Explore publications.

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.

An investigation to balance the sustainability of insurance companies and successful claims on insurance policies is important towards the economy, as well as for assisting the insured during financial difficulty when their items are damaged or lost. This paper examines some rejected short-term insurances where a misunderstanding of the contract might have caused the insurer to lose out on the claim. Data were collected from cases of courts and insurers on rejected claims due to misunderstandings by the insured. The first text data set was a court case. The second was a graph on the root causes of claims rejections from the insurer. The third was a matrix data set presenting the languages of the insured. The insurers who had refused to pay insurer claims supplied the latter two. Despite the three contexts being widely different, their results show that rejections of claims by the insurers were due to the insured failing to understand the conditions of the contracts, and thus not being able to figure out what could have caused the rejection. Hence, the use of complicated and legal language in insurance contracts results in the insured not fully understanding the contract, which leads to claims being rejected. The paper recommends the use of simple language and pinpointing in layman's form, those aspects that could lead to rejections of claims.

The pricing of insurance premiums in the non-life insurance sector remains a challenging and complex task. It demands a delicate balance between accurately estimating risk exposure and ensuring profitability for insurers. Generalised Linear Regression Models (GLMs) have become the preferred methods for premium price modelling in the motor insurance sector. While the approach of using a single superior model on which predictions are based ignores the use of robust estimator models. This paper examines various methodologies and sheds light on superiority of twenty-two models compared to each other for pricing automobile insurance. These methods vary from traditional actuarial methods to the modern statistical models such as machine learning algorithms. By using archival technique, their inferiority and superiority are explored, considering the ever-changing landscape of risk factors and market dynamics. Furthermore, it highlights the potential benefits of leveraging these methods and the mechanism for pricing short-term insurance, particularly in motor vehicle insurance. It also develops a framework that can be used in pricing to cater to risk analysis constituents to mitigate uncertainties and provide good services to clients. Our findings show that ANN, NN, XGB, random forest (RF) are superior models, and we conclude that the modern statistical methods can accurately estimate the risk exposure as compared to traditional methods such as the GLMs.