top of page


Home Credit - Credit Risk Modelling

Johann Thiel, Pushkar Sathe


Project Description
The Home Credit Group (HCG) is looking to develop a model with which they can determine the amount of risk associated with a particular loan for individuals with little or no credit history. Such models are referred to as 'scorecards'. The main concern is that scorecards may need to be updated constantly based on customer behavior over time, which means that they may lack any kind of long-term predictive ability. The goal is to develop a scorecard that can predict credit defaults prior to their occurrence and be 'stable' over time. The data provided encompass a wide variety of categories, but no credit scores.

The HCG is the primary stakeholder in this project. They, along with other lending institutions, are clearly interested in making loans for which there is a low chance of default. Borrowers can also benefit from this analysis as it may allow a lender to offer loans or loans with better terms to individuals lacking an extensive credit history.

Screen Shot 2022-06-03 at 11.31.35 AM.png
github URL
bottom of page