The Effect of Applying Sustainability (Maqasid Shariah) and Competition on Islamic Bank Financing

Abstract

In the last decade, the Islamic Bank (IB) has been contributing to global financial development. There are inconclusive results about the function of IB to meet Islamic Finance ethics, particularly sustainability from the IB perspective (called Maqasid Shariah-MS). This paper aims to contribute to the basic research question: does sustainability matter in improving competition in IB financing? This study developed an advanced empirical model from previous theoretical studies that separated IB behavior using MS or less MS. To distinguish this effect, this study employed causal inference with Instrumental Variable (IV) analysis. The IV-Probit was applied to conduct a simulated treatment effect analysis accordingly. The dataset was obtained and combined from two global IB sustainability indicators and comprehensive financial statements of IB across 35 banks in 12 countries. The results confirm that applying IB in different MS with similar increasing level of competition produce different growth of financing. The IB with MS has larger financing than less MS. These results confirm that sustainability matters for improving competition and financing growth.

1. Introduction

Sustainable development (SD) is becoming crucial and requires more concrete implementation in combating global climate change and inequality. The financial sector must expand the financing source and be more inclusive of other financial innovations through banking and non-banking institution to support existing sustainable financing that mostly relies on conventional financial systems. One of the most promising financial systems is Islamic Finance (IF). Assets Under the Management of IF will be more than USD 5 trillion or equal 3–4% of the world’s total assets in 2026 [1], compared to 1% in 2014. Several references, such as Abdelsalam et al. [2], Balcilar and Hammoudeh, [3] and Bedoui and Mansour [4] stated that IF is considered close to socially responsible and ethical investment (or called as moral capitalism in the conventional system). In addition, Campisi et al. [5] report that the government has not fully supported IF to stimulate this financing for SD in the world, albeit IF produces better profitability and bankability indicators. For instance, Morea and Poggi [6] report that IF schemes produce reasonable and fair remuneration for investment [7] rather than conventional, which require high interest that exhaust more incentives or subsidies from the government in renewable energy powerplant project. The IF has a potential prospect to support SD.

Due to this prospect, since 2015, the World Bank has initiated Global Islamic Finance Development Center (GIFDC) to develop IF. The GIFDC is a global umbrella signed as a Memorandum of Understanding between the World Bank and the General Council for Islamic Banks and Financial Institutions (CIBAFI) that consists of 130 countries members. As part of this initiative, World Bank [8] recommend that policy intervention of Banking in IF should strengthen the Basel III and the Islamic Financial Service Board (IFSB) framework such as improving Sharia’s good governance, risk sharing and entrepreneurship, enhancing the scale in low-income earner, human capital and Islamic finance literacy; the IF institution should focus on Islamic social finance for recognize and management practice of zakat, waqf and affordable financing for the poor as a sustainable means that called as Maqasid-Al-shariah. The Maqasid-Al-shariah channels (i.e., the way of justice, freedom of religion and protection of life, intellect, posterity and wealth) can meet the SDGs and enhance the shared prosperity of the people. Albeit this principle has been recognized across the IF community, applying this concept in the global arena requires more study and further investigation in the field.

According to the banking market structure and competition theory, increasing competition produces efficiency and market power; this hypothesis called as Relative Efficiency Hypothesis (REF) [9]. As Porter and Van Der Linde [10] noted, sustainability reduces risk and enhances the bank’s market power. This hypothesis supports Benabou and Tirole [11] that sustainability reduces risk and improves output. Although this hypothesis exists, the empirical finding investigating these aspects are still limited for Islamic banks (IB). The previous study in conventional banks has proxied bank sustainability into various indicators, such as Andrieș and Sprincean [12], reported across 493 banks in 39 countries that the Environment, Social and Governance (ESG) indicator as a proxy of bank sustainability reduces financing costs. Bannier et al. [13] stated that Corporate Social Responsibility (CSR) as another proxy of bank sustainability has negatively associated with credit risk. According to Ahmed et al. [14], bank sustainability is much closer to ESG indicators than CSR as part of ESG. To ensure whether the competition, credit loans, sustainability and bank references are the ultimate issues, this study conducts a systematic literature search through the Scopus database with competition, credit, loans and banks. There are 353 documents found. After the searched word was added to sustainability, only two papers remained. In addition, this paper searched the Scopus database to filter the documents that contain IB, but none of the records were found. This means that discussing issues related to this aspect is still absent.

Investigating the effect of sustainability on IB financing is inconclusive, some of studies support that IB meets sustainability objectives, such as Mergaliyev et al. [15], who reported that IB meets in micro indicators (i.e., shariah governance, population indicators and leverage) affect MS Performance. Otherwise, for macro indicators such as Gross Domestic Product (GDP), financial development and Human Development Index (HDI) do not affect MS Performance. Otherwise, Ahmed et al. [14] and Aliyu et al. [16] report that IB has failed in some countries such as the United Kingdom, Denmark, Egypt, South Africa, Switzerland, Turkey, Bahrain, Malaysia and the United Arab Emirates. In addition, Budiman et al. [17] found that about 60% of IB prefer to finance commercial financing rather than sustainability across 159 in 10 countries. Refining finding with the latest data and better modeling will improve the understanding of how the IB engage with sustainability or not.

Besides those issues, advancing the modeling process in structured ways will produce a better understanding of applying different MS on IB financing. Most models are based on index models, such as Mergaliyev et al. [15], or reduced-form models, such as Platonova et al. [18], that mostly ignore the microeconomic behavior that leads to missed in the analysis. There is urgent to develop an empirical model from the theoretical foundation considering the credit market. The analysis should distinguish two regimes: MS with less risk and MS with less risk ($M{S}^0$). These MS regimes produce separate continuum effects on competition toward IB financing. Usually, a lot of the literature that estimates using reduce form model fails to prove that IB cannot achieve sustainability. This article suspects that the reduced form model cannot capture how the microeconomic behavior of banks is operated. Therefore, building a foundation from microeconomic behavior that portrays MS as altruism motive from an Islamic perspective on how IB meets Islamic Finance (IF) business ethics in structural ways is necessary.

Therefore, this study proposes the research question: Does sustainability matters for improving competition in IB financing?

Hypothesis 1a.

The financing rate (profit-loss sharing rate—P) will negatively affect growth financing.

Hypothesis 1b.

MS improves the growth of financing.

Hypothesis 2.

Competition portrayed by Lerner Index (LI) increases financing growth.

Hypothesis 3.

IB with MS has less risk, is more competitive and improves financing growth. Otherwise, less MS is vice versa.

Therefore, to test those hypotheses, this study employed panel data that combined Bankscope (Financial Performance Indicators) and Thomson Reuters (Sustainability Indicators) from 35 IB across 12 countries from 2012–2019. This study’s empirical model is close to Weill [19], which discusses bank loans and competition in IB but with a single equation model. Another close empirical model applied in the conventional bank is Olivero and Li [20], which discusses bank loans and competition. Both studies were estimated without a structural model that could miss the aspect of the credit equilibrium model.

Consequently, this study advances the empirical modeling process from a theoretical model that requires causal inference using Instrumental Variable (IV) estimation. The causal inference will assess the effect of MS regime on two equations that is (1) the competition equation, and (2) the market financing equation. The MS regimes were tested using causal inference on the effect of delegated mandatory equity, such as Zakat and Qard Hassan (ZQH), on the MS regime. After the model passed the Wald exogeneity test. The IV model with ZQH is superior to the panel and IV without ZQH. The main hypothesis was conducted with two stages of analysis. First, the model was run as a baseline estimation with simulation increasing P for both MS and MS⁰ without increasing LI. Second, the model was re-run the baseline estimation by adding simulation with increasing LI. The results confirmed that applying MS has better financing growth than less MS.

