25 Chapter III Research Methodology This chapter presents the research methodology. This study will employ a quantitative methodology to study the impact of P2P on the profitability and stability of banks. The structure for the Research Methodology section of this study is as follows: (1) introduction, (2) data collection, (3) research method, including the measurement, (4) data analysis, including the preliminary, primary, and additional test, and (5) chapter summary. III.1 Introduction The appearance of P2P lending can be seen as an innovation in financial services. With technological advancement implemented and easy access for borrowers and lenders, this platform has a great potential market in the future. Not only in the future, they already have flourishing growth in their numbers and transaction. However, compared to banking transactions, P2P lending has yet to get beyond or even near them. This is the chance for the banking industry to understand the relationship between them and P2P lending. Is it a destructive or constructive relationship for them. Will the association make their performance increase or decrease. How about their stability. Will it be more risky for the banks. Using panel data regression, this study examines the relationship between P2P lending existence and banks’ performance and stability in Indonesia. III.2 Data Collection This study tries to use as many data banks as possible. Following the availability data of P2P lending from the OJK directory, this study used the time observation from January 2018 to August 2021. Using monthly data, this research collects the data of all banks in Indonesia, which is 165 banks that also include 15 Islamic banks. This sample collection is based on Amidu and Wolfe (2013) and Moudud- Ul-Huq (2021) that used commercial bank, development bank, and other banks as the sample of emerging and developing countries. These previous study not specifically used certain size of banks, in accordance to that, this study try to use as many bank as possible. Then, these banks are reassessed again to see if there is 26 any missing data. In conclusion, 98 banks can be used which is listed in Appendix A and summarized in Table III.1. However, after the data was cleaned accordingly to the P2P data availability, the time observation shrunk to 21 months from December 2018 to July 2020. So, this study’s final observation is 2,058 bank-month observations. Both banking and P2P lending data were secondary data obtained from the OJK directory. The banking data is available to the public. While the P2P data is not available to the public, thus, the data is acquired by asking directly to OJK, which is only available in aggregate of all P2P lending firms in Indonesia. Table III.1 Sample of the research No Bank Type Number of Bank 1 State Owned Bank 4 2 Regional Development Bank 25 3 National Private Bank 56 4 Branch office of overseas bank 1 5 Regional Islamic Development Bank 1 6 National Islamic Private Bank 11 Total Sample Bank 98 III.3 Research Method Looking at the research onion (Saunder et al., 2015) in Figure III.1, this study is applying positivism paradigm as we will try to understand the relationship between P2P existence with bank’s performance and stability. By looking at the previous research to understand the past findings and theories, this study use deduction approach to construct the hypothesis. This study applied mono quantitative approach as this study used secondary data from OJK. As explained in previous section, this study is using a long time period and incorporate many entities which as combination of longitudinal and cross-sectional data which called panel data. Thus this study will be using panel data regression that will be explained later in the next section. 27 Figure III.1 Research Onion (Saunder et al., 2015) In understanding the P2P impacts on banking performance and stability, this study examines a few variables. The first one is the dependent variable. This variable represents the outcome of the independent and the interest variables defined by the bank’s performance. Phan et al. (2020) and Almulla and Aljughaiman (2021) describe the bank’s performance using two variables. Following these previous studies, this research employed the dependent variables using return on asset (ROA) and returns on equity (ROE) as the banking profitability proxy. Higher ROA and ROE means higher profitability. While for the banking stability proxy, this study followed Almulla and Aljughaiman (2021) and Sudrajad and Hübner (2019) by utilizing loan loss provision (LLP) and Z-score (ZSCORE). The LLP represents the funds that spared by the bank to encounter its credit default. At the same time, the ZSCORE measured the reverse of insolvency probability. Thus, a lower LLP means the bank has less credit default and lower credit risk, and a higher ZSCORE means the bank is more stable. The calculation to form these measurements are described in Table III.2. 28 Table III.2 Measurement of Dependent Variables. Variable Description Calculation Dependent Variables (bank performance) Return On Assets (ROA) Net profit to total assets (Phan et al., 2020) ��������= ��� ����������� ��������������� ���������� Return On Equity (ROE) Net profit to shareholder’s equity (Phan et al., 2020) ���= ��� ����������� ��������������� ���������������� Dependent Variables (bank stability) Loan Loss Provision (LLP) Loan loss provision divided by total loans (Almulla & Aljughaiman, 2021) �������������= �������������� ��������� ������������������� ��������������� ��������������� Z-score (ZCORE) Financial stability measurement (Le, 2020) �−��������������������= ��������+ ��������������� ���������������� �������������� ����������� �� �������� The next variable is the independent variable or control variable. This variable consists of two types, banking-specific and macroeconomic variable. The banking characteristics of this study follow Nguyen et al. (2021), Phan et al. (2020), and Yudaruddin (2023) by measuring it using the cost-to-earnings ratio (CER), which is calculated from divided operational cost by operating income, bank size (SIZE), which is measured by the natural logarithm of total assets, and loan-to-deposit ratio (LDR), which is estimated by dividing total amount of loans by the total amount of deposits. While for the outside banking industry, initially, this study follows Antwi et al. (2020), Bejar et al.