Early warning early action for the banking solvency risk in the COVID-19 pandemic era: A case study of Indonesia

Hidayat, Taufiq; Masyita, Dian; Nidar, Sulaeman Rahman; Ahmad, Fauzan; Syarif, Muhammad Adrissa Nur Article Early warning early action for the banking solvency risk in the COVID-19 pandemic era: A case study of Indonesia Economies Provided in Cooperation with: MDPI – Multidisciplinary Digital Publishing Institute, Basel Suggested Citation: Hidayat, Taufiq; Masyita, Dian; Nidar, Sulaeman Rahman; Ahmad, Fauzan; Syarif, Muhammad Adrissa Nur (2022) : Early warning early action for the banking solvency risk in the COVID-19 pandemic era: A case study of Indonesia, Economies, ISSN 2227-7099, MDPI, Basel, Vol. 10, Iss. 1, pp. 1-21, https://doi.org/10.3390/economies10010006 This Version is available at: https://hdl.handle.net/10419/257369 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/   Citation: Hidayat, Taufiq, Dian Masyita, Sulaeman Rahman Nidar, Fauzan Ahmad, and Muhammad Adrissa Nur Syarif. 2022. Early Warning Early Action for the Banking Solvency Risk in the COVID-19 Pandemic Era: A Case Study of Indonesia. Economies 10: 6. https:// doi.org/10.3390/economies10010006 Academic Editor: ¸Stefan Cristian Gherghina Received: 23 November 2021 Accepted: 21 December 2021 Published: 24 December 2021 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). economies Article Early Warning Early Action for the Banking Solvency Risk in the COVID-19 Pandemic Era: A Case Study of Indonesia Taufiq Hidayat 1,* , Dian Masyita 2, Sulaeman Rahman Nidar 2, Fauzan Ahmad 3and Muhammad Adrissa Nur Syarif 3 1STIE Indonesia Banking School, Jakarta 12730, Indonesia 2Fakultas Ekonomi dan Bisnis, Universitas Padjadjaran, Bandung 40132, Indonesia; [email protected] (D.M.); [email protected] (S.R.N.) 3System Dynamics Bandung Bootcamp, Bandung 40534, Indonesia; [email protected] (F.A.); [email protected] (M.A.N.S.) *Correspondence: [email protected] Abstract: The COVID-19 pandemic has affected people’s lives and increased the banking solvency risk. This research aimed to build an early warning and early action simulation model to mitigate the solvency risk using the system dynamics methodology and the Powersim Studio 10 © software. The addition of an early action simulation updates the existing early warning model. Through this model, the effect of policy design and options on potential solvency risks is known before implementation. The trials conducted at Bank BRI (BBRI) and Bank Mandiri (BMRI) showed that the model had the ability to provide an early warning of the potential increase in bank solvency risk when the loan restructuring policy is revoked. It also simulates the effectiveness of management’s policy options to mitigate these risks. This research used publicly accessible banking data and analysis. Bank management could also take advantage of this model through a self-stimulation facility developed in this study to accommodate their needs using the internal data. Keywords: banking sector; COVID-19 outbreak; corporate insolvency; simulation; loan restructuring policy; system dynamics; early warning early action 1. Introduction The COVID-19 pandemic has impacted all sectors and the economic activities of people in Indonesia. Almost all economic activities experienced a slowdown, especially businesses supported by loans from banks, including large enterprises and small and medium enterprises (SMEs). Consequently, these conditions have impacted the banking sector, resulting in several risks. These include an increased risk of non-performing loans (Barua and Barua 2021), abundant bank liquidity due to slowing loan demand (Goodell 2020), and a decline in banking profitability (Knowles et al. 2020). Although the economic crisis due to the COVID-19 outbreak is different from previous crises, it has similar consequences, including widespread business bankruptcies, increased unemployment, and worsening banking solvency (Danielsson et al. 2020). The Financial Services Authority (OJK), as the supervisory authority and regulator of banking in Indonesia, issued a loan restructuring policy for debtors that were affected by the pandemic in March 2020 to reduce the pandemic’s impact on solvency risk. The non-performing loan (NPL) restructured based on this regulation can still be categorized as a performing loan, and the bank does not have to set aside any loan impairment expenses. The policy is a quick response to the impact of COVID-19, relaxing the rules for restructuring non-performing loans, and was enforced for the first time. It was extended several times, until 31 March 2023. The restructuring policy aims to curb the increase in NPL and allow time for banks to strengthen their reserves for impairment losses on loans and capital to avoid the solvency Economies 2022,10, 6. https://doi.org/10.3390/economies10010006 https://www.mdpi.com/journal/economies Economies 2022,10, 6 2 of 21 risk. However, the NPL ratio of banks in Indonesia reached 3.35% in June 2021, the highest level since January 2019. This could still increase if outstanding loans that were restructured with the OJK policy are not entirely repaid. As of 30 June 2021, the proportion of restructured loans to total loans was 17.32% (Bank Indonesia 2021). Siregar et al. (2021) estimated that the NPL potential of the loan restructuring would range from 10 to 30% when the OJK policy is revoked. In these estimations, the national banking industry experiences a potential NPL ratio level above 5% or exceeds the NPL standard set by the regulator. Subsequently, the capital adequacy ratio (CAR) would decline, increasing the bank solvency risk. In reference to this banking situation, the challenge is to estimate the bank solvency conditions based on each bank’s capital capacity and potential NPL risk. Considering the outstanding loan restructures, NPL volume and ratio, the simulation for each bank could be calculated to detect the risk of solvency at an early stage (early warning), and present some early action plans (early action) to prevent the risk (Leaning 2016;Lang et al. 2018). Al-Kharusi and Murthy (2020) and Pavlov and Katsamakas (2021) examined early warning financial risks during the COVID-19 pandemic using a financial statement simulation methodology. However, the research did not model an early action plan to prevent these potential risks. Therefore, this research developed the model’s function to not only provide some early warnings but also propose some early actions in one simulation, to prevent the bank solvency risk related to regulation. The system dynamics methodology was used to model the complex financial transactions to produce financial statement baselines. Additionally, the methodology helps to simulate the feedback effect of the early actions with some changes to the baseline financial statement behavior (Oladimeji et al. 2020). Subsequently, this research produces an early information simulation model of solvency risk per individual bank and preventive policy options for management of the potential risks. The model could practically become a self-simulation tool for early warning and early action regarding solvency risks in bank management. The research questions of this study are: (1) How to develop the simulation model to provide an early warning of solvency risk in the banking industry in Indonesia during the COVID-19 outbreak; (2) How to simulate early action to mitigate and prevent the solvency risk to response the revocation of the loan restructuring policy by OJK. Using system dynamic simulation, this research resulted in a model that can deliver early warnings for individual bank solvency risk and early actions for the individual bank management to prevent the solvency risks and increase the readiness for revocation of the loan restructuring policy. This research makes an important contribution to the study of the early warning system of bank bankruptcy risks for several reasons, namely: (1) the use of system dynamics simulation methods to modelling the complex, dynamic and ongoing bank risk behaviour during the pandemic COVID-19, (2) the research is to be able to produce an effective early warning of the bank solvency risks and (3) the research is a kind of a forward looking oriented simulation to predict the potential bank bankruptcy risks. The added value of this research is the existence of a dynamic bank balance sheet simulation so that the condition of the bank’s asset, liabilities, capital and profit and loss during the COVID-19 pandemic and the risk of bank solvency can be detected at any time. By monitoring the condition of bank’s loan performance through several main variables related to solvency risk, bank’s management can determine appropriate policies to reduce this risk and indirectly reduce the risk of bank’s bankruptcy. This research is able to produce simulation models in the form of early warnings and simulations of several policy options as early actions for the bank’s management to response the risk of non-performing loan, solvency risk and bank’s bankruptcy due to the COVID-19 pandemic. The structure of this paper consists of five parts, beginning with the introduction that contains the background of the research. Section 2discusses the literature review, namely the related literature and previous research, Section 3explains the research methodology followed by the research results and Section 4is a discussion of the results of the research. Economies 2022,10, 6 3 of 21 Lastly is the section of the conclusion and suggestions that also contains the implication of the research results. This study is ended with a reference of the studies that used in the research. 