The results of this study support the theoretical model that MS and Lerner Index (LI), as competition indicators, improved the growth of IB financing. The implication of these results beyond Weill [19] and Olivero and Li [20] findings indicates that applying a sustainability strategy for IB will improve the growth of bank loans, through IV variables such as ZQH. It confirmed Benabout and Tirole’s [21] arguments that sustainability attracts prosocial behavior because MS, as IB image concern, affects bank output.

This paper is structured into six sections. Section 1 includes an introduction that discusses global interest, research aims, research contribution, hypothesis and brief study results. Section 2 is the literature revision comprises various literature analyses that connect to the hypothesis building. Section 3 is a model building that describes how the empirical model developed from the theoretical model to answer the hypothesis. Section 4 is material and methods. Section 5 is the main results that answer the hypothesis. Section 6 is a discussion that consists of how the results confirm the previous study. Section 7 is the conclusion that concludes the results, implications of results, limitations of the study and further study for expanding both empirical and theoretical study.

2. Literature Review

The basic foundation of IB in the context of Islamic Finance (IF) consists of five principles such as the prohibition on riba (excessive interest), gharar (i.e., risk and uncertainty or speculation), financing for illicit sectors (i.e., weapons, drugs, alcohol and pork), the benefit should be profit and loss sharing principle, and all transaction have to be backed by a real economic transaction that involves intangible assets [22]. Accordingly, these five principles’ impact on IB business ethics could produce less risk than conventional banks [23]. This aspect has been practiced since the Qur’an was delivered 1400 years ago [24]. Another factor that IB has lesser risk is direct equity participation, such as Zakat, Infaq, Shadaqah, Waqf and Qard Hassan. For instance, according to the Pickup et al. (2018) [25] study in Indonesia, the Zakat potential in Indonesia is about 3.4% of the GDP, while 85% of the population is Muslim. Whereas about 22% percent of the global average, Islamic population complies with Zakat. This type of funding is close to delegated philanthropy that every Muslim should deliver to society [26]. In some countries, Muslims must share their equity (Zakat) according to annual growth equity [27]. Therefore, the IB should act as venture capital providers that collect this funding, and distribute it toward financing [28]. The IB is a worthy firm based on people’s focus rather than commercial motives. Hence, the IB is profits’ share rather than cash-flow and collateral banks [29]. If these references are connected with moral hazard assumptions such as Bester and Hellwig [30] and Freixas and Rochet [31]. The IB has a better risk which could indicate improving its efficiency by reducing the financing rate and better competition (Hypothesis 1a).

One of the reasons why, there is a large gap between conventional vs. IB is asymmetric information of IF practice in the society. If the people look at the figure according to Blease [32] the value of Islamic assets was about US$700 billion in 2008, and even the Muslim industry could control about $3.92 trillion of assets in 2021 and projected to $5.9 trillion in 2026 [1]. On the contrary, the Muslims could access that fund is limited. For instance, in Indonesia, according to Indonesia Financial Authority [33], the Islamic Financial Literacy (IFL) of Indonesia’s Muslims is about 12.2% compared to conventional banks, about 85%. This condition could be more beneficial for conventional banks that usually operate in asymmetric information, and most people stick to the techniques of conventional banks. In Malaysia, the IFL can minimize bankruptcy due to the large access offer by IF institutions such as IB [34]. In contrast, IB, which operates in the same environment, will be less competitive [23]. So, it is clear that achieving symmetric information with conventional banks, whereas IB financial literacy is low, is challenging, which could cause IB to struggle to compete with conventional banks.

Theoretically, the IF philosophy is similar to Bénabou and Tirole [11] (henceforth called BT). The prosocial behavior that BT mentioned is close to mandatory equity in Muslim society. In IF, mandatory equity has been recognized as Zakat [35]. Otherwise, voluntary equity is Infaq, Waqf and Shadaqat. The way to collect and distribute this participation funding to the people should comply with a similar standard practice that is Fiqh. In BT, prosocial behavior is multi-standard, which leads to partial participation for society to achieve incentives. Otherwise, in IF, prosocial behavior is mandatory and voluntary due to religious entities that produce full participation in incentives. This full participation is triggered by the social norms that cohere with endogenous interferences. In addition, applying the IF five principles on IB creates more rewards than conventional because of the risk in IB < conventional.

According to the IF perspective, prosocial behavior is a small part of Maqasid Shariah (MS). The MS was acknowledged and introduced in the modern literature by Bedoui and Mansour [4] when they defined how MS should be addressed to improve IF perspective to address global sustainability issues. Comparing the terminology of MS to the conventional economy, MS is more or less similar to moral capitalism. According to Young [36], moral capitalism is a business practice towards a more moral form of free market and entrepreneurial wealth creation. Moral capitalism aims to create optimal wealth for the global community with fair distribution outcomes resulting from the allocation of private sector capital investment coupled with minimal negative externalities for stakeholders, including the environment. The similar behavior of MS and moral capitalism regarding fair distributive outcomes and minimal negative externalities for stakeholders, including the environment. Otherwise, in MS, the delegated equity of assets is based upon the entrepreneur and individual as moral capitalism focus. Every individual owning productive assets should share its return equity according to Fiqh compliance. In addition, MS should concern about better equality for energy, water and natural resource conservation to ensure that equity will be shared between today and the next generation.

The MS aligns multiple objectives towards one Islamic business ethics vision [4]. The MS has five pillars: wealth, posterity, intellect, faith and human self. The objective of MS is improving prosperity through people and his/her environment. The basic instrument of MS that equity is distributed in society through finance. It started with the bottom-up level such as household, then moved from the corporate to the economic level. The corporate level is the lowest level that connects individuals and organizations, and the economic level is the aggregate level that consists of the macro level: social, environmental and economic indicators. This definition is then applied to the empirical model such as MS for IB conducted by Mergaliyev et al. [15].

At this moment, the issue of sustainability is a hot topic that requires more society to participate in this urgent agenda. Aligning another sustainability perspective to a global sustainability objective is important to ensure another perspective has similar pathways. The most persuasive study on how environment and competitiveness are promising was conducted by Porter [37]. How saving resources and efficiency or green behavior (i.e., reducing pollution and waste) creates competitiveness. The environmental improvement is not the cost but stimulates benefits for processes, and products that improve productivity significantly. On the other hand, this firm action should support by good regulation that focuses on the friendly aspect for the firm without concerning new investment or technology. This friendly regulation is defined as a focus on the outcome, strict regulation with real innovation, focus on end-user practice with similar standards, employ phase-in period, market incentives, stable and predictable and synchronize with other regulations. Adopting sustainable economics improves a firm’s competitive advantage. In the banking sector, sustainability is one of the necessary factors to be a successful player in global financial service [38]. According to evidence from conventional bank studies about bank sustainability, most empirical findings stated that, sustainability supports financial performance in single or global banking evidence such as Scholtens [39]. In fact, Scholtens and van’t Klooster [40] reported that sustainability reduces bank risk. Sustainability trigger bank’s efficiency and reduce risk.

In IB perspective, Ahmed et al. [14] stated that IB should contribute more to sustainability activity than conventional banks. This argument supports that MS will ensure that the IB fully adopts the Environment, Social and Government (ESG) [16]. Otherwise, if not, the IB failed to meet IF ethics. The struggling IB to achieve this objective has been recorded by Mergaliyev et al. [15], who reported that IB meets in micro indicators (i.e., shariah governance, population indicators and leverage) affect MS Performance, whereas macro indicators such as Gross Domestic Product (GDP), financial development, and Human Development Index have failed to effect MS performance. Platonova et al. [18] found that CSR in IB in the Gulf Cooperation Council (GCG) positively impacts financial performance. At the same time, Aliyu et al. [16] and Ahmed et al. [14] report that Islamic Bank still needs to meet its objectives. The somewhat effect of macro indicators on MS Performance means that IB is due to fierce competition with conventional banks [19]. Léon and Weill [41] found that about 14,000 firms financed from Islamic banks in 52 countries positively impact access to credit when conventional banking is low. According to Benabout and Tirole [11], when IB applies MS indicates an improving score of Governance and CSR score activity. The IB’s growth of financing will increase. The MS triggered image concerns that improve both investors’ and borrowers’ engagement with IB. So, when IB conducts MS, such as more governance, CSR and prosocial behavior (i.e., collecting mandatory equity: Zakat, or voluntarily equity: as Qard Al Hassan) as a lesser risk deposit fund that improves the growth financing (Hypothesis 1b).

