Dynamic Hedging Study Based On Structural GARCH-Copula Model

The developing of our futures market is providing tools of avoiding risk for many of enterprises. But from a detailed analysis of the China futures of hedging market can be found that many enterprises can not avoid risks by futures because of speculation,the ratio of hedging and the point of entering the hedging market. So the major aim of this paper is to provide more scientific and reasonable hedging scheme for hedgers from studying on the ratio of hedging.The mainstream econometric model is used in measuring hedging ratio both at home and abroad now, these models can depict persistent characteristics of volatility well. But the continuous volatility of spot and futures market may have occurred structure transfer because of the influence of new information, it will lead to the bias in predicting ratio of the optimal dynamic hedging. This paper will lead structure state transfer theory to hedging model in order to solve the problem.This paper elaborates the theoretical and practical significance of the futures hedging first, then discuss the mainstream hedging theory at home and abroad, such as least-square method,correct error model,the ARCH clan,GARCH clan model, and analysis the problem when they are used in hedging. This paper proposed structural GARCH- Copula model according to the existing problems, then depicted the nonlinear correlation of futures yields and spot yields by using Copula function.This paper used copper futures in Shanghai futures exchange for empirical after establishing the model, it proved that the hedging effectiveness of structure state transfer model was the best. The major aim of this paper is setting up new methods and model,providing scientific decision-making techniques,improving the effectiveness of risk management for hedging. Thesis mainly innovation points lie in: (1) introducing the state transition thought to hedging; (2) using the predicted data in volatility for estimating hedging ratio; (3) introducing Copula function to calculate the correlation between the spot and futures in the aspect of measuring correlations.