Exploratory Analysis of Borrower Characteristics and Loan Conditions

EDA & SQL Based Project

1. Project Description

An exploratory data analysis project examining relationships between borrower characteristics and loan attributes using Python-based EDA and SQL queries.

This project was a collaboration with Isis Hassan and project repo can be found here

2. Dataset Description

• Data Source: Predicting Loan Payback from kagglehub
• Observations: Loan-level records representation(means inidividual loan)
• Key Variables:
  • Borrower characteristics: credit_score, annual_income, employment_status
  • Loan attributes: loan_amount, interest_rate, loan_purpose
• Notes & Quirks:
  • Dataset is synthetic but designed to resemble real-world loan data
  • Data quality checks were performed for missing values, duplicates, and incorrect data types
  • An identifier column was removed as it was not analytically relevant

3. Research Goal / Questions

The goal of this project is to explore how borrower characteristics relate to loan conditions.

Main research questions:

• How does credit score influence loan attributes such as interest rate, loan amount, and loan purpose?
• How does a borrower’s annual income relate to the size of the loan issued?
• How do loan characteristics (amount, interest rate, purpose) vary across different employment statuses?

4. Steps Taken

• Data Cleaning
  • Inspected structure, summary statistics, and data types
  • Checked for duplicate rows and irrelevant columns
  • Dropped columns not needed to answer research questions. Columns kept:
    • id: loan id, set as index
    • *Borrower characteristics: credit_score, annual_income, employment_status
    • Loan attributes: loan_amount, interest_rate, loan_purpose
• Exploratory Data Analysis (EDA)
  • Univariate analysis of numerical and categorical variables
  • Bivariate analysis to examine relationships between borrower and loan variables
• SQL Analysis
  • Translated EDA insights into descriptive SQL questions
  • Used SQL queries to validate and summarize observed patterns
• Visualization
  • Histograms, boxplots, scatter plots, and bar charts
  • Each visualization followed by concise interpretive commentary

5. Main Findings

• Borrowers with higher credit scores tend to receive lower interest rates.
• No meaningful correlation between other variables

6. Next Steps / Ideas for Improvement

• Extend the analysis with multivariate techniques.
• Build a predictive model for loan amount or interest rate.
• Incorporate additional borrower attributes (e.g., debt-to-income ratio).
• Validate findings using a real-world (non-synthetic) dataset.

7. Repository Structure

• Python File
• SQL Queries
• Presentation
• Project Plan