Comparative analysis of bank performance between
Islamic banks and conventional banks with respect to
stability, asset quality and efficiency.
Bachelor Finance
Student: Shannon Sewnandan Number: 6148379
Field: Finance & Organization Supervisor: Mark Dijkstra Completion: July 2014
Abstract
This study aims to conduct a comparative analysis on the impact of the global financial crisis on the Islamic and conventional banks. Three performance indicators namely stability, asset quality and efficiency are being considered. The study covers a ten-year period from 2002 to 2012. In methodology, the study adopts the regression analysis on a sample of 230 Islamic banks and 1260 conventional banks is 23 Muslim countries. The study documents two main findings: (i) the Islamic banks are more stable than conventional banks during the crisis and (ii) there are no differences in asset quality and efficiency, among the two types of the banks due to the financial crisis. Because Islamic banks follow the rules according to the Sharia, interest is prohibited. Thus they are not exposed to the interest risk.
1. Introduction
As the financial crisis has been affecting banks worldwide, an unconventional banking system, Islamic banking, is introduced to reduce financial crisis. Islamic banking experienced an annual growth rate of 12% to 15% per year (Iman and Kpodar, 2010). Future projections of the IMF for Islamic banking also suggest that in many Muslim countries Islamic banking would become an alternative to the conventional banking sector and an important part of the global financial sector in the sense that banks would become relatively more stable than conventional banks.
In 1962 the first Islamic financial institution in Malaysia was established, called the Pilgrimage Fund. The establishment of the first Islamic bank, Dubai Islamic bank, started in in 1970. Three years after this, a first print of the Handbook of Islamic banking, appeared with prescriptions about the rules within Islamic banking. Islamic banks operate upon principles of Islamic law (Sharia) which requires risk sharing, through profit-and-loss sharing and prohibits interest (Riba) (Beck, Demirgüç-Kunt, & Merrouche , 2013).
This paper aims to clarify a comparative analysis between Islamic and conventional banks with respect to stability, asset quality and efficiency which defines bank performance in this paper. The impact of the financial crisis on the Islamic and conventional banks is also taken into account as contribution to the existing literature. With the use of different
indicators, explanatory and control variables a regression is conducted to see if there are any differences. The sample used for the model covers a period before and during the crisis 2002-2012, and is based on Islamic and conventional banks in Muslim countries selected on the basis of their asset value in million dollars. This leads to the following research question: How do Islamic banks perform regarding their stability, efficiency and asset quality, comparing to conventional banks?
Beck, Demirgüç-Kunt, & Merrouche (2013), compared bank performance of Islamic and conventional banks terms of asset quality, their results do not show any significant differences between Islamic banks and conventional banks. Whereas in terms of efficiency, Islamic banks have significantly lower cost-income ratios (6.4 %) and lower overhead costs (0,9%). Which implies lower efficiency for Islamic bank and small significant differences. Cihak & Hesse (2010) compared bank performance with the use of a z-score, an indicator for stability.Their results shows that Islamic banks (20,2%) are more stable than conventional banks (13,3%).
explains methodology and data. This chapter is the foundation of the actual analyses, which are shown in chapter 4. The hypotheses are tested in this chapter while chapter 5 contains the conclusion.
2. Literature
Islamic banks accord to the tenets of the Sharia, the Islamic jurisprudence. Chikak & Hesse (2010), define Islamic banking as the provision and use of financial products and services that are being offered conform Islamic religious practices and laws.
Financial transactions regarding to Islamic banking cannot include an interest payment (Riba). Literally Riba means increase, addition expansion or growth. Riba is an prohibition on the advanced determination of an positive return on a loan. Any form of
interest, fixed and variable, is prohibited regardless for which purpose the loan is made (Ariff, 1988). According the view of the Quran, the holy book of the Islam, Asad (1980) defines interest as any income ‘…obtained through interest-bearing loans involving an exploitation of
the economically weak by the strong and resourceful: an exploitation characterized by the fact that the lender, while retaining full ownership of the capital loaned and having no legal concern with the purpose for which it is to be used or with the manner of its use, remains contractually assured of gain irrespective of any loses which the borrower may suffer in consequence of this transaction’. The Sharia forbids Riba, however trade is allowed (Siddiqi, 2005).