2. Literature Review According to De Vany (1984), the main causes of bank bankruptcy are the information asymmetry, agency problem and moral hazard that occur together. Smith (2010) found evidence that there is a correlation between agency problems and bank bankruptcy in the crisis period of 2007 and 2008. The agency problem is the problem of mismatch of interests between shareholders as principals and management as agents (Jensen and Meckling 1976; Rose 1992). In the banking sector, agency relationships occur between bank management with shareholders and banking supervisory authorities (Henrard and Olieslagers 2004) as well as with depositors (Kuritzkes et al. 2003). Banking supervisory authorities play a role in protecting the interests of depositors by issuing various regulations that must be obeyed by bank management and shareholders (Donnellan and Rutledge 2016), including the issuance of policies for loan restructuring during the COVID-19 pandemic by the Financial Services Authority, then it is an agency relationship intervention in order to reduce the risk of bank bankruptcy and protect depositors (Hidayat et al. 2021). Bank bankruptcy can also arise due to changes in financial conditions both internally and externally of the bank. Bank bankruptcy can also arise from increased loan risk arising from debtor moral hazard, weak analysis of creditworthiness, external conditions such as the decline in the community’s economic capacity due to the COVID-19 pandemic or lending to high-risk sectors. Bank management as an agent in the agency theory needs to recognize weak signals in the economic environment that will affect loan risk and bank bankruptcy, such as high NPL levels and declining CAR ratios. A weak signal is a symptom of bank performance that provides the basis for managerial decision making to ensure that the bank’s strategy can be achieved. Meanwhile, based on the weak banking signal information, a system is needed to provide some early warnings for bank management to be aware of the potential risks that may arise. early warning systems are a key tool for bank management to anticipate and make policies to reduce the potential risks of bank bankruptcy (Gunnersen 2014). Most of the literature research on early warnings of bank solvency risk are backwardoriented. This means that the literature is based on historical financial statements to provide early warning indicators of these potential risks. For instance, Korzeb and Niedziółka (2020); Barua and Barua (2021); Hardiyanti and Aziz (2021) showed the phenomenon of increased NPL risk during the COVID-19 pandemic. The increase in NPL reduced cash flow, profit, and CAR (Mayes and Stremmel 2012;Donnellan and Rutledge 2016). Consequently, a decrease in CAR increased the bank solvency risk, as measured by Z-Score (Lepetit et al. 2020). This research used the NPL, CAR, and Z-Score ratios to show the solvency risk. Facing the COVID-19 pandemic requires a forward-looking approach for early warnings of potential bank solvency risks and early action to prevent these risks. For these reasons, the simulation methodology was used to develop baseline projections for financial statements. These baselines describe future financial risk conditions and changes in their behavior due to some unexpected events. Furthermore, the simulation results were used to develop several alternative early action policies to reduce solvency risk (Pavlov and Katsamakas 2021;Petropoulos et al. 2020). These include the promotion of loan growth, managing restructured loans, increasing bank operational cost efficiency, lowering interest expenses, and increasing loan interest (Bastana et al. 2016;Samorodov et al. 2019;Rahmi and Sumirat 2021). Those policies are expected to strengthen bank capital to avoid solvency risk. However, the early action policy should be simulated first to determine the feedback on potential changes in the solvency risk levels and the optimum policy options (Schuermann 2014). According to Wu (2014) and Kunc et al. (2018), the system dynamics methodology is a simulation modeling that accommodates the feedback process in managerial decision making. Economies 2022,10, 6 4 of 21 System Dynamics is a methodology to design strategies and policies with computer simulation tools (Sapiri et al. 2020), to produce better responses to the complex and dynamic problems in the social, managerial or economic fields (Sterman 2000;Morecroft 2015; Duggan 2016). To solve the complex problems, according to Bala et al. (2017), the structures, and the relationship between the structures in the problem, should be analyzed. The system dynamics model describes the structure of financial statement accounts based on stock, rate, auxiliary, and constant. The pattern of the relationships between the accounts in the financial statements are modeled through causal-loop and stock-flow diagrams (García 2019). The financial statements model of system dynamics was developed in some studies by Islam et al. (2013); Wu (2014); Istiaq (2015); Pierson (2020); Aksu and Tursun (2021); Pavlov and Katsamakas (2021); Hidayat et al. (2021) to analyze financial reporting, banking risk management, management control systems, and solvency stress-testing. 3. Methodology This research used the previous literature to accommodate the bank’s needs when dealing with the COVID-19 pandemic and the loan restructuring policy revocation. It aims to develop a simulation model to generate early warnings and identify early actions to mitigate bank solvency risks using Powersim Professional 10 © software. Furthermore, it intends to advance the modeling of bank financial statements and several important financial ratios developed by Islam et al. (2013); Pierson (2020) and Hidayat et al. (2021) through causal-loop and stock flow diagrams. The modeling begins by analyzing the structures of bank financial statements and growth assumptions to produce the baseline financial report, and then could be used as a comparison in further simulations. When the loan restructuring policy ends, the baseline financial report is simulated to obtain the early warning on solvency risk and prepare the policy response. The policy response is simulated first to determine its impact on solvency risk before implementation. The simulation modeling stages for early warning and early action for bank solvency risk are presented in Figure 1. Economies 2022, 10, x FOR PEER REVIEW 4 of 24 2019; Rahmi and Sumirat 2021). Those policies are expected to strengthen bank capital to avoid solvency risk. However, the early action policy should be simulated first to determine the feedback on potential changes in the solvency risk levels and the optimum policy options (Schuermann 2014). According to Wu (2014) and Kunc et al. (2018), the system dynamics methodology is a simulation modeling that accommodates the feedback process in managerial decision making. System Dynamics is a methodology to design strategies and policies with computer simulation tools (Sapiri et al. 2020), to produce better responses to the complex and dynamic problems in the social, managerial or economic fields (Sterman 2000; Morecroft 2015; Duggan 2016). To solve the complex problems, according to Bala et al. (2017), the structures, and the relationship between the structures in the problem, should be analyzed. The system dynamics model describes the structure of financial statement accounts based on stock, rate, auxiliary, and constant. The pattern of the relationships between the accounts in the financial statements are modeled through causal-loop and stock-flow diagrams (García 2019). The financial statements model of system dynamics was developed in some studies by Islam et al. (2013); Wu (2014); Istiaq (2015); Pierson (2020); Aksu and Tursun (2021); Pavlov and Katsamakas (2021); Hidayat et al. (2021) to analyze financial reporting, banking risk management, management control systems, and solvency stresstesting. 3. Methodology This research used the previous literature to accommodate the bank’s needs when dealing with the COVID-19 pandemic and the loan restructuring policy revocation. It aims to develop a simulation model to generate early warnings and identify early actions to mitigate bank solvency risks using Powersim Professional 10© software. Furthermore, it intends to advance the modeling of bank financial statements and several important financial ratios developed by Islam et al. (2013); Pierson (2020) and Hidayat et al. (2021) through causal-loop and stock flow diagrams. The modeling begins by analyzing the structures of bank financial statements and growth assumptions to produce the baseline financial report, and then could be used as a comparison in further simulations. When the loan restructuring policy ends, the baseline financial report is simulated to obtain the early warning on solvency risk and prepare the policy response. The policy response is simulated first to determine its impact on solvency risk before implementation. The simulation modeling stages for early warning and early action for bank solvency risk are presented in Figure 1. Figure 1. Stages of Early Warning, Early Action Simulation Modeling of Bank Solvency Risk (Schuermann 2014). Early Warning Early Action Simulation Beginning Balance Assumption of growth of loan, saving and interest rate System dynamics modelling process: 1. Causal loop diagram 2. Stockflow diagram Simulation of bank financial report based on scenario of loan restructured default rate Bank policies response option Bank financial accounting reporting Baseline of bank financial report & ratio Moratorium of credit restructuring policy Bankruptcy risk indicator: ROA, NPL, CAR, and Z-Score Figure 1. Stages of Early Warning, Early Action Simulation Modeling of Bank Solvency Risk (Schuermann 2014). The simulation model was applied to BBRI and BMRI, the two largest banks in Indonesia. Data were obtained on the initial balance of financial statements to model from the financial position statement balance as of 31 December 2019 (audited). The simulation modeling period is monthly, from the 1st period (January 2020) to the 48th period (December 2023). An early warning simulation of solvency risk was performed in the 21st period, or September 2021, to determine the early risk condition if the loan restructuring policy revocation is implemented in March 2023 (39th period). After obtaining the early warning information, an early action simulation was performed to determine the Economies 2022,10, 6 5 of 21 impact of each policy on changes in CAR and Z-Score until the 48th period. The bank is categorized as insolvable if the Z-Score < 0 and/or the CAR ratio less than its threshold based on the standard of each bank. The higher the CAR and Z-Score, the more solvable the bank (Lepetit et al. 2020). The Z-Score formula used was: Z-Score: ((ROA + (equity/total assets))/ROA standard deviation (1) The other financial ratios that become indicators of bank solvency are: Return on asset (ROA) = Net Profit Total Asset (2) Capital adequacy ratio (CAR) = Equity Total risk-weighted asset (3) Non-performing loan (NPL)ratio =Non-performing total loan (4) Loan loss provision (LLP)ratio =Loan loss provision Non-performing loan (5) 3.1. Causal Loop Diagram of Early Warning Early Action Simulation Model The structure of the simulation model was prepared based on the account components in the