In terms of competition, sustainability and bank output, the general hypothesis that connects Porter [37] and sustainability [38] according to Structure Conduct Performance (SCP) Hypothesis and The Relative Efficiency (REF) hypothesis [42]. The SCP hypothesis says that decreasing the deposit rate improves loan rates. Improving bank structure, such as market concentration, improve market power. Stiglitz and Weiss [43] show that higher interest rates on business loans may increase the riskiness of loans due to adverse selection and moral hazard problems. This means applying sustainability as a moral hazard reduces bank risk and credit rate. Olmo et al. [44] report about 1236 banks across 48 emerging and developed countries during the 2015–2019. They find that banks adopt sustainability leading to greater profitability, market power and efficiency. While other studies, such as Amidu and Wolfe’s [45], study with 978 banks across 55 countries in emerging market economies, find that the market structure, competition and lending channel increasing competition reduce bank loans. The Olmo et al. [44] studies conduct sustainability and competition without banks’ output such as loans. Otherwise, Amidu and Wolfe [45] has similar LI index, Banks Loan and market structure but without sustainability. The sustainability will improve competition (by reducing bank financing rate) as supported by Hypothesis 1a and 1b, the MS will help IB has more market power, and the financing growth will improve. Otherwise, the LI will lower the financing rate if competition strengthens, but the MS has a higher benefit than $M{S}^0$ (Hypothesis 2).

The REF hypothesis says efficiency improves higher output, lower price, market share and performance. In the context of IB, according to Beck et al. [22] IB is less efficient than conventional banks. In contrast, IB has better capital and more robust assets. This finding aligns with Khan [23] and Léon and Weill [41] that IB is less efficient than conventional due to market penetration cost to obtain more of the market. A study by Léon and Weill [41] informs if, in some regions with a lower conventional bank, IB can easily substitute the conventional easy. Otherwise, in tight competition, IB cannot improve the market segment. On the other hand, according to Bester and Hellwig [30] and James’ [46], the IB with MS has better competition, price, risk and growth of financing than less MS ($M{S}^0$) (Hypothesis 3).

3. Model Building

The IB assumes has the probability of financing with better risk management is $\left[p\left(MS\right)\right]=1$, whereas the IB adopts less MS that is, $M{S}^0=\left[1-p\left(MS\right)\right]=0$. The IB choice is associated with risk that contains MS (${\sigma }_i^{MS}$) and less risk $M{S}^0\left({\sigma }_i^{M{S}^0}\right)$. MS is the growth of the Total MS composite indicator comprising Social Indicators (10 sub-indicators), Governance Indicators (92 sub-indicators) and Zakat and Qard Hassan. For a complete definition of this indicator, please see the Supplementary Materials.

A decision that MS is

• p(MS) = 1 or Discrete Value = 1 is the positive growth of the MS total score, and assumed if IB complies with this indicator, IB will have better management and less risk. Positive growth of MS is increasing total score in the MS either CSR and Governance;
• p(MS) = 0 or Discrete Value = 0 is the negative growth of the MS total score, and assume if MS total score is decreasing, the IB is risky.

This study assume that the regime will confirm by the positive/negative growth when the IB has negative growth persistently, so the MS regime will shift to less MS regime ($M{S}^0$). Otherwise, if the IB has a persistent positive growth in the MS., meaning IB has a strong regime in MS.

If the IB adopting MS is determined by risk, such that:

$${\sigma }_i^{\lambda }\in \{{\sigma }_i^{MS},{\sigma }_i^{M{S}^0}\}\mathrm{where}{\sigma }_i^{MS}<{\sigma }_i^{M{S}^0};\lambda =\{MS,M{S}^0\}$$

(1)

Equation (1) states that IB face project with more risk, so ${\sigma }_i^{MS}<{\sigma }_i^{M{S}^0}$. The implication of Equation (1), produce two separated continuums that portrays a combination of the expected return that IB adopts MS (${\rho }^{\lambda })$ and profit loss sharing rate ($P^{\lambda }$). These continuum areas separated by two P optimal points, that is $P_c^{MS}$ and $P_0^{M{S}^0}$, where $P_c^{MS}<P_0^{M{S}^0}$. The P distribution is defined as $P\in \left\{P^{MS},\cdots ,P_c^{MS},P^{M{S}^0},\cdots ,P_0^{M{S}^0}\right\}$. These two separated distributions distinguish two benefit function $\pi \in \left\{{\pi }_{MS},{\pi }_{M{S}^0}\right\}$, where ${\pi }_{MS}>{\pi }_{M{S}^0}$ have two cases distribution according to 1, so the benefit function:

$$\pi \left\{\begin{array}{c}{\pi }^{MS}={\int }_0^{P_c^{MS}}{\rho }^{MS}\left({\sigma }_i^{MS}\right)dP\\ {\pi }^{M{S}^0}={\int }_{>P_c^{MS}}^{P_0^{M{S}^0}}{\rho }^{M{S}^0}\left({\sigma }_i^{M{S}^0}\right)dP\end{array}\right.\mathrm{where}{\pi }^{MS}>{\pi }^{M{S}^0}$$

(2)

The Equation (2) assumes that is monotonic behavior, hence the optimal point of benefit and risk according to Boyd and Nicolo [47] where profit loss sharing (P) as the function of deposit (Wadiah-W) is

$${\rho }^{\lambda }=P\left({\sigma }_i^{\lambda }\right)-W\left({\sigma }_i^{\lambda }\right)=0$$

(3)

Equation (3) states that the expected return will reach an optimal point where the P − W is no changes or 0. According to Equations (2) and (3), the total benefit for two continuum is,

$$\pi =\underset{p\left(MS\right)}{\underbrace{{\int }_0^{P_c^{MS}}\left\{P\left({\sigma }_i^{MS}\right)-W\left({\sigma }_i^{MS}\right)\right\}dP}}+\underset{p\left(M{S}^0\right)}{\underbrace{{\int }_{>P_c^{MS}}^{P_0^{M{S}^0}}\left\{P\left({\sigma }_i^{M{S}^0}\right)-W\left({\sigma }_i^{M{S}^0}\right)\right\}dP}}$$

(4)

Equation (4) is confirming the probability of financing behavior of IB (either $p\left(MS\right)$ or $p(M{S}^0)$) that determine by P and W. In Equation (4), $p\left(MS\right)$ is a total project that IB benefit. So, when the project financing has more benefit, the growth of financing is positive $\frac{\mathsf{\Delta }\rho }{\mathsf{\Delta }P}>0.$ Hence if $\left[p\left(MS\right)\right]=1\mathrm{so}\frac{\mathsf{\Delta }\rho }{\mathsf{\Delta }P}>0=\mathsf{\Delta }{F}_{MS}^{it}>0$, otherwise if $\left[1-p\left(MS\right)\right]=0\mathrm{so}\frac{\mathsf{\Delta }\rho }{\mathsf{\Delta }P}<0=\mathsf{\Delta }{F}_{M{S}^0}^{it}<0$. Simplification for $\frac{\mathsf{\Delta }\rho }{\mathsf{\Delta }P}=\mathsf{\Delta }{F}_{\lambda }^{it}$ that proxied if financing growth is positive or vice versa. The detailed behavior of Equations (1)–(4), is depicted in Figure 1 as follows:

Figure 1a portrays Equation (4) that consists of area with less risk project $A^{\prime }=p(MS)$, and risky project $B^{\prime }=p\left(M{S}^0\right)=p\left(M{S}^0\right)=1-p\left(MS\right)=0$. Area A’ and B’ are total benefit obtained by IB that larger than conventional banks A’ > A and B’ > B. (For a complete argument, please see Supplementary S1).