The Sharia describes two more principles to which Islamic banks accords to, profit-and-loss sharing principle and the mark-up principle. The profit-profit-and-loss sharing principle (PLS) suggest that the profits and losses are shared between the parties involved in a financial transaction, the borrower (investor) and the lender (bank). Thus, a bank will not receive a fixed rate of interest on their outstanding loan, on the contrary, they share in profits or in losses of the business owner to whom the loans are provided (Chikak & Hesse, 2010).
According to the PLS principle, Aggerwal & Yousef (2000) describe two instruments that are used by Islamic banks. The first instrument is Mudarabah financing. A Mudarabah contract is a certain kind of partnership where one party lends money to another party to be able to invest in a commercial enterprise. The first person, who is called Rabb-ul-mal, is the investor, while the second person, exerts skills and effort in the enterprise, the mudarib. The banks receives no return or negative return on its investment in case of a loss. Likewise the investor receives
no compensation for her effort. In case of a gain, returns are shared between the bank and the investor according to the negotiated percentage. Second instrument is Musharakah, which means partnership. The banks provides funds needed for the project from depositor’s savings while the entrepreneur offers labor and expertise. The profits (or losses) are then shared between the bank, the depositors and the entrepreneur at a fixed ratio. By investing their funds jointly with their customers, Islamic banks become partners and will have to share the risk with both depositors and entrepreneurs. This requirement of risk sharing makes return on equity of Islamic banks higher than for conventional banks. This method is mainly applied to the stock market, in which the public, banks and even Government may submit a tender (Taktak, Zouari & Boudriga, 2010).
The mark-up principle is associated with financing the purchase of an asset in
exchange for a negotiated profit margin. Based on these principles there are two instruments, namely Murabaha and Ijara financing. Murabaha financing, where the banks purchases an asset on behalf of the entrepreneur. The bank resells the asset to the entrepreneur for a higher, predetermined price. This predetermined price covers the original purchase prices and an added negotiated profit margin. The payment by the entrepreneur is made in the future. Ijara financing, where the bank purchases an asset and the entrepreneur is allowed to use the asset for a fixed charge. The ownership of the asset remains with the bank. Ijara financing is a contract known as leasing (Siddiqi, 2005).
Another principle that is prohibited by the Sharia is Gharar. Gharar means
uncertainty, risk or speculation. There is mention of gharar in a transaction when an essential element of a contract is uncertain or omitted, relating to the price, quality, quantity or delivery of the goods. However, it is not possible to compose all financial transactions without risk. The prohibition of gharar has therefore mainly relation to incomplete information at the contract formation, and does not cover business risks. By a desire to eliminate gharar, the financial Islamic banks want to achieve more transparency and fairness in markets (El-Gamal, 2006).
For Islamic banks equity financing ranged from 42% of total financing in 2002 to 55% in 2012 (Dixon, 2002). Islamic banks do not pay interest on customers deposit accounts. Instead, they place customer funds into profit-sharing investment accounts (PSIA). According to the profit-and-loss sharing principle, the bank provides its customers with returns on the assets in which customers funds are invested. And if these returns are negative, so are the returns to the customers. Whereas in conventional banking system, customers receive interest on their
deposit accounts. Since Islamic banks operate with interest free instruments, they are not exposed to interest risk. Because they determine prior to the contract a fixed rate, it makes the Islamic banks invulnerable to interest rate fluctuations (Karim & Ali, 1989).
Islamic banks seem to experience more credit risk due to long term financing through
mudarabah of musharaka, than conventional banks do. This is risk in the sense that the
borrower default in paying back the loan. Within Islamic banks, it is assumed that there is a trust relationship between the bank and the borrower. Therefore Islamic banks are required to gather information actively about the credit worthiness of the borrower. Otherwise, borrowers can easily manipulate gains and losses in their favor, because the relationship is only based on trust. Conventional banks however also know relationship lending, which is defined as the provision of financial services by a financial intermediary (bank) on the basis of long-term investment in obtaining firm-specific information or borrower-specific information through multiple interactions with the customer. This is not the same as in Islamic banking because, conventional banks receive interest returns whereby they can cover a part of the risk
(Elyasianie & Goldberg, 2004).