financial statements, transaction flows to earn profits, early information on bank solvency risks, and early actions to mitigate these risks. This was described as a causal-loop diagram in Figure 2to show the transaction flow of the banking business activities. The positive link shows a unidirectional or positive causality relationship between the two structures. In contrast, a negative link shows an inverse or negative causality relationship between the two structures. Economies 2022, 10, x FOR PEER REVIEW 6 of 24 Figure 2. Causal-loop Diagram of Early Warning Early Action Model (Source: Author). The + sign indicates a unidirectional relationship between two variables, while the - sign indicates an inverse relationship between the two variables The early warning indicators of banking solvency risk used in this model appear in terms of ROA, NPL Ratio, CAR, and Z-Score. Any weakening in these indicators forms the basis for determining early action policy scenarios to improve the conditions. Therefore, this research simulated early action scenarios to overcome the potential weakening of banking solvency risks when the loan restructuring policy is revoked in March 2023. The scenarios include: a. Promoting loan growth or new loan policy to increase performing loans, to obtain more interest income and strengthen ROA, CAR, and Z-Score. However, when economic growth is abnormal, new loans should be selectively added to avoid additional NPL. b. Interest management is carried out by adjusting the loan interest rate and the savings interest rate to obtain an optimum net interest margin. c. The efficiency of operating expenses, including bank overhead, employee costs, and other expenses could reduce the ratio of operating expenses to income. d. Combined policy of (a), (b) and (c) above. 3.2. Stock Flow Diagram of Early Warning, Early Action Simulation Model For simulation purposes, the relationship between the structure of financial transactions depicted in the causal-loop diagram in Figure 2 was operationalized into a stockflow model, whose symbol is presented in Table 1. The structure of the balance sheet in the financial statements, consisting of equity, performing loans, third party funds and others in Figure 2, are categorized as stock because they have a balance at a certain time. In this situation, changes in stock balance are determined by the rate of inflows and outflows per unit time. The rate is determined by the multiplication of the stock and a constant variable, either directly or through several calculation stages using the auxiliary function. Liquid Asset & Securities Tax - Equity Profit After Tax + Non Performing Loan + Third Parties Fund interest expense + - COVID19 Pandemic Credit restructuring policy + Performing loan + Loan write-off + Risk weighted asset Capital adequacy ratio - + Loan Restructured + Total loan ++ NPL Ratio + - + <Total loan> <Liquid Asset & Securities> Total Asset + + Return on asset + - Standard Deviation of ROA Z-Score Equity per Total Asset - + - + Interest income - + Loan impairment losses expense + - Early action scenario - + New Loan + + Operating expense - - - <NPL Ratio> Early warning ++ + + + + + + Loan loss provision LLP - + LLP ratio -+ + + + + Dividend - + Figure 2. Causal-loop Diagram of Early Warning Early Action Model (Source: Author). The + sign indicates a unidirectional relationship between two variables, while the − sign indicates an inverse relationship between the two variables Economies 2022,10, 6 6 of 21 When conducting the intermediary function, the banking business receives third-party funds (TPF) or savings that could be converted into loans to obtain an interest income. When interest income covers the interest expense, operating expense, loan impairment losses expenses and tax, the bank reports some net income as additional capital. During the COVID-19 pandemic crisis, interest income decreased, while loan impairment losses increased, meaning that banks faced an increased solvency risk. The early warning indicators of banking solvency risk used in this model appear in terms of ROA, NPL Ratio, CAR, and Z-Score. Any weakening in these indicators forms the basis for determining early action policy scenarios to improve the conditions. Therefore, this research simulated early action scenarios to overcome the potential weakening of banking solvency risks when the loan restructuring policy is revoked in March 2023. The scenarios include: a. Promoting loan growth or new loan policy to increase performing loans, to obtain more interest income and strengthen ROA, CAR, and Z-Score. However, when economic growth is abnormal, new loans should be selectively added to avoid additional NPL. b. Interest management is carried out by adjusting the loan interest rate and the savings interest rate to obtain an optimum net interest margin. c. The efficiency of operating expenses, including bank overhead, employee costs, and other expenses could reduce the ratio of operating expenses to income. d. Combined policy of (a), (b) and (c) above. 3.2. Stock Flow Diagram of Early Warning, Early Action Simulation Model For simulation purposes, the relationship between the structure of financial transactions depicted in the causal-loop diagram in Figure 2was operationalized into a stock-flow model, whose symbol is presented in Table 1. The structure of the balance sheet in the financial statements, consisting of equity, performing loans, third party funds and others in Figure 2, are categorized as stock because they have a balance at a certain time. In this situation, changes in stock balance are determined by the rate of inflows and outflows per unit time. The rate is determined by the multiplication of the stock and a constant variable, either directly or through several calculation stages using the auxiliary function. Table 1. Symbols of Stock-Flow Diagram in Powersim Professional 10© (Sterman 2000). Symbol Definition Economies 2022, 10, x FOR PEER REVIEW 7 of 24 Table 1. Symbols of Stock-Flow Diagram in Powersim Professional 10© (Sterman 2000). Symbol Definition The symbol of STOCK is to declare variables with an accumulation derived from the previous value plus the difference between inflows and outflows. Stock(t) ==   (Inflow(s) − Outflow(s))ds + Stock(t0) d(Stock)/dt = Inflow(t) − Outflow(t). The symbol of RATE states the formulation of the amount of the stock inflow and outflow in the system in a certain time unit. For example, the rate of loan market = $100/month. The symbol of AUXILIARY or AUX is used to formulate the equation rate by defining the determining factors of the rate equation separately. Additional equations are substituted for each other and several separate rate equations. For example, Aux = Aux B x Constant The symbol of CONSTANT is a function of a certain number, the input for the auxiliary or equation rate in the model, its value remains in the simulation period. It is used to simulate management policies, such as loan interest rate income and saving interest rate expense policies. The symbol of ARROW indicates the flow of information from one variable (auxiliary, stock, constant, level) to another. The symbol of GRAPH contains certain parameter functions to explain other parameters/quantities. Stock-flow diagrams are developed based on bank financial statement accounts, including the financial assets account group (Figure 3), liabilities account group (Figure 4), and equity and statement of profit or loss (Figure 5). The formulation used to calculate the changes in the account is given in Appendix A. The financial assets account diagram shows the cycle of investment transactions to earn interest income and maintain liquidity. In this situation, the bank prioritizes its liquid assets for investment in performing loans and generating high returns, although it is necessary to anticipate the risk of loan default. A defaulted loan then could be restructured and controlled as a restructured loan, while a defaulted loan that could not be restructured could then be administered as an NPL. Since eliminating NPL write-off reduces capital, the bank forms a loan lost provision (LLP). The NPL ratio is the early warning of loan risk, which must not exceed 5%. The bank maintains adequate levels of liquidity in the form of liquid assets and securities. The total balance between the two financial assets can meet the transaction payment needs for the next 1 month. Investment in securities generates interest income. In the research design, the impact of changes in loan interest rates on additional loans is explained in the function of the interest rate loan to its delayed effect ratio variable in the additional loan formula in Table A1. If the annual interest rate of loan is increased, it will have a negative impact on additional loans. The stock-flow diagram of liabilities illustrates the transaction flow of the third-party fund (TPF) originating from the addition and withdrawal of TPF funds from savers, as Rate of loan market AuxAux B Constant_1 C onstant_1 Target loan to deposit ratio LDR assumption The symbol of STOCK is to declare variables with an accumulation derived from the previous value plus the difference between inflows and outflows. Stock(t)=Rt1 t0=(Inflow(s) −Outflow(s))ds + Stock(t0) d(Stock)/dt = Inflow(t) −Outflow(t). Economies 2022, 10, x FOR PEER REVIEW 7 of 24 Table 1. Symbols of Stock-Flow Diagram in Powersim Professional 10© (Sterman 2000). Symbol Definition The symbol of STOCK is to declare variables with an accumulation derived from the previous value plus the difference between inflows and outflows. Stock(t) ==   (Inflow(s) − Outflow(s))ds + Stock(t0) d(Stock)/dt = Inflow(t) − Outflow(t). The symbol of RATE states the formulation of the amount of the stock inflow and outflow in the system in a certain time unit. For example, the rate of loan market = $100/month. The symbol of AUXILIARY or AUX is used to formulate the equation rate by defining the determining factors of the rate equation separately. Additional equations are substituted for each other and several separate rate equations. For example, Aux = Aux B x Constant The symbol of CONSTANT is a function of a certain number, the input for the auxiliary or equation rate in the model, its value remains in the simulation period. It is used to simulate management policies, such as loan interest rate income and saving interest rate expense policies. The symbol of ARROW indicates the flow of information from one variable (auxiliary, stock, constant, level) to another. The symbol