“Impact of internal and external factors on the net interest margin of banks in Indonesia”

This study aims to assess the impact of bank-specific factors and macroeconomic in- dicators on the net interest margin (NIM) of commercial banks in Indonesia. Data from Indonesian commercial banks are used. Data are collected from the banks’ an- nual reports and the Financial Services Authority (OJK) for the period 2008 to 2018. A panel data regression model is used to estimate the effect of bank-specific and mac- roeconomic factors. The results prove that the variables of Non-Performing Loans (NPL), Loan to Deposit Ratio (LDR), Return on Assets (ROA), Interest Rate (SBI), and Exchange Rate (FOREX) affect NIM. The exchange rate variable has a predominant effect, while the NPL factor has a less strong influence on NIM. The empirical evidence from this research is important for commercial banks in Indonesia to improve operational efficiency through NIM performance. Internal and external factors of a bank should be subject of attention of bank managers.


INTRODUCTION
The role of a financial intermediary is an important task for banks to support their operations both individually and in companies that affect economic growth. According to Werner (2016), banks function as a liaison institution to distribute funds from economic entities with excess funds to other parties with deficit funds. As a result, banks receive deposits and loans at interest rates to depositors and borrowers. Interest is paid to depositors, and interest charged to recipients of funds results in a distribution called interest differentials to banks, where ideally the bank gives lower interest to savers and charges higher fees to creditors. According to this definition, the net interest margin (NIM) is the difference between the income and expenditure of the bank interest divided by the total value of bank assets (Tarus et al., 2012). Increased competition drives banks to improve efficiency through lower net interest margins (Angori et al., 2019). Wide net interest margins make it difficult for banks to expand their functions as financial intermediary institutions, because low deposit rates reduce the motivation to save and vice versa, with high loan rates being a heavy burden for companies in investing (Claessens et al., 2017). Information related to the bank's NIM ratio is also part of the signal in investment decision making (Endri & Fathony, 2020). As a result, banks should be able to perform an intermediary function at the lowest possible cost to boost overall economic growth.
An interesting phenomenon in the national banking shows the fluctuating NIM ratio during the period 2008 to 2018. In 2008, the NIM ratio reached a value of 5.98% and increased to 6.28% in 2009, and thereafter until 2018, the NIM ratio has decreased. In 2017, the value of the NIM ratio was 5.26% (an increase of 5.28%). On average, NIM of banks experienced a decline in 2008-2018. The decline in the NIM ratio demonstrates that this condition shows higher banking operational efficiency and affects the improvement of national banks' performance. As for the NIM of national banks, which tends to be high, the Financial Services Authority (OJK) issued a new regulation in the form of an NIM. The highest limit of the NIM ratio, according to the determined rules, is 4% with the aim of increasing the competitiveness of national banks in facing competition with other banks in ASEAN countries, where the banks' NIM ratio ranges from 2% to 4%. BUKU 4 bank group is the national bank with the largest assets and has the highest NIM ratio (6.36%) as of December 2018, which grew by 126 bps (1.26%) from 5.10% in December 2017. Furthermore, followed by BUKU 1 bank groups that have an NIM ratio of 5.28%, there was an increase of 7 bps (0.07%) from 5.82% and a position above the banking industry average by 5.28%. As for the other two bank groups, BUKU 2 with NIM increased by 117 bps (1.17%) from 3.54% to 4.71% and BUKU III with NIM increased by 105 bps (1.05%) from 3.44% to 4.49% in the same time period. The NIM ratio of the two banks in the BUKU group is below the banking industry's average of 5.39%. This study aims to analyze the internal and external factors that determine the NIM ratio of Indonesian conventional banks by grouping two factors, namely bank specific factors and macroeconomic variables, and using a panel data regression model.

