부동산 비기초가격과 부동산 신문담론의 시계열 관계 분석 : 서울 아파트 매매가를 중심으로

This study investigates the relationship between non-fundamental real estate prices and newspaper discourse on real estate, based on the theory that media coverage leads public sentiment in certain ways and that public opinion has a substantial impact on market prices. Non-fundamental real estate prices variables were estimated through fully modified OLS (FM-OLS), and newspaper discourse on real estate variables was constructed through topic modeling. The time series relationship between non-fundamental prices and newspaper discourse was confirmed through Granger causal analysis, impulse response analysis, and variance decomposition analysis. The results were as follows. First, non-fundamental real estate prices were negatively associated with newspaper discourse on real estate policy. Second, a positive relationship existed between non-fundamental real estate prices and information on real estate advertisements. Third, non-fundamental real estate prices had a negative relationship with information on the nearby metropolitan area. This study suggested a method to construct variables that were difficult to quantify, such as policies and information, using big data from newspaper articles. Thereby, this study empirically examined the relationship that was previously assumed to be associated with non-fundamental real estate prices based on a priori or theoretical judgment.