论文部分内容阅读
采用随机系数马尔科夫体制转换(RCMRS)模型对中国铜期货市场套期保值比进行估计.RCMRS模型跳出GARCH类模型基于新息描述的研究框架,视最优套期保值比为随机系数,直接估计出依赖于市场体制状态的时变套期保值比.市场体制状态在模型中被视为潜在变量,和其它参数一起通过最大化似然函数估计出来.由于考虑了不同市场体制状态对套期保值比的影响,RCMRS模型估计的最小方差套期保值比波动范围要小于GARCH类模型估计结果的波动范围.均值—方差效用函数不仅反映了风险,还同时反映了收益率及风险厌恶程度.在采用方差下降百分比测度套期保值效率的同时,另外采用均值—方差效用最大化原则对RCMRS模型与GARCH、VECM、VAR及OLS模型的套期保值表现进行了样本内和样本外比较.样本内比较支持RCMRS模型,而样本外比较则不利于RCMRS模型.
The RCMRS model is used to estimate the hedging ratio of Chinese copper futures market.The RCMRS model jumps out of the research framework of GARCH model based on the description of interest rates and considers the optimal hedge ratio as a random coefficient, The time-varying hedging ratio that depends on the state of the market system is estimated.The state of the market system is treated as a latent variable in the model and estimated together with the other parameters through the maximization of the likelihood function.Because of the consideration of the state of different market systems, Hedge ratio, the variance range of the minimum variance hedge ratio estimated by the RCMRS model is less than the volatility range of the GARCH class model estimation result. The mean-variance utility function not only reflects the risk, but also reflects the rate of return and risk aversion. The hedging efficiency of the RCMS model and the GARCH, VECM, VAR and OLS models were compared both in-sample and out-of-sample using the principle of maximizing mean-variance utility, while using hedging percentage as a measure of hedging efficiency. The RCMRS model is supported, while the out-of-sample comparison is not good for the RCMRS model.