Research on Underwriting Risk Based on EWM-TOPSIS and SARIMA Models

Abstract

Recent climate change-induced extreme weather events have heightened property insurance claims, prompting the development of an insurance model utilizing advanced mathematical models. Utilizing the EWM-TOPSIS model, a comprehensive framework for underwriting risk assessment was developed and applied to determine the 20-year Insurance Claim Risk Index (ICRI) for both Beijing and Guizhou Province. Predictive modeling using the SARIMA model produced ICRI values for Beijing (0.134, 0.154, 0.149) and for Guizhou Province (0.265, 0.304, 0.169). Assuming the ICRI follows a normal distribution, a statistical analysis was conducted on 40 ICRI data points. Based on this, the ICRI was categorized into three levels, providing a foundation for the insurance company's decision-making. In Beijing, it is suggested to insure for the next three years, while in Guizhou, it is advised not to insure for the first two years or to start the business in the third year.