310Bulletin of Monetary Economics and Banking, Volume 20, Number 3, January 2018

I. INTRODUCTION

The implementation of the dual banking system, in 2008, stipulated that banks in Indonesia have now embraced two working systems that are based on conventional principles and islamic. A large number of banks will certainly increase banking competition in the country, not to mention the competition in Islamic banking. As of October 2014, there are 13 islamic banks consisting of four foreign exchange banks, six non-foreign exchange banks and three mixed commercial banks (www. ojk.go.id).

The size of competition and the market structure are seen from the market power, which is the ability of a company (seller) to raise its relative price compared to its competitors without losing all its sales. The market power is the difference between the price and the marginal cost expressed relative to the price, which is formulated as L = (P-MC)/P where L is the Lerner index which is an indicator of market power, P is the price at which the firm sells its output, and MC is the company’s marginal cost for the volume of the firm. When P = MC, the firm competes as in perfect competition where the product is sold homogeneously, then L = 0. If P> MC, then L> 0 means that monopolist profit is maximal (Pindyck and Rubinfeld, 2012).

This study uses the company’s share approach on the grounds that this approach is more specific. The theory that can give a clear picture as revealed by Cruch and Ware in Teguh (2013) mentioning that there are two groups of oligopoly companies, namely oligopolies that control some sales or all sales. The first group represents the 8 largest companies that control 75% of the total output in the market. The second represents the 8 companies that control at least 33% of the total market output, if 8 companies control less than 33% of the market share, the industry can be said not concentrated. This statement is supported by Lubis (2012) which uses 8 sample banks in Indonesia.

Based on these two studies, the researcher used samples of 8 islamic commercial banks using the percentage of finance share approach which is the result of the financing amount of each islamic bank proxy to total financing of all islamic banking.

Figure 1. The Rating of Islamic Banks Based on Financing in October 2014 (Processed)

47.78
22.14
9.19	7.61
		4.77
			3	2.53	1.74

Bank	Bank Syariah	Bank Rakyat	Bank Panin	Bank Negara	Bank	Bank Jabar	Bank Central
Muamalat	Mandiri	Indonesia	Syariah	Indonesia	Bukopin	Banten	Asia Syariah
Indonesia		Syariah		Syariah	Syariah

Market Structure and Competition of Islamic Banking in Indonesia	311

Figure 1 showed that the largest percentage of financing is owned by Bank Muamalat Indonesia (47.78%), followed by Bank Syariah Mandiri (22.14%), BRI Syariah (9.19%), Panin Syariah Bank (7.61%), BNI Syariah (4.77%), Bank Bukopin Syariah (3%), and BJB Syariah and BCA Syariah with 2.53% and 1.74%, respectively. The large difference in terms of financing between the top four banks and the four bottom banks indicates that the islamic banking industry is sufficiently concentrated.

The banking concentration will bring competition to some banks, leading to monopoly. Uncompetitive competition is based on structural industry approach i.e. Structure-Conduct-Performance (SCP). The SCP approach assumes a one- way approach between the structure and performance (Martin, 1988). This is a disadvantage of the SCP approach. Thus, a new approach to measuring the competition parameters, based on the behavior, is the New Empirical Industrial Organization (NEIO). One form of this method is the Panzar and Roose model or better-known as the model of H-statistics. Researcher uses this method as it has advantages over other models that are able to determine the broader market structures, estimates using linear regression, simple variables, and the use of individual data (cross section) which is more accurate in predicting market power (Panzar and Roose, 1987). This paper aims to determine both the competition and market structure of islamic banking based on Panzar and Roose parameters.

The second part explores the theories and basic test models. The third section is a research methodology and the fourth section reviews the estimation results. The last part is a conclusion.

