“Continued discussion on conventional versus Islamic banks: combining financial ratios and efficiency”

Two different types of banking systems, Islamic and conventional, dominate the bank- ing structure in Saudi Arabia. The purpose of this paper is to contribute to the on-going debate as to which of the two is better. Using data for the period 2014–2018, the study compares Islamic and conventional banks. It combines traditional financial ratios, Return on Assets (ROA) and Return on Equity (ROE), with Data Envelopment Analysis (DEA) to perform a comprehensive analysis. In terms of ROA, the performance of conventional banks is better than that of Islamic banks, but in terms of ROE, vice versa. DEA results show that conventional banks are more efficient than Islamic banks. In fact, in terms of ROA and ROE, Al Rajhi Bank, an Islamic bank, is the best performer. But in terms of efficiency scores from DEA, Al Rajhi ranks seventh among all banks, while NCB, a conventional bank, ranks first. Issuing shares and utilizing funds in profitable options, such as loans and advances to increase net income, are the policy recommendations for Islamic banks to further improve. In addition, as the study finds no correlation between the ratio and efficiency scores, it proposes to use a combined measure of ratio analysis and efficiency analysis for a comprehensive assess- ment of bank performance. etc. basic are Musharakah, Ijarah, Qard-e-Hasna and Murabaha’ (Salman Nawaz, These stand for the principle of profit and loss sharing, partnerships and joint ventures, sales contracts, leasing con-tracts, interest-free loans and trade with mark-up, respectively.


