全球通货膨胀率的国际联动效应研究——基于贝叶斯潜在多动态因子模型的分析  

Research on International Co-movement in Global Inflation: A Study Based on Bayesian Dynamic Latent Factor Model

作　　者：杨科[1] 田凤平[2] 林洪[3] 

机构地区：[1]华南农业大学经济管理学院 [2]中山大学国际商学院 [3]广东商学院

出　　处：《国际贸易问题》2013年第6期145-156,共12页Journal of International Trade

基　　金：国家自然科学基金项目"农产品期货市场波动率的预测以及预测精度评价研究"(项目编号:71203067);广东省高校优秀青年创新人才培育项目(育苗项目)"我国农产品期货市场波动率的预测研究"(项目编号:2012WYM_0033);广东省哲学社会科学规划项目"广东省居民消费平滑研究"(项目编号:GD11YLJ01);中央高校基本科研业务费专项资金中山大学青年教师培育项目"中国城乡居民消费行为差异与消费金融产品创新研究"(项目编号:13wkpy21)的阶段性成果

摘　　要：本文通过构建基于Gibbs抽样估算的贝叶斯潜在多动态因子模型对全球63个国家通货膨胀率的全球性联动效应和区域性联动效应进行了实证研究。研究结论表明:在整体样本期间上,通货膨胀率的全球性联动效应和区域性联动效应能解释各国通货膨胀率波动的36%和18%,而特定国家通货膨胀率的异质性成分对通货膨胀率波动的解释能力接近50%;通货膨胀率的全球性联动效应对工业化国家的通货膨胀率波动的解释能力达到60%以上;大部分国家通货膨胀率的国际联动效应的强度较稳健,但有些国家的通货膨胀率的国际联动效应强度在不同时间区间上变化较大,并能通过相关的历史事件加以解释。The study on international co-movements in global inflation claims theoretical and practical significance as it can help predict changes in the global economy and propose ways to deal with the complex global inflation,and prepare China with a long-term planning for the future.In this paper,we apply a Bayesian Dynamic Latent Factor Model to investigate global co-movements and regional co-movements with inflation rate data of 63 countries.The research results indicate that the global co-movement and the regional co-movement account for an average 36% and 18%,respectively,of inflation rate fluctuations among individual countries,with heterogeneity of national inflation rate accounting for approximately half the fluctuations.The global co-movements alone can explain above 60% of the inflation rate variation in industrialized countries.While the intensity of international co-movements in most national inflations appears reasonably stable through the subsamples,a wildly fluctuating intensity is detected among inflation rates of some countries,which can be explained through historical events that are relevant.

关 键 词：通货膨胀率 联动效应 贝叶斯潜在多动态因子模型 GIBBS抽样 

分 类 号：F822.5[经济管理—财政学]