II. THEORY

2.1. The Development of Islamic Banking

The development of industrial economic calculation was very rapid after the approach of industrial economic model formulated by Bain (1956) through a combination of deductive and empirical approaches. This development has also had an impact on the banking industry. Banks, as strategic institutions in the intermediation of funds, play a significant role in the economic growth. Currently, the banking industry is not only based on the conventional principles, but also on sharia principles. The development of Islamic banking, internationally, was first initiated by Egypt at the Session of the Minister of Foreign Affairs of the Organization of Islamic Conference Organizations (ICO) in Karachi, Pakistan, in December 1970. In subsequent developments in the 1970s, efforts to establish Islamic banks began to spread to many countries. Some countries such as Iran, Pakistan, and Sudan have even transformed their entire financial system into no interest system.

The development of islamic banking continues to the Southeast Asia countries, one of which is Indonesia. The establishment of an Indonesian Islamic bank began in 1980 through discussions on the theme of Islamic banks as pillars of Islamic economy. As a trial, the idea of Islamic banking was practiced on a relatively small scale such as in Bandung (Bait At-Tamwil Salman ITB) and Jakarta (Cooperative Ridho Gusti). Then, in 1990, the Indonesian Ulema Council initiated the

312Bulletin of Monetary Economics and Banking, Volume 20, Number 3, January 2018

establishment of the first Islamic bank, Bank Muamalat Indonesia, which officially operated on May 1, 1992.

Today, the number of islamic banks is increasing due to at least two reasons. Firstly, the crisis of 1998 where Bank Muamalat, in that year, was stronger in facing the crisis. Secondly, Central Bank of Indonesia regulations that have officially implemented dual banking system supported by the enactment of Law Number. 21 of 2008 concerning Islamic Banking.

Both of these causes encouraged the conventional banks to open branches of islamic banks from the initial sharia units, only in the form of development into an independent islamic bank. The increasing number of islamic banks operating in Indonesia certainly pushed the level of competition between banks.

The relevant sizes of banking competition are used i.e. Panzar and Roose models (1987). The usual model, called H statistics, is based on the behavior of individuals or companies in the economy. The PR-H statistic test, in banking, was first performed by Shaffer (1982) with a sample of banks in New York, where the general results of banking in New York operate on monopolistic competition. This is similar to that of Molyneux and Forbes (1995) for banking in Europe, in countries such as France, Germany, Spain, and Britain in the period of 1986-1989, but the banking sector in Italy is indicated as a monopoly market. In addition, many economists such as Panzar and Roose (1987; 1982), Nathan and Neave (1989), Perrakis (1991), Bikker et al (2006) and Goddard and Wilson (2008) recommend the H-statistic PR models. Their study results assumed that companies are free to enter or exit the market without losing capital, and between firms, may lead to competition since it was first established (Gasaymeh, et al, 2014).

2.2. Non-structural approach

If in a structural approach with the SCP model prioritizing one-way approach where the performance of a banking can be seen from its structure, this is certainly not in accordance with the actual conditions in which the banking performance will be more visible from its competing behavior. Then, came the New Empirical Industrial Organization model. The Panzar Roose model, also known as H-Statistics, is an approach commonly used by researchers, especially in the banking industry.

This model was developed by Panzar and Rose in 1987 to measure the degree of competition in an industry, especially banking and competition derivatives in long-term equilibrium, monopoly, and monopolistic competition industries. There are two factors in the acceptance of banking which are the input price and control variable. Assuming that the n - input itself and output production function. The empirical model of H-statistics can be written as follows:

(1)

where TR is the total revenue, wi as an input factor consisting of wage labor, cost of funds and fixed capital costs. While the CF represents the controlled variable including total capital to total assets and total debt to total assets.

(2)

Market Structure and Competition of Islamic Banking in Indonesia	313

The H-statistical value reflects the competitive market structure and is the sum of the elasticity coefficients in the control variables. The representation of the H-statistic values is equal to 0 or negative, then it includes monopolistic competition or collusion oligopoly. It ranges between 0 and 1 for the case of monopolistic competition and if it is 1, then the form of competition is perfect.

The H-statistic model test has different results, it relates to the panel model used whether through Common-Constant, Fixed Effect method or Random Effect method (Naylah, 2010).

Some of the advantages of using H-statistic model are: (1) able to see the broader market structure, (2) can be estimated using linear regression model, (3) only need some variable for testing. Therefore, the Panzar and Roose (PR H-Statistics) model is very comprehensive if used to analyze competition, especially in the banking industry (Panzar and Roose, 1987).