of GRAPH contains certain parameter functions to explain other parameters/quantities. Stock-flow diagrams are developed based on bank financial statement accounts, including the financial assets account group (Figure 3), liabilities account group (Figure 4), and equity and statement of profit or loss (Figure 5). The formulation used to calculate the changes in the account is given in Appendix A. The financial assets account diagram shows the cycle of investment transactions to earn interest income and maintain liquidity. In this situation, the bank prioritizes its liquid assets for investment in performing loans and generating high returns, although it is necessary to anticipate the risk of loan default. A defaulted loan then could be restructured and controlled as a restructured loan, while a defaulted loan that could not be restructured could then be administered as an NPL. Since eliminating NPL write-off reduces capital, the bank forms a loan lost provision (LLP). The NPL ratio is the early warning of loan risk, which must not exceed 5%. The bank maintains adequate levels of liquidity in the form of liquid assets and securities. The total balance between the two financial assets can meet the transaction payment needs for the next 1 month. Investment in securities generates interest income. In the research design, the impact of changes in loan interest rates on additional loans is explained in the function of the interest rate loan to its delayed effect ratio variable in the additional loan formula in Table A1. If the annual interest rate of loan is increased, it will have a negative impact on additional loans. The stock-flow diagram of liabilities illustrates the transaction flow of the third-party fund (TPF) originating from the addition and withdrawal of TPF funds from savers, as Rate of loan market AuxAux B Constant_1 C onstant_1 Target loan to deposit ratio LDR assumption The symbol of RATE states the formulation of the amount of the stock inflow and outflow in the system in a certain time unit. For example, the rate of loan market = $100/month. Economies 2022, 10, x FOR PEER REVIEW 7 of 24 Table 1. Symbols of Stock-Flow Diagram in Powersim Professional 10© (Sterman 2000). Symbol Definition The symbol of STOCK is to declare variables with an accumulation derived from the previous value plus the difference between inflows and outflows. Stock(t) ==   (Inflow(s) − Outflow(s))ds + Stock(t0) d(Stock)/dt = Inflow(t) − Outflow(t). The symbol of RATE states the formulation of the amount of the stock inflow and outflow in the system in a certain time unit. For example, the rate of loan market = $100/month. The symbol of AUXILIARY or AUX is used to formulate the equation rate by defining the determining factors of the rate equation separately. Additional equations are substituted for each other and several separate rate equations. For example, Aux = Aux B x Constant The symbol of CONSTANT is a function of a certain number, the input for the auxiliary or equation rate in the model, its value remains in the simulation period. It is used to simulate management policies, such as loan interest rate income and saving interest rate expense policies. The symbol of ARROW indicates the flow of information from one variable (auxiliary, stock, constant, level) to another. The symbol of GRAPH contains certain parameter functions to explain other parameters/quantities. Stock-flow diagrams are developed based on bank financial statement accounts, including the financial assets account group (Figure 3), liabilities account group (Figure 4), and equity and statement of profit or loss (Figure 5). The formulation used to calculate the changes in the account is given in Appendix A. The financial assets account diagram shows the cycle of investment transactions to earn interest income and maintain liquidity. In this situation, the bank prioritizes its liquid assets for investment in performing loans and generating high returns, although it is necessary to anticipate the risk of loan default. A defaulted loan then could be restructured and controlled as a restructured loan, while a defaulted loan that could not be restructured could then be administered as an NPL. Since eliminating NPL write-off reduces capital, the bank forms a loan lost provision (LLP). The NPL ratio is the early warning of loan risk, which must not exceed 5%. The bank maintains adequate levels of liquidity in the form of liquid assets and securities. The total balance between the two financial assets can meet the transaction payment needs for the next 1 month. Investment in securities generates interest income. In the research design, the impact of changes in loan interest rates on additional loans is explained in the function of the interest rate loan to its delayed effect ratio variable in the additional loan formula in Table A1. If the annual interest rate of loan is increased, it will have a negative impact on additional loans. The stock-flow diagram of liabilities illustrates the transaction flow of the third-party fund (TPF) originating from the addition and withdrawal of TPF funds from savers, as Rate of loan market AuxAux B Constant_1 C onstant_1 Target loan to deposit ratio LDR assumption The symbol of AUXILIARY or AUX is used to formulate the equation rate by defining the determining factors of the rate equation separately. Additional equations are substituted for each other and several separate rate equations. For example, Aux = Aux B x Constant Economies 2022,10, 6 7 of 21 Table 1. Cont. Symbol Definition Economies 2022, 10, x FOR PEER REVIEW 7 of 24 Table 1. Symbols of Stock-Flow Diagram in Powersim Professional 10© (Sterman 2000). Symbol Definition The symbol of STOCK is to declare variables with an accumulation derived from the previous value plus the difference between inflows and outflows. Stock(t) ==   (Inflow(s) − Outflow(s))ds + Stock(t0) d(Stock)/dt = Inflow(t) − Outflow(t). The symbol of RATE states the formulation of the amount of the stock inflow and outflow in the system in a certain time unit. For example, the rate of loan market = $100/month. The symbol of AUXILIARY or AUX is used to formulate the equation rate by defining the determining factors of the rate equation separately. Additional equations are substituted for each other and several separate rate equations. For example, Aux = Aux B x Constant The symbol of CONSTANT is a function of a certain number, the input for the auxiliary or equation rate in the model, its value remains in the simulation period. It is used to simulate management policies, such as loan interest rate income and saving interest rate expense policies. The symbol of ARROW indicates the flow of information from one variable (auxiliary, stock, constant, level) to another. The symbol of GRAPH contains certain parameter functions to explain other parameters/quantities. Stock-flow diagrams are developed based on bank financial statement accounts, including the financial assets account group (Figure 3), liabilities account group (Figure 4), and equity and statement of profit or loss (Figure 5). The formulation used to calculate the changes in the account is given in Appendix A. The financial assets account diagram shows the cycle of investment transactions to earn interest income and maintain liquidity. In this situation, the bank prioritizes its liquid assets for investment in performing loans and generating high returns, although it is necessary to anticipate the risk of loan default. A defaulted loan then could be restructured and controlled as a restructured loan, while a defaulted loan that could not be restructured could then be administered as an NPL. Since eliminating NPL write-off reduces capital, the bank forms a loan lost provision (LLP). The NPL ratio is the early warning of loan risk, which must not exceed 5%. The bank maintains adequate levels of liquidity in the form of liquid assets and securities. The total balance between the two financial assets can meet the transaction payment needs for the next 1 month. Investment in securities generates interest income. In the research design, the impact of changes in loan interest rates on additional loans is explained in the function of the interest rate loan to its delayed effect ratio variable in the additional loan formula in Table A1. If the annual interest rate of loan is increased, it will have a negative impact on additional loans. The stock-flow diagram of liabilities illustrates the transaction flow of the third-party fund (TPF) originating from the addition and withdrawal of TPF funds from savers, as Rate of loan market AuxAux B Constant_1 C onstant_1 Target loan to deposit ratio LDR assumption The symbol of CONSTANT is a function of a certain number, the input for the auxiliary or equation rate in the model, its value remains in the simulation period. It is used to simulate management policies, such as loan interest rate income and saving interest rate expense policies. Economies 2022, 10, x FOR PEER REVIEW 7 of 24 Table 1. Symbols of Stock-Flow Diagram in Powersim Professional 10© (Sterman 2000). Symbol Definition The symbol of STOCK is to declare variables with an accumulation derived from the previous value plus the difference between inflows and outflows. Stock(t) ==   (Inflow(s) − Outflow(s))ds + Stock(t0) d(Stock)/dt = Inflow(t) − Outflow(t). The symbol of RATE states the formulation of the amount of the stock inflow and outflow in the system in a certain time unit. For example, the rate of loan market = $100/month. The symbol of AUXILIARY or AUX is used to formulate the equation rate by defining the determining factors of the rate equation separately. Additional equations are substituted for each other and several separate rate equations. For example, Aux = Aux B x Constant The symbol of CONSTANT is a function of a certain number, the input for the auxiliary or equation rate in the model, its value remains in the simulation period. It is used to simulate management policies, such as loan interest rate income and saving interest rate expense policies. The symbol of ARROW indicates the flow of information from one variable (auxiliary, stock, constant, level) to another. The symbol of GRAPH contains certain parameter functions to explain other parameters/quantities. Stock-flow diagrams are developed based on bank financial statement accounts, including the financial assets account group (Figure 3), liabilities account group (Figure 4), and equity and statement of profit or loss (Figure 5). The formulation used to calculate