6] Azad, M. A. Kalam., and Â. M. R Sir, “Prospects analysis of an Islamic capital market in Bangladesh”, Global Journal of Management And Business Research, 2013. [7] Bader, M. I. Khaled, S. Mohamad, M. Ariff, and T. H. Shah, “Cost, revenue, and profit efficiency of Islamic versus conventional banks: international evidence using data envelopment analysis”, 2008. [8] F. Allen, D. Gale, “Competition and Financial Stability. Journal of Money, Credit and Banking”, vol. 36(3), pp. 453-480, 2004. [9] G. Bekaert, C. R. Harvey, C. Lundblad, and S. Siegel, “Global growth opportunities and market integration”, The Journal of Finance, vol. 62(3), pp. 1081-1137, 2007. [10] G. L. Kaminsky, and C. M. Reinhart, “The twin crises: the causes of banking and balance-of-payments problems”, American economic review, vol. 89(3), pp. 473-500, 1999. [11] H. A. Loghod, “Do Islamic banks perform better than conventional banks? Evidence from Gulf cooperation council countries”, Journal of Management, vol. 7(3), pp. 56-72, 2010. [12] H. I. Mobolaji, and A. R. A. Alaro, “Financial Engineering and Financial Stability: The Role of Islamic Financial System”, Journal of Islamic Economics, Banking and Finance, vol. 113(464), pp. 1-14, 2011. [13] I. S. Drissi, and K. Angade, “Islamic financial intermediation the emergence of a new model”, European Journal of Islamic Finance, vol. 12. pp. 1-7, 2019. [14] J. C. Rochet, “Capital requirements and the behaviour of commercial banks”, European Economic Review, vol. 36(5), pp. 1137-1170, 1992. [15] K. Hussein, “Bank-Level Stability Factors and Consumer Confidence—A Comparative Study of Islamic and Conventional Banks’ Product Mix”, In Islamic Finance, pp. 86-104. Palgrave Macmillan, Cham, 2016. [16] K. Mukminin, “Profit maximization in islamic banking: an assemblage of maqasid shariah conception”, European Journal of Islamic Finance, vol. 12, 2019. [17] S. Kwan, R. A. Eisenbeis, “Bank risk, capitalization, and operating efficiency”, Journal of financial services research, vo. 1;12 (2-3), pp. 117-31, 1997. [18] M. Ariff, “Islamic banking, a variation of conventional banking”, Monash Business Review, vol. 4(3), pp. 1-8, 2006. [19] M. Hasan, and J. Dridi, “The effects of the global crisis on Islamic and conventional banks: A comparative study”, Journal of International Commerce, Economics and Policy, vol. 2(02), pp. 163-200, 2011. [20] M. Iqbal, P. Molyneux, and S. Conermann, “Thirty years of Islamic banking. History, Performance and Prospects”, Bankhistorisches Archiv, 32(2), 155-158, 2006. [21] M. K. Hassan, M. and A. H. M. Bashir, 2003 “Determinants of Islamic banking profitability”, 10th ERF annual conference, Morocco, vol. 7(1), pp. 2-31, 2003. [22] M. K. Hassan, and M. F. Dicle, “Basel II and regulatory framework for Islamic banks”, Journal of Islamic Economics, Banking and Finance, vol. 1(1), pp. 1-16, 2005. [23] M. L. Rahman, and S. H. Banna, “Liquidity Risk Management: A Comparative Study between Conventional and Islamic Banks in Bangladesh”, Journal of Business and Technology (Dhaka), vol. 10(2), pp. 18-35, 2015 [24] M. Rahim, S. Rohaya, and R. H. Zakaria, “Comparison on stability between Islamic and conventional banks in http://www.ojs.unito.it/index.php/EJIF ISSN 2421-2172 7 EJIF – European Journal of Islamic Finance No13, August (2019) Malaysia” Journal of Islamic Economics, Banking and Finance, vol. 113(915), pp. 1-19, 2013. [25] M. U. Chapra, “The future of economics: An Islamic perspective”, Kube Publishing Ltd. vol. 21, 2016. [26] P. P. Biancone, and M. Radwan, “European companies: evaluation for sharia compliance “opportunities and challenges”, European Journal of Islamic Finance, vol. 5, 2016. [27] P. P. Biancone, and M. Radwan, “Social Finance And Unconventional Financing Alternatives: An Overview”, European Journal of Islamic Finance, vol. 10, pp. 1-6, 2018. [28] R. Ameer, R. Othman, and N. Mahzan, “Information asymmetry and regulatory shortcomings in profit sharing investment accounts”, International Journal of Islamic and Middle Eastern Finance and Management, vol. 5(4), pp. 371-387, 2012. [29] R. F. Engle, and C. W. Granger, “Co-integration and error correction: representation, estimation, and testing, Econometrica” journal of the Econometric Society, pp.251-276, 1987. [30] S. Kwan, and R. A. Eisenbeis, “Bank risk, capitalization, and operating efficiency”, Journal of financial services research, vol. 12(2-3), pp. 117-131, 1997. [31] T. Beck, Demirgüç-Kunt, A., and O. Merrouche, “Islamic vs. conventional banking: Business model, efficiency and stability”, The World Bank, 2010. [32] A. D. Dickey, D. P. Hasza, and W. A. Fuller, “Testing for unit roots in seasonal time series”, Journal of the American Statistical Association, vol. 79(386), pp.355-367, 1984. [33] P. V. Phillips, and P. Perron, “Testing for a unit root in time series regression”, Biometrika, Vol. 1;75(2), pp. 335- 46, 1988. http://www.ojs.unito.it/index.php/EJIF ISSN 2421-2172 8 EJIF – European Journal of Islamic Finance Editorial Team Editor in Chief Prof. Paolo Pietro Biancone, University of Turin, Italy Editorial Board Prof. Dian Masyita, University of Padjadjaran, Indonesia Prof. Abdulazeem Abozaid, Qatar Faculty of Islamic Studies, Qatar Prof. Ahmad Aref Almazari, King Saud University, Saudi Arabia Prof. Marco Meneguzzo, Università degli Studi di Roma "Tor Vergata", Italy Prof. Nidal A. Alsayyed, Inayah Islamic Finance Research Institute, USA Prof. Roberta Aluffi, University of Turin, Italy Prof. Ghassen Bouslama, NEOMA Business School, Campus de Reims, France Prof. Nazam Dzolkarnaini, Salford University, UK Prof. Kabir Hassan, University of New Orleans, USA Prof. Khaled Hussainey, University of Plymouth, UK Prof. Rifki Ismal, University of Indonesia Prof. Tariqullah Khan, Hamad bin Khalifa University, Qatar Prof. Ali Khorshid, ICMA Centre Reading University, UK Prof. Amir Kia, Utah Valley University, USA Prof. Laurent Marliere, Université Paris-Dauphine, France Prof. Federica Miglietta, University of Bari, Italy Prof. Hakim Ben Othman, University of Tunis, Tunisia Prof. Mohamed Ramady, King Fahd University of Petroleum and Minerals, Saudi Arabia Prof. Mamunur Rashid, Nottingham University, Malaysia Prof. Younes Soualhi, International Islamic University, Malaysia Prof. Laurent Weill, University of Strasbourg, France I. Introduction II. Literature Review III. Methodology A. Unit Root Test- Augmented Dickey Fuller B. Granger Causality IV. Results and Interpretations A. Descriptive Statistics 1) Islamic Banks 2) Conventional Banks B. Fixed Vs. Random Effect- Hausman Test C. Random Effect Regression D. Significance of Model V. Findings and Conclusion References