Determinants of Profitability in Indonesian Islamic Banking: Case Study in the COVID-19 Period

With a high level of COVID-19 virus spread throughout the world, Indonesia is one of the countries in Southeast Asia affected by the largest transmission chain. This affects various layers of the industry in this country, one of which is financial institutions and the banking system. This paper tries to look at the performance of Islamic banking in the face of the COVID-19 pandemic. Using the linear regression method, the authors use ROA as the dependent variable. Whereas CAR, NPF, FDR, and BOPO as independent variables. The results show that CAR and BOPO have significant results while the rest do not show satisfactory results. It can be concluded that during the pandemic, Islamic banking experienced a pretty good and convincing performance.


Introduction
In early 2020, the Corona Virus  shook the world with its impacts, both health and economically. This virus first appeared in Wuhan City, Hubei Province, and then spread ferociously to various parts of the world. Dong et al. (2020) stated that COVID-19 spreads more than 7,000 cases outside China and 3,000 people have died. Officially, Word Health Organization (WHO) later established COVID-19 as a global pandemic (WHO, 2020).
With the spread of the COVID-19 virus throughout the world, Indonesia is one of the countries in Southeast Asia that is affected by the biggest transmission chain. Djalante et al. (2020) estimated that Indonesia will be greatly affected for a long time. In addition to the health sector, the economic sector was highly affected by the outbreak of COVID-19; as a result, financial institutions and the banking system became very fragile.
Bank Indonesia as the central bank has responded by issuing a mixture of stimulus packages to support the risk mitigation efforts of the spread of Covid-19 and maintain stability in conventional or sharia financial systems (BI, 2020). The policy in question is the easing of the financing or credit structure for directly affected customers such as MSMEs and liquidity policies. The findings of previous researches conducted by Wahyudi et al. (2017) and Wahyudi  The main purpose of this research is to study the conditions experienced by Islamic banking during the Covid-19 pandemic. The Indonesian banking system is quite unique by adopting a dual banking system under the supervision of the Financial Services Authority (OJK) and has different characteristics. Banks in Indonesia have a dual system, Islamic banks and conventional banks. Furthermore, these findings are for policy recommendations that can be taken by stakeholders in banking. Haryati et al. (2019) examined the impact of bank age and bank performance towards the profitability of Islamic banks with intervening variable namely earnings distribution. The results showed that bank age had no significant effect on profitability, but NPF had a significant effect on revenue sharing. Furthermore, there was no significant effect of FDR and CI (Cost to Income) to revenue sharing. Finally, this study also revealed that profit sharing affected bank profitability significantly.

Literature Review
Ardana (2018) examined the influence of external factors (inflation and BI rate) and internal factors (Capital Adequacy Ratio, Return on Equity, Financing to Deposits Ratio, and Non-Performing Financing) on the profitability of Islamic banks in Indonesia. The results showed that the external and internal factors had a significant effect on the ROA simultaneously. Individually, the CAR, NPF, and inflation variables did not significantly affect ROA, while the FDR and ROE variables significantly effect on ROA both in short and long term. BI rate has no significant effect on ROA in the short term, but effects on the ROA variable in the long run significantly.
Another paper by Rahmah & Kusbandiyah (2018) examined the effect of performance on Islamic banks' profitability. Multiple regression analysis used in this research. The results show that FDR and operational costs to operating income have significant effect on profitability while NPF and CAR have no effect on the profitability of Islamic Banks.

Research Methods
This type of research is quantitative descriptive to examine the effect of independent variables on the dependent variable. Samples were selected using the type of purposive sampling in accordance with the objectives of this study, namely 11 Sharia Commercial Banks that meet the sample criteria which include BNI Syariah, BSM, BRI Syariah, BCA Syariah, Syariah Net Bank, Panin Dubai Syariah Bank, Bukopin Syariah, Mega Syariah Bank, Bank Victoria Syariah, BTPN Syariah and Bank Aceh Syariah. Data obtained from the publication of the first quarterly report of each Islamic bank in 2020.
Independent variables include CAR, FDR, NPF and operational costs to operating income. CAR is used to measure the strength of bank capital in resisting external shocks (Hassan and Bashir 2003). According to Berger (1995) the higher the ratio, the more stable and efficient the bank is. While the relationship between this variable and profitability can vary throughout all stages of the business cycle. NPF is used to measure the quality of productive assets calculated from nonperforming financing to total financing. The higher the ratio shows the low quality of productive assets and vice versa (Wahyudi, 2015). FDR is used to measure total financing against third party funds. The higher FDR shows good Islamic bank intermediation function (Wahyudi, 2015). NPF aims to measure the level of quality of productive assets. The higher this ratio shows the quality of Islamic bank financing getting worse and vice versa (Wahyudi, 2015). Operational costs to operating income measures the efficiency of operational costs compared to Islamic bank revenue (Wahyudi, 2015). All of these performance indicators have been used extensively by researchers to see the performance of Islamic banks.
The dependent variable in this study is profitability using ROA. According to Kumbirai & Webb (2010), ROA shows how much net profit is generated per rupiah of assets. The higher ROA, the more profit the bank. ROA was chosen because the bank's financial performance measurement tools have been widely used (Al-Tamimi, 2010).
Multiple linear regressions is used to answer the research problem formulation with test steps including classic assumption tests (normality test, multicollinearity test, autocorrelation test, and heteroscedasticity test). Next, the coefficient of determination test (R 2 ) and hypothesis testing are performed namely the F test and t test by using SPSS 22 software.