The size of the competition is reflected in the form of market structure. In policy making, banks will consider the form of market structure so that the policy is taken on target. In addition, banks will gain profit or loss based on where they operate. Figure 2 presents the forms of market structure in the economy.

Figure 2. Forms of Markets in Economics

						Perfect competition
Markets									Monopoly
according to
structures
									Monopsony
						Imperfect competition
									Oligopoly
									Oligopsony
									Monopolistic

Source: Pindyck and Rubinfeld (2010), processed.

Based on figure 2 above, it can be explained that the perfect competition on the market is an industry where there are many sellers and buyers, and no seller and buyer can influence the price on the market (price taker). Monopoly is a market that has only one seller and many buyers (price maker). Instead, the monopsony market is a market with many sellers but only one buyer. The monopoly and monopsony are closely related. As a sole producer, the position of monopoly is unique. If the monopolist raises the price then he should not worry about competition. This is because the monopolist controls all the output to be sold (Pindyck and Rubinfeld, 2010).

314Bulletin of Monetary Economics and Banking, Volume 20, Number 3, January 2018

The oligopoly market structure is a market structure where few firms compete with one another and the entry of other firms is inhibited, in contrast to the oligopsony which is the market structure with more sellers and more barriers to entry into the market. Basically, the monopolistic market lies between two types of markets namely perfectly competitive market and monopoly market with each company having unique products.

2.3. Empirical Review

Table 1.

Summary of Previous H-Statistical Research in The Banking Industry

No	Aspects	Description
1	Author	Muhamed Zulkhibri Abdul Majid and Fadzlan Sufian
	Title	Market Structure and Competition in Emerging Market: Evidence from
		Malaysian Islamic Banking Industry
	Period	2001 – 2005
	Results	H-Statistics ranged between 0.375 - 0.616 while in the equilibrium test the value
		ranged between 0824 - 0883 and rejected the Wald test, meaning that the market
		structure and competition of Islamic banking, in Malaysia, is in the market of
		monopolistic competition.
2	Author	Farhad khodadad Kashi and Jamal Zarein Beynabadi
	Title	The Degree of Competition in Iranian Banking Industry Panzar – Rosse
		Approach
	Period	2005 – 2010
	Results	H-Statistic of 0.7101, rejecting the Wald test for the monopoly market and
		perfect competition with a 1% significance, meaning that the Iranian banks
		compete in the monopolistic competition market.
3	Author	Rima Turk Ariss
	Title	Competitive Condition in Islamic and Conventional Banking: A Global
		Perspective
	Period	2000- 2006
	Results	H statistics showed that the banking in the Middle East operates on
		monopolistic market competition and the Lerner index value showed that
		Islamic banking is more competitive than the conventional.
4	Author	Maal Naylah
	Title	The Influence of Market Structure on the Indonesian Banking Performance
	Period	2004 – 2008
	Results	Testing of market structure (CR4): oligopolistic form of low moderate
		concentration (Type IV)
5	Author	Andi Fahmi Lubis
	Title	Indonesian Banking Market Power
	Period	1990 – 2004
	Results	Bresnahan-Lau model test showed that the level of competition on the credit
		market of Indonesian banking industry is still quite high based on the markup
		coefficient of 0.0223

316Bulletin of Monetary Economics and Banking, Volume 20, Number 3, January 2018

3.2. Variable Restrictions

The variables consist of three independent variables, including labor inputs, cost of funds, fixed capital cost, and two controlled variables i.e. primary ratio, the asset to loan ratio, while the dependent variable uses the income ratio and return on asset (ROA).

Table 3.