LITERATURE REVIEW
Net interest margin (NIM) is broadly defined as the difference in interest charged to loan recipients and interest income obtained by savers as part of the bank's total productive assets (Ho & Saunders, 1981;Maudos & de Guevara, 2004). It should be borne in mind that in a limited sense, interest income and expenses are included as a ratio of interest income assets. This definition of the NIM ratio has two drawbacks: 1) it does not consider bank income from fees and commissions; and 2) it deviates from the marginal distribution (Brock & Suarez, 2000).
Usually, NIM is a measure of the actual difference in bank interest or past differences (expost spread). According to Demirguc-Kunt and Huizinga (1999), this difference is preferred, since it is usually available periodically at comparable levels of consolidation.
Previous research on the influence of internal factors and macroeconomic indicators on the bank's NIM ratio provides contradictory findings. Certain bank characteristic factors have been shown to influence NIMs, such as operational expenses, loans, capital adequacy, liquidity, credit quality, bad credit, interest rate risk, opportunity costs for bank reserves, size of bank assets and ownership structure. For macroeconomic indicators, inflation rate and economic growth indicate the most influential NIM determinant. However, it is understood that the consensus is that high inflation contributes to an increase in NIM, and the impact of real GDP growth remains ambiguous (Carbó & Rodríguez, 2007). Conversely, real GDP growth has a negative impact on bank NIMs based on the fact that (i) creditworthiness of borrowers and net worth decrease during recessions and increase in lending rates (Berger et al., 2020), and (ii) economic growth increases, lowering bank standards (Carbó & Rodríguez, 2007). On the other hand, there is a positive effect of economic growth on NIM due to the fact that loan demand increases during the increase cycle.
Demirgüç-Kunt and Huizinga (1999) examine the factors that influence NIM ratios and profitability using a sample of banks in 80 countries for 1988-1995. The results showed the following determinants of NIM and bank profitability: bank specific factors, macroeconomic variables, bank taxation, regulations and institutions, and financial structure. These findings also reveal that a small market concentration causes low margins and profits. Previous studies conducted by Ho and Saunder (1981) have been developed by Angbazo (1997)  In addition, macroeconomic variables are an important factor in determining bank NIM performance, because macroeconomic factors are considered as external factors affecting operational processes in the banking industry; better economic conditions can increase credit demand, which drives an increase in loan interest expense when borrowers do not replace loans (Kashyap & Stein, 2000). A bank's decision to raise or lower lending and savings rates as a reaction to external shocks in the money market is the same as the company's decision regarding commodity prices as a reaction to changes in costs. Some studies prove that macroeconomic factors do not affect NIM (

RESEARCH METHODOLOGY
In accordance with the objectives, the study es- 2) banks were commercial banking companies operating conventionally; and 3) banks had complete data availability during the study period, both for internal and external factors.
Based on the sample selection criteria, eleven sample banks participate in this study, which will then be estimated using the panel data regression model represented by the following formula: where Ci -constant panel fixed effect data regression for the i-th bank, i = 1, ... 11.
From the equation model above, the hypothesis for each regression coefficient of the fixed effect panel data regression model is tested on the dependent variable performance ratio of the bank NIM using the t-test. T-test was conducted to empirically prove that each independent variable used in this study could partially determine the ratio of banking NIM performance as a variable that was explained (dependent variable) significantly with a confidence level of 95 percent or alpha equal to five percent (α = 0.05). T-test shows that NPL and exchange rates negatively and significantly affect the bank's NIM ratio, while the variables of LDR, ROA, and Bank Indonesia interest rates (SBI) positively and significantly affect the banking NIM ratio. CAR, BOPO, and inflation have no effect on the performance of the banking NIM. Based on the estimation results of the panel data regression method using the fixed effects model, of the eight independent variables, which are bank financial performance factors, and macroeconomic factors that affect the banking NIM ratio, there are five variables that have a significant effect, namely NPL, LDR, ROA, SBI variables, and FOREX. Out of the variables that have a significant impact, macroeconomic variables, such as FOREX, are the most dominant variables affecting the banking NIM with a coefficient of 1.60 with a negative direction sign. The NPL ratio is the smallest influencing factor with a coefficient of 0.06 with a negative direction sign.

