寻找中国房价的领头羊——基于20个大城市数据的实证分析

城市房价存在明显的时空关联性。本文通过构建一个简单的消费者均衡模型，并利用全局和局部Moran指数，论证了35个大中城市的房价在空间上的相关性，进一步利用交叉谱技术分析了18个大中城市房价变动周期的关联性和领先滞后关系。结果发现：1．我国35个大中城市的房价存在显著的空间自相关。2．采用Fisher’s Kapper统计量和Kolmogorov-Smirnov统计量检验发现20个城市中除重庆和厦门外，房价波动都具有显著的周期性。3．在18个大中城市中，并不存在完全同步的房价周期。同一区域不同城市之间房价周期的关联性强度存在显著差异。即使在具有较强相关性的城市之间，房价变动也并不同步而是存在多种领先滞后关系。

Looking for Bellwether of Chinese Housing Prices -An Empirical Analysis from 20 big Cities

There exists an obvious correlation of house prices in time and space. In this paper, by constructing a simple consumer equilibrium model and making use of global and local Moran index, we demonstrated the spatial relevance of housing prices among 35 large and medium-sized cities. Further using cross-spectrum analysis of 18 large and medium cities, we study the relevance of housing prices cycle and leading-lag relationship. The results showed that： 1. there are significant spatial autocorrelation of housing prices among Chinese 35 large and medium-sized cities. 2. By use of Fisher＇s Kapper statistic and Kolmogorov-Smirnov test statistic,we found that housing prices fluctuation have a significant cyclical among 20 cities in addition to the city of Chongqing and Xiamen. 3. There is no complete synchronization of the housing cycle in 18 large and medium-sized cities. The coherence is different among cities which locate in the same region