In order to linkage between the level of competition and P as depicted in Figure 1b, it can derive, according to the Learner Index formula $LI=\frac{P-MC}{P}$ as noted in Equation (4) about P as Price and MC is Marginal Cost. In Equation (3), that $P=P^{\lambda }({\sigma }_i^{\lambda })$ and $MC=W\left({\sigma }_i^{\lambda }\right)-\frac{\partial W}{\partial {\sigma }_i^{\lambda }}$, therefore:

$$LI=\frac{P^{\lambda }\left({\sigma }_i^{\lambda }\right)-W\left({\sigma }_i^{\lambda }\right)-\frac{\partial W}{\partial {\sigma }_i^{\lambda }}}{P^{\lambda }\left({\sigma }_i^{\lambda }\right)}$$

(5)

Equation (5) is the Lerner Index, that in line with $\lambda ;$ so Equation (5) has two cases, portrayed in Figure 1b as follows:

$$LI\left\{\begin{array}{c}L{I}^{MS}=\frac{P^{MS}\left({\sigma }_i^{MS}\right) - W\left({\sigma }_i^{MS}\right) - \frac{\partial W}{\partial {\sigma }_i^{MS}}}{P^{MS}\left({\sigma }_i^{MS}\right)}=f(P^{MS})\\ L{I}^{M{S}^0}=\frac{P^{M{S}^0}\left({\sigma }_i^{M{S}^0}\right) - W\left({\sigma }_i^{M{S}^0}\right) - \frac{\partial W}{\partial {\sigma }_i^{M{S}^0}}}{P^{M{S}^0}\left({\sigma }_i^{M{S}^0}\right)}=f(P^{M{S}^0})\end{array}\right.$$

(6)

Equation (6) says that LI location depends on $P^{MS}$ and $P^{M{S}^0}$, when $\frac{\partial LI}{\partial P}>0$ seems that level competition is getting relaxed towards market power. Otherwise, $\frac{\partial LI}{\partial P}<0$, that level of competition is getting tighter along with decreasing $P^{\lambda }$ towards 0.

In order to obtain the empirical model according to Equation (6) to portrayed Figure 1a the return function as noted by ${\rho }^{\lambda }=f\left(P^{\lambda }\right)$ separated by ${\sigma }_i^{\lambda }$, such that:

$$\pi ={\int }_{\left[1-p\left(MS\right)\right]=0}^{\left[p\left(MS\right)\right]=1}{\rho }^{MS}dP={\int }_{\left[1-p\left(MS\right)\right]=0}^{\left[p\left(MS\right)\right]=1}{P}^{MS}\left({\sigma }_i^{MS}\right)dP$$

(7)

Furthermore, while P is the key variable that is connected between level of competition and market financing. Hence, Equation (7) portrays Figure 1b as follows:

$${\int }_{\left[1-p\left(MS\right)\right]=0}^{\left[p\left(MS\right)\right]=1}{P}^{\lambda }\left({\sigma }_i^{\lambda }\right)dP=f\left(L{I}^{\lambda }\right)dP$$

(8)

To simplify the equation about choice, Equations (7) and (8) were substituted according to Equation (4), thats portray with Figure 1 to obtain a simultaneous equation solution as follows:

$$\underset{\mathrm{Figure}1\mathrm{panel}\left(\mathrm{a}\right):\mathrm{market}\mathrm{financing}}{\underbrace{\mathrm{Pr}(p\left(MS\right)=\mathsf{\Delta }{F}_{\lambda }^{it}=1|0)=f\left({\tilde{P}}^{\lambda },{\sigma }_i^{\lambda }\right)}};\underset{\mathrm{Figure}1\mathrm{panel}\left(\mathrm{b}\right):\mathrm{competition}}{\underbrace{{\tilde{P}}^{\lambda }=f(W^{\lambda },{\sigma }_i^{\lambda },L{I}^{\lambda })}}$$

(9)

In Equation (9), is a simultaneous equation that portrays two behaviors between Figure 1a as the market financing equation; and Figure 1b as the competition equation.

Where $\mathrm{Pr}(p\left(MS\right)=\mathsf{\Delta }{F}_{MS}^{rt}=1|0)$ that consist of 1 when $\mathsf{\Delta }{F}_{MS}^{rt}$ is positive ($\frac{\mathsf{\Delta }\rho }{\mathsf{\Delta }P}>0$) as this would indicate whether ${\sigma }_i^{MS}$ has improved. Otherwise, 0 is vice versa. ${\sigma }_i^{\lambda }$ is risk aggregate indicator consisting of MS index, Non-Performing Loan (NPL), Cost to Income—EFF, Profit Loss Sharing Income—PLSI, Non-Profit Loss Sharing Income (in conventional terms is Non-Interest Income)—PLSNI, Gross Domestic Product (GDP) and Inflation (INF). The MS index consists of governance indicators and social indicators. The ${\tilde{P}}^{MS}$ is the profit loss sharing rate (in conventional terms, Net Interest Income) as an endogenous variable. ${\tilde{P}}^{\lambda }$ is profit loss sharing is determined by Wadiah Rate-W, R is net income and Zakat-Z and Qard-Hassan-ZQH. LI is the Lerner Index that indicates a level of competition.

The empirical equation was derived according to Equation (9) as follows:

$$\mathrm{Pr}(\mathsf{\Delta }{F}_{\lambda }^{it}=1|0)=f({\tilde{P}}^{MS},MS,NPL,EFF,PLSIPLSNI,GDP,INF)$$

(10a)

$${\tilde{P}}^{\lambda }=f(W,R,LI,ZQH)$$

(10b)

where Equation (10a) states as the market financing equation, and Equation 10b is the competition equation. After those Equations (10a) and (10b) were estimated, analyzing the effect of MS on competition in IB financing conducted with the interaction effect of MS affects competition using two channels, as noted in Equations (6) and (7),

$$\rho \left\{\begin{array}{c}p\left(MS\right)=1=\mathsf{\Delta }{F}_{MS}^{it}=\frac{\partial \mathsf{\Delta }{F}_{MS}^{it}}{\left\{\partial P^{MS}\cdot \partial MS\right\}}=\frac{\partial \mathsf{\Delta }{F}_{MS}^{it}}{\partial P^{MS}}\cdot {\tilde{P}}^{MS}+\frac{\partial \mathsf{\Delta }{F}_{MS}^{it}}{\partial MS}\cdot MS;\frac{\partial P^{MS}}{\partial LI}\cdot L{I}^{MS}\\ p\left(M{S}^0\right)=0=\mathsf{\Delta }{F}_{M{S}^0}^{it}=\frac{\partial \mathsf{\Delta }{F}_{M{S}^0}^{it}}{\left\{\partial P^{M{S}^0}\cdot \partial M{S}^0\right\}}=\frac{\partial \mathsf{\Delta }{F}_{M{S}^0}^{it}}{\partial P^{M{S}^0}}\cdot {\tilde{P}}^{M{S}^0}+\frac{\partial \mathsf{\Delta }{F}_{M{S}^0}^{it}}{\partial M{S}^0}\cdot M{S}^0;\frac{\partial P^{M{S}^0}}{\partial L{I}^0}\cdot L{I}^{M{S}^0}\end{array}\right.;\mathsf{\Delta }{F}_{MS}^{it}>\mathsf{\Delta }{F}_{M{S}^0}^{it}$$

(11)

Equation (11) stated that sustainability is matters for improving competition by comparing the LI effect on both IB apply MS or Less MS ($M{S}^0$).