the changes in the account is given in Appendix A. The financial assets account diagram shows the cycle of investment transactions to earn interest income and maintain liquidity. In this situation, the bank prioritizes its liquid assets for investment in performing loans and generating high returns, although it is necessary to anticipate the risk of loan default. A defaulted loan then could be restructured and controlled as a restructured loan, while a defaulted loan that could not be restructured could then be administered as an NPL. Since eliminating NPL write-off reduces capital, the bank forms a loan lost provision (LLP). The NPL ratio is the early warning of loan risk, which must not exceed 5%. The bank maintains adequate levels of liquidity in the form of liquid assets and securities. The total balance between the two financial assets can meet the transaction payment needs for the next 1 month. Investment in securities generates interest income. In the research design, the impact of changes in loan interest rates on additional loans is explained in the function of the interest rate loan to its delayed effect ratio variable in the additional loan formula in Table A1. If the annual interest rate of loan is increased, it will have a negative impact on additional loans. The stock-flow diagram of liabilities illustrates the transaction flow of the third-party fund (TPF) originating from the addition and withdrawal of TPF funds from savers, as Rate of loan market AuxAux B Constant_1 C onstant_1 Target loan to deposit ratio LDR assumption The symbol of ARROW indicates the flow of information from one variable (auxiliary, stock, constant, level) to another. Economies 2022, 10, x FOR PEER REVIEW 7 of 24 Table 1. Symbols of Stock-Flow Diagram in Powersim Professional 10© (Sterman 2000). Symbol Definition The symbol of STOCK is to declare variables with an accumulation derived from the previous value plus the difference between inflows and outflows. Stock(t) ==   (Inflow(s) − Outflow(s))ds + Stock(t0) d(Stock)/dt = Inflow(t) − Outflow(t). The symbol of RATE states the formulation of the amount of the stock inflow and outflow in the system in a certain time unit. For example, the rate of loan market = $100/month. The symbol of AUXILIARY or AUX is used to formulate the equation rate by defining the determining factors of the rate equation separately. Additional equations are substituted for each other and several separate rate equations. For example, Aux = Aux B x Constant The symbol of CONSTANT is a function of a certain number, the input for the auxiliary or equation rate in the model, its value remains in the simulation period. It is used to simulate management policies, such as loan interest rate income and saving interest rate expense policies. The symbol of ARROW indicates the flow of information from one variable (auxiliary, stock, constant, level) to another. The symbol of GRAPH contains certain parameter functions to explain other parameters/quantities. Stock-flow diagrams are developed based on bank financial statement accounts, including the financial assets account group (Figure 3), liabilities account group (Figure 4), and equity and statement of profit or loss (Figure 5). The formulation used to calculate the changes in the account is given in Appendix A. The financial assets account diagram shows the cycle of investment transactions to earn interest income and maintain liquidity. In this situation, the bank prioritizes its liquid assets for investment in performing loans and generating high returns, although it is necessary to anticipate the risk of loan default. A defaulted loan then could be restructured and controlled as a restructured loan, while a defaulted loan that could not be restructured could then be administered as an NPL. Since eliminating NPL write-off reduces capital, the bank forms a loan lost provision (LLP). The NPL ratio is the early warning of loan risk, which must not exceed 5%. The bank maintains adequate levels of liquidity in the form of liquid assets and securities. The total balance between the two financial assets can meet the transaction payment needs for the next 1 month. Investment in securities generates interest income. In the research design, the impact of changes in loan interest rates on additional loans is explained in the function of the interest rate loan to its delayed effect ratio variable in the additional loan formula in Table A1. If the annual interest rate of loan is increased, it will have a negative impact on additional loans. The stock-flow diagram of liabilities illustrates the transaction flow of the third-party fund (TPF) originating from the addition and withdrawal of TPF funds from savers, as Rate of loan market AuxAux B Constant_1 C onstant_1 Target loan to deposit ratio LDR assumption The symbol of GRAPH contains certain parameter functions to explain other parameters/quantities. Stock-flow diagrams are developed based on bank financial statement accounts, including the financial assets account group (Figure 3), liabilities account group (Figure 4), and equity and statement of profit or loss (Figure 5). The formulation used to calculate the changes in the account is given in Appendix A. Economies 2022, 10, x FOR PEER REVIEW 8 of 24 presented in Figure 4. An additional TPF is determined based on the national savings fund growth and the bank’s market share. After deducing the allocation of statutory reserves requirements at Bank Indonesia, the remaining TPF funds became available funds for new loans to debtors. Furthermore, for deposit growth, we use the variable size of bank assets. This is based on the Indonesia Banking Survey 2017 conducted by PWC. It can be seen that the amount of bank assets has a strong correlation with additional third-party funds. Furthermore, additional loans at Bank BRI and Bank Mandiri are encouraged because they have more extensive networks and access to customers. Therefore, in this study the effect of market share from assets is used as a reinforcing variable for additional third-party funds. Moreover, most of the third-party funds of Bank BRI and Bank Mandiri come from the government institutions and state-owned companies which are not sensitive to the amount of interest rates on deposit. The explanation is shown in the Figure 4. Figure 3. Stock-flow Diagram of Financial Assets (Source: Author). Financial Assets Liquid Asset Model Securities Performing loan Sell MS Loan payment NPL write off Loan maturity NPL NPL Restructuring Loan Model Additional loan Additional NPL Initial of Securities Performing loan correction NPL ratio NPL Correction Target NPL Restructuring Seasonal maturity Reserve of Loan Impairment Model Loan impairment Expected liquid asset Loan Model Loan to deposit ratio LDR LDR to Securities Ratio Indicated to buy MS Indicated to sell MS Buy MS Liquid Asset Model Liquid asset to financing ratio correction Liquid asset to financing ratio Third party fund TPF Model Borrowing Model Time to liquid asset correction Adjustment of prior year transaction Adjustment of loan impairment Initial of Loan Losses Provision Gain of value Irreguler policy of investment Loan write off accumulation NPL write off paid Delay time paid NPL write off Payment fraction of write of loan Loan Model Comprehensi ve income rate Restructurized loan Increased of restructurized loan Restructurized loan in pandemic Restructurized loan payment Rest Loan to NPL Restructurized loan LDR to MS Ratio Sell Target liquid asset to financing ratio graph Target of loan impairment rate Time to change impairment Change Impairment Rate Adjustment delay impairment rate Change Impairment Rate delay Impairment Rate Credit Impact in Covid situation Third party fund TPF Model Securities Ratio interest rate loan to its delayed effect Max Liquid Asset oufflow Liquid Asset outflow Maximum Sell of Marketable Securities NPL write off NPL write off rate Maximum loan impairment available Coverage impairment Impairment outflow Adjustment for impairment rate Policy of loan impairment rate input Policy of loan impairment rate LCR Realized NPL Rate average by sectors Expected performing loan Decrease on total loan cumulative NPL Corrective Action NPL Max Liquid Asset oufflow Cash in flow Cash out flow Figure 3. Stock-flow Diagram of Financial Assets (Source: Author). The financial assets account diagram shows the cycle of investment transactions to earn interest income and maintain liquidity. In this situation, the bank prioritizes its liquid assets for investment in performing loans and generating high returns, although it is necessary to anticipate the risk of loan default. A defaulted loan then could be restructured and controlled as a restructured loan, while a defaulted loan that could not be restructured could then be administered as an NPL. Since eliminating NPL write-off reduces capital, the bank forms a loan lost provision (LLP). The NPL ratio is the early warning of loan risk, which must not exceed 5%. The bank maintains adequate levels of liquidity in the form of liquid assets and securities. The total balance between the two financial assets can meet the transaction payment needs for the next 1 month. Investment in securities generates interest income. Economies 2022,10, 6 8 of 21 Economies 2022, 10, x FOR PEER REVIEW 9 of 24 Figure 4. Stock-flow Diagram of Liabilities (Source: Author). The stock-flow diagram of profit, loss, and equity in Figure 5 is developed through the accounts that make up income and expenses, as well as other equity transactions. Income is calculated before and after tax, which regularly changes following the banking business model. However, there some irregularly transactions that change the equity, including buy-back of stock, adjustments of expenses for transactions in previous years, employment benefit adjustments, and gains or losses of financial assets. The indicators of bank solvency risk that were analyzed in this paper are CAR and Z-Score. Liabilities Third party fund TPF Model Withdrawal TPF Time of TPF Withdrawal BNI TPF 2020-2022 Additional TPF TPF National model TPF national growth rate Third Party Fund National Available fund for loan Seasonal TPF Seasonal withdrawal Central bank reserve rate Multiplier for TPF National Growth Rate Market Share Normal TFP Deposit in Central Bank TPF Bank Growth Expected loan to deposit ratio LDR TPF National growth projection Effect of market share from asset TPF National growth projection input TPF National growth projection assumption Market Share TPF of