Journal of Economics and Business

As intermediary institutions, Islamic banks gather funds from society and then transmit in the form of financing. In practice, the distribution of financing that is the main feature that runs Islamic banks is not as easy in practice as it is in theory since it involves a large number of troubled financing constraints. The reason for this is the lack of procedures to assess banks and supervise customers. This research thus aims to analyze the effect of third-party funds (DPK), Return On Assets (ROA), Non-Performing Financing (NPF) and Financing the Deposit Ratio (FDR) on the total financing provided to society by the Islamic banks in Indonesia. The data used was monthly data issued by Otoritas Jasa Keuangan (OJK) in Sharia Banking Statistics during the period of January 2009-2017. This research uses the Johansen cointegration test to investigate long-term relationships and use the model of error correction (Error Correction Model) to see the short-term relationship. The results show that third-party funds and ROA influence financing in the short term and long term. Other variable is the FDR, which had the influence of short-term financing. Non Performing Financing does not affect the financing either in a short-term or long-term period.


Background
Islamic Bank (in Indonesia known as Syariah or sharia) is a financial institution that puts an effort in providing financing and other services in traffic economy, national or international operations are adjusted based on the principle of Islamic jurisprudence.Islamic banking started in 1992 in Indonesia with the founding of Bank Muamalat.Islamic banks, as the main motor of financial institutions, have become a locomotive for the development of theory and practice of Islamic economy (Karim, 2004).At present, the development of Islamic banking in Indonesia is experiencing rapid growth.The emergence of Islamic banking in Indonesia began after the amendment of the banking law that was marked by the issuance of Law No. 10/1998.This is one of the contents of the provisions of Bank Indonesia concerning the licensing for the opening of the new Islamic bank as well as the establishment of Syariah Business Units (UUS).Up to the year 2017, there were 13 public sharia banks (BUS), 21 UUS and 99 sharia people's financing banks (BPRS).
By August 2017, the projected market share of Islamic banking in Indonesia is estimated to be 5.44% (Galvan, 2017).With this development, of course, the Islamic banking industry sector in Indonesia has been increasingly promising.The banking sector, as the intermediary, occupies a very important position in bridging the needs of working capital investments in the real sector with that of the owner of the funds.Similar to intermediary institutions, Islamic banks gather funds from society and then distribute them in the form of financing.The financing practices carried out by Islamic financial institutions are based on profit sharing.This profit sharing financing is divided into two, namely mudharabah financing and musyarakah financing.Other types of financing also exist in Islamic banks in the form of buying and selling contracts, namely financing murabaha, salam, and istishna.Financing is one of the main tasks of the bank, namely the provision of funds to meet those who need funding.Basically, the products offered by Islamic banks are product distribution (channeling) funds or financing and product gathering funds.In practice, the distribution of funds run by Islamic banks is not easy, because in reality there are quite a lot of financing problems.The non-performing loan ratio (NPF) is a ratio that measures the level of problematic financing.The higher the ratio of troubled financing, the lower is the total funding channeled.This will affect the bank's profitability.A healthy bank must have profitability with a minimum of 1% Return On Assets (ROA) according to Bank Indonesia.
In Islamic banking, funds disbursed are not a loan or credit but rather it is the financing.This function assesses how liquid the bank is in orde channel its funds to customers.The bank's liquidity is seen through the Finance to Deposit Ratio (FDR), which shows that the bank has run its functions as an intermediary financial institution, to gather and distribute funds.Most bank funds come from customer deposits in the form of savings, current accounts, and time deposits.These customer deposits are referred to as third-party funds (DPK).The high nominal value of third-party funds indicates that public confidence is increasing in saving funds in Islamic banks, while at the same time demonstrating that the sharia banking market in Indonesia is still large (Hamidi, 2003).
Several previous studies regarding the financing of Islamic banks in Indonesia included Pramita, Wuri Arianti Novi andHarjum Muharam (2011), Sri, Ratna Anggraini, Etty Gurendrawati, andNuramalia Hasanah (2013).They all used independent variables of third-party funds (DPK), NPF, Capital Adequacy Ratio (CAR) and ROA, but show different results.Pramita et al. (2011) observed that only TPF influences financing, while in Sri, et al. (2013) found that only NPF variables influence financing.Nugroho (2014) noted that DPK in the short term affects financing, ROA and NPF in the long run affect financing, FDR in the short and long term affects financing.
Table 1.shows the development of the total funding given to customers, deposits, ROA, NPF and FDR.The aim of the study Based on the information given above, the purpose of this study is to analyze the effect of short-term and longterm deposits, ROA, NPF and FDR on financing in Islamic banking in Indonesia in the period 2009-2017.Islamic banking intended in this study includes Islamic banks and sharia business units in Indonesia.

