This repository contains a pricing model for gas storage contracts, a loan default prediction model, and a FICO score quantization system for credit ratings.
Calculates contract value based on injection/withdrawal dates, storage limits, and market prices.
Trained a Random Forest Classifier on borrower data to predict the Probability of Default (PD). Computes Expected Loss using a fixed 10% recovery rate.
Divides FICO scores into quantile-based buckets. Generates a rating map where lower ratings correspond to better credit scores.
This project applies machine learning and financial modeling techniques to enhance decision-making in gas storage, loan risk assessment, and credit rating systems.