INTRODUCTION
Islamic banking is the banking structure that operates in accordance with the law of Islamic jurisprudence (Sharia). Islamic jurisprudence prohibits interest, uncertainty, gambling, pre-determined fixed returns; any types of overcharging, getting fixed money on loans are forbidden (Haram). A tangible/physical asset supports all transactions in an Islamic bank (Hassan, Khan, Amin, & Khokhar, 2018). Sharia forbids investing in businesses that are prohibited (Haram) in Islam, like gambling, alcohol, pornography, pork, etc. . Like any conventional bank, Islamic banks also perform basic functions. The main products of Islamic banks are 'Mudarabah, Musharakah, Salam, Ijarah, Qard-e-Hasna and Murabaha' (Salman & Nawaz, 2018). These names stand for the principle of profit and loss sharing, partnerships and joint ventures, sales contracts, leasing contracts, interest-free loans and trade with mark-up, respectively.
Islamic banking is steadily growing in the global financial market. The success of Islamic banking system is not limited only to Muslim countries. It is gaining popularity in countries such as UK and USA (Lone, Aldawood, & Bhat, 2017). The regulation of Islamic banking The basic difference between the two banking systems is that conventional banks depend on interest, while Islamic banks depend on the principles of profit and loss sharing. Islamic banking services are basically of three different categories. The first category, free services, includes current account functions, such as 'overdraft, demand loans' and the like. Here, the profit and loss sharing in not possible. Profit and loss sharing happens in the other category of services, such as 'equity financing, hire purchase, leasing' and the like. The second category involves fixed charge/fee/commissions. And the major revenue earning is the third category of services dealing with indirect investment in a bank's own projects, investment auctioning, commodity trade, spot transaction in foreign exchange and the like (Oseni & Hassan, 2010).
Islamic banking has the risk sharing, whereas in conventional banking, the customer transfers the risk to the bank and in turn receives a guaranteed fixed return. Theoretically, Islamic banking is resistant to surprise shocks (Alghfais, 2017). Besides, in conventional banking, customers pay interest that does not depend on the project return (Qian & Veluyathan, 2017). This is due to its focus on risk sharing, limited risk taking and strong association with real activities (Hussain, Shahmoradi, and Turk, 2015). Islamic banking works based on the unique concept of maximizing the 'equal distribution of income' (Alanazi & Lone, 2016). Islamic banks depend primarily on sales-type products, which have lower risk than conventional debt-based products (Kabir, Worthington, & Gupta, 2014 Ghassan and Taher (2015) compared the financial robustness and stability of Islamic banks and conventional banks. The study found Islamic banks to be more financially stable than conventional banks. It also revealed low competition amongst banks in Saudi Arabia that affected the financial stability. The study expressed concern that the lower share of Islamic banks in the overall banking sector of Saudi Arabia hinders any improvements in the financial stability of the sector.
Al-turki (2014) studied nine commercial banks and three Islamic banks in Saudi Arabia for 2005-2011 using a linear panel data model. According to his study, there is no significant difference between Islamic banks and conventional bank for all variables except for the bank cost.
Kabir, Worthington, and Gupta (2014) studied the credit risk of 37 Islamic and 156 conventional banks for the period 2000-2012. The banks were from 13 different countries, including Saudi Arabia. The authors acknowledged the issues of using only accounting ratios. They used accounting information of the Z-score and nonperforming loan (NPL) ratio and compared the results with Merton's distance-to-default (DD) model. This DD model is a market-based credit risk measure. The results indicated that, compared to conventional banks, Islamic banks had lower credit risk in terms of a DD model and higher risks in terms of the Z-score and NPL ratio.
Saif-Alyousfi, Saha, and Md-Rus (2017) compared Islamic and conventional banks in terms of the CAMEL model parameters for the period 2000-2015 using Tobin's Q, Market-to-book ratio, and ROE using pooled OLS and a random effect model. The study found that the Islamic banks contribution to shareholder value is greater than that of conventional banks. Higher capital ratio and credit risk in conventional banks led to weakened shareholder value, while Islamic banks did not have such a problem. Moreover, higher level of loans lowered shareholders value in conventional banks' shareholder value, while it resulted in an increase in Islamic banks. In addition, conventional banks with higher liquidity have lower shareholder value, while it was vice versa in Islamic banks.
Alghfais (2017) studied four Islamic and eight conventional banks in Saudi Arabia for the period 1988 to 2016 using a binary logistic regression method. The study suggests that Islamic banks are more profitable and have lower risk, their contribution in the economic growth is greater, and they have higher capitalization compared to conventional banks. However, conventional banks have better investment portfolios and are more efficient. Since Islamic banks have high capitalization, they are able to use it for high profitability, but they do not have standardized products. The researchers opined that it might be the reason why Islamic banks are less efficient compared to conventional banks.
Brown, Hassan, and Skully (2007) studied the operational efficiency of Islamic banks for the period 1998 to 2003 for 11 countries, including Saudi Arabia. They found that conventional banks were more stable in terms of loan portfolio. The reason may be more effective management practices, as conventional banks are much older. Islamic banks were better in terms of net financing to total assets. This could be due to the concept of equity financing, when Islamic banks work as joint partners in the projects.
After searching for further studies, the researchers found some works on the overall banking structure using a ratio analysis. Almazari and Almumani (2012) explored the profitability efficiency of banks in Saudi Arabia during the period 2006-2010. They used return on assets and operating income. One important finding of this study was that high total assets did not result in high profit.
Medabesh (2012) studied the determinants of Saudi banks' profitability during the period 1990-2008. The author used ROA and ROE in the analysis. It was found that the banks were decreasing the exposure to risk, had good international reputation, and generated profits from long-run investments. Due to these characteristics, the author commended the safety of banks in Saudi Arabia from bankruptcy and high credibility. Almumani (2013) applied DEA with constant returns to scale for the data on 10 Saudi banks for the period 2007-2011. The inputs were total deposits and total expenses, while the outputs were total loans and advances and net investment. Using constant returns to scale, the average technical efficiency ranged from 94.23% to 99.28%. Banque Saudi Fransi and ANB had the highest efficiency scores of more than 99%, while Riyadh Bank, Albilad and Saudi Holland Bank had the lower score (about 94%). Banque Saudi Fransi ranked first and SHP had the last position in terms of average efficiency for the period of study.
Hassan, Khan, Amin, and Khokhar (2018) used DEA, where the inputs were deposits, assets and capital and the outputs were investments, advances and income for the period 2008-2016. Using constant returns to scale, the study compared four Islamic banks operating in Saudi Arabia and found that Al Rajhi Bank was the best bank followed by Bank AlJazira, Alinma and Al Bilad. Hassan (2006) explored the relative efficiency of Islamic banks for the period 1995-2001, using panel data from 43 banks from 21 countries, including Saudi Arabia. The author used both parametric and non-parametric methods. The input variables were labor, fixed capital, customers and short-term funding. The output variables were total loans, other earning assets and off-balance sheet items. The study found that conventional banks were more efficient than Islamic banks. Technical efficiency was better than allocative efficiency. Also, scale efficiency was greater than technical efficiency. Another key finding was that the efficiency results were correlated with ROA and ROE and it was recommended that they be used simultaneously.
The above literature indicates that there are quite a few studies on the performance of Saudi Arabian banks. However, the debate on Islamic banks compared to conventional banks is still inconclusive, as studies show different results. This study aims to continue the discussion and provide conclusive evidence of which form of banking is best.

