基于DC-t-MSV模型的套期保值研究  

Research on Hedging Strategy Based on DC-t-MSV Model

作　　者：王宜峰[1] 王春发[1] 蒋一琛[1] 

机构地区：[1]浙江财经学院金融学院,杭州310018

出　　处：《西部论坛》2011年第5期104-108,共5页West Forum

基　　金：浙江财经学院校级研究生课题

摘　　要：利用多元随机波动率模型确定最优套期保值率:首先建立动态相关系数多元随机波动率模型(DC-MSV),再建立基于t分布的动态相关系数多元随机波动率模型(DC-t-MSV)。以沪深300股指期货数据为样本,利用以上两种模型分别结合方差最小套期保值模型计算最优套期保值率,结果表明,利用DC-t-MSV模型的套期保值效果优于DC-MSV模型。DC-t-MSV模型引入t分布,充分考虑了金融数据尖峰厚尾的特性;同时,参数估计部分采用马尔科夫链蒙特卡罗模拟方法(MCMC),克服了MSV模型参数估计困难的缺点。This paper studies optimal hedging ratio determination by using multivariate stochastic volatility model. First, build a dynamic correlation multivariate stochastic volatility model （ DC-MSV）, then introduce the multivariate t-distribution and establish DC- t-MSV model. After that, by taking Shanghai and Shenzhen 300 index futures data as samples, the previous two models are used to calculate optimal hedging ratio by combining minimum-variance hedge model respectively, the results show that the hedging effect by using DC-t-MSV Model is better than the effect by using DC-MSV. After t-distribution is introduced, DC-t-MSV Model sufficiently considers high-peak and fat-tail characteristics of financial data, meanwhile, its parameter estimation partly uses Markov Chain Monte Carlo imitation method, which overcomes the difficulty in parameter estimation by MSV Model.

关 键 词：最优套期保值率 多元随机波动率模型 DC-t—MSV模型 马尔科夫链蒙特卡罗模拟 T分布 

分 类 号：F830.91[经济管理—金融学] F224.0