Since mudarabah account holders are subject to the profit-and-loss sharing
arrangements, like shareholders, they need disclosure of sufficient information about their investments to protect their interests. Conversely, due to the nature of a mudarabah contract,
mudarabah account holders do not have any access to that information. Which leads to an
asymmetric information problem between a principal (mudarabah account holders) and an agent (Islamic bank). Likewise, they have no role in the corporate governance of a bank since they do not have any voting rights to elect the board of governors of the Islamic banks. Therefore, to discipline bank management, mudarabah account holders have to rely only on monitoring of the board on behalf of the shareholders. Mudarabah account holders can gain some benefit out of monitoring if interests of the shareholders and mudarabah account holders coincide. These issues may give rise to complex agency problems as well as it may lead to moral hazard on the part of the bank management. Management of an Islamic bank may protect its own interest or the interest of its shareholders instead of those of mudarabah account holders (Archer,Karim & Al-Deehani, 1998).
Cihak & Hesse (2010) compared bank performance of 77 Islamic and 397 conventional banks in 18 Islamic countries. The sample was based on the period 1993 to 2004. They used stability as a measure for bank performance. Stability is defined as the probability of a banks
(ROA) and the capital to asset ratio (CAR) divided by the standard deviation of the return on assets. In other words the probability that the value of a bank’s assets become lower than the value of its debt. This study finds that small Islamic banks with total assets below $1 billion more stable than small conventional banks. Whereas large conventional banks with total assets above $1 billion, are more stable than large Islamic banks. They also conclude that small Islamic banks are more stable than large Islamic banks.
Also Beck, Demirgüç-Kunt, & Merrouche (2013), compared bank performance of Islamic and conventional banks employing 22 Islamic countries in the period 1995-2007. They used asset quality as the first measure, which is defined as three indicators namely: loss reserves, loan loss provisions and non-performing loans. The second measure is efficiency for which two indicators were used, overhead costs and cost–income ratio. In terms of asset quality, their results do not show any significant differences between Islamic banks and conventional banks. Whereas in terms of efficiency, Islamic banks have significantly lower cost-income ratios (6.4 %) and lower overhead costs (0,9%). Which implies lower efficiency for Islamic bank and small significant differences.
Abedifar, Tarazi and Molyneux (2010), evaluate the risk and stability characteristics of Islamic banking. They used a sample of 553 banks from 24 countries between 1999 and 2009. Their study finds that small Islamic banks that are more leveraged and based in countries with predominantly Muslim population, have 13% lower credit risk than conventional banks. The z-score is defined in the same way as Cihak & Hesse did, return on assets (ROA) and capital to asset ratio (CAR) divided by the standard deviation of ROA. Results shows that on average the z-score is 4,67% higher for small Islamic banks than for similarly sized conventional banks.
Bonin and Hasan (2005) focus more on the effect of bank efficiency and ownership. They used a panel data of 220 banks and 830 observations from the period 1996 to 2000 and computed efficiency scores with the use of stochastic frontier estimations. Explanatory variables as ownership type, year and bank size were part of the regression with efficiency as the dependent variable.
3. Methodology
This paper aims to compare the performance of Islamic and conventional banks before and during the last financial crisis. In order to do so a regression analysis will be done. The model is as follows:
Where is one of the three measures of stability, asset quality and efficiency of bank i in year t.
Three different measures are used to define bank performance; stability, asset quality and efficiency.
Stability is defined as the z-score, which is z = (ROA+CAR)/SD(ROA) (where profits are normal distributed). The z-score is the first measure that measures the risk of insolvency or distance to default. Insolvency is defined as a state where bank’s capital to asset ratio (CAR) and return on assets (ROA) is smaller than zero, (CAR+ROA<0). Capital to asset ratio (CAR) is calculated by equity(E) divided by total assets(A), (CAR = E/A). Return on assets (ROA) is calculated by profit(π) divided by total asset(A), (ROA = π/A). Z indicates the number of standard deviation that a bank’s return on assets has to drop below its expected value before equity is depleted and the bank is insolvent. Therefore a higher z-score indicates that a bank is more stable (Chikak & Hesse, 2012).
Asset quality is the second measure that is used to evaluate bank performance. Three indicators are used to define asset quality: loss reserves, loan loss provisions and
non-performing loans, as percentage of total assets, all scaled by gross loans. Where loss reserves are loans that have not been paid yet or losses which have occurred during a certain time frame. The second indicator, loan loss provision describes an expense that is reserved for defaulted loans or credits. It is an amount set aside in the case that the loan defaults. A loan that is in or close to default is called a non-performing loan (Beck, Demirgüç-Kunt, & Merrouche 2013).
