An exploratory analysis of whether lenders are accurately pricing risk across loan grades and borrower intent.
This project explores whether lenders are accurately pricing loan risk. Specifically, whether the interest rates charged to borrowers reflect the actual risk of non-payment.
Loan grades (A–G) are assigned to reflect borrower risk. This analysis compares default rates and average interest rates across grades to assess whether the pricing ladder matches the risk ladder.
Examines whether the purpose of a loan (medical, education, personal, etc.) influences default rates and whether lenders adjust interest rates accordingly.
A heatmap combining both dimensions to identify which grade and intent combinations carry the highest default risk.
Loan grades are a reasonable risk signal, but loan intent is largely ignored in pricing. Riskier loan purposes carry similar interest rates to safer ones despite meaningfully higher default rates.
Credit Risk Dataset via Kaggle - 32,000+ loan records with borrower demographics, loan details, and default outcomes.
Python, Pandas, Seaborn, Matplotlib