摘要
针对Nelson(2008)的周期之谜给出如下定理:当序列的方差比大于1时,BN周期成分的符号应反号,蒙特卡罗仿真实验结果也证实了理论定理。本文进而分解我国GDP的周期,结果表明GDP的方差比为1.97,所以周期成分的符号应反号,样本期内共经历6轮周期,这6轮周期源于主要的随机冲击,具有波动幅度较大、持续时间存在明显差异的特征。
Based on the cycle puzzle of Theorem: When the variance ratio of an I (1) the BN cycle component should be reversed. Nelson (2008), this paper proves a series is greater than one, the sign of The Monte-Carlo simulations verify the theorem. We decompose the BN cycle for Chinese GDP, decomposition results show that the variance ratio of the GDP is 1.97, the cycle component should be re- versed to form the real cycle of Chinese GDP. In the sample period, there are six cycles which are not only consistent with Chinese real economic fluctuations, but also coincident with the main shocks. The cycles have high fluctuations and different durations.
出处
《数量经济技术经济研究》
CSSCI
北大核心
2013年第4期46-56,84,共12页
Journal of Quantitative & Technological Economics
关键词
BN分解
周期反号
持久性
方差比
BN Decomposition
Sign Reversion
Persistence
Variance Ratio