Research On The Scale Of Catastrophe Insurance Fund Based On GlueVaR

China is the third largest country in the world,with a vast territory.We are encountering more and more catastrophic disasters.With the development of search and rescue technology,casualties and disappearances have decreased,but the economic losses are increasing because of the economy booming.Under the effect of global warming,the frequency and intensity of flood disasters are increasing.The impact and the affected area are huge,causing a lot of casualties and economic losses every year.China’s post disaster relief compensation is insufficient and the insurance compensation is too low,so the frequent occurrence of catastrophe causes the attention of the society to catastrophe management,which requires determining the design and implementation of catastrophe insurance system,and establishing the corresponding catastrophe insurance fund system,in which the scale is one of the important parameters of catastrophe insurance fund.Taking the flood sub fund as an example,this paper makes a distribution fitting according to the direct economic losses of flood disasters suffered by the whole country and each province.The log-normal can best describe the characteristics of the losses.And the volatility of each province is greater than the total volatility of the losses of the whole country.After fitting,the insurance fund adopts the way of excess of event basis reinsurance to calculate the fund size.Taking Glue VaR as the risk measure,on the basis of meeting the sub-additivity and tail sub-additivity,it divides the risk attitude into four risk tolerance types,and then calculates the fund size.It also explores the influence of premium rate,α and β on the scale of insurance fund.When establishing the scale of flood disaster insurance fund in each province,compare the proportion of the insurance fund model established under each province’s distribution and the insurance fund model calculated under the national loss distribution,then it is found that with imperfect data and model,the latter measurement method is much fitter for our national conditions.