We explore the higher order linkages between commodity markets and global financial markets. We focus on spillovers of realized good and bad volatilities, realized sign jump variation, realized skewness, and realized kurtosis. Our results show that the measurement of risk spillovers is sensitive to the definition of risk used in their construction. Asymmetries between good and bad volatility transmission matter, and results when jumps and higher order risk measures are considered are substantially different from those obtained when traditional volatility measures are used. We provide empirical support for theoretical asset pricing models that conduct the optimization required for portfolio balancing in the mean-variance-skewness space by showing that risk diversification opportunities vary greatly when one considers variance or skewness as the fundamental proxy for risk.
