The Multiple Performance Analysis of Indonesian Islamic Banks: The Stakeholder Theory Approach

This study aimed to develop an alternative measurement model based on the stakeholder theory that answers the critiques toward the Balance Score Card (BSC). It synthesized variables from previous studies, including Job Satisfaction (JSA), Business Process (BPR), Customer Satisfaction (CSA), and Financial Performance (FPE). Moreover, the study set nine hypotheses representing the relationship between JSA, BPR, CSA, and FPE. The hypotheses were accepted after the Structural Equation Modelling (SEM) analysis of the financial data of the four biggest Indonesian Islamic banks and responses from 200 respondents. This finding confirmed the robustness of the model proposed in this study. The model denotes that the FPE results from the excellent implementation of JSA, BPR, CSA, and the interaction between these variables, which directly impact financial performance.


Introduction
Banks are essential in the economic system as financial intermediaries where funds could be most efficient (Hasan et al., 2015). The main challenge for the banks' management is that they need to monitor performance persistently because unnoticed problems could be detrimental in the future (Tursoy, 2018). However, the performance of banks usually stated from the financial perspective cannot picture the internal management and governance quality (Atkinson et al., 1997). The result is incomplete and oversimplifies stakeholders' roles in success (Jensen, 2001).
Performance assessment using stakeholder theory may capture information beyond financial figures, including the constraints and factors influencing the firm (Freeman, 2017). The theory also examines the stakeholders' involvement in creating the firm's performance (Harrison and Wicks, 2013). The most popular measurement model is the Balanced Scorecard, which uses the stakeholder theory approach. However, the model has received many critiques, such as its inability to count the Learning and Growth variable that influences the company's financial performance (Pessanha and Prochnik, 2011

Measurement Model Based on Stakeholder Theory
Stakeholders are the parties affected by the company's existence (Laplume et al., 2008). They include the employees, management, shareholders, consumers, the surrounding community, NGO, press, and government (Miles, 2012). Each party has a different direct or indirect interest in the company, which may have an extra level of interest in their stakeholders. A higher stakeholders' bargaining power increases the company's effort to adapt (Atkinson et al., 1997). According to the stakeholder theory, a successful company delivers the value needed by its stakeholders. The value guides the company in its business operations (Freeman, 1984).
The stakeholder theory accepts the role of management quality, customer satisfaction, and process effectiveness in making value (Atkinson et al., 1997). Based on the theory, the performance assessment gives comprehensive information about the company.
One measurement model of a company's performance is the Balanced Scorecard (BSC) (Kaplan and Norton, 1996), the managerial equivalent of stakeholder theory (Jensen, 2001). Companies worldwide have applied BSC (Awadallah and Allam, 2015) because it has been developed based on the practices of prominent world-class companies (Cooper and Ezzamel, 2016). The model provides information about previous achievements and prospects, the company's productive assets, human resources, and customer satisfaction. The original BSC model is shown in Figure  1.  (Kaplan and Norton, 1996) The BSC model has the ability to become a strategic management tool, including performance appraisal, resource allocation, and staff training (Martinsons et al., 1999). It promotes an innovative spirit within the company (Bose and Thomas, 2007), and helps the project management convert tasks into measurable performance indicators (Basu et al., 2009).
The critiques toward the BSC model come from many scholars. According to Ghalayini and Noble (1996), the model does not provide operational performance indicators. Although it should be a constructive framework suggesting areas to be measured, the result cannot be compared between the current and prevailing period (Klimczak, 2005).  Pessanha and Prochnik (2011) showed the weakness of BSC in incorporating employee engagement in the objectivesetting process. As a result, the model is unfamiliar to the employees. Jensen (2001) stated that BSC is a multiple performance measurement, making the company's management unable to maximize all dimensions because they cannot make trade-offs between the measures. Although all of its variables are important, there is no direct connection between the learning and growth and financial indicators. Antonsen (2013) criticized BSC due to the data complexity and the high resource requirements during implementation. The model requires the company to acquire new data, causing work overload for the working unit and resulting in employee cynicism and refusal. Consequently, it becomes pressure on the management because more complex data requirements could ruin the employee-management relationship. The many critics from scholars toward BSC necessitate its modification.

Stakeholder Theory within Islamic Banking
Dusuki (2008) (Mohammed et al., 2016). This implies that the maqasid shari'ah connects to all human life elements on personal and community levels (Laldin, 2008). Therefore, Islamic banks must comprise these principles in their business strategy and operation.
The stakeholder theory holds that an organization's success depends on the capability to create benefits for internal and external stakeholders. The company's management must consider the stakeholders' perspective and their activities (Freeman, 2017). In comparison, the maqasid shari'ah focuses on the value under Islamic principles, which must be incorporated into the bank's operations by the management. This shows that the stakeholder theory and maqasid shari'ah are congruent. Consequently, the Islamic banks' business responsibility is to achieve the annual target revenues and create the stakeholders' well-being (Shah and Hassan, 2019).