METHODS
Data on banks' deposits, loans and assets are from the websites of respective banks and Saudi Arabian Monetary Agency. The process of measuring bank performance is divided into two categories, namely traditional and frontier approaches (Varadi, Pradeep, & Boppana, 2006). The traditional approach uses accounting ratios to measure overall financial performance (Karim & Alam, 2013). Return on Assets (ROA) and Return on Equity (ROE) are some of the common used items for the ratio analysis. The second category, a frontier approach, includes both parametric and non-parametric methods to establish a mathematical relationship between variables, Stochastic Frontier Approach (SFA) and Data Envelopment Analysis (DEA) ( ROA is used to analyze the secondary data and information to fulfill the research requirement. Return on total assets is the relationship between net income and total assets of an organization; it reveals the operational efficiency of the business organization. This ratio shows the use of the resources, i.e. how efficiently a business utilizes its resources to generate income. A high return on assets ratio shows the maximum utilization of the resources and considers it favorable for the business organization. .

Profit before tax ROA
Total assets = Return on equity is the relationship between net income from business operations and shareholder equity and reveals the profitability or return for the real owner of the corporation. A high return on equity is favorable for the business organization and indicates how shareholders are getting returns on their invested funds in the business organization. .

Profit after tax ROE
Equity capital = (2) Next, Data Envelopment Analysis (DEA) is used to measure efficiency. DEA compares different units based on inputs and outputs and is used to find relative efficiency of banks. DEA can be performed using DEA Online Software (DEAOS), Basic Radial Models (Envelopment Forms), Input Oriented, and Constant Returns to Scale (CRS). The inputs will be the deposits and capital, and the output will be income. The model for DEA is given below.

For maximization problem:
DMU m uses x ij of input n, and n = deposits, capital; t n is weights assigned to n th input; DMU m yields y om of output t, and t = net income; q t is weights assigned to t th output, where m = 1, 2, 3…, 12. Kabir, Worthington, and Gupta (2014) have shown that assessments of bank performance change using different methods, therefore, it is proposed to combine and use different methods to measure bank performance. In addition, after an extensive review of existing studies, Hassan and Sirajo (2016) conclude that Islamic and conventional banking systems work according to different principles and, therefore, the method for studying one system may not be suitable for studying another. The study recommended methodological research.

ROA (return on assets)
ROA of twelve banks of Saudi Arabia is shown in

Efficiency analysis
The efficiency scores of the twelve banks of Saudi Arabia are calculated using Data Envelopment Analysis. The inputs are deposits and capital and the output is income. Out of four Islamic banks, only Al Rajhi (25% of the total Islamic banks) has an efficiency score higher than the overall industry average, while four of the eight banks, namely National Commercial Bank, Saudi Investment Bank, Arab National Bank and Samba Financial Group (50% of the total conventional banks), have an efficiency score higher than overall. In addition, in the last year included in the study (that is, 2018), only two out of four banks had a 100% efficiency score, while six out of eight banks had an efficiency score of hundred percent (Table 3).

