Ripple effect and predictive power of housing market returns: evidence from the Gangnam area

Abstract

We measure the ripple effect from the Gangnam district and construct the "Gangnam ripple index", which represents the influential role of the Gangnam area in the Korean housing market. We present empirical evidence that a ripple effect can serve as a reliable predictor of Korean housing market returns and propose a plausible explanation for its predictive power. To address endogeneity issues when using the ripple index as a predictor variable in a regression model, we apply a difference-in-differences method. The Gangnam ripple index positively correlates with housing market returns across various forecasting horizons. Its predictive power becomes stronger for a subsample of small apartments owing to their high exchangeability and for the bullish markets where investor sentiment drives housing prices. Moreover, our findings suggest that an investor sentiment-based approach accounts for part of the predictive power of the Gangnam ripple variable.