“The influence of financial inclusion on banks’ performance and risk: new evidence from MENAP”

This study seeks to investigate the relationship between financial inclusion factors and banks’ performance and risk among MENAP countries. The sample includes 271 banks located in 24 countries in the region that are interconnected, and micro- and macro-variables that affect the performance and risk levels of these banks. The results indicate that enhancing the level of financial inclusion in the region can increase banks’ performance and decrease their risk. They also point out where these banks could benefit more from financial inclusion in terms of reducing their risks. Future research may include investigating financial inclusion tools to explore the relationship between financial inclusion and banks’ performance and risk in developing countries. More re- search can be conducted on each MEANP country to analyze their characteristics and the influence of financial inclusion on bank performance and risk. In addition, future research should be conducted to study the relationship between regulations, rules and financial inclusion across countries and economies. revenues earned and profits made from banking operations; the other is in carrying the risks associated with this process. The relationship between profits and risk is a direct one in that if banks are looking to earn more profits, then they will incur more risk. When a bank expands its banking network by opening more branches, this is considered a capital investment. Making the decision to venture into this type of investment is considered a critical This study examines the influence of financial inclusion on banks’ performance and risk in the MENAP region. The sample consists of 271 banks in 24 countries of the region. Two regression techniques are used to achieve robust results when examining two levels of financial inclusion: the bank level and the national level. The variables used to measure financial inclusion are the number of bank branches and the national financial inclusion index. Micro- and macro-control variables are used to measure banks’ performance and risk. The findings show that increased financial inclusion through banking penetration in MENAP can reduce bank risk measured by the value of non-performing loans, while there is no relationship between the national financial inclusion index and bank risk at the country level. For the basic regression used to find banks’ performance measured by ROAA and ROAE, the analysis in-dicates that banking penetration, measured by the number of branches, significantly influences ROAE as a performance indicator at the bank level. Although there is no significant relationship between banking penetration and ROAA, the relationship is still positive. Furthermore, there is no significant relationship between where banks rank by the national financial inclusion index and their financial performance in MENAP. For the quantile regression, the results show that banking penetration can significantly increase the ROAA in the first quantile for MENAP banks, while there is no correlation between banking penetration in the second and third quantiles and banks’ ROAA. Furthermore, broader banking penetration through branches may enhance the second and third quantile of ROAE, while there is no relationship between the region’s first-quantile banks. Although no relationship was found between the national financial inclusion index and banks’ performance indicators among their quantiles, these results could be due to the low level of financial inclusion in the region. The results show that banking penetration is significantly related to the risk for banks in the first quantile. This means that opening more branches can reduce banks’ risk in the region. Further, the national financial inclusion index is significantly linked to the risk of banks in the second and third quantiles. These results suggest that increased financial inclusion in


INTRODUCTION
Banks expand their networks through branches, ATMs, points of sale (POS) and other electronic terminals (Shihadeh & Liu, 2019). They are aimed at attracting deposits and reaching more customers by offering direct and indirect credit and other services. The opening of new branches allows banks to invest in technology and equipment and hire more staff. This benefits local economies because it enables individuals to access formal financial services, especially in disadvantaged areas. Through branch banking, these institutions can offer their services and receive feedback from their customers. This enables them to implement innovative services targeting young people, women, craftsmen, farmers, and other groups of people who are usually marginalized in access to credit. By offering these services through their branches, banks can earn more profits, especially if this type of extension comes with innovative services and appropriate loans conditions, such as reasonable collateral requirements and credit cost.
There are two sides to the investment process: one is in the revenues earned and profits made from banking operations; the other is in carrying the risks associated with this process. The relationship between profits and risk is a direct one in that if banks are looking to earn more profits, then they will incur more risk. When a bank expands its banking network by opening more branches, this is considered a capital investment. Making the decision to venture into this type of investment is considered a critical