To ensure that $\mathsf{\Delta }{F}_{MS}^{it}>\mathsf{\Delta }{F}_{M{S}^0}^{it}$, several assumptions should follow according to Equation (11).

Assumption 1.

${\alpha }^{MS}=\frac{\partial \mathsf{\Delta }{F}_{MS}^{it}}{\partial P^{MS}}<{\alpha }^{M{S}^0}=\frac{\partial \mathsf{\Delta }{F}_{M{S}^0}^{it}}{\partial P^{M{S}^0}}$ where ${\alpha }^{MS}<0;{\alpha }^{MS0}<0$; this assumption as define in Hypothesis 1a;

Assumption 2.

${\beta }^{MS}=\frac{\partial \mathsf{\Delta }{F}_{MS}^{it}}{\partial MS}>{\beta }^{M{S}^0}=\frac{\partial \mathsf{\Delta }{F}_{M{S}^0}^{it}}{\partial M{S}^0}$ where ${\beta }^{MS}>0;{\beta }^{M{S}^0}>0$; this assumption applies for Hypothesis 1b;

Assumption 3.

${\delta }^{MS}=\frac{\partial P^{MS}}{\partial L{I}^{MS}}>{\delta }^{M{S}^0}=\frac{\partial P^{M{S}^0}}{\partial L{I}^{M{S}^0}}$; ${\delta }^{MS}>0;{\delta }^{M{S}^0}>0$ applies for Hypothesis 2;

Assumption 4.

$\mathsf{\Delta }{F}_{\lambda }^{it}=\frac{\partial \mathsf{\Delta }{F}_{\lambda }^{it}}{\left\{\partial P^{\lambda }\cdot \partial M{S}^{\lambda }\right\}}=\frac{\partial \mathsf{\Delta }{F}_{\lambda }^{it}}{\partial P^{\lambda }}\cdot {\tilde{P}}^{\lambda }+\frac{\partial \mathsf{\Delta }{F}_{\lambda }^{it}}{\partial M{S}^{\lambda }}\cdot MS;\frac{\partial P^{\lambda }}{\partial L{I}^{\lambda }}\cdot L{I}^{\lambda }$; $\mathsf{\Delta }{F}_{MS}^{it}$ > $\mathsf{\Delta }{F}_{MS}^{it}$ if Assumptions 1–3 are compliant and apply in Hypothesis 3.

The sustainability matters when $\mathsf{\Delta }{F}_{MS}^{it}$> $\mathsf{\Delta }{F}_{M{S}^0}^{it}$ (Assumption 4) that should comply with Assumptions 1 to 3. Assessing the impact of these assumptions applying simulation of the interaction effect that is indicated with ${\left(\cdot \right)}^{*}$

$${\left(\mathsf{\Delta }{F}_{\lambda }^{it}\right)}^{*}=\frac{\partial \mathsf{\Delta }{F}_{\lambda }^{it}}{\left\{\partial P^{\lambda }\cdot \partial M{S}^{\lambda }\right\}}=\frac{\partial \mathsf{\Delta }{F}_{\lambda }^{it}}{\partial P^{\lambda }}\cdot {\left({\tilde{P}}^{\lambda }\right)}^{*}+\frac{\partial \mathsf{\Delta }{F}_{\lambda }^{it}}{\partial M{S}^{\lambda }}\cdot MS;\frac{\partial P^{\lambda }}{\partial L{I}^{\lambda }}\cdot {\left(L{I}^{\lambda }\right)}^{*}$$

(12)

The hypothesis testing will evaluate whether baseline model of 12 (b =$\mathsf{\Delta }{F}_{MS}^{it}$ > $\mathsf{\Delta }{F}_{M{S}^0}^{it}$) vs. simulation model with increasing a similar level of competitiveness ($L{I}^{\lambda }$): (s = ${\left(\mathsf{\Delta }{F}_{MS}^{it}\right)}^{*}$ > ${\left(\mathsf{\Delta }{F}_{M{S}^0}^{it}\right)}^{*}$) on both MS are consistent (b > s). In IB with MS increasing competition (LI), will improve the price and reduce the growth of financing and profit, but if IB applies strong MS: s < b.

So, this analysis ensures how the empirical model assists the theoretical model to confirm the definition in Equation (11).

4. Material and Methods

4.1. Materials

This study using panel data that retrieved from Bankscope and Thomson Reuters. The dedicated sustainability data were obtained from Thomson Reuters due to the only institution that provides a large consistent dataset of bank data for sustainability in Islamic Bank. In addition, the Bankscope is one of the largest consistent dataset that contains holistic financial statement that requires by the empirical model.

In the first stage, those datasets were combined in bank level across countries that contain financial performance indicators from Bankscope and Thomson Reuters for sustainability indicators for IB. The total sample period was retrieved from 2012–2020 is about 160 banks across 35 countries. To obtain a consistent dataset, the data for year 2020 were removed to avoid inconsistency due to the pandemic. The data were filtered according to sustainability indicators on Thomson Reuters that remain 12 countries and 35 banks, as noted in

Table 1. Data Sample and Data Source.

No.	Country	Number of Banks	Period	Database
1	Arab Saudi	4	2012–2019	Bankscope: (Financial Performance Non-Performing Loan-NPL, Operating Expenses, Total Assets, Net Income, Gross Loans, Profit Loss Sharing Income—PLSI, Profit Loss Sharing Non-Income-PLSNI) Thomson Reuters: (Sustainability Performance: CSR Activities Score; Financial Reporting Disclosure Index (FRDI) Score, Zakat and Charity, Qard Al Hasan).
2	Bahrain	4	2012–2019
3	Indonesia	4	2012–2019
4	The United Kingdom	1	2012–2019
5	Jordan	1	2012–2019
6	Kuwait	5	2012–2019
7	Malaysia	6	2012–2019
8	Oman	1	2012–2019
9	Qatar	4	2012–2019
10	Thailand	1	2012–2019
11	Turkey	1	2012–2019
12	United Arab Emirates	3	2012–2019
	Total	35	2012–2019

Source: Author compilation, 2023.

.

If the data are incomplete due to missing data, whereas forcing balance panel data are crucial in IV-Probit estimation. The data were interpolated using the moving average technique to fulfill the missing data. The strong balance panel data are crucial to avoid non-robust results, as Wooldridge [48] noted.

Table 2. Statistics Descriptive of Selected Variables.

Variable	Obs	Mean	Std. Dev.	Min	Max	Concept Adaptation
Total Cost = w1 + w2 + w3	217	0.056	0.025	0.022	0.257	Bikker and Spierdijk [9]
MS	217	51.412	16.013	0	117	Mergaliyev et al. [15]
w1 = Interest Expense/Deposit	217	0.027	0.020	0.000	0.187	Bikker and Spierdijk [9]
w2 = Interest Expense/Total Assets	217	0.018	0.012	0.000	0.064	Bikker and Spierdijk [9]
w3 = Other Operating Exp./Total Fix Assets	217	6.910	30.567	0.000	301.512	Bikker and Spierdijk [9]
MC = Marginal Cost	217	0.021	0.012	0.000	0.054	Bikker and Spierdijk [9]
Lerner Index (LI)	217	0.611	0.223	0.019	0.997	Bikker and Spierdijk [9]
Profit Loss Sharing Rate—in %	217	2.839	1.886	0.115	9.084	Boyd and Nicolo [47]; Beck et al. [22]
Wadiah Rate—in %	217	2.705	2.129	0.096	18.697	Boyd and Nicolo [47]; Beck et al. [22]
Dummy of Growth of Financing ($\mathsf{\Delta }{F}^{it})$	217	0.842	0.365	0.000	1.000	Solow’s [49]
Dummy of Growth of Maqasid Shariah ($M{S}^{it}$)	217	0.618	0.487	0.000	1.000	Solow’s [49]
Cost to Income (EFF)	217	0.025	0.021	0.005	0.206	Beck et al. [22]
Non-Performing Loan (NPL)—in%	217	3.344	3.725	0.000	20.811	Beck et al. [22]
Profit Loss Sharing Income (PLSI)	217	1115.291	3359.788	(796.745)	36,650.000	Beck et al. [22]
Profit Loss Sharing Non-Income (PLSNI)	217	537.962	1445.816	(451.299)	14,066.710	Beck et al. [22]
Gross Domestic Product (GDP)	217	3.033	2.634	(4.700)	16.300	Beck et al. [22]
Inflation (Inf)—in %	217	2.400	2.362	(2.100)	16.300	Beck et al. [22]
Net Income ($R^{it}$)	217	10.287	3.309	(0.636)	14.812	Beck et al. [22]
Zakat and Qard Hassan (ZQH)	217	3.214445	6.032481	0	19.31852	Mergaliyev et al. [15]