Efficiency is measured with the cost-income ratio, which is overhead costs divided by gross revenues. Overhead costs are also known as operating expenses and refers to an ongoing expense of operating a business. A decrease in the cost-income ratio indicates either
decreasing costs or increasing revenues. Therefore the lower this ratio the more efficient the bank is (Foster & Gupta, 1990).
To distinguish the impact of bank type, a dummy variable is included in the regression that takes the value of 1 if the bank in question is an Islamic bank, and 0 if the bank is a conventional bank. Selection is based on Islamic banks which are 100% Islamic and conventional banks which are 100% conventional.
million dollars. The natural logarithm is used in this model. This is because the natural logarithm of assets is skewed, so that outliers might not influence to much (Stock & Watson, 2012). Skewness for bank size, measured in assets is 22,57.
Profit before tax, measured in dollars, is also included in the regression as an
explanatory variable. According to Chikak & Hesse (2008) the more profitable a bank is the more it can retain for reserves, which serves as a buffer to make banks less vulnerable for economic shocks. Because, the tax regulation in every country is different, the profit before tax is used.
The second dummy is for controlling the period before the financial crisis (2002-2007) and during the financial crisis period (2007-2012). The dummy takes the value 1 for the period during the crisis and the value 0 for the period before the crisis.
Table 1
Summary statistics
Reports the summary statistics of the variables used in the analysis for the whole sample of countries that host both Islamic bank and conventional banks.
Variable Obs Mean Standard
Deviation Minimum Maximum
Stability
Z-score (%) 12455 6,445 8,585 -3,449 43,725
Asset quality
Loss loan provisions (%) 781 6,509 10,936 0 22,14
Loss loan reserves (%) 778 0,014 0,044 0,008 1
Non-performing loans (%) 452 12,026 18,731 0 50,33 Efficiency Cost-income ratio (%) 12456 78,910 24,827 0 91,47 Explanatory Variables (mln $) Assets 12456 18,681 62,44 10 783,82 lnAssets 12456 6,568 3,686 2,332 17,179 Profit 12456 93,167 1.098,999 -13.350,87 30.808,91 Control variables Employees (mln) 635 3.714,79 17.647,81 0 148.5 lnEmployees 635 5,135 2,174 0,693 11,908
Table 2
Reports the averages of all variables used in the analysis for the Islamic banks and conventional banks in the whole sample. Variable Islamic banks Conventional Banks Stability Z-score (%) 8,868 6,299 Asset quality
Loss loan provisions (%) 0,021 0,013
Loss loan reserves (%) 9,753 6,278
Non-perfoming loans (%) 8,776 12,333 Efficiency Cost-income ratio (%) 78,967 78,907 Explanatory Variables (mln $) Assets 52.145,5 144.859,3 lnAssets 6,811 6,554 Profit 35,280 96,660 Control variables Employees (mln) 707,75 3.853,281 lnEmployees 6,070 5,092
Leverage ratio (D/E) (%) 75,920 82,868
Yeh (1996) used the number of employees working at a bank as a control variable to evaluate bank performance. The expansion of Islamic banking word wide, requires an increasing number of employees. For this variable the natural logarithm is used for the same reason that it is used for assets.
The second control variable that is added to the regression model is the debt to equity ratio. Which is defined as Debt (D) divided by Equity (E), D/E, and is calculated by total liabilities divided by equity. According to Yeh (1996) lower leverage ratios indicates that a bank is less dependent on borrowing for its operations.
4. Data
To construct and compare measures for bank performance, data for all variables is obtained from Bankscope.
The sample comprises 27 countries and 3235 banks, out of which 232 Islamic banks and 3005 conventional banks. In table 3, the number of Islamic banks and conventional banks of 27 countries in Bankscope that host both Islamic banks and conventional banks are presented. The sample is selected on the basis of countries that have at least one Islamic bank and is restricted to countries that have at least two years of reporting data in Bankscope. Only Islamic banks which are 100% Islamic and conventional banks which are 100% conventional are used for the sample. Also commercial conventional banks are not included in the sample because Islamic banks are not commercial as well. This reduced the sample by 230 Islamic banks and 1260 conventional banks operating in 23 countries over the period 2002-2012. All banks in the sample are ranked by their asset value. The asset value of the smallest Islamic bank is 10 million dollar. By using 10 million asset value as the cutoff point only conventional banks with an asset value of at least 10 million are selected. Not all data concerning banks that have less than 10 million asset value is available in Bankscope. The sample contains over the period 2002 to 2012, thus pre-dating (2002-2007) and during (2008-2012) the recent global financial crises.