Literature review
Financial inclusion studies examine how limited access and use of formal financial services are. Many financial inclusion studies focus on individual characteristics and obstacles and analyze the effects of financial inclusion (or the lack thereof) on the poor and the determinants of financial inclusion across countries, economies, and regions. These studies cover the demand side and, until now, there is little research on supply. Therefore, there is a need for new evidence that will help banks increase their financial inclusion by investing in additional branches, ATMs, and online payment methods and increasing their lending.
Shihadeh and Liu (2019) presented global evidence that enhancing financial inclusion would improve banks' performance and reduce their risk. They used data on 701 banks in 189 countries and economies to examine the relationship between financial inclusion indicators such as branches, savings, loans, credit cards, debit cards, and formal accounts. They also used non-performance loans, return on average assets (ROAA) and return on average equity (ROAE) as indicators of banks' risk and performance, respectively. The result show that increased financial inclusion will positively affect banks' performance and risk. Shihadeh, Hannon, Guan, Ul Haq, and Wang (2018) examined the relationship between financial inclusion and banks' performance among Jordanian banks. Here, financial inclusion indicators consisted of the total value of credit extended to SMEs, the number of ATMs and ATM services, the number of credit cards issued, and the number of new services offered. The return on assets (ROA) and gross income were used as these banks' performance indicators. The study has found that financial inclusion has a positive impact on performance of Jordanian banks.
Berger, Leusner, and Mingo (1997) studied branch efficiency for commercial banks in the USA The authors used panel data on 761 observations for the banks. Thus, they used intermediation approaches to evaluate the branches' profits and expenses. Furthermore, the study used 769 observations for its production approaches, thus using operating costs and accounts opened, loan origination, debits, number of deposit accounts, credits as inputs, and closed accounts as output factors to determine output and costs for these branches. The findings show that banks can earn more revenue if they provide suitable places for customers to access their services and when they invest more in branch banking. The findings also show that more branches can cover more costs, as the num-ber of branches increases and the costs per branch decrease as some costs are shared across branches. Furthermore, bank management cannot control branch performance, while branch managers can play an important role in the efficiency of their branch. In addition, these banks were unable to achieve cost savings from mergers and acquisitions. Hirtle (2007) studied the effect of the size of bank networks on banks' performance through branch networks for commercial banks in the USA for the period 1995 to 2003. The results show that as deposits grow, costs decrease as branch networks expand. Further, network size has no significant effect on banks' performance. Harimaya and Kondo (2012) used data on regional and second-tier regional banks under the Japanese regional financial institutions classification to study the impact of expanding bank branch networks on costs and profits. The study found that if banks concentrated on activities throughout their own areas, they would improve their cost-efficiency. In addition, if regional banks do not expand their branches, they will achieve low-profit efficiency. While these findings are interesting, it is noteworthy that this study excludes banks operating in urban areas.
Nguyen (2014) used US bank branches closed and merged with other bank branches to examine whether the physical existence of a branch is still important in terms of obtaining financing and access to the financial sources. The data covers the years 1999 to 2012 and measures the impact of closed and merged branches on the volume of local loans. The study concluded that bank branches were still very important, especially for disadvantaged communities, and the closure of branches leads to a lack of lending, especially for small firms. Skillern (2002) reviewed the literature on the benefits of bank branch availability to low-income people and those living in rural areas and, thus, its impact on economic development. The findings indicate that when the bank branches are closed and it becomes difficult to obtain credit from the traditional bank branches, the alternative financial services offered come with higher fees. In these situations, low-income individuals will have less access to financial services, while these same services are typically required for economic development. Humphrey (1994) used cross-sectional data on 161 banks for the period 1991 to 1992 to study the impact of the increased use of ATMs on banks' performance. The study found that ATMs offer easy and convenient services to customers, but for banks, the costs appear to be slightly higher. The increased number of ATMs reduces the cost of each deposit transaction, but as depositors increase their transactions, this results in total costs being relatively equal to or slightly higher than before ATMs were introduced. The provision of additional banking access achieved through ATMs has raised revenues and bank profits, but this change was found to be weak and consistently insignificant. . These deposits allow banks to offer services to their clients and, therefore, earn profits from the difference between the interest paid on these accounts and the interest received from loans. These studies are based on banking theory, especially the financial intermediation theory of banking, which states that banks generate deposits so that they can lend this money to those seeking credit. This principle is in line with the financial inclusion mechanism that gives citizens of the country access to its formal financial system (i.e., through its banks). Here, citizens can open accounts, make deposits, and borrow to cover their needs and set up their businesses. This theory argues that in order for these banks to lend money, they need deposits.

