Is volatility spillover enough for investor decisions? A new viewpoint from higher moments

Abstract

This paper provides a new viewpoint on the time and frequency dynamics of the spillover effects among eight major world equity market indexes. We extend the Diebold–Yilmaz approach and the Barunilk and Krehik methodology to estimate and measure the skewness spillover. Our empirical results indicate that the total skewness spillover is far smaller than the total volatility spillover among all markets. Although both volatility spillover and skewness spillover vary with time, the skewness remains relatively smooth and varies gradually when extreme events occur, while the total volatility spillover changes more rapidly and dramatically. Moreover, we observed that most skewness spillover is generated in the short term (1–5 days), while most volatility spillover is produced over the long term (over 21 days).