Juma Kasozi

The low adoption of insurance by potential policyholders in developing countries like Kenya is a cause for concern for insurers, regulators, and other marketing stakeholders. To effectively design targeted marketing strategies to boost insurance adoption, it is crucial to determine the factors that affect insurance uptake among potential policyholders. In this study, the 2021 FinAccess Survey, which interviewed sampled individuals above 16 years in Kenya and machine learning techniques, including Random Forest, XGBoost, and Logistic Regression, were utilized to uncover the factors driving insurance uptake and the reasons for the low adoption of insurance among potential policyholders. Random Forest was the most robust model of the three classifiers based on Kappa score, recall score, F1 score, precision, and area under the operating characteristic curve (approaching 1). The paper explores eight reasons why people currently do not have insurance policies. The results indicated that affordability was the primary driver of uptake with 68.67% of having expressed a desire to possess insurance but are unable to afford it. The highest level of education being the next most significant factor. Cultural and religious beliefs and mistrust of insurance providers were found to have a minimal impact on uptake. These findings imply that offering affordable insurance products and conducting awareness campaigns are critical to increase insurance adoption.

Trends offer direction and momentum. However, trends in mortality are affected by trend breaks, which are a consequence of mortality shocks. Additionally, insufficient historical data challenge the credibility of the forecasted trends, which are useful for actuaries in pricing, reserving, and valuing life insurance products. To address these challenges, the study aims to determine and incorporate trend breaks among individual causes of death and coherently forecast them by applying the bottom-up hierarchical forecasting approach for life insurance models. The models used are categorized as base (linear model), auto-statistical (Arima, Exponential-Smoothing, and Prophet), and auto-machine learning. The data from the World Health Organization consisted of annualized mortality quantities by cause, gender, age, and period for Kenya. Results based on the mean absolute percentage error criteria across the causes of death showed that all the models apart from the base model showed significant improvement after accounting for the trend breaks with the best being the auto machine learning approach leading with seven causes of death. Updating forecasts based on the computed trend breakpoints that varied between 2007 to 2011 generally improved forecast accuracy. These results suggest that forecasting errors may be reduced after accounting for trend breaks and model specifications. Furthermore, this implies that insufficient data do not necessarily produce deficient forecasts. The study’s contribution involved applying approaches that enhance the accuracy of forecasting models to prevent adverse effects of mortality shocks in actuarial modeling.