Results
The test results of multiple linear regression models from this study have a classic assumption that has accuracy in estimation, unbiased and consistent. The results of normality tests using Kolmogorov-Smirnov shows that the data is normally distributed (the sig value is 0.20 or greater than 0.05). The statistical result also shows that there is no autocorrelation symptom in this research variable with a distribution of 1266 values with Durbin-Watson method.
The presence of multicollinearity symptoms occurs in CAR variable that indicated by tolerance and VIF values of 0.074 and 13.439, respectively. The presence of multicollinearity symptoms in the test results is considered to be influenced by a small number of observations. Other performance variables do not occur with multicollinearity symptoms in accordance with assumptions with tolerance values> 0.100 and VIF values <10.00. The Glejser test result shows there is no symptom of heteroscedasticity. Thus, the classical assumption requirements have been met because the Sig. value is more than 0.05.
After conducting classical assumption test, then regression analysis result can be done to see the influence of independent variables to dependent variable. The regression test result can be seen from Table 1 and Table 2:  Source: Primary data, 2020 (Processed) Table 2 shows the results of the partial independent influence test. The partial test results are two performance ratios that have a significant influence on the ROA ratio, namely CAR and BOPO, respectively 0.001 and 0,000. The ratio of FDR and NPF shows the opposite results which have no effect on ROA, respectively 0.243 and 0.258.
Based on Table 2, CAR has a positive impact on ROA, this shows that the condition of Islamic bank capital is able to withstand the level of possible risk of loss caused by the COVID-19 pandemic. Based on the result we can conclude that 1% increases in CAR will be caused the change in ROA by 0.138%. This positive result is related to research by Shiang Liu (2013)  According to Table 2, BOPO appears with negative sign, 1% increases in BOPO will decrease the ROA about 0.252%. This result is similiar with the research of Purwoko & Sudiyatno (2013) and Prasanjaya & Ramantha (2013) which stated that operational costs from operating income have a negative effect on asset returns. If BOPO ratio is more than 1, it means the bank is less healthy. The higher the cost of the bank's income means that its operational activities are increasingly inefficient, so its income is also small. In other words BOPO affected negative significantly related to bank profitability.

Conclusion and Suggestion
According to the results shown in this study, some of insights can be used as conclusions. In seeking investments in Capital Adequacy Ratio (CAR), Financing Deposit Ratio (FDR), Cost-to-Income (BOPO), Non-Performing Financing (NPF) to profitability of Islamic banks. Then the conclusion can be explained as follows: First, Capital Adequacy Ratio (CAR) significantly positive effect on ROA. This result can be seen from the t test with a significance level of 0.001 (< 0.05). Then, it can be said that CAR has an influence on ROA. Second, Financing to Deposit Ratio (FDR) appears with insignificant results on ROA. The significance value is 0.243 which means it is greater than 0.05. Therefore, it can be said that FRD has no effect on ROA.
Third, Non-Performing Financing (NPF) variables also appear with insignificant results on ROA. The significance value is 0.258 which means it is greater than 0.05. Hence, it can be said that NPF has no effect on ROA. Fourth, Cost to Income (BOPO) appears with a negative sign towards ROA. With a significance value of 0,000 (< 0.005), it can be said that BOPO has a significant effect on ROA. During the COVID-19, Islamic bank performance indicators continued to show quality and aggressive performance growth.
These findings have practical implications for the community, making it possible to continue using Islamic banking services despite the COVID-19 pandemic. Macroeconomic factors such as inflation and Gross Domestic Product (GDP), or microprudential factors such as Third-Party Funds (TPF), Net Operating Margin (NOM), Return On-Equity (ROE) and Net Rewards (NI) are highly recommended for future research.