Customer Segmentation and Probit Regression Analysis of Financial Behaviour

Analysis conducted as part of coursework for course 125788 Big Data in Finance and Banking (Master of Analytics 2024, at Massey University - Auckland Campus)

Abstract

This case study investigates the interplay between gambling behaviour, financial capability, and financial hardship using advanced segmentation and regression analysis. Leveraging SAS Viya and SAS Miner, the study identifies distinct consumer segments and evaluates the factors influencing financial distress through logistic regression (Probit). Key insights include the protective role of employment and homeownership against financial hardship and the detrimental impact of gambling-related harm and low financial capability. The findings highlight the importance of targeted interventions to improve financial literacy, promote resilience, and mitigate the effects of financial shocks and gambling harm on vulnerable populations.

Overview

This repository contains the data analysis and Stata code used to explore the relationships between gambling behaviour, financial capability, and financial hardship. Using SAS Viya and SAS Miner for clustering and segmentation analysis, and SAS Viya for logistic regression (Probit), the study identifies key drivers of financial distress and provides actionable insights for targeted interventions.

Key Objectives

1. Segment populations based on financial behaviours and attitudes.
2. Examine the role of employment, homeownership, financial capability, and gambling behaviour in financial hardship.
3. Provide actionable insights to mitigate financial distress through policy and education.

Software Used

• SAS Viya for Learners 4
• SAS Miner 15.3

Data

• The synthetic dataset is based on Australian HILDA survey data and includes:
  • Financial Capability: Scores for financial literacy, goal-setting, and management.
  • Financial Hardship: Indicators of distress (e.g., missed payments, food insecurity).
  • Gambling Behaviour: Harm scores and behavioural patterns.
  • Demographics: Age, marital status, employment, income, and education.

Key Findings

1. Segments Identified:

   • Financially Secure: High capability, stable income, minimal hardship.
   • Highly Vulnerable: Severe hardship, older age, low education.
   • Seeking Help: Financial distress, proactive recovery efforts.
   • Financial Shock Responders: Struggling with sudden financial challenges.

2. Probit Regression Insights:

   • Employment and homeownership are protective against financial distress.
   • Low financial capability and gambling harm increase vulnerability.
   • Income, marital status, and education levels affect segment membership differently.

Contact

For queries, please contact Luis Vieira.

License

This project is licensed under the Apache License 2.0.