Designing Citizen Business Loan Model to Reduce Non-Performing Loan: An Agent-based Modeling and Simulation Approach in Regional Development

Abstract

Citizen Business Loan (CBL) constitutes a program poverty alleviation based on economic empowerment of small and medium enterprise. This study focuses on implementation of CBL at Regional Development Bank branch X. The problem is the existing of interdependencies between CBL’s implements (Bank) and the uncertainty of debtor’s capability in returning the credit. The impact of this circumstance is non-performing loan (NPL) becomes relatively high (22%). The ultimate objective is to minimize NPL by designing the model based on the agent that can represent the problem through a simulation using agent-based modeling and simulation (ABMS). The model is considered by managing the probability of the debtor to pay or not based on 5 C categories, they are: character, capacity, capital, condition, and collateral that inherent to each debtor. There are two improvement scenarios proposed in this model. The first scenario only involves the first category of debtor in simulation. The result of this scenario is NPL value as 0%. The second scenario includes the first and second of debtor’s category in simulation and resulting NPL value between 4.6% and 11.4%.