上海房地产市场的多尺度周期波动特征——基于集合经验模态分解和周期相位识别的分析  

The Multi-scale Characteristics of Cyclical Fluctuation of Real Estate Market in Shanghai——Based on the Analysis of EEMD and Cyclical Phase Identification

作　　者：陆长玮 LU Chang-wei(School of Economics,Fudan University,200433)

机构地区：[1]复旦大学经济学院,200433

出　　处：《上海经济研究》2020年第8期46-57,共12页Shanghai Journal of Economics

摘　　要：本文利用集合经验模态分解(EEMD)方法,基于2005年7月-2017年12月的月频数据探究了上海房地产市场周期波动的多尺度特征。研究结果表明,上海房地产市场存在5个显著的周期性波动分量,既包括平均时长为季度和半年的年内房地产周期,也包括平均时长为2年和3年的年际房地产周期以及11年长度的年代际房地产周期;而且这5个房地产周期分量都显著通过了白噪音检验,对上海房地产市场变动有不同的贡献。我们还对各个房地产周期进行了周期转折点识别和周期相位分布分析,发现不同尺度房地产周期的相位分布既有重叠加强,也存在交错对冲。Using the approach of ensemble empirical mode decomposition and the monthly data from July 2005 to December 2017,this paper investigates the multi-scale characteristics of cyclical fluctuation of real estate market in Shanghai.It turns out there are five significant multiscale cycles which include two intra-year real estate cycles with the length of quarterly and halfyear,two inter-year real estate cycles with the length of 2 years and 3 years,an inter-decadal real estate cycle with the length of 11 years.These real estate cycles all reject the hypothesis of white noise and have significant physical information and different contribution for the variations in the real estate market.Furthermore,we explore the turning points of real estate cycles above and identify the cyclical phases of different times respectively.It is found that the cyclical phases under different scales overlap and reinforce each other sometimes,but contradicting and hedging each other sometimes.

关 键 词：集合经验模态分解 房地产周期 白噪音检验 相位分布 多尺度特征 

分 类 号：F061.5[经济管理—政治经济学]