Note: Obs = Number of observations, Std.Dev = Standard Deviation; Min = Minimum; Max = Maximum. Source: Author calculation, 2023.

depicts descriptive statistics for selected variables that employ in the empirical model that will discuss in the next subsection.

4.2. Methods

Equations (10a) and (10b) are the main model that tests the hypothesis using IV-Probit techniques. Accordingly, the empirical model is estimated as follows.

$$\begin{array}{c}\mathrm{Pr}\left(\mathsf{\Delta }{F}^{it}=1|0\right)={\alpha }_0^{it}+{\alpha }_1^{it}\cdot {\tilde{P}}^{it}+{\alpha }_2^{it}\cdot M{S}^{it}+{\alpha }_3^{it}\cdot EF{F}^{it}+{\alpha }_4^{it}\cdot {NPL}^{it}+{\alpha }_5^{it}\cdot PLS{I}^{it}\\ +{\alpha }_6^{it}\cdot PLSN{I}^{it}+{\alpha }_7^{it}\cdot GD{P}^{rt}+{\alpha }_8^{it}\cdot IN{F}^{rt}+{ϵ}_1^{it}\\ {\tilde{P}}^{it}={\beta }_0^{it}+{\beta }_1^{it}\cdot R^{it}+{\beta }_2^{it}\cdot L{I}^{it}+{\beta }_3^{it}\cdot W^{it}+{\beta }_4^{it}\cdot ZAK+{\beta }_5^{it}\cdot ZQH+{ϵ}_2^{it}\end{array}$$

(13)

Equation (13) states that i is a banking firm, where t is period, and r is a country in the sample, where $\lambda \in \{I{B}^{1t},I{B}^{2t},\cdots ,I{B}^{it}\}$ so the it represents the heterogeneity of IB across the world adopt $MS$ and $M{S}^0.$

The Learner Index (LI) was estimated using the econometric method, which Marginal Cost (MC) estimates as follows:

$$M{C}^{it}=\frac{T{C}^{it}}{Q^{it}}({\mathsf{\gamma }}_1\cdot \ln Q^{it}+\sum {\beta }^{it}{w}^{it})$$

(14)

where Q is Gross Loan, and other variables are defined in Table 1. The LI was estimated using constraint regression that produced MC as follows:

$$L{I}^{it}=\frac{P^{it}-M{C}^{it}}{P^{it}}$$

(15)

where P is Profit Loss Sharing Income (PLSI) + Non-Profit Loss Sharing Income (PLSNI)/Total Assets. The MC was calculated using Equation (14).

In order to ensure whether IV-Probit is a proper technique that requires endogeneity, The exogeneity test was conducted. The test using Wald test to ensure the IV-Probit is adequate to estimate Equation (13). This procedure was suggested by Smith [38] and Newey [50] to test the null hypothesis of no endogeneity. The null hypothesis is accepted where the model does not require an IV equation, such that ${\tilde{P}}_{it}^{\lambda }$ is no endogeneity variable. Otherwise, if the null hypothesis is rejected, the Wald statistic and p-value of the test are jointly zero.

In order to test the Hypotheses 1–2, Equation (13) was employed until the Wald exogeneity test was passed [51]. After it passed, each parameter in Equation (13) was tested with a p-value from the marginal effect of IV-Probit estimation. In particular, for Hypothesis 3, the test was conducted using causal inference with IV as suggested by Tan [52]. The IV can be applied by simulating the interaction treatment effect based on Equation (13). The estimation starts with baseline estimation and followed by simulation estimation. The results of this interaction effect are depicted in the graph as suggested by Long and Freese [53]. The simulation demonstrated that increased similar competition by about 10% and increased P (from 1% to 6%) in IB with different MS producing different financing growth.

5. Results

This section will report the estimation results to test each hypothesis according to Equation (13). Prior to describing the results, the Wald Exogeneity test was conducted to obtain the most proper model that reflects IV equation behavior to support the theoretical model, as noted in Figure 1. Ideally, the Wald test should reject the null hypothesis by adding IV variable (i.e., $R^{it}$, $L{I}^{it},W^{it}$) into the baseline model, as depicted in

Table 3. Marginal Effect of MS and Competition on IB Financing: Panel Probit vs. IV-Probit.

	(Panel Probit)	T-Stat	(IV-Probit 1)	T-Stat	(IV-Probit 2)	T-Stat
VARIABLES	$\mathsf{\Delta }{\mathit{F}}_{\mathit{\lambda }}^{\mathit{i}\mathit{t}}$		$\mathsf{\Delta }{\mathit{F}}_{\mathit{\lambda }}^{\mathit{i}\mathit{t}}$		$\mathsf{\Delta }{\mathit{F}}_{\mathit{\lambda }}^{\mathit{i}\mathit{t}}$
Market Financing Equation:
Profit Loss Sharing Rate (${\tilde{P}}^{it}$)	−0.0556 ***	−2.598	−0.0658 ***	−3.118	−0.0764 ***	−3.457
	(0.0214)		(0.0211)		(0.0221)
Growth for MS Score ($M{S}^{it}$)	0.0838 **	2.229	0.0842 *	1.927	0.0889 **	2.053
	(0.0376)		(0.0437)		(0.0433)
Efficiency ($EF{F}^{it}$)	0.00450	0.006	0.0155	0.014	−0.0300	−0.028
	(0.764)		(1.101)		(1.090)
Non-Performing Loan ($NP{L}^{it}$)	−0.00484	−0.639	−0.00430	−0.788	−0.00419	−0.765
	(0.00758)		(0.00546)		(0.00548)
Net Income from $\tilde{P}$ ($PLS{I}^{it}$)	−1.19 × 10−6	−0.220	−3.61 × 10−7	−0.063	−1.32 × 10−6	−0.240
	(5.40 × 10−6)		(5.76 × 10−6)		(5.49 × 10−6)
Income from Non-$\tilde{P}$ (PLSNI)	−2.64 × 10−5 ***	−2.927	−2.33 × 10−5	−1.363	−2.51 × 10−5	−1.426
	(9.02 × 10−6)		(1.71 × 10−5)		(1.76 × 10−5)
Growth of GDP ($GD{P}^{rt}$)	−0.00721	−0.407	−0.00538	−0.371	−0.00475	−0.321
	(0.0177)		(0.0145)		(0.0148)
Inflation (${INF}^{rt}$)	−0.00385	−0.228	−0.00782	−0.575	−0.00864	−0.640
	(0.0169)		(0.0136)		(0.0135)
	${\tilde{P}}^{it}$		${\tilde{P}}^{it}$		${\tilde{P}}^{it}$
Competition Equation:
Net Income ($R^{it}$)	−0.0184 **	−2.035	−0.0139 ***	−3.983	−0.0187 ***	−3.970
	(0.00904)		(0.00349)		(0.00471)
Lerner Index ($L{I}^{it}$)	−0.0138	−0.124	0.0526 **	2.338	0.0720 **	2.449
	(0.111)		(0.0225)		(0.0294)
Wadiah Rate ($W^{it}$)	0.0156	0.891	0.0305 ***	3.109	0.0379 ***	3.081
	(0.0175)		(0.00981)		(0.0123)
Zakat and Qard Hassan ($ZQ{H}^{it}$)					−0.00184 **	−2.147
					(0.000857)
Number of Observations	217		217		217
Wald Exogenity Test: ${\chi }^2$ p-value			3.80 (0.051)		4.85 (0.028)

Standard errors in parentheses: *** p < 0.01, ** p < 0.05, and * p < 0.1. Source: Author calculation, 2023.

