The Forecasting Models Research Of Residential Market

As a major industry of the national economy, abnormal fluctuations in real estate often affect the country's economic lifeline, and even the daily life of common families. So, it is very important to establish the information collection and forecasting system on real estate industry, especially on residential market, and the forecasting system is critical. Therefore, it is necessary that we should establish a generalized and effective forecasting system using the effective mathematical statistics methods. The ARIMA model, which is mainly applied to estimate the quantitative relationship between economic variables, is the essence of the theory on financial time series. As we will see, the data on residential market has obvious periodic features, so the ARIMA model is credibility being adopted here.This study selected the following main indexes: the trading volume of new houses in Shanghai, Beijing, Guangzhou, and Shenzhen; the national new house trading price index and chain index over the previous year; the area of residential lands and the volume of being traded in Shanghai and Beijing. We will establish the model, analysis the data using the mixed auto regressive-moving average model (ARIMA), and modify it with Bayes theory. Without considering government actions, we finally obtain the forecasting model. Using the model we will forecast the trading volume and new house prices in the next 20 years, etc.In this paper, first, we describe the current research situation on this project, and introduce the active models researchers mostly adopt. Based on the analysis of the existing models, we summarize the advantages and disadvantages of various methods. Next, we amend the original data. Then, we explain the ARIMA model theory and Bayes theory. Finally, we establish the forecasting model with the amended data, and make a forecast for the next 20 years using the forecasting model. At last we make summarizations and conclusions of the whole paper.