利用多变量马尔科夫转移因子模型对我国经济周期波动的经验研究  

The Empirical Research for Business Cycle In China Using the Multivariate Markov Regime-switching Model

作　　者：石柱鲜[1] 刘俊生[2] 吴泰岳[2] 

机构地区：[1]吉林大学数量经济研究中心,长春130012 [2]吉林大学商学院,长春130012

出　　处：《数理统计与管理》2007年第5期821-829,共9页Journal of Applied Statistics and Management

基　　金：教育部人文社会科学博士点基金项目<我国外汇风险预警模型研究>的资助(项目号:03JB790043);教育部人文社会科学重点研究基地重大课题<中国经济转轨时期增长轨迹与特征的实证研究>项目(项目号:05JJD790006)的资助

摘　　要：本文应用多变量动态马尔科丈转移因子模型对我国1991年1月以来的经济周期波动进行研究。通过选取两组与经济景气一致的宏观经济指标进行实证分析,结果表明多变量动态马尔科夫转移因子模型对不同组指标的分析是一致的;根据模型所构造出的景气指数与一致合成指数的对比分析,我们发现这两个指数不论从变动趋势和峰谷转折点,还是波动幅度上都极其相似;通过对经济周期转折点测定,并与我国经济运行状况对比,我们认为用多变量动态马尔科夫转移因子模型刻画经济周期的特征是有效的。In this paper a dynamic factor model with Markov - switching is developed to investigate business cycle in China since 1991. After empirical analysis of two groups of macroeconomic indicators that are consistent with economic cycle, we find that multivariate dynamic factor model with Markov - switching is consistent in analyzing different groups of indicators; by comparing cycle index developed from model to coincident composite index, we finds that these two indexes are very similar in trend, turning points and level of fluctuation. Measuring the turning - point and comparing economic running - status, we conclude that applying the multivariate Markov regime - switching model to business cycle researches in China is a success.

关 键 词：经济周期 协同运动 非对称性 局面转移模型 动态因子模型 

分 类 号：O212[理学—概率论与数理统计]