Table 1 shows that the z-score varies from -3,45 to 43,73 in the sample, with an average of 6,45. Table 2 shows that Islamic banks have a significantly higher z-score (8,87), suggesting that they are more stable than conventional banks (6,30).
Loan loss provisions range from 0% to 22%, with an average of 6,51%. Loan loss reserves range from 0,008 % to 1%, with a mean of 0,014%. Non-performing loans, finally, range from 0% to 50,33%, with an average of 12.03%. Islamic banks have significantly higher loss loans provisions (0,021%) and loss loan reserves (9,75%) than conventional banks (0.013%) respectively (6,28%). Non-performing loans are significantly higher for
conventional banks (12,33%) compared to Islamic banks (8,78). As loss loan reserves are determined ex ante and non-performing loans ex post, this can lead to difference in results. This is because an under – or overestimation regarding to loss loan reserves could be made in the sense that loss loans reserves are determined before a loan has been paid. While non-performing loans are determined after knowing if the loan is paid or not. Non-non-performing loans are therefore more accurate than loss loan reserves.
The cost-income ratio ranges from 0% to 91,47%, with an average of 78,91%. This ratio is 78,97 for is Islamic banks and 78,91 for conventional banks, which shows that there are no significant differences in efficiency.
5. Results
This chapter discusses the results from the model, which estimates bank stability, asset quality and efficiency of Islamic banks compared with conventional banks.
Table 3 shows that there is a significant difference in stability between Islamic banks and conventional banks as the Islamic bank dummy is significant at the 1% level with a coefficient of 0,26. All variables are statistically significant at the 5% level except for the variable leverage ratio. Reason for this could be that investments of Islamic banks are
primarily based on equity financing. Which results in lower leverage ratios for Islamic banks. Conventional banks use both debt and equity for their investments. The R² in the model is 44%. Which means that 44% of the model is explained by the explanatory variables. Based on the dataset used and analysis done on this dataset, Islamic banks are more stable than
conventional banks.
Table 3
This table reports regression results for the dependent variable Stability. The regression coefficients are reported without brackets. The standard errors clustered by bank are reported in the brackets.
Z-score Z-score Z-score Z-score Z-score
dum_islam 2,57** 2,81*** 2,85*** 2,42* 0,26** [1,07] [0,97] [0,97] [1,30] [0,13] lnAssets -0,98*** -0,99*** -0,37*** 0,02** [0,06] [0,06] [0,10] [0,01] Profit 0,0004* 0,0002 0,05** [0,0002] [0,0002] [0,25] lnEmploy -0,88*** -0,01 [0,15] [0,02] Leverage -0,37*** [0,001] Constant 6,30*** 12,69*** 12,81*** 12,09*** 36,55*** [0,26] [0,46] [0,47] [0,73] [0,12] N 12455 12455 12455 12455 12455 R2 0,08 0,18 0,30 0,31 0,42 Adjusted R2 0,07 0,17 0,29 0,29 0,42
* Two sided significance at10% ** Two sided significance at 5% *** Two sided significance at 1%
Table 4 provide information about the three indicators to measure bank asset quality: loss reserves (Panel A), loan loss provisions (Panel B) and non-performing loans (Panel C). Regression results shows that there are no significant differences in asset quality between Islamic banks and conventional banks. The Islamic bank dummy is insignificant for the regressions loan loss reserves and loan loss provisions. For non-performing loans, the Islamic bank dummy is significant at the 10% level. The R² in this model is 12% for loss loan
reserves, 5% for loan loss provisions and 11% for non-performing loans. Though the R² is not high for all indicators, almost all the variables are statistically significant at 10%, 5% of 1% level. Except for the variable profit before tax in the loss loan reserve regression, the natural logarithm of assets and profit before tax in regression for the loan loss provisions and the natural logarithm of employees in the non-performing loan regression. According to the regression results, there is no difference in asset quality between Islamic banks and
conventional banks in the model.
