门限协整套利：理论与实证研究  

Threshold Cointegration Arbitrage:Theory and Application

作　　者：葛翔宇[1,2] 吴洋[2] 周艳丽 

机构地区：[1]中国数量经济学会 [2]中南财经政法大学统计与数学学院 [3]澳大利亚科廷科技大学

出　　处：《统计研究》2012年第3期79-87,共9页Statistical Research

摘　　要：不同市场上的同质或相似商品的价格存在长期均衡关系,当价格偏离均衡时,由于套利交易的存在,偏离会迅速回到均衡。在一定的门限值以外,二者服从协整关系,在门限值以内,二者没有协整关系,这种关系称为门限协整。本文在Balke,Fomby(1997)[1]和Hasen(1996)[6]的基础上提出了基于门限向量误差修正模型(T-VECM)的sup-Wald检验,用Bootstrap方法模拟统计量的渐进分布,验证了英国富时指数期货(uk100)和德国法兰克福指数期货(ger30)的门限协整关系,用Hasen、Seo(2002)[11]提出的极大似然估计方法(MLE)同时估计出门限参数和协整向量,并给出了在这种门限协整关系下进行跨市场无风险套利的策略。There exists a long-run stability relation between similar commodities traded in different markets.When the price deviates from equilibrium,it will return to the no-arbitrage region.They are cointegrated outside a threshold,whereas aren＇t cointegrated inside the threshold,which is defined as threshold cointegration.In this paper,we put forward a method of sup-Wald test based on threshold vector ECM model on the work of Balke,Fomby and Hasen,simulate the asymptotic distribution of the statistic using the Bootstrap method,give the presence of threshold effect between uk100 and ger30 index future,estimate the threshold parameter and cointegrating vector simultaneously,and promote an arbitrage strategy under this cointegration relation.

关 键 词：门限协整 跨市场套利 股指期货 门限误差修正模型 BOOTSTRAP 

分 类 号：F222.3[经济管理—国民经济]