我国货币供应时序结构的奇异谱分析  

Singular Spectrum Analysis on the Structure of the Time Series of Money Supply in China

作　　者：徐海云[1,2] 陈黎明[1] 向书坚[1] 

机构地区：[1]中南财经政法大学信息学院,湖北武汉430060 [2]桂林理工大学理学院,广西桂林541004

出　　处：《财经理论与实践》2010年第1期7-12,共6页The Theory and Practice of Finance and Economics

基　　金：教育部新世纪优秀人才支持计划项目(NCET20720855)

摘　　要：时间序列的结构分析是深入研究原始序列的重要前提。应用奇异谱分析并以极大熵谱估计为辅助,对我国广义货币供应量M2进行时间序列结构分析,结果显示:改革开放以来,我国广义货币供应量M2除了趋势项外,还具有周期分别约为10年、4～5年和3年,方差解释能力依次为23.22%、7.48%和3.44%的主周期波动成分;所有的周期波动成分的振幅均随时间而增大;长期趋势在改革开放前期增长速度较慢,而在中后期增长速度较快。The analysis of time series structure is an important premise for studying the original series. This paper performs the analysis of the time series structure of Chinese broad money （M2） supply using singular spectrum analysis with comparing to maximum entropy spectral estimation （MESE）. The results show that since Chinas reform and opening up, the time series of M2 supply have the oscillatory components with the periods of about 10, 4-5 and 3 years except for the long term trend, whose ratios of variances contribution are 23. 22% ,7.48 and 3. 44% to original time series respectively; and the amplitudes of all oscillatory components are increasing with time＇s passing; moreover, the increase speed of the time trend is relatively slow in earlier reforming and opening up, and relatively faster in later.

关 键 词：广义货币供应 周期波动 奇异谱分析 极大熵谱估计 

分 类 号：F822.2[经济管理—财政学] O177.7[理学—数学]