Literature Review and Similar Research Studies
Literature Review According to the provisions set forth in the regulation of Bank Indonesia No. 2/8/PBI/2000, article 1, Islamic banks are commercial banks, Act No. 7 (1992) regarding banking has been amended by Act No. 10 (1998).This state that those who conduct business activities based on the principles of the Islamic Sharia, including sharia business unit and branch offices of foreign banks, should conduct business activities based on the principles of Islamic jurisprudence.
According to Karim (2010), in general, Islamic finance products are divided into four categories according to their intended use, namely: (1) financing with the principle of buying and selling; (2) financing under the lease principle; (3) financing under the profit sharing principle and (4) financing with supplementary contracts.According to Sharia Banking Statistics (Bank Indonesia, 2017), Islamic bank financing is the main choice of placement of Islamic banking funds compared to other placements such as placement in securities.According to Bank Indonesia Regulation No. 10/19 / PBI / 2008, third-party funds, hereinafter referred to as DPK are bank obligations to residents in rupiah and foreign exchange.Third party funds in Islamic banks are sources of funds originating from the community collected through wadiah demand deposits, mudharabah savings, and mudharabah deposits.Deposits are then channeled to various types of financing.

Similar Research Studies
Soedarto ( 2004) found that interest rates, the level of capital adequacy of rural banks, the amount of public deposits, and the amount of non-current loans affect BPR lending.Donna, Duddy Roesmara and Nurul Chotimah (2008) found that profit sharing, DPK, capital per asset and income affected financing, while NPF did not affect the financing of Islamic banks in Indonesia.Hasanudin, Mohammad and Prihatiningsih (2010) observed that deposits affect the lending of BPR.Firaldi, Mufqi (2013) noted that TPF has a short-term effect on total financing, NPF has a short-term effect on total financing and the inflation rate does not have an influence on the total financing provided by BPRS in Indonesia.Putro (2013) in his study indicated that deposits, NPF, labor, and GDP of MSMEs affected total financing.Nugroho (2013) found that TPF affects financing in the short term, FDR influences financing in the short and long term, while ROA and NPF affect financing in the long run.Husnaeni (2016) shows that DPK, CAR, FDR, and ROA affect Murabaha financing.Based on the theoretical description and previous research studies, it can be the hypothesized that: H1: Third party funds, ROA, NPF, and FDR affect the financing of Islamic banks in Indonesia

Research Method
In this study, we used Islamic bank financing (PBS) as the dependent variable, and Return on Assets (ROA), Non-Performing Financing (NPF) and Financing to Deposit Ratio (FDR) as independent variables.The data in this study are from 2009 to 2017 taken monthly.This study uses the Error Correction Model (ECM) approach to see short-term relationships and cointegration tests to see indications of long-term relationships.ECM testing is done if there are indications of a long-term relationship using cointegration.Variables are said to be cointegrated, when stationary in the same order.To test the stationarity of data, this study used the Phillips-Platform (PP) test.In the Phillips-Peron test, it is necessary to determine the number of truncation lags for Newey-West correction, by using the formula N1 / 3 = 60 1/3 = 3.91 which is then rounded off to the closest unit value, namely 4 (Yahya Hamja, 2008).
Normality test is a test that is conducted to see whether the residual value is normally distributed or not (Widarjono, 2007).A good regression model is to have a residual value that is normally distributed.The normality test uses the Jarque-Bera coefficient and its probability value because both numbers support each other (Winarno, 2011).Linearity test is a test developed by J.B Ramsey known as Ramsey RESET Test.This test is to find out whether an explanatory variable is right or not included in an estimation model (Gujarati, 2003).The test developed by J. B Ramsey is used to test whether the function forms of a model are linear or not.
Data from a random process is said to be stationary if it satisfies the criteria, that is if the average and constant variants throughout the ovarian time between the two-time series data depend only on the lag between two specific time periods (Widarjono, 2007).In conducting stationarity tests, unit root tests and integration degree tests are carried out.The procedure for determining, whether the data is stationary or not, is to compare the Philips-Perron statistical value with its critical value, namely the MacKinnon statistical distribution.If the absolute value of the PP statistic is greater than the critical value, the observed data is displayed as stationary.On the other hand, if the absolute statistic PP value is smaller than the critical value, the data is not stationary.In the root of the unit PP, if it concludes that the data is not stationary, a process of data differentiation is needed.Stationary test data through this differentiation process is called the integration degree test.If the absolute value of the statistical PP is greater than the critical value at the first level of differentiation, then the data is said to be stationary at first degree.If the value is smaller, then the degree of integration test needs to be continued at a higher differentiation so that the stationary data is obtained.The next stage, after the ECM model is formed, is testing the ECT.Error Correction Term is part of ECM.This ECT value is obtained from the sum of the independent variables of the previous month reduced by the dependent variable in the previous month, and the resulting value is the adjustment value of the imbalance of the dependent and independent variables in the short and long term.The ECT Model in this study: ECT = DPK t (-1) + ROA t (-1) + NPF t (-1) + FDR t (-1) -PBS t (-1) The result of the ECT probability value will determine whether the model can be analyzed in both the short and long term.If the ECT variable is positive and significant at α = 5%, then the model specification is valid and can explain the dependent variable.