. If the model has fulfilled endogeneity in the IV estimation, the p-value is decreasing along with adding IV variables. In addition, the model also has smaller standard error and more significant variables than previous model such IV-Probit, and Panel Probit. The results will confirm that model is robust and ready for hypothesis testing.

Table 3 demonstrates three marginal effects from Panel Probit, IV-Probit 1 that indicate without including IV of altruism motives such as Zakat and Qard Al Hassan. In contrast, IV-Probit 2 indicates adding IV altruism motives such as Zakat and Qard Al Hasan in the model to improve better endogeniety and model robustness that portray Equation (9). In Table 3, column IV-Probit 2, indicates that the p-value is passed Wald exogeneity test (i.e., the p-value = 0.028 < 0.05) that conclude model IV-Probit 2 is the proper model to test the hypothesis and in line with a theoretical model of Equation (9).

Addressing Hypothesis 1a, whether ${\tilde{P}}^{it}$ has negative slope on growth of financing ($\mathsf{\Delta }{F}_{\lambda }^{it}$). Table 3 column IV-Probit 2 indicated at parameter ${\tilde{P}}^{it}$ that the probability increasing of 1% of ${\tilde{P}}^{it}$ reduces the probability of growth financing by more than 7% with a 99% significance level. The parameter of ${\tilde{P}}^{it}$ is negative ($\frac{\partial \mathsf{\Delta }{F}_{\lambda }^{it}}{\partial P^{\lambda }}<{\alpha }^{\lambda }$) and in line with Hypothesis 1a and support with Figure 1a that increasing growth of financing reduce ${\tilde{P}}^{it}$. The negative parameters confirm the existing hypothesis that improving the price of financing decreases the growth of IB financing.

In Hypothesis 1b, whether $M{S}^{it}$ improves financing growth. Table 3 column IV-Probit 2 indicated on parameter $M{S}^{it}$ that improving 1% of $M{S}^{it}$ increases the probability of growth financing by more than 8% with a 95% significance level. During 2012–2019, the $M{S}^{it}$ parameter indicates a significant effect on the growth of financing ($\mathsf{\Delta }{F}_{\lambda }^{it}$). This empirical evidence of $M{S}^{it}$ that $\frac{\partial \mathsf{\Delta }{F}_{\lambda }^{it}}{\partial M{S}^{\lambda }}>0$ is confirmed with the hypothesis.

In Hypothesis 2, the LI positive effect on $\mathsf{\Delta }{F}^{it}$. According to Table 3, columns IV-Probit 2, at parameter $L{I}^{it}$ positively affects $\mathsf{\Delta }{F}^{it}$ by 95% of the significance level. Increasing by 1% of the level of competition improves the probability of improving growth financing by about 7%. Now, the parameter confirms that empirical estimation is in line with the theoretical model that $\frac{\partial P^{\lambda }}{\partial L{I}^{\lambda }}>0$.

Other control variables, such as The Net Income ($R^{it}$), affect IB financing growth ($\mathsf{\Delta }{F}^{it}$) with negative parameters. Wadiah Rate ($W^{it}$) and Zakat and Qard-Hassan ($ZQ{H}^{it}$) positively affect ${\tilde{P}}^{it}$. Those variables are significant, with a 95–99% of the statistical significance level. Other risk variables such as NPL, Efficiency, PLSI, NPLSI, GDP and INF do not statistically affect the $\mathsf{\Delta }{F}^{it}$. Only $R^{it}$ parameter is opposite to the theory; otherwise, $W^{it}$ and $ZQ{H}^{it}$ are as expected with the theory.

The last hypothesis, whether sustainability matters for improving competition in IB finance. In

Table 4. Marginal Effect of Treatment Interaction Effect of Different MS and Competition.

	Simulation	T-Stat
Islamic Bank (IB) with $M{S}^0$	1.0541 ***	2.740
	(0.3846)
Islamic Bank (IB) with $MS$	1.4754 ***	4.153
	(0.3552)

Standard errors in parentheses: *** p < 0.01. Source: Author calculation, 2023.

, when IB applies MS: in competition equation: $L{I}^{it}$ increase ${\tilde{P}}^{it}$. Accordingly, in the market financing equation: ${\tilde{P}}^{it}$ reduces $\mathsf{\Delta }{F}^{it}$ and $M{S}^{it}$ increases $\mathsf{\Delta }{F}^{it}$. When IB adopts $MS$, the competition level improves by 10%, and ${\tilde{P}}^{it}$ increase from 1% to 5% producing $\mathsf{\Delta }{F}^{it}$ increase of 47.54% more than $M{S}^0$ with a 99% significance level. In $M{S}^0$, increases $\mathsf{\Delta }{F}^{it}$ about 5.41% with 99% of the significance level. Now, it is proved that adopting MS increases competitiveness and financing growth compared to less MS.

In order to conduct the above simulation, the results in Table 4 were employed to indicate the separated distribution of MS and $M{S}^0$to illustrate Figure 1a. Whether a transition probability shifted across MS or $M{S}^0$, interchangeably. Surprisingly, the value matches MS distribution, and the transition probability is constant for both MS and $M{S}^0$, there is no shifting between them, that in line with Equation (4), $\mathsf{\Delta }{F}^{it}=p\left(MS=1\right);\left(1-p\left(MS\right)\right)=0)$. The interaction effect parameter of different MS with increasing P produces persistence gap between MS and $M{S}^0$, as depicted in Figure 2.

Figure 2 is close to Figure 1a. Figure 2 is the curve that indicates the various combination of ${\rho }^{it}=\mathsf{\Delta }{F}^{it}$ as the function of increasing of ${\tilde{P}}^{it}$ on both regimes, that is, MS (red line) and $M{S}^0$ (blue line). Figure 2a is the baseline estimation without LI treatment. In Figure 2b is the simulation estimation with increasing levels of $L{I}^{it}$ by 10% in both regimes. The results show that increasing LI decreases $\mathsf{\Delta }{F}^{it}$ both MS and $M{S}^0$. The red line in panel b is shifting toward lower financing growth than in panel a. The results have consistent transition probability for both red and blue lines that MS is larger than $M{S}^0$. The IB attempt to maintain applying MS accordingly. The results state that sustainability matter to stay competitive for dominant market power.

6. Discussion

According to this study, the results add new findings from the perspective of IB that the REF hypothesis exists. If IB conducts MS, lower prices, improve market share and financing growth. Both theoretical and empirical models have aligned and produced similar behavior, as Benabout and Tirole [11] argue. In addition, the empirical model supports Bester and Hellwig [30] argument that improving moral hazards such as $MS$ produces more financing growth than less moral hazard. The $MS$improving image concern will help banks obtain more incentives from people. Delegated philanthropy managed by Islamic banks such as Zakat and Qard Hassan responds to a widespread demand by stakeholders that Islamic banks interact with good corporate ummah (people). The evidence also supports Stiglitz and Weiss [43] and James [46] finding that prosocial behavior conducted by an IB, such as $MS$ will attract financing activity that is not available from other lenders. In terms of theoretical and empirical, my model has supported the previous finding and extended the results for prosocial behavior and altruism—the MS claim also another alternative way to support sustainability besides conventional Bank sustainability, such as ESG.