Proposed Conceptual Model and Hypotheses Development
Previous studies have agreed on the importance of an integrated and multidimensional performance measurement model. They have acknowledged the need to prioritize variables according to their strategic importance, an area that has invited minimal attention.
This paper develops the conceptual model, a modification of BSC, shown in Picture 2. Based on the model, the companies' management should compromise between indicators (Jensen, 2001). The model answers BSC's criticism by prioritizing financial performance's ultimate objective, becoming a dependent variable.

Financial Performance
Previous studies introduced many approaches regarding banks' financial performance, with the CAMELS system the most popular. During the examination, the CAMELS system requires data from financial, funding, and cost and expense reports, economic analysis, cash flow outlook, and human resources operation. This model assesses the bank's holistic condition based on its strengths and weaknesses (Canbas et al., 2005).
The CAMELS system comprises the following six components:  Capital adequacy determines a bank's capability to withstand financial distress, such as market, credit, and interest rate risks (Christopoulos et al., 2011).  Asset quality evaluates how a bank assures to cover bad and doubtful loans. This parameter also shows the funds retained by the banks in the event of a non-performing loan (Masood et al., 2016).  Management quality exposes the soundness of the bank's management in reducing risks, improving efficiency, as well as increasing production and profits (Masood et al., 2016).  Earning measures the profitability and assets productivity, growth, and sustainability of earnings capacity (Christopoulos et al., 2011).
 Liquidity measures the capacity for short-term commitments. A good liquidity position is when a company receives sufficient cash (Masood et al., 2016).  Sensitivity to Market Risk concerns the bank's ability to identify, manage and monitor the market risk and recognize the possible problem. The market risk comprises foreign currency, interest rate, product purchase, and selling prices of the bank's assets (Masood et al., 2016).

Job Satisfaction
The Learning and Growth variable in the original BSC model is not directly related to performance. According to Kaplan (1998), Learning and Growth support the other variables because it is how a company develops effective employee behavior. However, Devie and Widjaja (2012) stated that the variable is a translation of employees' satisfaction with their job. In this case, job satisfaction is how employees feel contented with their work based on different aspects (Wexley and Yuki, 2005).
Job Satisfaction directly connects with the company's financial success. This relationship was also confirmed by Judge et al. (2001) after conducting a meta-analysis of around 300 studies on the subject. According to Necowitz and Roznowski (1994), the impact of Job Satisfaction on a company's performance is more significant

Customer Satisfaction
The Customer Satisfaction variable is the main success criterion in today's highly competitive banking industry (Mandal, 2015). It is the overall consumer's perception of a product or service before, during, and after consumption (E. W. Anderson and Sullivan, 1993). Moreover, customer satisfaction indicates the buyer's pleasure or disappointment regarding the products (Oliver, 1980). It is determined by the value customers receive from the transaction (Hallowell, 1996).
Customer satisfaction influences the purchasing and post-purchase processes, including attitude change, repeat purchases, and brand loyalty (Churchill et al., 2013).
There are five dimensions of high Customer Satisfaction levels. The first is tangibility, implying the company's physical environment that builds up impressions. This dimension increases the probability of clients returning in the future. The second is reliability, referring to how a firm completes its guaranteed benefit, quality, and precision within the agreed-upon parameters between the company and its clients. The third is responsiveness, alluding to a company's eagerness to help its consumers receive good, quality, and timely service. The fourth is assurance, which refers to the employees' ability to acquire customers' trust and confidence in the company's products. Customers may not return when uncomfortable with the employee. The fifth is empathy, alluding to how the company gives individualized attention and concern to make customers feel highly appreciated and special (Delgado-Ballester, 2004;Lau et al., 2013). Nelson et al. (1992) may be the first to coin the connection between Customer Satisfaction and the company's performance. This study was followed by many others that confirmed the result, such as Keisidou et al. (2013), Nayebzadeh et al. (2013), and Ndubisi and Nwankwo (2019). Subsequently, studies on Job Satisfaction and Customer Satisfaction have found a strong and positive relationship between the two variables (Kermani, 2013;Mafini and Pooe, 2013). Mendoza and Maldonado (2014) stated that this linking is stronger in the environment of personal than nonpersonal services.
This study conceptualized that Customer Relationship is an intervening variable between Job Satisfaction and the Performance of the bank.

