Lapsation Logistic Regression Model: A Case in Life Insurance

Life insurance lapse poses significant financial challenges for insurers and policyholders, yet the determinants of lapse behaviour remain inconsistent across studies. This study identifies key factors influencing lapse prediction in life insurance contracts using logistic regression analysis on data from 499 policyholders in Malaysia. The analysis reveals that smoking status, payment mode, and face amount significantly impact lapse rates, while age and gender show no statistically significant influence. Smokers exhibit higher lapse odds compared to non-smokers, and quarterly premium payments correlate with increased lapse likelihood relative to monthly payments. Policies with lower face amounts demonstrate markedly higher lapse rates. The predictive model achieves 79.6% accuracy, underscoring its robustness in classifying lapse behaviour. These findings highlight the critical role of behavioural and financial factors over traditional sociodemographic variables in lapse prediction. The results provide actionable insights for insurers to design targeted retention strategies, such as flexible payment structures and tailored policy features, to mitigate lapse risks. Future research should expand the model to include additional variables like income and economic conditions to enhance predictive power. This study contributes to the literature by clarifying inconsistent findings and offering empirical evidence to refine lapse risk management in the life insurance industry.