Ripple effect and return predictability in the Korean housing market

Abstract

We provide empirical evidence that the ripple effect in the Korean urban housing market can be a good predictor of housing market returns and suggest a possible explanation for the predictability of the ripple effect. We calculated a measure for ripple effect using Diebold and Yilmaz spillover methodology and denoted it as gangnam ripple index. We apply difference-in-difference methodology to gain robustness for the measure, ripple index. Our study show that gangnam ripple positively predicts housing market returns in various forecasting horizons. We find that gangnam ripple predicts better for small-size house samples because of high exchangeability. Also, gangnam ripple predicts well in the bullish market. Furthermore, we examine a possible explanation based on expectation sentiment.