Table 4 Panel A
This table reports regression results for the dependent variable Asset quality for all three indicators. Panel A reports results for loss loan reserves, Panel B for loan loss provisions and Panel C reports for non-performing loans. The regression coefficients are reported without brackets. The standard errors clustered by bank are reported in the brackets.
Loss loan reserves Loss loan reserves Loss Loan reserves Loss loan reserves Loss loan reserves dum_islam 3,48** 3,88** 3,89** -0,85 -1,41 [1,57] [1,55] [1,55] [1,75] [1,69] lnAssets -0,43*** -0,44*** -0,67*** -0,63*** [0,11] [0,11] [0,14] [0,14] Profit 0,0002 0,0003 0,0002 [0,0003] [0,0002] [0,003] lnEmploy 0,65*** 0,87*** [0,22] [0,22] Leverage -0,18*** [0,03] Constant 6,28*** 8,99*** 9,04*** 5,57*** 19,74*** [0,40] [0,77] [0,78] [1,14] [2,71] N 781 781 781 781 781 R2 0,0063 0,03 0,03 0,05 0,12 Adjusted R2 0,0050 0,025 0,027 0,04 0,11
* Two sided significance at10% ** Two sided significance at 5% *** Two sided significance at 1%
Panel B Loan loss provisions Loan loss provisions Loan loss provisions Loan loss provisions Loan loss provisions dum_islam 0,007 0,007 0,007 0,007 0,006 [0,006] [0,006] [0,006] [0,005] [0,005] lnAssets 0,00035 0,00035 -0,00055 -0,0005 [0,00044] [0,00045] [0,0004] [0,0004]
Profit 7.74E-08 5.31E-07 4.78e-07
[1.42e-06 ] [7.28e-07 ] [ 7.23e-07]
lnEmploy 0,002*** 0,002*** [0,0006] [0,0006] Leverage -0,0002*** [0,00008] Constant 0,013*** 0,011*** 0,011*** 0,0027 0,2*** [0,002] [0,003] [0,003] [0,003] [0,007] N 778 778 778 778 778 R2 0,002 0,003 0,003 0,03 0,05 Adjusted R2 0,001 0,001 0,001 0,02 0,034
* Two sided significance at10% ** Two sided significance at 5% *** Two sided significance at 1% Panel C NPL NPL NPL NPL NPL dum_islam -3,56 -2,64 -2,62 -7,1 -7,63* [3,14] [3,09] [3,09] [4,54] [4,55] lnAssets -0,9*** -0,91*** -1,86*** -1,91*** [0,21] [0,22] [0,42] [0,42] Profit 0,0002 0,0004 0,0004** [0,0006] [0,0006] [0,0006] lnEmploy 0,6 0,99 [0,72] [0,77] Leverage -0,2* [0,14] Constant 12,33*** 18,14*** 18,19*** 23,18*** 38,66*** [0,92] [1,65] [1,66] [4,25] [11,93] N 452 452 452 452 452 R2 0,003 0,041 0,04 0,1 0,11 Adjusted R2 0,001 0,37 0,03 0,09 0,09
* Two sided significance at10% ** Two sided significance at 5% *** Two sided significance at 1%
The results in Table 5 shows that Islamic banks and conventional banks do not differ in efficiency as the Islamic bank dummy enters insignificantly at the 10% level in the regression of cost-income ratio. This contradicts Beck, Demirgüç-Kunt, & Merrouche (2013), where Islamic banks have higher efficiency due to lower agency problems and thus lower
monitoring and screening costs. They also used variables that controls for agency problems e.g. governance structure. The R² for the cost-income ratio is low, whereas only the natural logarithm of employees and the leverage ratio are significant at the 5% respectively 1% level. This could be the reason for not controlling for other factors that influence efficiency.
According to the model, regression results shows no difference in efficiency between Islamic banks and conventional banks
Table 5
This table reports regression results for the dependent variable Efficiency. The regression coefficients are reported without brackets. The standard errors clustered by bank are reported in the brackets.
