Abstract: Housing prices in China have grown rapidly and dramatically over the past two decades; at the same time, the housing sector has contributed greatly to China’s economy. Thus, the importance of exploring China’s housing question cannot be overemphasized. To better understand the dynamics of housing prices in China, we try to forecast housing prices in China’s first-tier cities: Beijing, Shanghai, Guangzhou, and Shenzhen, by means of rolling ARIMA models and Grey BR-AGM (1,1) model in order to compare their forecasting performances. The empirical results demonstrate that Grey BR-AGM (1,1) model outperforms other models with a quicker reaction to external policy shocks. 

Key words:  , Housing Price Forecasting, Rolling Windows, ARIMA, Grey BR-AGM (1,1)