Bank Market Share TPF of Bank assumption Expected loan to deposit ratio LDR assumption Figure 4. Stock-flow Diagram of Liabilities (Source: Author). In the research design, the impact of changes in loan interest rates on additional loans is explained in the function of the interest rate loan to its delayed effect ratio variable in the additional loan formula in Table A1. If the annual interest rate of loan is increased, it will have a negative impact on additional loans. The stock-flow diagram of liabilities illustrates the transaction flow of the third-party fund (TPF) originating from the addition and withdrawal of TPF funds from savers, as presented in Figure 4. An additional TPF is determined based on the national savings fund growth and the bank’s market share. After deducing the allocation of statutory reserves requirements at Bank Indonesia, the remaining TPF funds became available funds for new loans to debtors. Furthermore, for deposit growth, we use the variable size of bank assets. This is based on the Indonesia Banking Survey 2017 conducted by PWC. It can be seen that the amount of bank assets has a strong correlation with additional third-party funds. Furthermore, additional loans at Bank BRI and Bank Mandiri are encouraged because they have more extensive networks and access to customers. Therefore, in this study the effect of market share from assets is used as a reinforcing variable for additional third-party funds. Moreover, most of the third-party funds of Bank BRI and Bank Mandiri come from the government institutions and state-owned companies which are not sensitive to the amount of interest rates on deposit. The explanation is shown in the Figure 4. The stock-flow diagram of profit, loss, and equity in Figure 5is developed through the accounts that make up income and expenses, as well as other equity transactions. Income is calculated before and after tax, which regularly changes following the banking business model. However, there some irregularly transactions that change the equity, including buy-back of stock, adjustments of expenses for transactions in previous years, employment benefit adjustments, and gains or losses of financial assets. The indicators of bank solvency risk that were analyzed in this paper are CAR and Z-Score. Economies 2022,10, 6 15 of 21 Economies 2022, 10, x FOR PEER REVIEW 18 of 24 Figure 11. Early Action Simulation of BMRI. 4.4. Discussion Several important points should be discussed regarding the simulation of the loan restructuring policy revocation that began in March 2020. The first is to evaluate the potential of the default on the restructured loans. This evaluation allows banks to control the collectability of restructured loans to avoid them deteriorating or becoming NPLs (Bauer et al. 2021). However, evaluation should be monitored through the bank guidance to the debtors with the potentially non-performing loans. This could be implemented by visiting the debtors and their businesses, or conducting some on-site monitoring, to analyze the respected debtors’ ability to pay. There should be some concerns from banks regarding the provision of advice to debtors regarding business management to maintain and improve the payment capacity of the restructured loans. Therefore, loan monitoring by banks suppresses non-performing loans and improves financial performance (Duong et al. 2020; Hidayat et al. 2021). In a normal situation (not a pandemic), the increase in loan interest rates and a decrease in deposit rates will greatly depend on the level of price elasticity of each product. In this case, if the loan interest rate decreases, the demand for loan from the prospective debtors will increase, and if the deposit interest rate decreases, deposit placement activities will decrease. The COVID-19 condition has led to government intervention in handling the situation through restrictions on community activities. This is the main reason that causes less than optimal new loan growth or in other words, although loan interest rates decreases, the demand of prospective debtors obtaining the new loan does not increase. This is because people still in doubt of their ability to repay loans. However, the design of this study has considered the factors that influence the rate of loan growth (additional loan) with the following explanation: • The increase in new loans is not only influenced by loan interest rates, which tend to decline during the pandemic, but is greatly influenced by the COVID-19 condition with the level of public and business trust as potential debtors being quite low due to the tightening economic activities and doubts about their ability to loan repayment. • The increase in new loans is influenced by the level of bank liquidity, which was quite abundant during a pandemic. However, with the level of trust from the public and business that had not recovered as well as the economic activity that had not yet recovered, the bank could not carry out the new loan growth optimally. 23.0% 24.0% 25.0% 26.0% 27.0% 28.0% 1 5 10 15 20 25 30 35 40 41 42 43 44 45 46 47 48 CAR BMRI Baseline New Loan Interest Management Cost Management Combined Policy Default rate 20% 250 260 270 280 290 300 310 1 5 10 15 20 25 30 35 40 41 42 43 44 45 46 47 48 Z-Score BMRI Baseline New Loan Interest Management Cost Management Combined Policy Default rate 20% Figure 11. Early Action Simulation of BMRI. At BBRI, the policy simulation of interest rate management, new loan policy, the policy of a decrease in operating expense, and combined policy scenarios resulted in higher CAR and Z-Score levels than the position at default rate 20% (Figure 10). Similarly, the Z-Score for the four policy scenario options at BBRI has the same pattern, with changes in CAR compared to the baseline level. The combined policy pushed the CAR and Z-Score ratio of BBRI from 26.46% and 689 to 28.78% and 757, or increased by 8.76% and 9.9%, respectively, above the baseline position at the 48 th month. The policies that resulted in the highest CAR and Z-Scores, after the combined policy, are the policy of decreasing operating expense, interest management policy, and new loan policy. The combined policy of increased revenue and saved costs is the best option to effectively boost CAR and Z-Score in BBRI under the recovering economic conditions caused by the COVID-19 pandemic. For BMRI, based on the two parameters of CAR and Z-Score on Figure 11, the combined policy causes the highest changes in CAR and Z-Score values, followed by the policy of decreasing operating expenses, interest management policy, and new loan policy. Moreover, the combined policies strengthen the solvency of BMRI compared to other policies. Therefore, the combination of policies to increase revenue and save costs is the best option to effectively boost CAR in BMRI under abnormal economic conditions, due to the impact of the pandemic outbreak. The combined policy also has a high sensitivity level to boost BMRI’s CAR and Z-Score by 20.23% and 24.56% from the baseline position for the 48th period. In contrast, policies with low leverage in the current economic conditions are new loan policies. This is because the demand for new loans is still low, due to the COVID-19 pandemic. 4.4. Discussion Several important points should be discussed regarding the simulation of the loan restructuring policy revocation that began in March 2020. The first is to evaluate the potential of the default on the restructured loans. This evaluation allows banks to control the collectability of restructured loans to avoid them deteriorating or becoming NPLs (Bauer et al. 2021). However, evaluation should be monitored through the bank guidance to the debtors with the potentially non-performing loans. This could be implemented by visiting the debtors and their businesses, or conducting some on-site monitoring, to analyze the respected debtors’ ability to pay. There should be some concerns from banks regarding the provision of advice to debtors regarding business management to maintain and improve the payment capacity of the restructured loans. Therefore, loan monitoring by Economies 2022,10, 6 16 of 21 banks suppresses non-performing loans and improves financial performance (Duong et al. 2020;Hidayat et al. 2021). In a normal situation (not a pandemic), the increase in loan interest rates and a decrease in deposit rates will greatly depend on the level of price elasticity of each product. In this case, if the loan interest rate decreases, the demand for loan from the prospective debtors will increase, and if the deposit interest rate decreases, deposit placement activities will decrease. The COVID-19 condition has led to government intervention in handling the situation through restrictions on community activities. This is the main reason that causes less than optimal new loan growth or in other words, although loan interest rates decreases, the demand of prospective debtors obtaining the new loan does not increase. This is because people still in doubt of their ability to repay loans. However, the design of this study has considered the factors that influence the rate of loan growth (additional loan) with the following explanation: • The increase in new loans is not only influenced by loan interest rates, which tend to decline during the pandemic, but is greatly influenced by the COVID-19 condition with the level of public and business trust as potential debtors being quite low due to the tightening economic activities and doubts about their ability to loan repayment. • The increase in new loans is influenced by the level of bank liquidity, which was quite abundant during a pandemic. However, with the level of trust from the public and business that had not recovered as well as the economic activity that had not yet recovered, the bank could not carry out the new loan growth optimally. Taking into account the simulation results, it can be conveyed that in the COVID19 period there are conditions, namely: abundant bank liquidity, declining loan interest rates and declining deposit interest rates. This can then be adjusted to post-COVID-19 conditions, where the loan interest rate can be increased, namely from a declining interest rate to the original interest rate, in line with increasing public and business confidence in the conditions of post COVID-19 and the recovery in economic activity. The simulation results also show that the banks have a fairly good level of LLP and CAR to absorb the solvency risk. Therefore, although there is an increase in