China, as the world second largest economy, is one of the hotspots in both Economics field and Statistics field, so our research’s target country is China. In the Economics field we conclude that there are three gharries pulling one country’s economy, they are Investment, Consumption and Export. This theory can also be applied into Chinese economy, in most of time, we think that the most powerful gharry that pulling Chinese economy is investment. In the investment sector, the biggest proportion is real estate investment, “Real estate investment grew at about a 20 percent average annual rate from 1999 to 2018, more than twice the rate of GDP growth over the same period in China”. In 2014 (Jan-Jun), the real estate investment was about 14% of the total GDP in that period. Connecting with the real estate investment, the commodity residential building selling price is what we concentrate on because it is highly relevant with the people’s life and Chinese economy. People need to live in the house and they need to have enough money to buy a house, real estate sector is an important sector not only in China but for the whole world. A country cannot develop well without a stable housing market, this can be proved through every financial crisis in past decades. Actually, through this we can see that the fluctuation of the housing price in China is connecting with the economy of the whole world. Furthermore, one of China’s special economic patterns is playing an important role for the commodity building selling price which is macro-control and policy effect in economy, this gives the real estate sector in China a large amount of non-determinacy but also gives political economists a lot of chances to research about it. So based on all of the reasons above, prediction for the commodity building selling price and making a statistical model for that is a really attractive and challenging research for all of the statisticians and econometricians, but we think it is worth to have a try. Since there is a big difference between every cities’ economy and policy in China, we decided to use the capital of China, Beijing, as the sample target and try to use the research methods that we have conducted based on this sample to other cities in the future. The data is collected from different official statistical department websites: “China Statistical Yearbook”(2000-2016) , People’s Bank of China, National Bureau of Statistics, Ministry of Finance, Ministry of Education, CEIC Data. There are totally 28 variables in our dataset. In our research, we will make an appropriate statistical model for the commodity residential building sold in Beijing(total) against several other covariates using several different statistical modelling methods such as: multiple linear regression, weighted least square regression, time series model assumptions and ARIMA model, all of the mission is completed by R language and data is cleaning by Python language. In these models we will also test whether Chinese government real estate policies will have a large impact on the residential building sold. Furthermore, we will try to include the important Chinese festival date or other official important date to see the impact of these dates to the model. Finally, we will make a prediction for the commodity residential building sold for the next 12 months and based on these predictions we will make a decision about our future works.
Author: Mingshi Cui
Faculty Advisor: Dr. Jing Zhang, Department of Statistics
You must be logged in to post a comment.