As the housing market is liberalized the market movement comes to the more important signal of agents’ activities, that can be used in establishing the housing policy. Housing market in Korea has been heavily regulated in order to distribute the scarce new housing units to the excessive buyers with lower price. As the problems of shortage in housing stock and rapid increase in housing price became alleviated, the government bean to gradually deregulate the housing market from the mid-1990s. Thus, the importance of monitoring the market movement has been growing. There are many indices to reveal the market movement. However, we choose the housing price as the target of our analysis since it comprehensively represents the results of agents’ market behavior.
The purpose of this study is to establish the short-term forecasting model for the housing prices which can be used in monitoring the housing market for the present and for the future.
We firstly adopted the time series model named ARIMA since it allowed us to forecast without any further information. ARIMA works well in the monotonic trend of movement. However, it has some limit to reflect the turning trend or extraordinary movement like financial crisis in Korea. Thus, we developed our model to the “intervention model” which added a dummy named intervention variable into the ARIMA model. Intervention model was found to be better in fitting and in forecasting. However, the weakness of the model still remains since it does not use any further information that can be valuable in forecasting. Thus, “transfer function model” was attempted to use more information that raised the forecasting power of the model. Several additional variables such as gross domestic product, money supply, consumer price index, and residential building permits were selected to be inserted into the given model.
Sales price index and Jonsei deposit index reported from the Korea Housing Bank were used as the housing prices. They are classified by the region (whole nation, Seoul, 6 large cities) and by the house type (overall, apartment). Monthly and quarterly data during 1986.1-2006.6 were used in identification, estimation, diagnosis, and prediction of the model. Employing the results of the above work, we finally predict the sales price in 2001 will be increased by 0.6% for the whole nation and by 2.8% for Seoul. And for the Jonsei deposit, it will be increased by 2.9% for the whole nation and 5.4 % for Seoul.
- 주택시장 경기동향 및 단기전망 연구(Short-term forecasting model for the housing market)
- 윤주현; 김혜승
주택시장 경기동향 및 단기전망 연구(Short-term forecasting model for the housing market)
경기도 : 국토연구원
|Series Title; No||국토연 / 2000-53|
|Subject Country||South Korea(Asia and Pacific)|
|Subject||Territorial Development < General|