| China is a country with high earthquake risk,and earthquake disasters are characterized by large loss and uneven distribution.According to historical data,the areas with high earthquake incidence in China include the southeast coastal area of Fujian,the northern area of North China,the southwest area and the northwest area.Among them,Yunnan is located in the southwest border of China,with 8 seismic zones distributed,and the earthquake disaster is particularly serious.According to the survey,Yunnan Province accounts for a particularly large proportion of the losses caused by earthquakes in all provinces of China.The earthquake disaster has seriously threatened the life and property safety of the people in the disaster-affected areas.For this reason,the former China Insurance Regulatory Commission and the Ministry of Finance issued the Implementation Plan for the Establishment of the Earthquake Catastrophe Insurance System for Urban and Rural Residents in 2016,proposing to build a multi-level risk sharing system formed by policyholders,insurance companies,reinsurance companies,special reserves for earthquake catastrophe insurance,financial support and other multiple entities.Yunnan Province also piloted the earthquake special index insurance in Dali for the first time in 2015,and has now been extended to Yuxi.Lincang also began to pilot the comprehensive insurance for earthquake catastrophe of rural houses in 2018,extending the insurance liability to other natural disasters,household appliances and accident relief.The implementation of earthquake catastrophe insurance has provided financial support for social governance and ensuring people’s livelihood,reduced the risk of farmers returning to poverty due to disasters,and played a supporting role in the high-quality development of regional economy.However,from the current operation of the earthquake catastrophe index insurance,the insurance compensation is far lower than the actual loss,that is,the basis risk is large,which leads to the need to rely on the government for post-earthquake relief,making the government’s financial pressure too large,and the catastrophe insurance can not fully play its role in social governance.To this end,it is necessary to formulate a reasonable compensation standard to reduce the basis risk of index insurance,and at the same time to give full play to the risk protection function of insurance to help the high-quality economic development and rural revitalization strategic deployment during the "Fourteenth Five-Year Plan" period.The formulation of the insurance compensation standard for the earthquake catastrophe index depends on two aspects: one is to set a reasonable trigger index(or parameter).A reasonable trigger index(or parameter)needs to be highly related to the actual loss.For natural disasters,the intensity of the disaster directly affects the size of the loss.The magnitude that measures the intensity of the earthquake disaster can be used as the trigger index of the earthquake index insurance;The second is the accuracy of loss assessment.Index insurance generally determines a compensation standard based on the specific value of the selected trigger index(parameter),which requires a more accurate estimation of the loss size corresponding to different indexes,that is,in the case of taking the magnitude as the trigger index,it is necessary to estimate the loss within the range of different magnitudes,and set a reasonable compensation mark according to the estimated value of the loss.Therefore,when the estimated loss is reasonable,the deviation between the determined compensation standard and the actual loss is small,and the corresponding basis risk is small.To sum up,in order to reduce the basis risk and give full play to the risk protection function of catastrophe insurance,it is necessary to make a more accurate assessment of earthquake catastrophe losses and formulate a reasonable compensation standard.Therefore,this paper will discuss the estimation methods of earthquake catastrophe losses,compare common statistical models and machine learning algorithms,and introduce XGBoost algorithm to compare the advantages of different models and methods in the estimation and prediction of earthquake catastrophe losses through NMSE and other indicators.The results show that XGBoost can significantly improve the accuracy of earthquake catastrophe loss estimation.Finally,taking Dali Earthquake Catastrophe Insurance as an example,a more reasonable compensation standard is formulated based on the loss estimation results obtained by XGBoost algorithm,which improves the protection level of catastrophe insurance,enables the earthquake catastrophe index insurance to give full play to its risk protection function,and helps the high-level economic construction and rural revitalization strategy of Yunnan Province during the 14 th Five-Year Plan period.The structure of this article is divided into six chapters.We can see the structure of the article: At the beginning of the article,the introduction demonstrates the research status,research significance and research background;The second chapter demonstrates the current situation of earthquake catastrophe insurance in Yunnan Province,and analyzes the existing problems;The third chapter introduces the commonly used earthquake catastrophe risk assessment methods,including extreme value theory and machine learning algorithm;The fourth chapter is empirical research.We collected 247 earthquake data from 1974 to 2020 in China,including eight explanatory variables,such as magnitude,population density in the disaster area,and the number of earthquake casualties.We used XGBoost algorithm and POT algorithm to model and predict earthquake losses,and compared their prediction results.Finally,based on the optimal model,the random simulation of Dali earthquake disaster is carried out,and the loss of Dali earthquake disaster is predicted,and a reasonable compensation standard is formulated according to the prediction results;The fifth chapter is the research conclusion and prospect of this paper. |
PDF Full Text Request