NPL, this does not have a significant impact on CAR and Z-Score (Agenor and Zilberman 2015). A high level of LLP reduces the bank solvency risk but increases the loan impairment losses, and suppresses profit and capital growth. Therefore, banks face a trade-off between solvency and profitability (Zheng et al. 2019). There is a challenge in the simulation of policy scenarios by banks to prepare better action plans in response to worsening NPLs when the restructuring policy is revoked. This allows for the impact of the policy on the bank solvency to be known before it is implemented (Paiva et al. 2020;Asadollahi et al. 2021). Moreover, it enables banks to prepare for the existing policy scenario options (Morecroft 2015). The model has been developed into a self-simulation facility that could be implemented by the bank management to enable banks to deal with the potential solvency risks using their own data. Banks could calculate the level of solvency and other risks as a consequence of the COVID-19 pandemic through a simple interface that generates early warning and early action. Figure 12 shows the Bank Solvency Risk Self-Simulation Interface developed in this research. Three parts were prepared in the application. The first part shows data input instruments for potential defaults on restructured loans. The second part shows input instruments for bank policy parameters to manage solvency risk through four policy scenarios. The third part displays the simulation results of bank solvency indicators, including NPL Ratio, CAR, ROA and Z-Score. Economies 2022,10, 6 17 of 21 Economies 2022, 10, x FOR PEER REVIEW 19 of 24 Taking into account the simulation results, it can be conveyed that in the COVID-19 period there are conditions, namely: abundant bank liquidity, declining loan interest rates and declining deposit interest rates. This can then be adjusted to post-COVID-19 conditions, where the loan interest rate can be increased, namely from a declining interest rate to the original interest rate, in line with increasing public and business confidence in the conditions of post COVID-19 and the recovery in economic activity. The simulation results also show that the banks have a fairly good level of LLP and CAR to absorb the solvency risk. Therefore, although there is an increase in NPL, this does not have a significant impact on CAR and Z-Score (Agenor and Zilberman 2015). A high level of LLP reduces the bank solvency risk but increases the loan impairment losses, and suppresses profit and capital growth. Therefore, banks face a trade-off between solvency and profitability (Zheng et al. 2019). There is a challenge in the simulation of policy scenarios by banks to prepare better action plans in response to worsening NPLs when the restructuring policy is revoked. This allows for the impact of the policy on the bank solvency to be known before it is implemented (Paiva et al. 2020; Asadollahi et al. 2021). Moreover, it enables banks to prepare for the existing policy scenario options (Morecroft 2015). The model has been developed into a self-simulation facility that could be implemented by the bank management to enable banks to deal with the potential solvency risks using their own data. Banks could calculate the level of solvency and other risks as a consequence of the COVID-19 pandemic through a simple interface that generates early warning and early action. Figure 12 shows the Bank Solvency Risk Self-Simulation Interface developed in this research. Three parts were prepared in the application. The first part shows data input instruments for potential defaults on restructured loans. The second part shows input instruments for bank policy parameters to manage solvency risk through four policy scenarios. The third part displays the simulation results of bank solvency indicators, including NPL Ratio, CAR, ROA and Z-Score. Figure 12. Bank Solvency Risk Self-Simulation Interface Developed in the Research. The implementation of a self-simulation facility requires bank officers to provide input data or assumptions regarding the estimated percentage of restructured loan defaults and the potential non-performing loans. These inputs produce an early warning about solvency risk. Furthermore, bank management could prepare inputs for each solvency Figure 12. Bank Solvency Risk Self-Simulation Interface Developed in the Research. The implementation of a self-simulation facility requires bank officers to provide input data or assumptions regarding the estimated percentage of restructured loan defaults and the potential non-performing loans. These inputs produce an early warning about solvency risk. Furthermore, bank management could prepare inputs for each solvency risk management policy provided by applying the required parameters. After obtaining the input data for the four policies, the application generates new conditions, providing the policy by displaying the NPL, CAR, and Z-Score as early actions for the bank management. Based on the early warning and early action information, using self-simulation, the bank could implement the policy that produces the highest CAR and Z-Score at the first opportunity. The earlier the best policy is implemented, the better prepared banks become to face the loan restructuring policy revocation that will take place in March 2023. After the policy is implemented, and during the lead up to the policy revocation, banks could still perform self-simulation. However, they should consider the current conditions of loan performance through the NPL ratio. They could also continuously monitor or update bank’s readiness to face solvency risk after the revocation of the OJK loan restructuring policy. 5. Conclusions This article develops an early warning system model for bank bankruptcy risk and a simulation of management policy response. The development of the model with a system dynamics methodology adds theoretical knowledge about the flow of the bank financial transactions that caused the bank bankruptcy risk, and then produces simulations and management policy choices to reduce the potential risks. By identifying the flow in the system dynamics, bank management can develop appropriate risk mitigation policy choices according to the symptoms that have arisen. System dynamics methodology is used to build models and perform early warning and early action simulations to help banks identify solvency risks earlier and design policies and strategies to mitigate the risk. Early warnings regarding bank solvency risk could be identified through simulations and early actions could be taken to overcome the potential risks and problems after the restructuring policy revocation. To deliver the robust model of solvency risk, this research used the publicly accessed data from two banks and analyzed some financial ratios based on the financial statements. To enable banks to use their data, Economies 2022,10, 6 18 of 21 this research provided a self-simulation facility, so that bank management could obtain a more accurate solvency risk prediction. It is suggested that future research could consider not only solvency risk but also the liquidity and market risks. The limitations of the analysis that carried out in this paper, among others, that it only discussed the potential loan risk as the beginning of the bankruptcy risk and has not discussed the other causes of the loan risks such as the debtor moral hazard, the weak analysis of creditworthiness, the declining of the payment capacity of the debtors due to the pandemic COVID-19 and the specific economic sectors that experienced the high level of loan risks. It is suggested that future research could consider not only the loan and solvency risk but also the liquidity and market risks in the model. Further, the upcoming research is suggested to discuss the causes of the bankruptcy risks that have not been analyzed in this article. Author Contributions: T.H. carried out the conception, design of the work, and drafting the article. D.M. reviewed the literature, methodology and gave final approval of the article. S.R.N. reviewed data analysis and interpretation. F.A. worked on data collection and the modelling. M.A.N.S. worked on the simulation. All authors have read and agreed to the published version of the manuscript. Funding: The research presented in this scientific article was self-funded. Data Availability Statement: The data presented in this study are available on request from the corresponding author. Conflicts of Interest: The authors declare no conflict of interest. Appendix A Table A1. Formula of Variable in Stock-flow Diagram. Variabel Formula Performing Loan Initial Performing Loan + (Loan Payment − Additional Loan) + (NPL Restructuring − Additional NPL) − Performing Loan to Restructured Loan Loan Payment (Loan Maturity ∗Performing Loan) ∗(1 + Seasonal maturity) Additional Loan (MIN (Max Liquid Asset Outflow, If ((Marketable Securities Model + Liquid Asset Model)>Expected liquid Asset, Performing loan correction, 0<<IDR Million/Month>>) ∗ (Credit Impact in COVID situation/Ratio interest rate loan to its delayed effect))) Additional NPL Performing loan ∗Realized NPL Rate Average by Sectors NPL Restructuring (MAX (Target NPL Restructuring, (Target NPL Restructuring + NPL Correction))/Time to Restructuring) ∗ Loan Model NPL Initial NPL + (Additional NPL −NPL Restructuring) + Rest Loan to NPL −NPL Write off Rest Loan to NPL MIN (Maximum Restructurized Loan Outflow, decrease on total loan cumulative) NPL write off (Non performing loan NPL ∗NPL write-off rate) Restructurized Loan Initial Restructurized Loan + Performing Loan to Restructured Loan − Rest Loan to NPL − Restructurized Loan Payment Restructurized Loan Payment MAX(0<<IDR Million/month>>,MIN(Maximum Restructurized Loan Outflow, Loan Maturity/Multiplier Maturity Rate in Restructurized Implemented ∗Restructurized Loan)) Liquid Asset Model Initial Liquid Asset + Loan Payment −Additional Loan + Restructurized Loan Payment + Cash Inflow − Cash Outflow −Buy Marketable Securities (MS) −Sell MS Cash Inflow Interest income + Operating income + New borrowing + Additional TPF + NPL Write Off Paid + Fixed Asset Disposals Economies 2022,10, 6 19 of 21 Table A1. Cont. Variabel Formula Cash Inflow (Operating expense − Depreciation)+ Additions of Fixed Asset + Borrowing Payment + Withdrawal TPF + Interest Expense + Dividen Payment + Tax expense + Buyback Stock Sell MS MIN (Maximum Sell of Marketable Securities, Indicated to sell MS ∗LDR to MS Ratio Sell) Buy MS MIN ((Indicated to buy MS ∗LDR to MS Ratio Buy)∗1+Irreguler policy of investment, Max Liquid Asset oufflow) Securities Initial Securities + Buy MS −Sell MS + Gain of Value Reserve of Loan Impairment Model Initial of Reserve of Loan Impairment Model + Loan Impairment −Impairment Outflow + Adjustment of Loan Impairment Loan Impairment (Loan Model ∗Impairment Rate) Impairment Outflow (MIN (Maximum Loan Impairment Available, NPL write off)) NPL Ratio Non performing loan NPL/Loan Model Loan Loss Provision (LLP) Reserve of Loan Impairment Model/Non performing loan NPL Third Party Fund (TPF) Initial TPF + Additional TPF −Withdrawal TPF Additional TPF TPF national growth rate ∗ Seasonal TPF ∗ (Market Share Normal TPF ∗ Effect of market share from asset) Withdrawal TPF (Third party fund TPF Model/Time of TPF Withdrawal) ∗(Seasonal withdrawal) Equity Model Initial Equity + Net Profit −Equity Adjustment −Dividend Payment −Buy back of Stock Net Profit (Interest Income −Interest Expense + Operating Income −Operating Expense −Loan Impairment -Tax expense) Equity Adjustment Adjustment of prior year transaction + Employment Benefit Adjustment Dividend Payment Profit After Tax ∗Dividend Payout Ratio Buy Back of Stock Buy back decision or event Capital Adequacy Ratio (CAR) Equity Model/Risk Weighted Asset ∗100<<%>> ROA Profit After Tax/Asset Model Z-Score (ROA + (Equity Model/Asset Model))/Standard Deviation of ROA References Agenor, Pierre-Richard, and Roy Zilberman. 