关键词: 房地产市场, “北上广深”, 有向无环图, 溢出指数

Abstract: This article separates the housing prices in Beijing, Shanghai, Guangzhou and Shenzhen by using the HP filter method. Directed Acyclic Graph (DAG) and the Information Spillover Index methods are implemented innovatively to analyze the contemporaneous linkage effect and the information spillover effect among the housing prices and Rolling Window Estimate method to analyze the spillover to the external information and regulation policies. The results show that there are close linkage effect and relatively high market information spillover effect among the housing prices of the first-tier cities. In terms of the contemporaneous linkage effect, the price trend of Shen Zhen shows the most obvious external linkage effect, the price trend of Beijing is most easily affected by other cities over the same period; Shanghai’s price fluctuation has strong external linkage effect on other cities, while Shenzhen’s price fluctuation is influenced more quickly and clearly by the other cities. In terms of the information spillover effect, Shenzhen’s price trend plays a leading role with the strongest spillover effects; Shanghai’s price has the most influence on other markets in medium and long terms. Spillover index among the house prices trend rises with the favorable policies and information, and declines with the restriction policies and unfavorable information. Spillover effect of house price volatility is more sensitive to external information, and the restrictions on real estate market policy increases the information spillover scale among price fluctuations of core cities.

Key words: housing market, Bei.-Shang.-Guang.-Shen. cities, directed acyclic graph, spillover index