Cost income ratio Cost income ratio Cost income ratio Cost income ratio Cost income ratio dum_islam 0,06 0,35 0,41 0,3 -0,78 [3,1] [3,06] [3,06] [3,88] [3,83] lnAssets -1,11*** -1,16*** -0,31 -0,11 [0,19] [0,2] [0,28] [0,28] Profit 0,0008 0,0003 0,0002 [0,0007] [0,0005] [0,0005] lnEmploy -1,52*** -1,08** [0,45] [0,45] Leverage -0,18*** [0,04] Constant 78,91*** 86,21*** 86,45*** 91,79*** 104,09*** [0,74] [1,46] [1,47] [2,18] [3,59] N 12456 12456 12456 12456 12456 R2 0,00 0,027 0,029 0,04 0,07 Adjusted R2 0,0008 0,026 0,026 0,03 0,06
* Two sided significance at10% ** Two sided significance at 5% *** Two sided significance at 1%
In the regression in table 6 a crisis dummy and an interaction term with the crisis dummy is added. The latter is included to distinguish between the effect of the crisis on any difference between Islamic and conventional banks. This table shows no differences in asset quality between Islamic and conventional banks neither before nor during the crisis, as none of the
included dummies as well as the interaction term is significant. The same goes for efficiency. No significant difference before or during the crisis for Islamic and conventional banks. However significant differences are found regarding to stability. Islamic banks during the crisis are more stable than conventional banks during the crisis as the dummy for Islamic banks and the dummy for the crisis period are both significant at the level of 1%. Also the interaction term is significant at 1%.
Table 6
This table reports the performance of Islamic banks and conventional banks before and during the crisis by adding a crisis dummy and an interaction term with this dummy. The sample period is 2002-2012. Regression coefficients are reported without brackets. Standard errors clustered by bank are reported in brackets.
Z-score Loan loss reserves Loan loss provisions Non-performing loans Cost income ratio dum_islam 0,88*** -0,47 -0,005 -11,68 -3,13 [0,23] [3,93] [0,009] [10,28] [7,12] dum_crisis 0,27*** -3,96*** -0,002 -10,46*** 1,50 [0,06] [0,94] [0,003] [3,47] [1,76] dum_islam * dum_crisis -0,88*** -0,64 0,013 6,04 3,19 [0,28] [4,34] [0,011] [11,44] [8,40] lnAsset 0,01 -0,47*** -0,0005 -1,42*** -0,15 [0,01] [0,14] [0,0004] [0,45] [0,29]
Profit 2.64E-06 0,0002 4.72E-07 0,0004 0,0002
[0,00002] [0,0002] [7.23e-07 ] [0,0007] [0,0005] lnEmploy -0,009 0,71*** 0,002*** 0,76 -1,06** [0,015] [0,22] [0,0006] [0,76] [0,45] Leverage -0,37*** -0,17*** -0,0002*** -0,26* -0,18*** [0,001] [0,03] [0,00008] [0,14] [0,04] Constant 36,38*** 22,33*** 0,02*** 49,46*** 103,26*** [0,12] [2,73] [0,007] [12,28] [3,76] N 12455 781 778 452 12456 R2 0,42 0,16 0,05 0,15 0,07 Adjusted R2 0,42 0,14 0,03 0,12 0,06
* Two sided significance at10% ** Two sided significance at 5% *** Two sided significance at 1%
6. Conclusion
This paper aims to compare bank performance between Islamic banks and conventional banks. Islamic banks accord to the tenets of the Sharia, the Islamic jurisprudence. Financial transactions regarding to Islamic banking cannot include an interest payment (Riba), and share risks.
Bank performance of Islamic and conventional banks is compared through stability, asset quality and efficiency. The impact of the financial crisis on the Islamic and
conventional banks is also taken into account, therefore the sample was chosen for a period before and during the crisis (2002-2012). Banks were selected on the basis of their asset value, years of reporting data and only 100% Islamic and 100% conventional banks are included. In the model Islamic banks are more stable than conventional banks as the dummy is significant at the 5% level. A possible explanation is that Islamic banks are not exposed to interest rate risk. According to asset quality and efficiency there are no differences between Islamic and conventional banks as the dummy for Islamic banks is insignificant. However, this model apparently does not capture more important factors that explain asset quality and efficiency due to the very low R² in the model. Also no difference before or during the crisis for Islamic and conventional banks are found for asset quality and efficiency. However
Islamic banks during the crisis are more stable than conventional banks during the crisis as the dummy for Islamic banks and the dummy for the crisis period are both significant at the level of 1%. Also the interaction term is significant at 1%. The financial crisis effected mostly conventional banks in a way that they went insolvable and were not able to pay back their outstanding debt.
