Navigating catastrophic risks: A comparative study of predictive models for vehicle insurance pricing in South Africa

Abstract

The South African insurance industry faces pricing challenges due to increasing catastrophic events driven by climate change and socio-economic uncertainties. The purpose is to find the best practices for vehicle pricing in South Africa. Five leading predictive models: Generalised Linear Models, Random Forest, XGBoost, Gradient Boosting Regressor, and Artificial Neural Networks are compared for estimating motor insurance premiums under normal and optimised conditions. Furthermore, analyses of catastrophic risk distribution are conducted. Results indicate that machine learning techniques, particularly XGBoost, achieve superior predictive accuracy under optimised conditions, along with the lowest RMSE and MAE. Traditional models like GLM ensure interpretability and regulatory compliance but struggle with non-linear complexities. However, integrating GLM with XGBoost enhances predictive performance. XGBoost is superior and GLM highly complies with the regulations. Gauteng is susceptible to malicious damage, KwaZulu-Natal is mostly exposed to floods and storms, and Western Cape is more vulnerable to droughts and earthquakes. These insights highlight the necessity of adaptive, data-driven risk management tailored to regional vulnerabilities. This research contributes to advancements in insurance pricing methodologies, offering actionable insights for policy development, disaster preparedness, and resilience-building strategies to address South Africa’s evolving risk landscape.

Keywords: {'name': 'Insurance Pricing, Predictive Modelling, Machine Learning, XGBoost, Catastrophic Events, Optimisation.'}

Authors & affiliations

Sandile Buthelezi
Sefako Makgatho University of Health Science, South Africa

Taurai Hungwe
Sefako Makgatho University of Health Science, South Africa

Solly Matshonisa
Sefako Makgatho University of Health Science, South Africa

Vimbai Mbirimi-Hungwe
Sefako Makgatho University of Health Science, South Africa

Deposited references

Supplied references, not a list of publications citing this article.

DOI: 10.1017/S1748499524000289

Abdelhadi, S., Elbahnasy, K., & Abdelsalam, M. (2020). A proposed model to predict auto insurance claims using machine learning techniques. Journal of Theoretical and Applied Information Technology, 98(22), 3428–3437.

Bertsimas, D., & Zeng, C. (2024). Catastrophe insurance: An adaptive robust optimization approach (arXiv:2405.07068). arXiv. https://arxiv.org/abs/2405.07068

DOI: 10.1007/s11166-006-0171-z

Bosman, W. (2023). The future of short-term insurance in the face of climate change. PPS. https://www.pps.co.za/future-short-term-insurance-face-climate-change

DOI: 10.20525/ijrbs.v13i2.3211

Gartenstein, D. (2018). Advantages and disadvantages of pricing strategies. Chron. https://smallbusiness.chron.com/advantages-disadvantages-pricing-strategies-22271.html

Greberg, F., & Rylander, A. (2022). Using gradient boosting to identify pricing errors in GLM-based tariffs for non-life insurance.

DOI: 10.3390/risks7020070

DOI: 10.1016/j.worlddev.2017.01.010

DOI: 10.3390/risks11090164

DOI: 10.1111/jori.12063

DOI: 10.1007/978-3-030-86211-4_22

Vosloo, B. (2022). KPMG insurance survey 2022: The insurance industry sees strong growth despite tough market conditions. South African Financial Markets Journal. https://financialmarketsjournal.co.za/kpmg-insurance-survey-2022-the-insurance-industry-sees-strong-growth-despite-tough-market-conditions/

DOI: 10.1177/1079063213478202

DOI: 10.3390/risks12040062

DOI: 10.1088/1757-899X/551/1/012075