Classification-Based Credit Risk Analysis: The Case of Lending Club

In this paper, we perform a data-driven credit risk analysis, using the data from loan applicants made to a company named Lending Club. The approach adopted in the work required the use of exploratory data analysis and machine learning classification algorithms, namely Logistic Regression and Random Forest. We further made use of the calculated probability of default in order to design a credit derivative (Credit Default Swap) in order to achieve hedging against an event of credit default. The results on the test set are presented using various performance measures.