Comparative analysis of housing and stock prices using a regime switching model

This study uses a Markov Switching Intercept Autoregressive Heteroscedasticity (MSIAH) AR model to fit and compare the time series dynamics of housing prices and stock prices in Korea. High and low volatility regimes are considered in the model. Data covers the period from January 1986 to June 2019. The analysis using estimation results reveals that housing price has lower RCM statistics than the stock price, indicating a better regime classification of the housing market. The high-volatility regime is also found to be associated with high-return in the housing market, whereas it is associated with low-return in the stock market. The average regime duration is longer in the housing market than the stock market. The impulse response function analysis indicates that dynamic effects of price changing shocks are stronger and longer in the housing market than the stock market. The analysis also shows that the shock impact on the price is getting larger when the market is in a high-volatility regime.