Business Process
The business process is how the company accomplishes its target performance. Although a company has many processes running simultaneously, only a few provide significant value to customers (Chisambara, 2014). Kaplan and Norton (1992) stated that the company's higher-level business processes positively impact performance. This was later backed by Kairu et al. (2013), Rintari et al. (2018), and Tibbs and Langat (2016). The benefits of a good business process GJAT I JULY 2022 I VOL 12 ISSUE 1 I ISSN : 2232-0474 I E-ISSN : 2232.0482 www.gjat.my include profitability, value-added per employee, work spirit, leadership trust, reputation, and employee turnover. Other advantages include raising the employees' positive perceptions and the customers' perceptions of the company (Škrinjar et al., 2008).
Companies must integrate and automate their corporate activities due to fierce competition. They must integrate and automate their processes with the rise of ecommerce, globalization, and outsourcing activities to third-party vendors (Dayal et al., 2001).
The connection between Job Satisfaction to the Business Process was indicated by Škrinjar et al. (2008), though studies in this field are rare. Yee et al. (2008) found a positive interaction between employee satisfaction, profitability, and quality, a business process indicator (Van Looy and Shafagatova, 2016). Therefore, Job Satisfaction and Business Process should have a positive relationship.
Hypotheses 5 to 9 were formulated as follow:

Data Analysis
This study uses previously validated instruments to validate the conceptual model, as shown in Table 1. The questionnaire instruments used a fivepoint Likert scale, with 1 = strongly disagree and 5 = strongly agree.
The original English instruments were translated into Bahasa Indonesia. To ensure the target respondents' comprehension, a pilot test was conducted, involving 15 bank employees and 15 customers. Some minor revision was taken based on the participants' participants.
The study's object comprised four fullfledged Islamic banks, including Bank Syariah Mandiri, Bank Muamalat Indonesia, Bank BRI Syariah, and Bank BNI Syariah. These banks represent the industry because they accounted for 49% of Islamic banks' assets in 2019 (Otoritas Jasa Keuangan, 2020). The respondents comprised the banks' 200 employees and customers chosen by simple random sampling in Jakarta capital city.

This study used Structural Equation
Modelling (SEM) to assess the hypotheses. It adopted the two-step approach (Anderson and Gerbing, 1988), where the first step estimated the measurement model using Confirmatory Factor Analysis (CFA). The second step determined the best-fitting model and tested the casual relationship using the Structural Equation Modelling (SEM).

Reliability and Validity
Reliability was measured by calculating the Composite Reliability (C.R.) and examining the Average Variance Extracted (AVE). Following Hair et al. (2010), this study used the threshold of C.R.> 0.7 and AVE > 0.5. It employed the second-order CFA for construct validity, where items with 0.5 or higher loadings were retained for a further process.  Table 2 shows that the factor loading exceeds 0.5, meaning all instruments are valid. The Composite Reliability (C.R.) and AVE also exceed the minimum, implying that the model is valid and reliable.

Model Testing
The first test was for the data normality, where the critical ratio (c.r) for kurtosis and skewness is 0.724, ranging from -2.58 to +2.58. The test for the data outlier using the AMOS Mahalanobis Distance, with p <0.001, showed that no data has a value > 48.267. Therefore, the data comply with the normal distribution and has no outlier.
This study's conceptual model was converted into the SEM shown in Figure 3. The model's degree of freedom (DOF) was found to be 203, implying it was overidentified. The goodness of fit test results in Table 3 show that the observed data fit with the model.

Discussion
This study aimed to develop a performance measurement model for Islamic banking based on the stakeholder theory. The measurement supports Maqashid Shari'ah as the leading business direction for Islamic banking.
The study tested the nine hypotheses representing the inter-relationship between the JSA, BPR, CSA, and FPE variables.
The SEM procedure showed that all hypotheses are accepted, meaning the model proposed in this study is sound and visible.
The results show that the proposed model could be an alternative to BSC. Data is easily collected and applicable at the operational level, enabling banks to conduct measurements.
This study's model addresses the claims that all BSC variables are equally important (Jensen, 2001). The model puts FPE as the ultimate result of JSA, BPR, CSA, and the interaction between these variables. The study answers Pessanha and Prochnik's (2011) critique regarding the lack of direct connection of the Learning and Growth variable with performance in the original BSC. It proposes using the JSA variable that relates positively to the other three constructs. Additionally, this study stresses the importance of employee involvement to the company's financial achievement.

Conclusion
This study proposed a measurement model based on the modified Balanced Scorecard (BSC) because of some weaknesses in the original BSC. Some variables were synthesized from previous studies to develop a new measurement model consisting of Job Satisfaction (JSA), Business Process (BPR), Customer Satisfaction (CSA), and Financial Performance (FPE). Data were collected from the financial reports of four Indonesian Islamic banks and surveys of 200 respondents. The data were analyzed using Structural Equation Modelling (SEM) to test nine hypotheses representing the interconnection of JSA, BPR, CSA, and FPE. The test showed that the nine hypotheses were accepted.
The proposed model is more practical than the original BSC since the banks' operational level is able to collect and undertake the measurements efficiently. The model also indicates that the FPE variable is the ultimate result of sound JSA, BPR, and CSA, as well as the interaction between these variables. All JSA, BPR, and CSA variables have a direct relationship with FPE. Therefore, this study emphasizes employee involvement, represented by JSA, in the company's financial success.