The Minimum Cvar Model Based On Dc - Msv Dynamic Hedging Model And Empirical Research

The main functions of the futures markets is risk transfer, after putting out HS300stock index futures, the hedgers can control risks and make full use of the effective measure of assets allocation by stock index futures hedging. The most core issue of hedging is calculating the optimal hedge ratio. Through researching hedging models and setting some constraint conditions to estimate the optimal hedge ration can greatly enhances hedging effectiveness.This paper is dividing into five parts. The first chapter is about the instruction, domestic and foreign study state, and their existing deficiencies. The second chapter is mainly about the existing theories and hedging models in common. The third chapter is mainly about the CVaR’s principles and the dynamic hedging model of futures. Then based on minimum CVaR,we built the dynamic hedging model of futures. The fourth chapter is mainly about the empirical study and the comparison analysis by using the model which builds by the paper in HS300stock index futures. The fifth chapter is about summary and analysis of prospects. According to the feature of stochastic volatilities of futures and spot price, the main coverage of the paper is that it calculating the dynamic hedge ratios based on DC-MSV model which basing the major feature of this model are shown as follows, firstly, it reveals the feature of optimal hedge ratio which is time-varying. Secondly, using minimum CVaR of the hedged portfolio return as optimal function, the tail loss of hedged portfolio has been fully considered and the hedgers’ expectation return and risk aversion has been synthesize considered. Finally, this paper takes positivistic researches on dynamic hedging model which calculates the hedging ratio of spot and futures of HS300stock index based on the minimum CVaR, and the results show that this model performs on hedging effectiveness better than naive, OLS, GARCH models and dynamic hedging models of minimum CVaR which based on DC-MSV model on both within-sample and out-sample.