ANALYSIS OF BANK-SPESIFIC FACTORS TO DETERMINE THE PROFITABILITY OF ISLAMIC BANKS IN INDONESIA : A PANEL REGRESSION APPROACH

Analysis Of Bank-Spesific Factors To Determine The Profitability Of Islamic Banks In Indonesia : A Panel Regression Approach. Profitability is an important indicator of a bank’s performance. In 2014, profits of Islamics bank in Indonesia decreased by 19.7 percent. This paper aims to analyze the impact of bank-specific factors to the Islamic banking system in Indonesia has shown better development Islamic bank’s profitability. This study observed 11 Islamic banks in the Indonesia banking system in the period between 2010 2014. The quarterly data are taken from the Indonesian Banking Directory, published by the Financial Service Authority (OJK). Using panel data regression, the Fixed Effect Model with cross-sectional correlation (SUR) has selected as the best model. According to the obtained results, among internal factors of bank profitability, the most important one is the operating efficiency ratio. Furthermore, profitability is influenced negatively by liquidity risk, solvency risk, credit risk, and bank size.


Indonesia's first Islamic bank namely Bank
Muamalat Indonesia (BMI) in 1992 (Sari, et al, 2016), the However, ROA is a major indicator that more appropriate to describe profitability, because it is focusing on the bank capability to get earning in operational activities with utilizing bank's assets (Paulin and Sudarso, 2015). It shows how bank can convert its assets to maximize its profit.
In this paper we try to identify the     (Abduh and Yameen, 2013).
In panel data regression techniques, there are three models, such as pooled effect models, fixed effect models, and random effect models (Greene, 2007:183 In addition, to analyze the determinants of profitability, this paper also analyzes the dominant variable in influence profitability. Standardized variable can be used to obtain this objective, since it does not matter in what unit regressand and regressors are measured (Gujarati and Porter, 2008). The standardized regression coefficients, also sometimes called beta coefficients, have been proposed to facilitate comparisons between regression coefficients (Neter, 1983).
This research adopt the following regression equation to study the determinants of profitability of Islamic banks in Indonesia,

RESULT AND DISCUSSION
The  This study found that liquidity, which used the total financing to deposits ratio as a proxy, had a significant relationship with profitability. Interestingly, negative relationship was found between FDR and ROA, which suggested that an increase in deposits do not generate more profit to the bank. This could be due to increased the bad debt, so an increase in financing can not followed by an increase in profits. Furthermore, if banks provides many financing funds, actually it is really dangerous to the bank's suistainability and increase liquidity risk (Paulin and Sudarso, 2015). This findings is similar with study reported by Alhamdita dan This study shows a negative relationship between solvency risk, which proxied by capital adequacy ratio (CAR), with profitability. A negative relationship between CAR and ROA suggest that an increasing among of capital does not lead to an increase in profit of the banks. Thus, although higher level of the bank capital provides safety, over caution in banking bussiness reduce the profitability (Curak, et al, 2012). In addition, there is also an increase in risky capital such as transfers of funds Hajj from conventional banks, so the cost to bear even gretaer and deacrease the profitability. This results also supported by studies that have been conducted by Curak, et al (2012), Alhamdita andHeykal (2013), and Febrianto and Maskur (2015) which says that CAR has a negative effect to ROA.
Higher OER means smaller ROA. This can be due to the rising of bank operationt costs such as renew facilities and services provided (Paulin and Sudarso, 2015). When the increase in revenues is not as much as its operational cost, the profitability would be declining. The high expense can be due to the increasing in provision cost that must be provided by the banks. This result also in line with researches by Curak, et al (2012), Petria, et al (2013), and Rahaman (2015) which found that the effect of OER is significantly negative to profitability.
Credit risk (NPF) also has a negative and statisticaly signficant impact on bank profitability. Lack of monitoring programs could also lead to the high of non-performing finance. Based on the result, it has been known tha higher non-performance finance would make profitability goes down. It happens due to the lack of monitoring programs that make the returns are not collected on time. So, it can hamper the income of operational activities.
Bank size was found to be a significant but has inverse relationship with profitability. This results gives support to the recent papers from Almazari (2014), Rahaman (2015), and Havidz and Chandra (2015), who stated that there is a negative and significant effect between bank size and profitability. The reason is because the bank could not maintain the assets as they growing into bigger bank. The recent papers also mention that the diseconomies of scale occur when the level of size increase. Growing banks may face diminishing marginal returns, so the average profit could decline with size (Almazari, 2014).
The result in table 3 shows that the operating expense ratio (OER) has the greatest absolute standardized regression coefficients among other variabels. Its exhibits the very strong explanatory power of operating expense towards the level of profitability. It can be said that this variable has the most important effect on bank profitability among other variables.
Thus, the banks should focus more on managing the expenses in order to improve the efficiency and increase the bank profitability (Curak, et al, 2012). It indicates that if the (standardized) OER increase by one standar deviation, on average, the (standardized) ROA decreases by about 0,72761 standar deviations. Table 3.

Standardized Coefficient
Meanwhile, the lowest explanatory power of coefficient standardized in explaining ROA is FDR, which the result is -0,08782. It shows that the increasing of (standardized) FDR by ine standar deviation, on average, then only will decrease (standardized) ROA for about 0,08782.
This finding also in line with the study by Havidz and Chandra (2015), which found that the weakest explanatory power of coefficient is FDR in explaining ROA.

CONCLUSION AND SUGGESTION
This paper analyzed determinants profitability of 11 Islamic banks in Indonesia in period between the fourth quarter of year 2010 and the fourth quarter of year 2014 using panel regression analysis. According to our results, the profitability of the banks is determinate by bank-spesific factors, namely liquidity risk, solvency risk, operating expense ratio, nonperforming finance, and bank size. Since the bank-spesific factors are the results of bank policies and managements, thus the banks have abilility to influence them. This study shows those independent variables have negative and significant effect to profitability, which is the return on assets (ROA). In addition, the most important bank-specific determinant of profitability is operating expense ratio (OER). Therefore, banks should undertake the activities, whether it is necessary or not, to reduce management cost in order to increase the cost efficiency.
For the next study, it recommends to add more observation unit, use more set of explanatory variables such as externalmacroeconomic variable, and analyze using longer periods of time.