Hypotheses development
Banks seek to extend their networks to reach clients in rural areas and, thus, attract more clients and more deposits to increase their credit portfolios. Banks' investment processes are based on their credit portfolios and their diversification. By distributing their various credit components across individuals, SMEs, companies, and banks can increase their financial performance and decrease their risk. It is noteworthy that the philosophy of financial inclusion is based on including all citizens in the formal financial system and, thus, ensuring that all citizens have access to and can use financial resources. Thus, the philosophy of financial inclusion is consistent with banks' targets and can help to earn more profit and reduce their risk.
To develop a hypothesis that examines the relationship between financial inclusion and banks' performance and risk in MENAP countries, this analysis uses the following dependent variables: the ROAA and ROAE as the performance indicators, where the value of non-performance loans is used as the risk indicator. The analysis also includes two levels of independent variables: First, micro-independent variables, which include the national financial-inclusion index, the value of banks' net assets, the value of loans to the value of total deposits ratio, and the number of branches for each bank among the 26 countries reviewed in this study. Second, macro-independent variables include the annual growth of GDP, the growth of broad money, and the asset concentration of the five major banks as a competition indicator.
From all of the above and based on the findings of previous studies discussed, the study hypotheses can be written as follows: H0: Financial inclusion has no significant influence on banks' performance and risk in MENAP.
H1: Financial inclusion has a significant influence on banks' performance and risk in MENAP.

Data
This study obtained data from BankScope, which provides data on banks around the world. This data includes the number of branches, asset values, capital to asset ratios, return on equity, return on assets, asset concentrations of the top five banks, To examine the relationship between the number of bank branches and the national financial inclusion index (NFI) and link this data to bank performance, this study eliminated all banks for which data about their number of branches could not be found. Finally, the study includes complete data for 271 banks across 24 MENAP countries (see Appendix A, Table A1). Table 1 presents the definitions of variables and data sources.

Independent variables
The data includes the financial inclusion index of MENAP countries', which is the number of bank branches per 100,000 people, and the value of their formal accounts, formal saving accounts, formal loans, credit card balances, and the number of debit cards as a percentage of the population aged 15 years and over were used as the NFI, which indicates financial inclusion at the country level. The principal component analysis was used to reduce these six national-level financial inclusion indicators to one factor. As the study focuses on banking penetration as a financial-inclusion indicator, the number of branches was used as a key variable for financial inclusion at the bank level.
At the bank level, the study used control variables that may affect banks' performance and risk; these variables include banks' total assets, capital ratio (%), and loans to deposits ratio (%).
In contrast, at the country level (as this study is based on cross-sectional data), the study used the asset concentration of the five main banks and the percentage growth in GDP and broad money.
To test whether the multicollinearity problem exists in the study models, the variance inflation factor (VIF) test was used. The VIF result indicated that the mean VIF was 1.58 (see Appendix A,  Table A3). Therefore, the study employed the OLS robustness test as a basic regression to solve the issue.

Empirical models
To continue the empirical analysis, the study used OLS (robustness) and the quantile technique by Shihadeh   . OLS technique is also used in these equations as it is more suitable for data. Therefore, the relationship between dependent and independent variables is as follows:  where BP represents bank performance indicators, ROAA, ROAE, and NPL represent its non-performance loans (risk indicator). Branches indicate the number of branches per bank; Assets are a bank's total assets; LTD is the percentage of loans to de-posits. NFI is the national financial inclusion index; Banks 5 represents the asset concentration of the top five banks; GDP is the percentage GDP growth; and M2 is the percentage growth of broad money. Table 2 shows the descriptive statistics, including the means, standard deviations, and the minimum and maximum observations for the outcomes and the predictor variables.  Table 3 shows the correlation matrix to obtain the initial indicator of the relationship between the dependent and independent variables. Table  3 shows a large ratio between the number of branches and ROAE and ROAA as banks' performance indicators. The results also indicate that the correlation between the number of branches and NPLs is very small, but negative, which means that a bank having more branches can reduce the number of non-performing loans. Meanwhile, the number of bank branches also has a greater impact on banks' performance than their risk indicators.

Descriptive statistics
The national financial inclusion index moves in the opposite direction to ROAE, since the coefficient of banks' performance is small. Further, the national financial inclusion index consist-ently moves according to ROAA, but the coefficient is small here. For the risk indicator, the results indicate that non-performing loans move in the opposite direction (decrease) to the number of branches and the national financial inclusion index. Therefore, more banking penetration by increasing the number of branches and enhancing the country's financial inclusion index, by increasing its financial inclusion tools, can reduce the number of bank non-performing loans in MENAP, thus enhancing its profits and performance.

