Amadu-Jawara/Sierra-Leone-borrower-stress-model

A Python-based macroeconomic borrower-stress prediction model for Sierra Leone using economic indicators, engineered risk scoring, machine learning, and visual analytics. # Sierra Leone Borrower Stress Prediction Model ## BY AMADU JAWARA ## Project Overview This project develops a Python-based macroeconomic borrower-stress modelling framework for Sierra Leone. The model examines how difficult economic conditions may increase repayment pressure among borrowers. The analysis focuses on major macroeconomic stress factors, including: - Political and economic instability - Rising interest rates - Inflationary pressure - Rising unemployment - Public health and crisis-related economic shocks The project combines economic reasoning, data analysis, engineered risk scoring, machine learning, and visual analytics to produce a structured borrower-stress assessment. ## Project Motivation Borrower repayment behaviour is strongly influenced by the wider economic environment. In a fragile or unstable economy, households and businesses may face income shocks, rising borrowing costs, inflation, unemployment, and reduced business activity. For a country such as Sierra Leone, understanding borrower stress is important for: - Credit risk monitoring - Banking sector stability - Loan portfolio management - Early warning analysis - Financial inclusion planning - Policy and development decision-making This project was created as a portfolio project to demonstrate how Python and machine learning can be applied to real-world financial risk and macroeconomic analysis. ## Research Question How can macroeconomic indicators be used to estimate borrower stress in Sierra Leone under conditions of rising interest rates, unemployment, inflation, and economic instability? ## Objectives The objectives of this project are to: 1. Collect and organize relevant macroeconomic indicators for Sierra Leone. 2. Engineer a borrower stress score using selected risk indicators. 3. Classify periods into different borrower risk categories. 4. Train a machine learning model to predict borrower stress categories. 5. Identify the most important drivers of borrower stres …