DISCUSSION
Bank specific factors, such as the CAR ratio, have no effect on the performance of the national banking NIM ratio. The CAR ratio is a comparison of bank performance used to calculate the bank's capital adequacy in order to support assets vulnerable to risk. In other words, the CAR ratio reflects the ability of banks to avoid risk on risk assets (risk aversion). This means that the higher the CAR ratio's value, the stronger the bank's capital strength in maintaining the possibility of loss risk arising from risk operation. Maudos  The credit risk variable measured by the NPL ratio negatively and significantly affects the NIM ratio. These results prove that the higher the NPL ratio, the lower the banking NIM ratio. This result also shows that banks may prefer profits with lower margins, when the financial situation of individuals and businesses worsens. The results of this study are in line with Zhou and Wong (2008) and Fungacova and Poghosyan (2011) who argue that the NPL ratio has a negative effect on the NIM ratio. Depositors require higher premiums to store their savings in risk banks (i.e. banks with high NPL ratios). An increase in interest on savings, assuming other factors are constant, will contribute to a decrease in NIMs, building the opposite relationship between problem loans and NIMs. Fungacova and Poghosyan (2011) prove that credit risk (NPL) negatively affects the NIM ratio. An increase in bank credit risk is caused by an increase in savings interest rates, which causes the amount of bad loans to increase, thereby reducing bank interest receipts. Given the reduced bank income from the credit side and the increased burden of funds from savings in terms of direct costs, the bank's profit margins have experienced a reduction, which affects a decrease in efficiency. Different results are proven by Maudos and Solis (2009) who state that the NPL ratio (credit risk) has a positive influence on the NIM ratio. Positive sign obtained in the variable means the bank avoids risk (risk aversion) such as interest rate risk and credit rate risk. The bank will set a higher interest rate to get a larger margin.
The banking intermediary function using the LDR ratio has a positive effect on the NIM ratio of banks. The LDR ratio is used to determine how much a bank can rely on lending as an important source of liquidity in meeting the provisions of its short-term obligations, such as withdrawal of funds made by savers and the interest expense that must be given to its customers. The findings of this study are in line with previous research conducted by Raharjo et al. (2014) who prove the significant effect of LDR on the NIM ratio. The findings of this study prove that the greater LDR ratio of a bank causes the bank's NIM ratio to increase. LDR ratio is the ratio between all the volume of credits channeled by banks and funds received by banks from third parties. As the volume of lending activities provided by banks increases, the performance of the NIM ratio will improve due to the benefits of interest income or the benefits of lending.
The efficiency measured by the BOPO ratio variable has no effect on the bank's NIM ratio. Empirical evidence shows that inflation as a macroeconomic variable has no effect on banks' NIM. The inflation rate itself can positively or negatively affect the bank's NIM performance. Inflation is usually a factor that has a great impact on the revenue side rather than the cost side, and can be ended with an increase in bank performance (NIM). The impact of changes in inflation itself is determined by whether inflation can be anticipated or not by banks (Pasiouras & Kosmidou, 2007).
If the inflation rate is fully anticipated, the interest rate charged by the bank will increase to cover the inflation risk. Thus, revenues will increase faster than costs increase, so that it has a positive impact on bank performance, especially on the level of bank profitability. The results of this study differ from the findings of Tennant and Folawewo (2009), which prove that inflation has a negative effect on banks' NIM performance. Abid et al. (2014) and Hamadi and Awdeh (2012), on the contrary, prove that inflation has a positive effect on the bank's NIM ratio. Raharjo et al. (2014) show that inflation is the only external factor influencing interest margins. Saad and El-Moussawi (2012) found the opposite that inflation has no effect on NIM.
The exchange rate (FOREX) as the last macroeconomic variable in this study negatively and significantly affects the bank's NIN ratio performance. The appreciation of the rupiah against the US dollar has an impact on the lower NIM ratio. This means that when the exchange rate appreciates, it increases the loss, and in this condition the NIM ratio decreases. The results of the study reveal that if the local currency depreciates, the quality of loans may deteriorate and the bank's margins will decline. Empirical evidence from this study is in line with Ugur and Erkus (2010) and De Bock and Demyanets (2012) who state that exchange rate devaluation has an opposite effect on the NIM ratio. Different findings are presented in the study by Festic and Beco (2008), which shows that the nominal exchange rate has a positive effect on the NIM ratio. Babalola (2012) concludes that foreign exchange rates significantly influence the NIM ratio.

CONCLUSION
This study estimates and analyzes the determinants of NIM of banks listed on the Indonesia Stock Exchange in 2008-2018 using the fixed effects panel data regression method. This is because the research data used is a combination of time series data, which has an eleven-year period (2008-2018), and cross-sectional data, in which 11 banks were selected as research samples. The research findings prove that the ratio of bad loans (NPL) and the exchange rate negatively and significantly affect the performance of the bank's NIM ratio, while the LDR variable, return on assets (ROA), and SBI interest rates have a positive effect on the bank's NIM ratio. CAR, BOPO, and inflation ratios have no effect on the banking NIM performance. With regard to the joint influence, it is said that all determinants observed in the study affect the performance of the banks' NIM. Of the factors that have a significant effect, the exchange rate, macroeconomic variables are the dominant factors affecting the NIM ratio of banks, while the NPL ratio has the lowest effect.