Basic estimation -robust OLS
The analysis is as follows: model (M1.1) includes bank branches as a financial inclusion factor and the national financial inclusion index as an overall measure of financial inclusion. These key variables were added to the other control variables so one can examine whether financial inclusion (branches and NFI) has a significant impact on banks' performance and risk in MENAP. Table  4 presents the OLS estimation results for models (M1.1) and (M2.1). These models present the regression results for banks' performance and risk as an outcome. The regression results are presented in Table 4.
The estimation results for the model (M1.1) in Table 4 show that the number of branches can enhance banks' performance indicators. However, an unstable political regime can impact the business cycle in the region and, thus, NPLs can be significantly influenced by other factors not related to financial inclusion and the national financial inclusion index. The results also show that more banking penetration through the increased number of branches can decrease a bank's NPL as a risk indicator by about 150%. This analysis also shows that there is no relationship between NPLs of the largest five banks, GDP and broad money in MENAP.  Note: * p < 0.05, ** p < 0.01, *** p < 0.001. Note: Robust standard errors are in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

Robust estimationquantile regression
This section examines how banks can reduce risk and enhance performance by increasing their penetration through their increased number of branches and moving up the national financial inclusion index. Therefore, quantile regression for the number of NPLs, and ROAA and ROAE are used, based on the basic equations (M1.2) and (M2.2). Quantile regression allows us to study the impact of independent variables on different quantiles of the dependent variable distribution. Thus, it gives a complete picture of the relationship between NPLs, ROAA, and ROAE as outcomes and micro-and macro-interpreters. Quantile regression also considers the robustness of outliers in observations and does not distribute estimation and conclusions. It is noted that banks in the first quintile of ROAA and the second and third quintiles of ROAE could enhance their performance by opening additional branches: i.e., by enhancing their banking penetration through branching and, thereby, by enhancing their financial inclusion. Branching could enhance the performance of banks with  Table 6 shows that there is no link between the number of branches and banks' risk in all quintiles, except in the first quintile, which means that banks with fewer non-performance loans could benefit from having more branches at that particular risk level. There is a relationship between the level of national financial inclusion and bank risk. Consequently, financial inclusion may reduce banks' risk in the second and third risk quintiles, with advantages for these banks in the third quintile. These results are in line with Shihadeh and Liu (2019).

CONCLUSION
This study examines the influence of financial inclusion on banks' performance and risk in the MENAP region. The sample consists of 271 banks in 24 countries of the region. Two regression techniques are used to achieve robust results when examining two levels of financial inclusion: the bank level and the national level. The variables used to measure financial inclusion are the number of bank branches and the national financial inclusion index. Micro-and macro-control variables are used to measure banks' performance and risk. The findings show that increased financial inclusion through banking penetration in MENAP can reduce bank risk measured by the value of non-performing loans, while there is no relationship between the national financial inclusion index and bank risk at the country level.
For the basic regression used to find banks' performance measured by ROAA and ROAE, the analysis indicates that banking penetration, measured by the number of branches, significantly influences ROAE as a performance indicator at the bank level. Although there is no significant relationship between banking penetration and ROAA, the relationship is still positive. Furthermore, there is no significant relationship between where banks rank by the national financial inclusion index and their financial performance in MENAP.
For the quantile regression, the results show that banking penetration can significantly increase the ROAA in the first quantile for MENAP banks, while there is no correlation between banking penetration in the second and third quantiles and banks' ROAA. Furthermore, broader banking penetration through branches may enhance the second and third quantile of ROAE, while there is no relationship between the region's first-quantile banks. Although no relationship was found between the national financial inclusion index and banks' performance indicators among their quantiles, these results could be due to the low level of financial inclusion in the region.
The results show that banking penetration is significantly related to the risk for banks in the first quantile. This means that opening more branches can reduce banks' risk in the region. Further, the national financial inclusion index is significantly linked to the risk of banks in the second and third quantiles. These results suggest that increased financial inclusion in the region may reduce banks' risk. From these results, it can be concluded that enhanced financial inclusion through banking penetration can increase banks' performance and reduce their risk, with the advantage being a reduced risk.

RECOMMENDATIONS
Based on the above results, further studies can be conducted on the links between banks' performance and risk at the country level among MENAP banks. In addition, more financial tools can be added to