PR-H Statistic Variables

	Names of Data	Types of	Time Frame	Units of	Sources of
		Variables		Measurement	Data
	Total Revenue (TR)	Bound	Monthly	Ratio	BI and FSA
	Return on Asset (ROA)	Bound	Monthly	Ratio	BI and FSA
	Wages of labor (WL)	Free	Monthly	Ratio	BI and FSA
	Cost of funds (WF)	Free	Monthly	Ratio	BI and FSA
	Fixed capital costs (WK)	Free	Monthly	Ratio	BI and FSA
	Primary ratio (Y1)	Free	Monthly	Ratio	BI and FSA
	Asset to loan ratio (Y2)	Free	Monthly	Ratio	BI and FSA

Description of variable constraints:

1.Total revenue (TR) is total bank income derived from non-interest operational income for islamic banking which is proxy to total asset

2.Return on Assets (ROA) is the ability of a bank to make the profit through the use of its assets. The ROA is in proxy from the profit before tax to total assets

3.Wages of Labor (WL) is the operational cost of the bank in terms of labor wages. The wage of labor is the burden the bank must pay to its workers. It uses the ratio of wages to total assets

4.The cost of funds (WF) is the burden borne by banks for bonuses on third parties (customers)

5.The Fixed capital cost (WK) is an administrative and promotional burden to be borne by a bank that is proxied using owned assets.

6.The Primary ratio (Y1) is a ratio of capital health which measures the extent to which the decrease in total incoming assets can be covered by capital

7.The asset to loan ratio (Y2) is the ratio to measure the amount of financing disbursed by the total assets owned by islamic banks.

3.3. Models and Methods of Analysis

Data analysis is the process of simplifying the data into a form that is easier to read and interpret. The econometric model for panel data was used in the present research. The form of the equation was as follows:

Yit = β0 + β1X1it + β2X2it + β3X3it + βnX3it + εit

(3)

The stages of panel data test consist of Pooled Least Square modeling, Fixed Effect, and Random Effect.

Market Structure and Competition of Islamic Banking in Indonesia	317

3.4. PR H-Statistic Method

The PR H-statistic method is one form of market power measurement in the banking industry with the un-structural approach. The un-structural approach is more of determining the company’s behavioral aspects in influencing the market conditions (Widyastuti and Armanto, 2013). This method was first developed by Panzar and Roose in 1987 and used to measure the degree of market power in the banking industry. The measure of market power through the PR H-statistics is derived from the sum of the input of price elasticity coefficients i.e. labor costs (WL), cost of funds (WF) and fixed capital costs (WK), which respond to the total revenue. The representation of the sum of input variables can be used to determine the market structure in which the firm operates.

Some of the advantages of the PR H-statistic method are: (1) able to see the broader market structure, (2) can be estimated using econometric model with regression, (3) the variable used is quite simple, (4) using the individual data (cross-section) as a form of competition among banks.

The PR-H Statistics method is applied to companies with one type of product. Thus, banks are treated as producers with output in the form of loans. The assumptions of this method are long-run equilibrium and the maximization of profit earning (Panzar and Roose, 1987).

Before estimating the PR H-statistic, a long-run equilibrium test is intended, for, in case of balance, the research can proceed to market power, whereas if not, the research is stopped. However, Shaffer (1982) in Widyastuti and Armanto (2013), said that if the study shows a disequilibrium condition, it indicates that the banking industry is developing dynamically during research observation, thus, the research can proceed.

The equation of long-run equilibrium test of the PR-H statistic method is as follows:

LnROAit = β 0 + β1Ln(WL,it)+ β2Ln(WF,it) + β3	Ln(WK,it) + β4Ln( Y1,it)
+ β5Ln(Y2,it) + εit	(4)

The long-run equilibrium is interpreted as follows:

If PR - H statisticROA < 0 there is no equilibrium

If PR – H statisticROA = 0 there is equilibrium

PR – H statisticROA is the sum of the elasticities of β1, β2, and β3. Whereas, the equation to see the degree of market power is only by replacing the variable Return on Asset (ROA) with total bank income to total assets (TR).

