基于Spearman ρ的时变Copula模型的模拟及应用  

The Simulation and Application of Time-Varying Copula Model Based on Spearman ρ

作　　者：王沁[1,2] 王璐[1] 何平[1] 

机构地区：[1]西南交通大学数学学院,四川成都610031 [2]中国科学技术大学统计与金融系,安徽合肥230026

出　　处：《数理统计与管理》2011年第1期76-84,共9页Journal of Applied Statistics and Management

基　　金：2009教育部人文社会科学研究项目基金资助(编号09YJCZH104);中央高校基本科研业务费专项资金资助(编号SWJTU09CX075;SWJTU09ZT37)

摘　　要：本文从Spearmanρ入手,利用Spearmanρ在非线性单调变换的情况下保持不变的特点,以及与条件期望预测机制存在的非线性的关系,提出建立时变Copula的模型的新方法;通过建立时变FGM-Copula模型的实例分析表明,这种构建Copula模型的方法较好捕捉了相依机制的时变性,预测了随机变量的趋势,具有一定的优越性。In this paper, based on the Spearman p which remains invariability in non-linear monotonic transformation and has non-linear relationship with the prediction mechanism of conditional expectation, the new method of establishing the time-varying Copula model is proposed. Through the example of establishing the time-varying FGM-Copula model and the analysis, it is shown that this method of establishing a time-varying Copula model can capture the time-varying dependent mechanism, and predicted the trend of random variables. The method has certain superiority.

关 键 词：时变COPULA 条件期望预测机制 Spearman ρ回归函数 

分 类 号：O211.3[理学—概率论与数理统计] F224[理学—数学]