| The Stock Index Futures has attracted many investors because its traction process is convenient and it offers a high leverage. Those who hold stock may want to reduce the risk that the stock price may decline in the future, so they usually sell the same amount of futures in the future market. The consistent trend in both markets will ensure that the risk of portfolio can be largely reduced. However, this is just a kind of naive hedging strategy. Because the price basis between spot and future is not fixed, they appear to be deviant most of the time. So if we simply set the hedging ratio to be one, the portfolio will still face the holding risk. So, many scholars make deep researches on the hedging ratio. The main methods include static methods and dynamic methods, and many models are developed basing on these methods.As the HS300Stock Index Future has been launched, many scholars in our country use foreign modes to study the domestic market and they draw different conclusion. However, as the stock future market expands in a high speed, the domestic investors urgently need a proper series of theory to guide their practical operation. Based on this consideration, this thesis compares the performance of several classical models to study their applicability in HS300Stock Index Future Market. Besides static models such as OLS and VECM, we also use dynamic model BGARCH. In addition, with the development of computer technology, the research on volatility has come to a high frequency period, and scholars put forward a new concept of Realized Volatility. When we are devising a hedging strategy, the volatility of spot and future are the most important variables. However, in the domestic thesis about hedging, the high frequency data is rarely used. So, this thesis introduces the RV variable into the conventional BGARCH model, trying to make a deeper study on the relationship between the volatility of spot and future, and evaluate the performance of the new model in hedging.This thesis uses rolling-over method and introduces the Edrington Efficiency index and the Value at Risk to assess the performance of the models used in this thesis. The empirical result shows that the introduction of realized volatility in the Garch model really explains the dynamic relationship between spot and future more accurately, and it significantly improve the efficiency of the conventional BGARCH model. In addition, when we consider the degree of the risk reduced, the RV-GARCH performs best; when we take the value at risk into account, both the VECM and the RV-GARCH can provide a good result. So, in the HS300Stock Index market, the RV-GARCH model is superior to other models on average. |
