Predicting take-up of home loan offers using tree-based ensemble models: A South African case study

Abstract

This paper investigates different take-up rates on home loans when banks offer different interest rates. If a bank could increase the take-up rates, the bank could possible improve the bank’s market share. The article explores empirical home loan price elasticity, the effect of loan-to-value (LTV) on the responsiveness of home loan customers and whether it is possible to predict home loan take-up rates. Different regression models are employed to predict take-up rates, and tree-based ensemble models (bagging and boosting) are found to outperform logistic regression models on a South African home loan data set. The outcome of the paper is that the higher the interest rate offered, the lower the take-up rate was (as expected). In addition, the higher the LTV offered, the higher the take-up rate (but to a much lesser extent than the interest rates). Models were constructed to estimate take-up rates, with various modelling techniques achieving validation Gini values of up to 46.7%. Banks could use these models to positively influence their market share and profitability. This paper addresses a subset of a bigger question: What is the optimal offer that a bank could make to a home loan client to ensure the maximum profit for the bank while still taking risk into account? To answer this question, one of the first factors that a bank needs to understand and to predict is take-up rates

More information

Identifiers

ORCID for Bart Baesens: orcid.org/0000-0002-5831-5668

Catalogue record

Export record