A Research On The Optimal Hedging Ratio Based On The CVaR

This paper introduce the Value at Risk indicator(VaR) and the Conditional Valueat Risk indicator (CVaR) into the field of hedging.As to the marginal distribution formof the return array of the future and spot prices,this paper use the Extrem Valuedistribution to fit the upper and lower tail of the return array separately,and thenonparametric kernel density estimation is used to fit the intermediate part of thedata.Then this paper adopt five kinds of Copula functions as the connectivity functionof the return array between the future and spot prices,and find the optimal Copulafunction using the least squares of euclidean distance.At last this paper employ theMonte carlo simulation to get the CVaR minimize hedge ratio under three differentconfidence levels,and make a comparison between the model used in this paper andthe traditional model which based on normal distribution,the conclusion is that themodel used in this paper perform better.The most common indicator been used in the former researches to measure risk isthe variance of the portfolio.But variance is an imperfect risk measurement indicatorbecause it is a two-way measurement and dissatisfies the consistencyprinciple.When it comes to the distribution form,previous studies mianly assume thatthe return array of the future and the spot price obey the normal distribution and thejoint distribution of the two array obey the joint normal distribution too.But asempirical results indicate,the return array often exist high kurtosis and fat tailfeatures,as wall as volatility clustering,which dissatisfy the hypothesis of thenormal distribution.This paper improve the traditional model both from the riskmeasurement indicator and the distribution form.The model perfprm well in capturingthe tail risk of the hedge portofolio,and is contribute to help the investor to know wellabout the risk of the hedge portofolio and also contribute to improve the theoreticalsystem of futures investment risk management.