摘要
研究目标:分析混频Granger因果关系检验的功效及其稳健性。研究方法:以货币与产出间的因果关系为例,比较分析不同情形下,混频和同低频Granger因果关系检验功效的差异性和检验结果的稳健性。研究发现:货币与产出间的因果关系不是一成不变的,混频Granger因果关系检验的水平和功效随着不同层次的货币供给、模型滞后阶数、预测期长度,变量的多寡,以及样本量的大小和所处的经济阶段的变化而呈现异质性和非对称性。研究创新:将混频技术与Granger因果关系结合起来,从不同角度全面系统地检验了货币与产出的因果关系,并分析了该方法的稳健性。研究价值:进行货币政策调控时,须评估不同情形下货币政策的作用效果,选取恰当的时机,采用更加合理的调控手段和方法。
Research Objectives： To analyze the efficacy and robustness of the mixed frequency Granger causality test. Research Methods： Take the causal relationship between money and output as an example, we compared and analyzed the difference and the robustness of the mixed frequency Granger causality test results under different circumstances. Research Find- ings： The causal relationship between money and output is not unchangeable, and the level and power of the mixed frequency Granger causality tests are varying with different scale of money supply, the lag order, the length of forecast period, the amount of variables, the sample size and the economic stages, which results have the characteristic of heterogeneity and asymmetry. Research Innovations： Combine the mixed frequency technique with Granger causality test, we examine the causality between money and output systematically and sys- tematically with a different viewpoint, and analyze the robustness of the method. Research Value： It is necessary to assess the effect of monetary policy under different circumstances, then select the right time, and use the suitable control methods when we adopt the monetary policy control.
出处
《数量经济技术经济研究》
CSSCI
CSCD
北大核心
2017年第10期144-160,F0003,共18页
Journal of Quantitative & Technological Economics
基金
教育部人文社会科学研究青年基金项目"宏观经济预测与分析的混频定量研究"(15YJC790055)
中国博士后科学基金资助项目"中国经济周期波动分析与预测的混频计量研究"(2014M551161)
吉林大学基本科研业务费项目"大规模不规则抽样数据的混频因子模型及其应用研究"(2015zz003)的资助
