A Study on the Lapse Rate of Insurance Products using Machine Learning Methods

Key Points

• The study aims to evaluate the predictive ability of machine learning for the lapse rate of insurance products compared to traditional methods.
• Employs machine learning techniques to predict the lapse rates of whole life and variable annuity insurance products.
• Compares predictive accuracy with traditional time series forecasting models through empirical analysis.
• Focuses on single-variable predictive models for variable annuity lapse rates.
• Machine learning models outperformed traditional time series models in predicting lapse rates.
• The results indicate that predictive models using non-linear assumptions can provide superior accuracy.
• Findings suggest a need for developing systems that combine big data and machine learning for lapse rate forecasting.