Abstract: The correct calculation of stock index futures' hedge ratios depends on a number of parameters including: the investors' objectives and level of risk aversion; the models that are used to estimate empirically the optimal hedgers; and whether in-sample or out-of-sample horizons. A framework is present in this paper that can be utilized to investigate the influence of above factors on the determination of Hushen 300 stock index futures' hedge ratios by empirical research. Firstly, asymmetric basis is inserted into the VECM-GARCH model to deal with the asymmetric effect of basis on the time-varying variance-covariance of index futures and spot returns and its impacts on dynamic hedging. Then the hedging effect of six models including VECM-GARCH-X is studied by the standards of minimum variance and utility maximizing both in-sample and out-of-samples. The effect of risk-aversion coefficients on the choice of the most suitable model is also investigated. The data sets used consist of daily (545 prices) cash and futures prices of the Hushen 300 stock index futures market from 16 April 2010 to 16 July 2012. Results indicate that the basis effect is asymmetric, and the model with the asymmetric effect provides greater risk reduction in general. But under the general risk-aversion coefficient the enhanced utility cannot make up the additional transaction costs from the dynamic hedging strategies, so constant hedging model like OLS can perform better in reality. Moreover, the choice of the best hedging model has great relation with the investors' risk-aversion coefficients. The higher the risk-aversion coefficients is, the better the dynamic hedging models perform, which is testified by the research of optimal rebalancing frequencies at the end of paper. The main contribution of this study is to present more realistic thread of analysis and theoretical direction, thus such research results possess the great practical significance and practical value.

Key words: stock index futures, hedging, risk-aversion coefficient, optimal rebalancing frequencies