So, the equation becomes:

Ln(TRit) = β 0 + β1Ln(WL,it)+ β 2Ln(WF,it) + β3		Ln(WK,it) + β4Ln( Y1,it)
+ β5Ln(Y2,it) + εit		(5)
Description:
lnTRit	= Non-interest income/total assets
lnROAit	= Profit before tax/total assets
lnWF,it	= Bonus expense (wadiah)/total assets
lnWL,it	= Labor load/ total assets

318Bulletin of Monetary Economics and Banking, Volume 20, Number 3, January 2018

lnWK,it	= Operational load/total assets
lnY1,it	= Total capital/total assets
lnY2,it	= Total debt/total assets
The interpretation of the market power is as follows.
PR-H statisticTR		= 0 monopolistic competition
PR-H statisticTR		= 1 perfect competition
PR-H statisticTR		0 < HTR < 1 = Monopolistic Competition

All of the variables in the PR H-statistic method use the addition of natural logarithm (ln) to show elasticity, since the market power values range between 0 and 1 so that the degree of the market indicator is immediately detected.

IV. RESULTS AND ANALYSIS

4.1. Panzar and Roose (PR H- Statistics) analysis

The panel data estimation is done on Eviews 7 with the timeframe of July 2010-September 2014, which is grouped every 3 months. Thus, there are 17 test result data of test phase, consisting of the best method selection (PLS, FEM, or REM), equilibrium test, and PR H-Statistic test. The results and discussion of each test are described as follows.

4.2. The Best Model of Islamic Bank Equilibrium Test

Table 4.

Selection of The Best Model of Islamic Bank Equilibrium Test

Periods	Test		Model Selection
		Prob	Significance	Best Model	Conclusion
			0.05
July –	Chow	0.0000	0.05	FEM	The best model is
September	Housman	0.0507	0.05	REM	REM, because 0.0507
2010					> 0.05
October –	Chow	0.0000	0.05	FEM	The best model is
December	Housman	0.0000	0.05	FEM	FEM, because 0.0000
2010					< 0.05
January –	Chow	0.0000	0.05	FEM	The best model is
March 2011	Housman	0.3875	0.05	REM	REM, because 0.3875
					> 0.05
April – June	Chow	0.0000	0.05	FEM	The best model is
2011	Housman	0.0320	0.05	FEM	FEM, because 0.0320
					< 0.05
July –	Chow	0.0000	0.05	FEM	The best model is
September	Housman	0.0479	0.05	FEM	FEM, because 0.0479
2011					< 0.05
October –	Chow	0.0000	0.05	FEM	The best model is
December	Housman	0.9381	0.05	REM	REM, because 0.9381
2011					> 0.05
January –	Chow	0.0001	0.05	FEM	The best model is
March 2012	Housman	0.9839	0.05	REM	REM, because 0.9839
					> 0.05

Market Structure and Competition of Islamic Banking in Indonesia	319

Table 4.

Selection of The Best Model of Islamic Bank Equilibrium Test (Continued)

Periods	Test		Model Selection
		Prob	Significance	Best Model	Conclusion
			0.05
April – June	Chow	0.0000	0.05	FEM	The best model
2012	Housman	0.8299	0.05	REM	is REM because
					0.82>0.05
July –	Chow	0.0000	0.05	FEM	The best model is
September	Housman	0.0000	0.05	FEM	FEM, because 0.0000
2012					< 0.05
October –	Chow	0.0000	0.05	FEM	The best model is
December	Housman	0.4376	0.05	REM	REM, because 0.4376
2012					> 0.05
January –	Chow	0.0000	0.05	FEM	The best model is
March 2013	Housman	0.0000	0.05	FEM	REM, because 0.0000
					< 0.05
April – June	Chow	0.0028	0.05	FEM	The best model is
2013	Housman	0.8007	0.05	REM	REM, because 0.8007
					< 0.05
July –	Chow	0.0000	0.05	FEM	The best model is
September	Housman	0.5859	0.05	REM	REM, because 0.5859
2013					> 0.05
October –	Chow	0.0000	0.05	FEM	The best model is
December	Housman	0.2848	0.05	REM	REM, because 0.2848
2013					> 0.05
January –	Chow	0.0000	0.05	FEM	The best model is
March 2014	Housman	0.6506	0.05	REM	REM, because 0.6506
					> 0.05
April – June	Chow	0.0000	0.05	FEM	The best model is
2014	Housman	0.7878	0.05	REM	REM, because 0.7878
					> 0.05
July –	Chow	0.0000	0.05	FEM	The best model is
September	Housman	0.8698	0.05	REM	REM, because 0.8698
2014					> 0.05

Source: Appendix 2.

The selection of the best model for islamic bank equilibrium test, as in Table 12, is Fixed effect model and Random Effect Model (REM). The FEM modeling took place in the period of October-December 2010, April-June 2011, July- September 2011, July-September 2012, and January-March 2013, while the other twelve periods used the REM test. However, this test often raises the problem of correlation between interference variables (autocorrelation), and to overcome it, the REM test method of Generalized Least Square (GLS) is used.

320Bulletin of Monetary Economics and Banking, Volume 20, Number 3, January 2018

4.3. Best Model of PR H-Statistics Test in Islamic Banking

Table 5.

Selection of The Best Models of PR H-Statistics Test in Islamic Banking

Periods	Test		Pemilihan Model
		Prob	Significance	Best model	Conclusion
			0,05
July –	Chow	0.0000	0.05	FEM	The best model is
September	Housman	0.3360	0.05	REM	REM, because 0.3360
2010					> 0.05
October –	Chow	0.0000	0.05	FEM	The best model is
December	Housman	0.7003	0.05	REM	REM, because 0.7003
2010					> 0.05
January –	Chow	0.0000	0.05	FEM	The best model is
March 2011	Housman	0.0201	0.05	FEM	FEM, because 0.0201
					< 0.05
April – June	Chow	0.0000	0.05	FEM	The best model is
2011	Housman	0.0044	0.05	FEM	FEM, because 0.0044
					< 0.05
July –	Chow	0.0000	0.05	FEM	The best model is
September	Housman	0.0048	0.05	FEM	FEM, because 0.0048
2011					< 0.05
October –	Chow	0.0005	0.05	FEM	The best model is
December					REM, because 0.5280
2011					> 0.05
January –	Housman	0.5280	0.05	REM	The best model is
March 2012	Chow	0.0001	0.05	FEM	REM, because 0.1144
					> 0.05
April – June	Housman	0.1144	0.05	REM	The best model is
2012	Chow	0.0001	0.05	FEM	FEM, because 0.0000
					< 0.05
July –	Chow	0.0000	0.05	FEM	The best model is
September	Housman	0.0000	0.05	FEM	FEM, because 0.0000
2012					< 0.05
October –	Chow	0.0002	0.05	FEM	The best model is
December	Housman	0.6590	0.05	REM	REM, because 0.6590
2012					> 0.05
January –	Chow	0.0000	0.05	FEM	The best model is
March 2013	Housman	0.0000	0.05	FEM	REM, because 0.0000
					< 0.05
April – June	Chow	0.0000	0.05	FEM	The best model is
2013	Housman	0.0015	0.05	FEM	FEM, because 0.0015
					< 0.05
July –	Chow	0.0000	0.05	FEM	The best model is
September	Housman	0.0000	0.05	FEM	REM, because 0.0000
2013					< 0.05
October –	Chow	0.4769	0.05	PLS	The best model is
December					PLS, because 0.4769
2013					> 0.05
January –	Chow	0.0000	0.05	FEM	The best model is
March 2014	Housman	0.0000	0.05	FEM	FEM, because 0.0000
					< 0.05

322Bulletin of Monetary Economics and Banking, Volume 20, Number 3, January 2018

Islamic banking appears to have a tendency to approach the equilibrium level indicated by the value close to 0. However, the movement is quite extreme in the period of July-September with (1.16 points), July-September 2012 (1.43 points), and July-September 2014 (1.64 points).

Islamic banking is in disequilibrium condition and tends to approach 0. This is similar to the research conducted by Widyastuti and Armanto (2013) and Sahut, et al. (2012).

This condition indicates that islamic banking is in a developing condition which means that islamic banking is not yet dominating the economy. This is in line with the Financial Services Authority survey in 2013 on financial literacy, which showed that the community’s understanding of banking was only 21.80%, which was lower compared to the Philippines (27%), Malaysia (66%), Thailand (73%) and Singapore with the highest level of understanding of the banking sector i.e. 98%.

The low level of banking literacy, in Indonesia, indicates that many people do not use banking as a financial intermediary institution. This has an impact on the low DPK owned by islamic banks and the hampering growth of the real sector.

The continuous development of islamic banking, in Indonesia, is caused by the visibility of its new market share of about 5% of the total banking assets in Indonesia. So, it takes hard work to add customers. One of the efforts made by the Financial Services Authority and Bank Indonesia is the promotion of Islamic finance, policy-making, and legislation. However, their efforts are certainly not enough to increase the market share of islamic banking. Therefore, there is a need for the role of the banking itself and the community. The efforts that need to be done by islamic banking is the increase in the number of branches and ATMs in the regions. In 2017, FSA recorded the number of operational headquarters (KPO) of islamic banks as follows: 152 offices, 136 sub-branches and 53 cash offices (still lacking).

Another effort that can be carried out is through the role of ulama in providing education on Islamic finance through Islamic study materials in mosques. With the collaboration of the government, banks, and community, it is expected that the future of islamic banking will be more advanced, growing and stable.

Market Structure and Competition of Islamic Banking in Indonesia	323

4.5. PR H-Statistics Test Results of Islamic Banking

Figure 4. PR H-Statistics Results Islamic Banking from July-September 2010 to July-September 2014

H-statistic parameter

1.2

1

0.8

0.6

0.4

0.2

0

Jul -

Okt -

Jan -

Apr -

Jul -

Okt -

Jan - Apr - Jul -

Okt -

Jan - Apr - Jul -

Okt - Jan - Apr -

Jul -

Sep

Des

Mar

Jun

Sep

Des

Mar Jun

Sep

Des

Mar Jun

Sep

Des Mar Jun

Sep

2010

2011

2012

2013

2014

Figure 4 shows that the pattern of the strength of islamic banking always falls within a range of more than 0 to more than 1, indicating that it is in the perfect market and monopolistic competitions.

Perfect competition conditions occur twice i.e. in the period of July-September 2013 (1.05) and July-September 2014 (1.06). Meanwhile, the other periods operate in the monopolistic market. This means that the diversity of products and advertising in islamic banking during this period is quite dominant in order to increase the number of customers. The strong competition amongst islamic banks is caused by the dominance of conventional banking that opened the branch of sharia after the endurance of Bank Muamalat Indonesia during the economic crisis of 1998 (Faiz, 2010).

Another factor is the regulation of the Ministry of Religious Affairs of the Republic of Indonesia related to the pilgrimage funds that must be deposited into Islamic banking, which reflect the government’s trust in Islamic banks in the management of pilgrim funds. In addition, islamic state securities products (retail Sukuk) issued by the government was recently quite enthused by the community. However, the products of Islamic banks themselves, in practice, are still dominated by the sale and purchase agreement (Murabaha) (Natadipurba, 2015).

Generally present in a monopolistic market and perfect competition means that although the product differentiation is quite good, there is still a perfect competition condition. This shows that islamic banking has not been consistent in making innovative financial products. Therefore, in the future, islamic banking needs to make superior financial products that will become his trademark. Thus, each islamic bank has its own market share and can compete from the non-price side.

324Bulletin of Monetary Economics and Banking, Volume 20, Number 3, January 2018

V.CONCLUSIONS 5.1. Conclusions

Based on the description and discussion, it can be concluded that:

1.The equilibrium test results show the value close to 0, meaning that the condition of islamic banking, in the long run, does not show equilibrium. This means that islamic banking is in a developing condition. Therefore, the collaboration between the government, especially FSA and BI, banking and community is needed to increase the growth of islamic banking in the future.

2.The H-statistic test results show that islamic banking operates on the monopolistic market and perfect competition. This means that islamic banking has not been consistent in developing its products. So, it requires the development of excellent products with specific characteristics and market share.

5.2. Suggestions

The suggestions submitted by the author for further research improvements are as follows:

1.In order for the results to be more comprehensive, there is a need for additional variables and period of study.

