SQL • Python • Power BI • DAX

Credit Risk Analytics
& Policy Scenario Analysis

End-to-end credit risk analytics focusing on customer behaviour, affordability indicators, and policy trade-offs.

GitHub Repo View Full Report

Project Overview

This project analyses consumer credit risk at a portfolio level to identify high-risk customer segments and quantify how credit policy decisions impact approval volumes and default outcomes.

A SQL-first workflow was used for data preparation and feature engineering, followed by exploratory analysis and interactive dashboard development in Power BI.

KPI Snapshot

Default Rate

~8%

Portfolio risk health

High-Risk Segments

~35%

Risk concentration

CTI Trade-off

Approvals ↑ / Risk ↑

Policy optimisation

Behavioural Risk

Refusals ↑ defaults

Decision feature

Analytical Approach

SQL joins & aggregations for customer-level datasets
Feature engineering of affordability & behavioural indicators
Exploratory analysis & validation
Policy-driven scenario analysis using CTI thresholds

Dashboard Highlights

Executive Risk Overview

Customer Risk Segmentation

Policy & Scenario Analysis

Key Insights & Decisions

Portfolio averages hide concentrated segment-level risk
Younger, lower-income customers drive disproportionate defaults
Previous refusals strongly predict future default risk
CTI relaxation yields diminishing returns beyond optimal levels

SQL Feature Engineering Example


SELECT
  age_band,
  income_band,
  COUNT(*) AS applications,
  SUM(CASE WHEN default_flag = 1 THEN 1 ELSE 0 END) AS defaults,
  ROUND(
    SUM(CASE WHEN default_flag = 1 THEN 1 ELSE 0 END)::DECIMAL
    / COUNT(*) * 100, 2
  ) AS default_rate
FROM credit_applications
GROUP BY age_band, income_band
ORDER BY default_rate DESC;

This query underpins segment-level default analysis used to evaluate credit policy decisions.

Credit Risk Analytics& Policy Scenario Analysis