Results and Discussion
The Growth of PBS, Third Party Funds, ROA NPF and FDR This section begins with a discussion of the development of variables during the study period, namely the growth of the total Islamic bank financing, the growth of Third Party Funds (DPK), the growth of ROA, FDR, and NPF.Financing is funding provided by a party to another party to support the investment that has been planned, both individually and by an institution.The performance of Islamic banks in financing during the study period has increased.The grwoth of financing funds can be seen in Figure 1.Return on Assets is used to measure the ability of a bank's management to obtain overall profits.The greater is the ROA of a bank, greater is the level of profit achieved by the bank, and in turn, the better is the bank's position in terms of asset use.During the study period, ROA tended to fluctuate.The lowest number occurred in May 2016, which was 0.16%, whereas the highest number occurred in March 2009, which was 2.44% (see Fig. 3).Profitbased Sharia financing, such as mudharabah and musyarakah, cannot be separated from the conditions of the real sector.In 2016, when Indonesia's economic growth was weakening at around 5 percent, in the world industry in general, was sluggish.This has an impact on profits earned by Islamic Banks.The profits in May 2016 were only 0.16%.Based on Fig. 4. it can be seen that the NPF level was fluctuating.In May of 2016, the NPF was 6.17%, which was the highest throughout the study period (2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017).activity of a financial institution, there will be problems with financing.NPF in a bank is not something that should not happen because basically, every economic activity has obstacles, which generally lead to the return of troubled loans.This is due to the varied financial conditions of bank customers.However, the condition of the NPF will force banks to form a number of reserves to maintain bank liquidity and solvency in order to protect customers.The greater the NPF, the greater is the opportunity cost that must be borne by the bank.Thus, NPF should be sought as low as possible (Hasanudin and Prihatiningsih, 2010).
Figure 5.The Growth of Financing to Deposit Ratio in the period 2009-2017 Based on Fig. 5. it can be seen that FDR also experienced fluctuating changes.In December 2017, the FDR rate was the lowest at 79.05%, while the highest FDR rate was observed in July 2013.In a conventional bank, a term more commonly known is the Loan to Deposit Ratio (LDR) stating the level of a bank's ability to pay back withdrawals made by depositors by controlling credit given as source liquidity.A higher ratio gives an indication that the chances for the concerned bank to be liquidated are low (Dendawijaya, 2009).The increase in FDR can mean that the distribution of funds to financing is getting bigger so that profits will increase.If the FDR value of a bank is above or below 85% -110%, it can be said that the bank does not function as a financial intermediary or can also be called disintermediary.
The next discussion is testing the data.This includes the classic assumption test, which consists of a normality test, linearity test, autocorrelation test, and heteroscedasticity.The next process is the stationarity test (unit root test) and cointegration test.The normality test used in this study uses the Jarque-Berra (J-B) method.Results show that the J-B value is 5.9388 with a probability of 0.07089, thus it can be concluded that the data is normally distributed.Next is the linearity test to determine whether or not an explanatory variable is suitable to be included in an estimation model (Gujarati, 2003).The probability value obtained is 0.7891 greater than α = 5%, which means there are no linearity problems.Autocorrelation and heteroscedasticity tests were also carried out using LM-test of 0.5976 and the value of heteroscedasticity test with a White test of 0.2197, both of which were greater than α = 5%.This indicates that the data is free from the problem of autocorrelation and heteroscedasticity.
After The regression results show that the coefficients of the error correction variable (ECT) are positive and statistically significant, meaning that the ECM specification model of Islamic Bank financing used in this study is valid.To examine see the long-term effects, a calculation is made by summing up the long-term coefficients of LDPK (-1), ROA (-1), NPF (-1) and FDR (-1), and the ECT coefficient, then it is divided by the ECT coefficient.The longterm coefficient formula is as follows: LDPK (-1) = (C5 + C9)/C9 NPF(-1) = (C7+C9)/C9 ROA(-1) = (C6 + C9)/C9 FDR(-1) = (C8+C9)/C9 The above results show that deposits and ROA, in the short and long term affect the financing of Islamic banks in Indonesia.This is evident from equation ( 1), which shows that the probability value of the DPK variable is 0,000 smaller than the significance level used which is 5%.This result is in line with the study of Pratami and Harjum Muharam (2011) and Firaldi (2013) that TPF affects financing.FDR affects short-term financing, but in the long run, it does not.This result is in line with Nugroho's (2013) research that FDR influences financing in the short term.Variables that do not affect financing, both in the short and long term are the NPF variables.This result is different from Sri, et al. 's (2013) study where they found that NPF influences financing.