This study’s theoretical model developed from Equations (1)–(10) confirms Khan [23] that MS in IB advocates economically lesser risk, more efficient than conventional Banks and promotes more significant equity and justice. On the other hand, the control variables such as parameter $EF{F}^{it}$ is no significant effect on $\mathsf{\Delta }{F}^{it}$. Those parameters that indicates efficiency unable to reveal in the market financing equation. However, in competitive financing equation, the $R^{it}$ has a negative slope at ${\tilde{P}}^{it}$. When $R^{it}$ increases ${\tilde{P}}^{it}$ decreases: this finding supports Andrieș and Sprincean [12] who state that ESG in conventional banks reduces financing costs.

The confirmation of first hypothesis (Hypothesis 1a) confirms previous theoretical findings. It adds new information about the MS concept as part of the existing strategic banking industry that is also of important concern. This statement aligns with Benabou and Tirole [21] who state that applying MS improves IB’s financing. As the estimation parameter is negative, this means applying MS reduces $P^{MS}$. Otherwise, along with reducing MS towards $M{S}^0$, the $P^{MS}$ reaches an optimal financing rate close to $P_c^{MS}$. This behavior is portrayed in Figure 1a. The $P^{MS}$ < 5–7% within MS indicates the optimal level of return that the IB will produce benefit, whereas if $P^{MS}$ > 5–7% indicates the project has a higher risk and moved to $P^{M{S}^0}$. Unfortunately, at this moment, reducing of $P^{MS}$does not improve efficiency along with this shifting. Otherwise, IB rely on managing risk by MS, which compensates reduction $R^{it}$. This means that IB has been struggling to compete with conventional banks, as supported above.

Following the Hypothesis 1b, the finding confirms that MS increases $\mathsf{\Delta }{F}^{it}$. This finding supports Hypothesis 1a, as stated by Porter [37] and Porter and Van Der Linde [10] argument that concerning sustainability improves competitiveness. MS even offers with a stronger message than a conventional bank. Albeit IB is less efficient than conventional banks, business ethics in IB are still stronger due to religious motives that embed in IB. This finding is close to James [46], report that MS, as a unique feature of IB, attract investor and borrower.

The confirmation of second hypothesis states that $L{I}^{it}$ improves ${\tilde{P}}^{it}$. Increasing the level of competition of IB towards market power improves ${\tilde{P}}^{it}$. This study contradicts with Weill [19] that Islamic bank has no market power to compete with conventional banks. Albeit this analysis does not compare with conventional banks empirically, according to theoretical model, the IB at this moment is against Weill [19] finding. In fact, support Léon and Weill [41] finding that IB could substitute conventional banks due to lowering risk, as Beck et al. [22] supported.

In the latest finding, it is confirmed by Figure 2, that applying MS proved to have a stronger effect compared to $M{S}^0$ within a similar level of competitiveness to increase $\mathsf{\Delta }{F}^{it}$. In market financing equation, decreasing the financing rate improves financing growth. This study supports the previous finding, such as Bannier et al. [13] also support that CSR reduces credit risk. Where applying MS reduces credit risk and improves IB financing.

7. Conclusions

Answering that sustainability matter for improving competition in IB financing? Yes, sustainability matters for IB. This statement has been supported by three consecutive hypotheses testing. From the third hypothesis, when IB applies MS, IB with MS has larger market power and improves IB financing compared to IB with $M{S}^0$. The simulation indicates that growth of financing in MS is consistently greater compared to $M{S}^0$ along with improving financing rate. The reason why the behavior of the third hypothesis is in line with the theory, the two following hypotheses support the empirical model, such as the second hypothesis, stated that in the competition equation: competition increases the profit loss sharing rate. While in the market financing equation: MS increases the growth of financing (Hypothesis 1b). Hypothesis 1a states that the financing rate has a negative effect on IB financing. All the hypotheses are aligned with both theoretical and empirical evidence. MS is crucial for IB to compete with conventional banks. When IB complies with MS, it attracts more customers and Wadiah rate and increases delegated philanthropy funding such as Zakat and Qard Hassan. The advantage of this finding is that applying MS in IB will improve its image concern and provide special service for the consumers. These aspects will attract other consumers or borrowers from banks with less MS or even conventional banking to invest in IB.

The implication of the results contributes to advancing knowledge on the REF hypothesis but with different indicators. This study has extended another perspective for existing efficiency analysis using MS indicators as an implicit proxy of efficiency through price behavior that portrays better risk management. The first implication of this paper, it is added new information that needs to be included about sustainability, competition and bank loan. MS as a sustainability perspective in IB has at least a similar impact of improving financing growth compared to conventional banking, while the existing literature was absent. Second, it is added that sustainability proxy by MS in IB has a similar sign effect with conventional banks. The MS improves competitiveness, and more financing growth leads to more profit compared to less MS. Another important implication that has been missed in the literature that prosocial behavior in IB is proven, similar to conventional banks that has been solidly proven. The prosocial behavior in IB is fully adaptable, even IF has delegated mandatory equity such as Zakat and Qard Hasan with less cost than deposit. This delegated mandatory equity not only corporate but individuals as well. Otherwise, this funding relies on corporate rather than individuals in conventional sustainability. The MS concept has financial prospects that will extend beyond current sustainability banking.

The policy implications of this study propose that government should improve symmetric information to achieve more transparency for consumers about basic principles of IF ethics. This policy is crucial to closing the gap in financial literacy between IB and conventional banks. Hence the concept of MS will be spread correctly. From a banking perspective, applying sustainability or MS is crucial to getting deeper into financial service and financial development. The bank should have strategic pathways for how their activity start with energy conservation (applying Renewable Energy in the Office and improving electric appliances with energy efficiency), reducing waste and less material (using the paperless system in the office and expansion of digital transaction), more governance in the banking system and applying more Corporate Social Responsibility (CSR) and Islamic prosocial behavior to the society such as Zakat, Infaq, Shadaqat and Qard Al Hassan and achieve financing priority for vulnerable group and conserving natural resources such as energy conservation and energy saving, or material saving.

This paper has limitations for both theoretical that ignore financial literacy, financial deepening and credit rationing. If those aspects are included, the model will be more realistic to indicate how IB works in the economy. The analysis has not yet compared conventional and IB empirically. At this moment, the impact is based on a theoretical model. Analyzing these aspects in future empirical research would be interesting and challenging and should extend the theoretical model into simpler analysis and expand the sample.

Albeit this paper has some caveats especially for sustainability data that compare the IB vs. conventional banks, this study proposes further research to expand both sample and period of the data. There is a need extension of the theoretical model to capture the financial literacy and credit rationing hypothesis, so further study can compare how the asymmetric information that is proxy by financial literacy enhances the performance of the current model using new data from conventional banks. This extension model will need more advanced techniques to estimate the theoretical model. However, at this stage, the empirical model can capture the theoretical model, without estimating conventional banks. Applying this model for adding conventional banks requires more assumption and model advancement. Especially balancing IB and conventional bank samples the reduced variance to avoid large standard deviation that hampered model robustness. Relaxing this problem, while the IV-Probit is likely to be used for further analysis, another potential method is the multinomial logit that directly estimates for estimating for IB and conventional that compare for Bank with sustainability and less sustainability. In addition, if further study connects MS with the current situation in digital banking. Complying with MS in practice will be easier and quicker access. Every bank could assess the entire company using digital platforms such as applications, or any other platform and then assessed it on time. The decision process becomes more efficient and productive.