Bank ranking
When the banks are compared in terms of broad categories of Islamic banks and conventional banks (Table 4), the composite score for conventional banks (5.79) is higher than that of the Islamic banks (7.92). According to ROA and ROE analysis, performance of Al Rajhi Bank, Saudi British Bank and National Commercial Bank is satisfactory, while the performance of Bank AlJazira, Saudi Investment Bank and Alinma Bank is below the industry average. There is a difference in the performance of banks according to ROA and ROE analysis. This difference is because the difference of profit before tax and total assets is included in ROA and that of profit after tax and shareholder funds is included in ROE. The reason

Rank correlation
Two distinct cases are evident from the difference in ranks from ratios and DEA. First, Saudi Investment Bank properly uses its deposits to receive advances and generate income on its advances distributed among their beneficiaries, as it is fourth in DEA. However, the bank is not able to properly utilize its total assets, as it ranks 11th in terms of ROA and ROE. Meanwhile, Al Rajhi Bank properly uses its total assets, as it is first in ROA and ROE, but the return on input, such as equity capital and deposits, is lower than desired. Better ROA performance means good returns on a bank's total assets. Better ROE performance means better returns on equity capital. Better performance in DEA means more income from equity capital and deposits.
In addition, an attempt was made to correlate between ranks of ROA, ROE and DEA using Spearman's rank correlation coefficient ( Table  5). The high degree of a significant positive correlation (0.692) between ROA and ROE is quite expected, as both of them are accounting ratios of more or less similar nature. However, no significant correlation was found between the ranks of ROA and ROE and ranks of DEA. In addition, this justifies the usage of two tools (ratios and efficiency) to measure performance. This would give a more comprehensive view of bank performance. Note: ** Correlation is significant at the 0.01 level (1-tailed).
A similar idea has been presented earlier by Hassan (2006) but a correlation was found between the two different types of measures and it was recommended to use them simultaneously. But this study finds no correlation between accounting ratios and efficiency scores. The difference in the results of these two studies may be due to the study period or the sample of countries or the choice of inputs and outputs in DEA. Although this study finds no correlation between accounting ratios and efficiency scores, it still recommends the adoption of both measures for a more comprehensive measurement of performance.

DISCUSSION
The present study indicates that the performance of conventional banks is better than that of Islamic banks. Despite using different variables and methods, the results of this study contradict Alghfais Investment Bank and AlJazira Bank should try to increase their net income by reducing operating expenses and managing funds from cheaper sources. These banks have performed poorly in terms of ROE.
To improve financial performance, Islamic banks, such as Bank AlJazira and Alinma Bank, have to minimize their interest expenses by managing funds form cheaper sources or issuing shares. However, tax liability of a bank cannot be minimized and issuing of shares will be the option in the hands of banks to manage monetary requirement in future. Besides, Al Rajhi Bank must utilize its deposits or available funds to profitably distribute funds as loans and advances to increase net profit.

CONCLUSION
Islamic and conventional baking systems are two different types systems common in Saudi Arabia. Therefore, this study attempts to establish which of the two systems is better for banks. The study innovates by combining two different approaches to measure bank performance, namely: financial ratios and efficiency analysis. Financial ratios, such as ROA and ROE, give conflicting results, while efficiency analysis using DEA indicates that conventional banks are more efficient than Islamic banks. Finally, the study recommends Islamic banks issue more shares and use their funds in profitable ventures. The study also advocates combining different measures to assess bank performance.
Though applying different principles of lending and borrowing, both the Islamic and conventional banking systems have the same goal. The results show that Islamic banks are better in terms of managing their returns on equity, while conventional banks are better in maintaining efficiency; therefore, both banking systems can learn from each other for further improvement. Islamic banks can learn in terms of managing their assets from conventional banks as they are more experienced and more widespread. Conventional banks can provide an option of Islamic banking products to those who are interested, which will increase their customer base. This can lead to further diversification of banking services, which ultimately will benefit everyone.