Conclusion
The regression results of the ECM model show that all the independent variables in this study are DPK, FDR, and short-term ROA that affect financing.Banking requires a source of funds that can be channeled to carry out business financing.Short-term financing is generally sourced from third-party funds, namely funds from customers.Any increase in third-party funds in Islamic banks will increase the amount of financing provided.NPF does not affect financing both short and long term, meaning that the problematic credit conditions do not affect banks in channeling financing as long as there is an increase in the number of third-party funds.An increase in the amount of third-party funds reflects public confidence, which tends to increase to saving money in Islamic banks.

Figure 1 .
Figure 1.The Growth of Sharia Financing

Figure 2 .
Figure 2. The Growth of Third Party Fund in the period 2009-2017

Table 1 .
The Growth of Islamic Bank Financing, Third Party Fund, ROA, NPF and FDR In this study, the ECM model is used primarily to overcome problems in time series data that are not stationary.The ECM model is used after the data normality test, linearity, stationarity, cointegration integration degree and classical assumption test, and is free from all problems from the test so that the ECM model used is feasible to use and analyze.The ECM models used in this study are as follows:PBS t = β0 + β1DPK t + β2 ROA t + β3 NPF t + β4 FDR t + e Model in this research: D(PBS) t = β0 + β1 D(DPK) t + β2 D(ROA) t + β3 D(NPF) t + β4 D(FDR) t + β5 DPK t-1 + β6 ROA t-1 + β7 NPF t-1 + β8 FDR t-1 + β9 ECT fulfilling normality and linearity, the next test is the unit root test to see whether the data is stationary or not.Data is said to be stationary if the absolute value of Philips-Platform test (PPtest) is greater than the critical value of 5%, whereas if the absolute value of PPtest is smaller than the critical value, then the data is not stationary.The results of unit root tests can be seen in table 2. The following:The results of unit root tests show that the NPF variable does not have a root unit problem because the absolute value of PPtest is greater than the critical value of 5%.Other variables that experience unit root problem need to be continued with the first integration test (first level difference).The following are the results of first level integration degree testing:Based on table 3, it can be seen that all variables in this study have been stationary at the first integration (first difference).The next step is the cointegration test.If the variables in this study are cointegrated, then there will be a stable relationship in the long run.Phillips-Perron method is used to find out whether there is cointegration or not, while the Error Correction Model (ECM) is used to derive the long-term equation.The following are the results of the Phillips-Perron cointegration test: Table.4 it can be seen that the t-statistic value is greater than the critical value of Phillips-Perron, meaning the residual of the equation has been stationary at zero or I(0) integration degrees.It can indicate longterm or cointegrated relationships.The indication of a long-term balance relationship cannot be used as evidence to say that there is a relationship in the short term.To determine which variable causes changes in other variables and to determine the effect in the short term, ECM calculation is performed.Error correction model (ECM) is used to see if there is or whether the relationships exist between variables in the short term.This model is one approach to analyze time series model in order to determine the consistency of short-term relationships with longterm relationships from the variables tested.The following is the model equation of ECM and the p-value: