---

Credit Risk Analysis for Loan Applicants 📊🏦

A complete SQL + Power BI project focused on analyzing loan applicant risk profiles and approval patterns using a real-world style BFSI dataset.

This project demonstrates data cleaning, preprocessing, feature engineering, exploratory analysis, and dashboard reporting — all done using MySQL and Power BI Service.

It highlights strong applied skills in BFSI analytics, SQL EDA, and data visualization.

---

📌 Table of Contents

• Project Overview
• Goals
• Tech Stack
• Repository Structure
• Dataset Description
• Data Cleaning & SQL Preprocessing
• Feature Engineering
• Exploratory Data Analysis (SQL Insights)
• Power BI Dashboard
• Key Insights
• How to Reproduce
• Project Documentation
• Resume Value
• Author

---

Project Overview 🚀

Banks and financial institutions must evaluate whether a loan applicant is likely to be approved or rejected. This project performs a complete risk analysis workflow:

✔ SQL-based cleaning & preprocessing ✔ Feature engineering (DTI, income bands, total income) ✔ Exploratory SQL analysis ✔ Power BI dashboards & KPIs ✔ Findings relevant to credit underwriting

This project is a perfect demonstration of BFSI domain knowledge, data analytics, and business reporting.

---

Goals 🎯

The primary objectives are:

• Clean and transform raw loan application data using SQL.
• Engineer new risk-related features: Total Income, Debt-to-Income (DTI), Income Bands, etc.
• Understand approval patterns through SQL-based EDA.
• Build Power BI dashboards for business insights.
• Provide a replicable and professional BFSI analytics pipeline.

---

Tech Stack 🧰

🗄 SQL (MySQL Workbench)

• Data cleaning
• Data validation
• Feature engineering
• Exploratory analysis

📊 Power BI Service

• Interactive dashboards
• KPI cards
• Segmentation analysis
• Risk visualization

📁 File Formats

• CSV (raw + cleaned data)
• SQL script (project_credit_risk_analysis.sql)
• Power BI exported PDF

---

Repository Structure 📂

Credit-Risk-Analysis/
│
├── Data/
│   ├── loan.csv
│   ├── loan_cleaned_data.csv
│
├── SQL_Queries/
│   └── project_credit_risk_analysis.sql
│
├── PowerBI/
│   └── credit_risk_analysis_HM.pdf
│
├── Documentation/
│   └── Credit_Risk_Analysis_Project_Documentation.doc
│
└── README.md

---

Dataset Description 📙

The dataset includes demographic, financial, and loan-related information such as:

• ApplicantIncome
• CoapplicantIncome
• LoanAmount
• Loan_Amount_Term
• Credit_History
• Property_Area
• Dependents
• Education
• Self_Employed
• Loan_Status (Approved / Rejected)

These variables are used to evaluate the risk of default and loan approval likelihood.

---

Data Cleaning & SQL Preprocessing 🧹

Performed completely in MySQL Workbench, documented in:

📄 SQL_Queries/project_credit_risk_analysis.sql

Key steps:

• Handling missing values (LoanAmount, Loan_Term, Credit_History)
• Cleaning categorical fields (Dependents = "3+" → 3)
• Standardizing strings
• Converting numeric text to integer/float
• Extracting engineered fields for risk analysis

---

Feature Engineering ✨

The following new fields were created using SQL:

➤ Total_Income

ApplicantIncome + CoapplicantIncome

➤ DTI (Debt-to-Income Ratio)

(LoanAmount * 1000) / Total_Income

➤ Income Bands

• Low
• Medium
• High

➤ DTI Bands

• Low Risk
• Medium Risk
• High Risk

These transformations significantly improved insight into credit behavior.

---

Exploratory Data Analysis (SQL Insights) 🔍

Examples of SQL insights performed:

✔ Overall Approval Rate

SELECT 
  COUNT(*) AS total,
  SUM(Loan_Status = 'Y') AS approved,
  ROUND(SUM(Loan_Status = 'Y') / COUNT(*) * 100, 2) AS approval_rate
FROM loan_cleaned;

✔ Approval by Credit History

SELECT Credit_History,
       COUNT(*) AS total,
       ROUND(SUM(Loan_Status='Y') / COUNT(*) * 100, 2) AS approval_pct
FROM loan_cleaned
GROUP BY Credit_History;

✔ Income Band Analysis

Applicants with higher income bands consistently show better approval chances.

✔ DTI Risk Segmentation

Lower DTI → Higher approval Higher DTI → Higher risk → Lower approval

---

Power BI Dashboard 📊

The full visualization is available here:

👉 Credit Risk Dashboard (PDF)

Dashboard Highlights:

• Overall approval rate

• Average loan amount

• Approval by:

  • Credit history
  • Income band
  • Property area
  • Dependents

• DTI distribution

• Drill-down for individual applicants

---

Key Insights 💡

🔹 Credit history is the strongest predictor of approval

Applicants with Credit_History = 1 show very high approval percentages.

🔹 Higher income → Higher approval probability

Low-income groups show significantly lower approval rates.

🔹 DTI (Debt-to-Income) is a critical risk indicator

Applicants with high DTI are more frequently rejected.

🔹 Property area and dependents influence approval

Urban areas and applicants with fewer dependents show higher approval chances.

🔹 Simple rule-based screening model can automate initial filtering

---

How to Reproduce 🧪

1️⃣ Run SQL Preprocessing

Load the dataset and execute:

project_credit_risk_analysis.sql

This generates loan_cleaned table.

2️⃣ Export Cleaned Dataset

Export the SQL result to:

Data/loan_cleaned_data.csv

3️⃣ Load into Power BI

Import the cleaned CSV and recreate visuals (or view exported PDF).

---

Project Documentation 📁

📘 Full Report

Documentation/Credit_Risk_Analysis_Project_Documentation.doc Contains methodology, screenshots, queries, and explanations.

📊 Dashboard PDF

PowerBI/credit_risk_analysis_HM.pdf

---

Resume Value ⭐

This project highlights your strength in:

• ✔ SQL Data Cleaning & EDA
• ✔ Power BI Dashboard Development
• ✔ BFSI Domain Understanding
• ✔ Credit Risk Analytics
• ✔ Feature Engineering
• ✔ Realistic Approval Pattern Insights

This is a perfect portfolio project for roles in:

• Data Analysis
• Business Intelligence
• Risk Analytics
• Banking & Financial Analytics

---

Author 👨‍💻

Harinath Makka BFSI Analyst • SQL Developer • Power BI Specialist (From Hyderabad, India)