基于M-Copula-GJR-VaR模型的黄金市场套期保值比率研究

发布时间：2018-01-12 15:20

本文关键词：基于M-Copula-GJR-VaR模型的黄金市场套期保值比率研究　出处：《湖南大学》2013年硕士论文　论文类型：学位论文

【摘要】：套期保值策略是金融风险研究领域极为重要的课题之一，其效果的好坏主要取决于最优套期保值比率估计的精度。基于VaR的风险测度方法既侧重收益的负向波动风险，又可通过置信水平的设定满足抱有不同风险偏好的投资者的需求。本文以具有金融和商品双重属性的黄金为实证对象，，充分考虑现货和期货市场的非对称性、两者之间的协整关系以及非线性相关的特征，并以VaR风险最小化为原则，建立一个M-Copula-GJR-VaR动态套期保值比率模型，用以估计黄金市场最优套期保值比率。本文首先根据黄金现货和黄金期货收益率的有偏、尖峰厚尾分布以及非对称波动特征，构建带误差修正项的GJR模型，估计黄金现货和期货收益率的条件波动率，并根据拟合情况确定具体的边际分布；然后，选择合适的单一Copula函数进行线性组合以构建M-Copula函数，利用各市场累积分布函数值序列对M-Copula函数模型进行参数估计，从而得到基于M-Copula-GJR模型的黄金现货和黄金期货收益率的联合分布，进而计算出基于M-Copula-GJR-VaR模型的最优套期保值比率。最后，对比分析M-Copula-GJR-VaR模型与CCC-GARCH-VaR模型、DCC-GARCH-VaR模型、Clayton Copula-GJR-VaR模型和Gumbel Copula-GJR-VaR模型的套期保值比率和套期保值效果。研究结果表明，采用M-Copula-GJR-VaR模型估计的套期保值比率最小且其套期保值效率最高，应用该模型进行黄金市场套期保值，可达到以相对较少的套期保值成本较大程度地规避现货市场价格风险的目的。同时本文还发现，经过5年多的发展，套期保值效率处于0.672~0.704之间的中国黄金期货市场还不成熟，套期保值功能的发挥还有待提高。
[Abstract]:Hedging strategy is one of the most important topics in the field of financial risk research. The effectiveness of the method mainly depends on the accuracy of the optimal hedge ratio estimation. The risk measurement method based on VaR focuses on the negative volatility risk of income. We can also meet the needs of investors with different risk preferences through the setting of confidence level. This paper takes gold with the dual attributes of finance and commodities as the empirical object and fully considers the asymmetry of spot and futures markets. Based on the principle of minimization of VaR risk, a dynamic hedging ratio model of M-Copula-GJR-VaR is established. Used to estimate the best hedge ratio in the gold market. In this paper, firstly, according to the bias of gold spot and gold futures yield, the distribution of peak and thick tail and asymmetric volatility, the GJR model with error correction term is constructed. The conditional volatility of gold spot and futures yield is estimated, and the specific marginal distribution is determined according to the fitting condition. Then, the appropriate single Copula function is selected for linear combination to construct M-Copula function. The M-Copula function model is estimated by using the value series of cumulative distribution function of each market. The joint distribution of gold spot and gold futures yield is obtained based on M-Copula-GJR model. Then calculate the optimal hedging ratio based on M-Copula-GJR-VaR model. Finally. Comparing M-Copula-GJR-VaR model with CCC-GARCH-VaR model and DCC-GARCH-VaR model. The hedge ratio and hedging effect of Clayton Copula-GJR-VaR model and Gumbel Copula-GJR-VaR model. The results show that M-Copula-GJR-VaR model has the smallest hedge ratio and the highest hedging efficiency. Can achieve the purpose of avoiding the spot market price risk with relatively less hedging cost. At the same time, this paper also found that after more than 5 years of development. China's gold futures market, whose hedging efficiency is between 0.672 and 0.704, is still immature, and the function of hedging needs to be improved.
【学位授予单位】：湖南大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F224;F830.94

【参考文献】