2.The high level of competition in islamic banking demands the government, in this case, the Financial Services Authority and related parties, to improve the level of interbank competition.

REFERENCES

Ariss, Rima Turk. (2010). Competitive Condition in Islamic and Conventional Banking : A Global Perspective. Journal Review of Financial Economics.

Athoillah, Moh. (2010). Struktur Pasar Industri Perbankan Indonesia: Rosse – Panzar Test. Journal of Indonesian Applied Economics.

Bain, J.S. (1956). Barriers to New Competition: Their Character and Consequences in Manufacturing Industries. Harvard University Press.

Bikker, B., Spierdijk, L dan Finnie, P., (2006). Misspesification of the Panzar Rosse Model: Assesing Competition in the Banking Industry. Working Paper no 114/2006. de Nederlansche Bank.

Church, Jeffrey. R., and Ware, Roger. (2000). Industrial Organization: A Strategic Approach. McGraw-Hill Companies, Inch. USA.

Faiz, Ihda. (2010). Ketahanan Kredit Perbankan Syariah Terhadap Krisis Keuangan Global. Jurnal Ekonomi Islam La_Riba. Vol. IV no.2.

Gasaymeh, A, Karim Abdul Z, dkk. (2014). Competition and Market Structure of Banking Sector: A Panel Study of Jordan and Gulf Council Cooperation Countries (GCC). Jurnal Ekonomi Malaysia. 48: 23- 24.

Goddard, J., & Wilson, J. (2008). Measuring Competition in Banking: A Disequilibrium Approach. EIEF Working Paper Series 080.

Kashi dan Jamal. (2013). The Degree of Competition in Iranian Banking Industry Panzar – Rosse Approach.African Journal of Management. Vol.7(43).pp.4380 – 4385.

Market Structure and Competition of Islamic Banking in Indonesia	325

Otoritas Jasa Keuangan. Laporan Keuangan Publikasi Bank. from http://www.ojk. go.id.

Lubis, Fahmi Andi. (2012). Market Power Perbankan Indonesia. Buletin Ekonomi Moneter dan Perbankan Bank Indonesia.

Molyneux, P. dan Forbes, W. (1995). Market Structure and Performance in Eropean Banking. Applied Economics. 27: 155 – 159.

Natadipurba, Chandra. (2015). Ekonomi Islam 101. Bandung : Mobidelta Indonesia. Nathan, A,. & Neave, E.H. (1989). Competition and Contestability in Canada’s Financial System: Empirical Results. Canadian Journal of Economics. 22(3): 576-

594.

Naylah, Maal. (2010). Pengaruh Struktur Pasar Terhadap Kinerja Perbankan Indonesia. (Tesis). Universitas Diponegoro. Semarang.

Panzar, J. and Roose, J. (1987). Testing for Monopoly Equilibrium. Journal of Industrial Economics.

Pindyck and Rubinfeld. (2010). Mikroekonomi. Edisi Keenam. Jilid 2. Jakarta: Indeks. Pindyck and Rubinfeld. (2012). Mikroekonomi. Edisi Kedelapan. Jilid 1. Jakarta:

Erlangga.

Perrakis, S. (1991). Assessing Competition in Canada’s Financial System: A Note. Canadian Journal of Economics. 24(3): 727-732.

Sahut Michel, Mili Mehdi, Krir Ben. (2012). Factors of Competitiveness of Islamic Banks in the New Financial Order. 8th International Conference on Islamic Economics and Finance.

Shaffer, S. (1982). A Non – Structural Test for Competition in Financial Markets. Bank Structure and Competition, Conference Proccedings. 225 – 224.

Teguh, Muhammad.(2013). Ekonomi Industri. Jakarta : RajaGrafindo Persada. Widyastuti, Sri. R dan Armanto, Boedi. (2013). Kompetisi Industri Perbankan

Indonesia. Buletin Ekonomi Moneter dan Perbankan Bank Indonesia. Zulkhibri.et.al. (2007). Market Structure and Competition in Emerging Market:

Evidence from Malaysian Islamic Banking Industry.MPRA Paper No.12126