개별 경제지표에 의한 부동산 경기전망에 관한 연구 - 건물유형별 및 토지거래건수를 중심으로 -

Abstract

The purpose of this study is to predict the business cycle of the Korean Real Market by the transaction volume of real estate forecasted by seasonal-ARIMA model. In order to this purpose, this study reviews the concepts of seasonal-ARIMA model, the advantages of ARMA model and limitations of this model. Also to find the best model, the seasonal-ARIMA procedures, the unit root test by using Augmented Dickey Fuller Test(ADF Test), identification of Auto Correlation Function(ACF) and Partial Correlation Function(PACF) are proceeded. By using the these procedures, model is selected as a best one and used to forecast volume of real estate transaction in Korea. The findings of this study are as following ; The transaction volume of real estate, which are classified by building types and land, show the seasonal effect for the period of 12 months and rapidly decrease last year, Also, according to the forecasted results, the volumes of single family home, apartment and land will decrease compared to the those before one year. On the contrary, it is forecasted the transaction volume of retail and office building will increase.