基于动态计量经济学模型的房地产周期研究  

Real estate cycles based on dynamic econometrics model

作　　者：张红[1] 马进军[1] 孔沛[1] 

机构地区：[1]清华大学建设管理系,北京100084

出　　处：《清华大学学报（自然科学版）》2007年第12期2111-2113,2118,共4页Journal of Tsinghua University(Science and Technology)

基　　金：国家自然科学基金资助项目(70673050)

摘　　要：为正确判断房地产发展趋势，以动态计量经济学模型为基础，科学地识别并预测房地产市场周期。采用北京1989-2004年的时间序列数据，将先验经济理论与数据统计分析结合，建立自回归分布滞后的ARMAX模型。对变量进行单整ADF检验和多重协整JJ检验，求出误差修正序列。用包含误差修正项的模型来预测市场周期，诈补中国房地产市场广泛存在的非理性因素影响和统计数据的缺陷。研究表明：北京房地产市场的周期约为4～5a；2005年的房地产市场正处于扩展阶段；2006年北京房地产市场仍将呈现稳步上升的态势。This paper presents a forecasting method for real estate market cycles to help market participants make reasonable decisions on real estate development trends. An autoregressive distributed lag ARMAX model was developed using the time series data of Beijing from 1989 -2004 by integrating transcendental economic theory and statistical analyses. The model uses the unit root test ADF and the multi-cointegration test JJ, with the error correction mode （ECM）. The dynamie eeonometries model eontaining the ECM term used to forecast market cycles includes the influence of common irrational factors in Chinese real estate markets and data defects. The results indieate that eaeh eycle in the Beijing real estate market lasts 4 - 5 years. The market in 2005 was in the expansion phase and will continue to be ascending steadily in 2006.

关 键 词：房地产周期 动态计量经济学 ARMAX模型 ADF检验 JJ检验 误差修正项(ECM) 

分 类 号：F293.35[经济管理—国民经济]