주택가격 변동성의 비대칭적 반응에 관한 실증적 연구

An Empirical Study on the Asymmetric Response of Housing Sales Price Volatility

This paper empirically analyzes the asymmetric effect of housing price volatility to the type of information using GJR-GARCH model. The data used for the empirical analysis is the monthly housing price indexes of five metropolitan cities including Seoul. The empirical results are as follows: Firstly, the GJR-GARCH model was found to be more suitable than the GARCH model for analyzing the asymmetric effect of housing price volatility to the type of information. This means that the GJR-GARCH model can be usefully applied to analyze the asymmetric response of housing price volatility as well as the financial time series. Secondly, according to the type of information, all five major cities including Seoul showed time-varying asymmetric housing price volatility. Seoul, Busan, Daegu and Daejeon showed greater volatility in unexpected good news than unexpected bad news, but Gwangju showed more volatility in unexpected bad news than unexpected good news. This asymmetric effect was found not to be uniform due to the locality and specificity of the housing market but to be different for each region. Thirdly, the sensitivity of volatility was also higher in Gwangju than in other regions. Depending on the type of information, there is a need to establish sophisticated housing price volatility prediction models and portfolio management strategies.