Modelling credit risk using system dynamics: The case of licensed credit reference bureaus in Kenya

Abstract

Credit reference bureaus (CRBs) have been operational in Kenya for many years owing to the large number of borrowers who fail to repay their loans. However, regulating how credit risk will be quantified by these CRBs is often based on standards and assumptions that are not practical to the real-world scenario. This study models credit risk to discover more effective and practical measures which relate to the borrowers and their operating environment. Data was collected from annual default reports from the Central bank of Kenya, CRBs and major financial institutions over a period of three years (2018, 2019, and 2020). The study also used focus group discussions to establish the key default factors and their baseline values. A sample of 29 participants was drawn from the population of CRB staff members who undertake the core functions of credit risk determination. Using the system dynamic modeling and simulation approach, the study identified faithful representations of default risk measurements. First, descriptive analysis was conducted using tabled summaries and bar charts and results identified customer income, issued loans and collateral amount as the most influential factors for credit risk. Explorative analysis applied causal loop diagrams (CLDs). Simulation analysis was then conducted after generating stock-and-flow diagrams and three important variables were identified, i.e., loan repayment, performing loans, and credit risk. The information gained from this study will benefit the government, the Central bank of Kenya (CBK), research scholars and other major financial institutions around the country.