Appendix A Table 3
Number of Islamic and conventional banks in the period 2002 to 2012 per county. Banks with at least two years of observation are included in the sample.
Country Number Islamic banks Number conventional banks Bahrain 32 35 Bangladesh 13 65 Brunei Darussalam 5 2 Egypt 2 39 Gambia 1 4 Indonesia 11 96 Iraq 8 20 Jordan 4 16 Kuwait 13 27 Lebanon 3 52 Malaysia 40 236 Maldives 1 2 Mauritania 2 8 Pakistan 12 84 Palestinian Territory 2 5 Philippines 2 100 Qatar 7 11 Russian Federation 2 1255 Saudi Arabia 7 14 Singapore 1 46 Sudan 15 13
Syrian Arab Republic 3 12
Tunisia 1 49
Turkey 14 154
United Arab Emirates 15 27
United Kingdom 9 625
Yemen 5 8
References:
Abedifar, P., Molyneux, P., & Tarazi, A. (2011). Risk and stability in islamic banking. Bangor Business School, University of Wales, Bangor.
Aggarwal, R. K., & Yousef, T. (2000). Islamic banks and investment financing. Journal of
money, credit and banking, 93-120.
Akacem, M., & Gilliam, L. (2002). Principles of Islamic banking: Debt versus equity financing. Middle East Policy, 9(1), 124-138.
Alpay, S. and M. K. Hassan (2007): A comparative efficiency analysis of interest free financial institutions and conventional banks: a case study on Turkey. Economic research
forum, working paper 0714.
Archer, S., Karim, R. A. A., & Al-Deehani, T. (1998). Financial contracting, governance structures and the accounting regulation of Islamic banks: an analysis in terms of agency theory and transaction cost economics. Journal of Management and Governance, 2(2), 149-
170.
Ariff, M. (1988). Islamic banking. Asian‐Pacific Economic Literature, 2(2), 48-64.
Bonin, J. P., Hasan, I., & Wachtel, P. (2005). Bank performance, efficiency and ownership in transition countries. Journal of Banking & Finance, 29(1), 31-53.
Dixon, R. (1992). Islamic banking. International Journal of Bank Marketing, 10(6), 32-37. El‐Gamal, M. A. (2007). Mutuality as an antidote to rent‐seeking Shariah arbitrage in Islamic finance. Thunderbird International Business Review, 49(2), 187-202.
El Hawary, D., Grais, W., Iqbal, Z., 2004. Regulating Islamic financial institutions: The nature of the regulated. World Bank Policy Research Working Paper #3227. El Qorchi, M. (2005). Islamic finance gears up. Finance and Development, 42(4), 46. Elyasiani, E., & Goldberg, L. G. (2004). Relationship lending: a survey of the literature. Journal of Economics and Business, 56(4), 315-330.
Foster, G., & Gupta, M. (1990). Manufacturing overhead cost driver analysis. Journal of
Accounting and Economics, 12(1), 309-337.
Karim, R., & Ali, A. (1989). Determinants of the financial strategy of Islamic banks. Journal of Business Finance and Accounting, 16(2), 193−212.
Khan, F. (2010). How ‘Islamic’is Islamic Banking?. Journal of Economic Behavior &
organization, 76(3), 805-820.
Khan, M. A., Mirakhor, A. (1987). Theoretical Studies in Islamic Banking and Finance, Houston, Texas: The Institute for Research and Islamic Studies.
Siddiqi, M. N. (2005). Riba, bank interest and the rationale of its prohibition. Markazi Maktaba Islami Publishers.
Taktak, N. B., Zouari, S. B. S., & Boudriga, A. (2010). Do Islamic banks use loan loss provisions to smooth their results?. Journal of Islamic Accounting and Business Research, 1(2), 114-127.
Warde, I. (2000): Islamic finance in the global economy. Chapter 4 The Evolution of Modern
Islamic Finance. Edinburgh University Press, Edinburgh, pp. 73-89.
Yeh, Q. J. (1996). The application of data envelopment analysis in conjunction with financial ratios for bank performance evaluation. Journal of the Operational Research Society, 980-
