A study based on the impact of extreme weather on the insurance industry

Abstract

Extreme weather has become a significant crisis in property insurance, necessitating the avoidance of high-risk areas for investment and asset protection. This paper establishes the Underwriting Model (UM) and Underwriting Decision Model (UDM) to develop strategies that mitigate extreme weather impacts. For UM, Munich Re Worldwide statistics from 2016 to 2023 were combined with Bayes-LSTM modeling to predict the likelihood of weather and disasters on each continent over the next decade. Using the Bayes-TOPSIS model, the risk of each continent was rated, revealing Europe (0.33), North America (0.21), Asia (0.57), Oceania (0.55), Africa (0.51), and South America (0.73) as varying risk levels. The combined Bayes-TOPSIS scores inform insurers' decisions on underwriting. The XgBoost Algorithm was then applied to formulate insurance strategies for the United States and Chile. For UDM, the ARIMA algorithm projected the global population to reach 10.124 billion in 50 years, with an average annual growth rate of 0.49%. Disaster frequency predictions, using the Random Forest Algorithm, and cost-benefit analysis informed the developer's decision-making model, emphasizing earthquake and lightning strike data.