Financial risk modelling in vehicle credit portfolio

Luxury cars are a segment of vehicles which are usually bought by people with a higher purchasing power. Still, majority of people make this luxury investment through vehicle finance services. The people from this segment tend to have a good credit record and thus are granted credit by vehicle finance service providers. Despite the good credit record and high purchasing power, a certain amount of risk is associated with these credit portfolios. This study deals with the analysis of a data set comprising of opulent vehicle credit portfolios characterized by relevant variables. It aims at assessing the risk associated with these portfolios and finally presents a predictive model which highlights the important variables and depicts the combination of those variables that classify a client under defaulter or non-defaulter. The study starts with the use of conventional statistical techniques and subsequently presents machine learning approach using three different decision tree classifiers.