Storm Surge Loss Distribution Fitting And Bond Pricing Based On Extreme Value Theory

Storm surge is a serious Marine natural disaster.In recent years,the global weather has become more unstable and the losses caused by storm surge disasters have become more serious.Although the frequency of storm surge is not high,it only happens occasionally,but once it happens,it will bring serious losses.In addition,the loss data over the years have not been completely recorded,and the loss data have obvious characteristics of peak and thick tail,so the analysis is usually based on the extreme value theory.At present,China's catastrophe risk market is not mature because of its late start.The main way to prevent catastrophe risk is to use catastrophe insurance.However,due to the non-insurability of catastrophe risk,its occurrence probability is small and the loss is large.It is difficult for insurance companies to bear such a high loss by themselves.In order to transfer and diversify the pressure of insurance market to bear catastrophe risk,it is necessary to not only seek the help of reinsurance market,but also transfer to capital market.With the help of catastrophe bonds and other catastrophe risk securitization tools,it can attract the purchasing power of the capital market,so as to further insurable catastrophe risks,which are generally considered uninsurable,and thus make China's catastrophe risk management more scientific and reasonable.In order to study storm surge disasters,this paper collected data on direct economic losses of storm surge in China from 1989 to 2018,and selected data of 154 disasters for research.POT model based on extreme value theory was adopted to analyze the thick tail characteristics of direct economic loss caused by storm surge disaster,and GPD was used to fit the data of direct economic loss caused by storm surge in China.Fitting of GDP is the most important link is the selection of threshold value,this paper summarizes seven kinds of threshold selection method,in which the selection of the Hill diagram method,beyond the function diagram method combined with fitting residual error method to get the two threshold,the test on the stability of two thresholds,two distribution functions are obtained by maximum likelihood estimation,according to the fitting results diagnosis contrast,determine the final fitting of GDP.According to the distribution of direct economic losses from storm surge disasters,the VaR of the value at risk under different confidence levels was calculated,and the catastrophe risk allocation mechanism for storm surge disasters was established to alleviate the situation that the compensation for catastrophe losses depended too much on government assistance.Then the pure reinsurance premium was calculated on this basis.After constructing the risk allocation mechanism,two kinds of storm surge bonds are designed.One,from the perspective of investors,constructs ashort-term partial principal security layered bond model based on hierarchical pricing and Wang transformation,and carries out numerical analysis.The other,from the perspective of the issuer,calculates the interest rate term structure for the long-term partial principal security bonds and uses the discounted cash flow method to price the bonds.Finally,it provides feasible Suggestions for China's catastrophe risk market and provides ideas for the research and development of China's catastrophe bond.