Lessons Learned on Credit Risk Grading Model Designing

Tahrima Faruq, Tanjila Mahboob, Mohiuddin Rasti Morshed and Syed Ehsan Quadir

The purpose of this paper is to share United Finance Limited’s1 experience of overcoming pitfalls that impact the effectiveness of Credit Risk Grading Model2. It was observed from a sample of 1609 clients of year 2011 and 2012 that the model is not estimating credit risk precisely. Further review of the model reveals that the construction of attributes of the model gives scopes to include biased information emphasizing the significance of formatting attributes properly. Besides, the logic of adopting multi-model approach in designing the model is under scrutiny which deters presenting universal model. These learning imply that to make the model effective in decision making, continuous development is needed.