摘要
在最优期货套期保值比的确定中,"最小下偏矩套期保值比"比传统的方差最小化模型能更好地反映投资者风险厌恶态度,但相应的计算不易实现。一种新的研究思路是运用由惩罚似然函数筛选出的Copula函数计算下偏矩,进而估计相应的套期保值比。基于S&P500指数期现货的空头套期保值实证结果表明,在"最小下偏矩套期保值比"的估计中,联合正态分布方法是不适用的,而混合copula方法的预测效果要优于非参方法和单一copula方法,是一种比较符合市场实际的"最小下偏矩套期保值比"计算方法。
In estimating of optimal hedging ratios, the minimum lower partial moment （LPM） hedging ratio model has a better ability to reflect the risk attitude of investors than the traditional minimum variance hedging ratio model. But the computation of the LPM is not easy. A new method is to calculate the minimum LPM hedging ratio based on mixed copulas selected by a penalized likelihood approach with shrinkage selection criterion. The empirical results based on the analysis of the S＆P 500 index spots and futures conclude that when estimating the minimum LPM hedging ratio, it is inappropriate to use the joint normal distribution methods and the mixed copula method is better than the nonparametric and the single copula methods.
出处
《厦门大学学报（哲学社会科学版）》
CSSCI
北大核心
2009年第3期34-40,53,共8页
Journal of Xiamen University(A Bimonthly for Studies in Arts & Social Sciences)
基金
国家自然科学基金项目"金融资产变结构波动的非参数GARCH建模及其应用研究"(70871003)
教育部人文社科一般项目"市场有效性
价格发现与定价权争夺:基于人民币即期汇率和远期汇率的研究"(07JA790077)
教育部"留学回国人员科研启动基金"(教外司留[2008]890)