2015. Loan Loss Provisioning Rules, Procyclicality, and Financial Volatility. Journal of Banking and Finance 61: 301–15. [CrossRef] Aksu, ˙ Ibrahim, and Mehmet Tursun. 2021. Analysis of Responsibility Centers Performance in Businesses By System Dynamics Method. Journal of Accounting and Taxation Studies 14: 949–66. [CrossRef] Al-Kharusi, Sami, and Sree Rama Murthy. 2020. Financial stability of GCC banks in the COVID-19 Crisis: A simulation approach. The Journal of Asian Finance, Economics, and Business 7: 337–44. [CrossRef] Asadollahi, Seyyed Yahya, Ali Asghar Taherabadib, Farhad Shahvisic, and Farshid Kheirollahi. 2021. Modeling Liquidity Risk Management in Banking Using System Dynamics Approach. Advances in Mathematical Finance and Applications 6: 493–514. [CrossRef] Bala, Atkinson, Arshad Bilash Kanti, Fatimah Mohamed, and Kusairi Mohd Noh. 2017. System Dynamics: Modelling and Simulation. Singapore: Springer Science & Business Media. Bank Indonesia. 2021. Financial Stability Review No. 37. September. Available online: https://www.bi.go.id/id/publikasi/kajian/ Pages/KSK_3721.aspx# (accessed on 24 October 2021). Barua, Bipasha, and Soburna Barua. 2021. COVID-19 Implications for Banks: Evidence from an Emerging Economy. SN Business & Economics 1: 19. Economies 2022,10, 6 20 of 21 Bastana, Mahdi, Mohammad Bagheri Mazraeha, and Ali Mohammad Ahmadvand. 2016. Dynamics of Banking Soundness Based on CAMELS Rating System. Paper presented at the 34th International Conference of the System Dynamics Society, Delft, The Netherlands, July 17–21. Bauer, W. Andreas, R. Sean Craig, JoséGarrido, Kenneth H. Kang, Kashiwase Kenichiro, Sung Jin Kim, Yan Liu, and Sohrab Rafiq. 2021. Flattening the Insolvency Curve: Promoting Corporate Restructuring in Asia and the Pacific in the Post-C19 Recovery. IMF Working Paper 2021/016. Washington, DC: International Monetary Fund. Danielsson, Jon, Robert Macrae, Dimitri Vayanos, and Jean-Pierre dan Zigrand. 2020. We Shouldn’t Be Comparing the Coronavirus Crisis to 2008—This Is Why? Available online: https://www.weforum.org/agenda/2020/04/the-coronavirus-crisis-is-no-2008 (accessed on 10 October 2021). De Vany, Arthur S. 1984. Modeling the banking firm: Comment. Journal of Money, Credit and Banking 16: 603–9. [CrossRef] Donnellan, John, and Wanda Rutledge. 2016. Agency Theory in Banking—‘Lessons from the 2007–2010 Financial Crisis’. International Journal of Business and Applied Social Science 2: 1–12. Available online: https://ssrn.com/abstract=2940060 (accessed on 11 October 2021). Duggan, Jim. 2016. System Dynamics Modeling with R. Cham: Springer International Publishing Switzerland. Duong, Tam Thanh Nguyen, Hai Thanh Phan, Tien Ngoc Hoang, and Tien Thuy Thi Vo. 2020. The Effect of Financial Restructuring on the Overall Financial Performance of the Commercial Banks in Vietnam. The Journal of Asian Finance, Economics, and Business 7: 75–84. [CrossRef] García, Juan Martín. 2019. Common Mistakes in System Dynamics. Chicago: Independently Publishers, ISBN 9781791578787. Goodell, John W. 2020. COVID-19 and Finance: Agendas for Future Research. Finance Research Letters 35: 101512. [CrossRef] Gunnersen, Sverre Edvard. 2014. Predicting Corporate Failure Through a Combination of Intelligent Techniques. Ph.D. dissertation, Monash University, Melbourne, Australia. Hardiyanti, Siti Epa, and Lukmanul Hakim Aziz. 2021. The Case of COVID-19 Impact on the Level of Non-Performing Loans of Conventional Commercial Banks in Indonesia. Banks and Bank Systems 16: 62–68. [CrossRef] Henrard, Luc, and Ruben Olieslagers. 2004. Risk management of a financial conglomerate. In Forum Financier-Revue Bancaire et Financiaire Bank en Financiewezen. Brussels: Larcier, pp. 52–64. Hidayat, Taufiq, Dian Masyita, Sulaeman Rahman Nidar, Erie Febrian, and Fauzan Ahmad. 2021. The Effect of COVID-19 to Credit Risk and Capital Risk of State-Owned Bank in Indonesia: A System Dynamics Model. WSEAS Transactions on Business and Economics 18: 1121–36. [CrossRef] Islam, Tuahith, Christos Vasilopoulos, and Erik Pruyt. 2013. Stress-Testing Banks under Deep Uncertainty. Paper presented at 31st International Conference of the System Dynamics Society, Cambridge, MA, USA, July 21–25. Istiaq, Muhammad. 2015. Risk Management in Banks: Determination of Practices and Relationship with Performance. Ph.D. dissertation, The University of Bedfordshire, Luton, UK. Jensen, Michael C., and William H. Meckling. 1976. Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure. Journal of Financial Economics 3: 305–60. [CrossRef] Knowles, John, Ken Nicholls, and Chris Bloor. 2020. Outcomes from a COVID-19 stress test of New Zealand banks. Reserve Bank of New Zealand Bulletin 83: 1–12. Korzeb, Zbigniew, and Pawel Niedziółka. 2020. Resistance of commercial banks to the crisis caused by the COVID-19 pandemic: The case of Poland. Equilibrium. Quarterly Journal of Economics and Economic Policy 15: 205–34. [CrossRef] Kunc, Martin, Michael J. Mortenson, and Richard Vidgen. 2018. A computational literature review of the field of System Dynamics from 1974 to 2017. Journal of Simulation 12: 115–27. [CrossRef] Kuritzkes, Andrew, Til Schuermann, and Scott M. Weiner. 2003. Risk Measurement, Risk Management and Capital Adequacy in Financial Conglomerates. Brookings-Wharton Papers on Financial Services 2003: 141–93. [CrossRef] Lang, Jan Hannes, Tuomas A. Peltonen, and Peter Sarlin. 2018. A Framework for Early-warning Modelling with an Application to Banks. European Central Bank Working Paper Series, No. 2182/October 2018. Frankfurt: European Central Bank. Leaning, Jennifer. 2016. The Path to Last Resort: The Role of Early Warning & Early Action. Daedalus. Ethics, Technology & War 145: 101–12. Lepetit, Laetitia, Frank Strobel, and Thu Ha Tran. 2020. An Alternative Z-score Measure for Downside Bank Insolvency Risk. Applied Economics Letters 28: 137–42. [CrossRef] Mayes, David G., and Hanno Stremmel. 2012. The Effectiveness of Capital Adequacy Measures in Predicting Bank Distress. Financial Markets & Corporate Governance Conference. Available online: https://ssrn.com/abstract=2191861 (accessed on 10 October 2021). Morecroft, John D. W. 2015. Strategic Modeling and Business Dynamics: A Feedback Systems Approach, 2nd ed. Chichester: John Wiley and Sons Ltd. Oladimeji, Olufunke Olufunmi, Heather Keathley-Herring, and Jennifer A. Cross. 2020. System dynamics applications in performance measurement research: A systematic literature review. Journal of Productivity and Performance Management 69: 1541–78. [CrossRef] Paiva, Bernardo M. R., Fernando A. F. Ferreira, Elias G. Carayannis, Constantin Zopounidis, João J. M. Ferreira, Leandro F. Pereira, and Paulo J. V. L. Dias. 2020. Strategizing sustainability in the banking industry using fuzzy cognitive maps and system dynamics. International Journal of Sustainable Development & World Ecology 28: 93–108. [CrossRef] Economies 2022,10, 6 21 of 21 Pavlov, Oleg V., and Evangelos Katsamakas. 2021. COVID-19 and Financial Sustainability of Academic Institutions. Sustainability 13: 3903. [CrossRef] Petropoulos, Anastasios, Vassilis Siakoulis, Konstantinos P. Panousis, Theodoros Christophides, and Sotirios Chatzis. 2020. A Deep Learning Approach for Dynamic Balance Sheet Stress Testing. Available online: https://arxiv.org/abs/2009.11075 (accessed on 15 September 2021). Pierson, Kawika. 2020. Operationalizing Accounting Reporting in System Dynamics Models. Systems 8: 9. [CrossRef] Rahmi, Yulia, and Erman Sumirat. 2021. A Study of The Impact of Alma on Profitability During the COVID-19 Pandemic. International Journal of Business, Economics, and Law 24: 54–65. Rose, Peter S. 1992. Agency Theory and Entry Barriers in Banking. The Financial Review 27: 323–54. [CrossRef] Samorodov, Borys, Halyna Azarenkova, Olena Holovko, Kateryna Oriekhova, and Maksym Babenko. 2019. Financial stability management in banks: Strategy maps. Banks and Bank Systems 14: 10–21. [CrossRef] Sapiri, Hasimah, Jafri Zulkepli, Norazura Ahmad, Norhaslinda Zainal Abidin, and Nurul Nazihah Hawari. 2020. Introduction to System Dynamics Modelling and Vensim Software. Sintok: UUM Press. Schuermann, Til. 2014. Stress Testing Banks. International Journal of Forecasting 30: 717–28. [CrossRef] Siregar, Reza Y., Anton H. Gunawan, and Adhi N. Saputro. 2021. Impact of The COVID-19 Shock on Banking and Corporate Sector Vulnerabilities in Indonesia. Bulletin of Indonesian Economic Studies 57: 147–73. [CrossRef] Smith, Alan D. 2010. Agency Theory and The Financial Crisis from a Strategic Perspective. International Journal of Business Information Systems 5: 248–67. [CrossRef] Sterman, Jhon D. 2000. Business Dynamics: Systems Thinking and Modeling for a Complex World. Boston: McGraw-Hill. Wu, Xiaoyu. 2014. Banking Crisis Dynamics and Financial Risk Management. Doctor of. Philosophy dissertation, The City University of Hong Kong, Hong Kong, China. Zheng, Changjun, Shumaila Meer Perhiar, Naeem Gul Gilal, and Faheem Gul Gilal. 2019. Loan loss provision and risk-taking behavior of commercial banks in Pakistan: A dynamic GMM approach. Sustainability 11: 5209. [CrossRef]