The purpose of this study is to use vector error-correction modeling together with directed acyclic graphs (DAG) for analyzing dynamic relations among monthly retail property price indices of 10 major cities in China from 2005–2021.
This paper apply both the PC and Linear Non-Gaussian Acyclic Model (LiNGAM) algorithms for inference of the DAG, with the former leading to the causal pattern and the latter leading to the causal path. This paper carry out innovation accounting analysis based on the causal path according to the LiNGAM algorithm.
Their results show sophisticated dynamics among processes of price adjustments following shocks. The results do not reveal clear evidence that supports dominance of the price series of the top-tier cities.
These results suggest that it could be beneficial to design policies at granular levels regarding regional retail property prices in China.
