Fig. 4: Net-pairwise directional connectedness in the full-sample period.

This figure shows the 66 pairs of carbon, energy and metals markets. The nodes represent each market, and the thickness of the edge shows the degree of the net-pairwise directional connectedness. The arrows going from markets i to j represent net spillovers, that is, the contribution of market i to market j is greater than that of market j to market i. “Net-pairwise spillovers” is the aggregate net-pairwise directional spillovers of the DY model. The “short-term” and “long-term” is the net-pairwise directional spillover in the short and long-term horizons of BK model, respectively. The top panels represent the volatility, skewness and kurtosis network in different frequencies, such as “Volatility (short-term)” represents the network drawn by the net-pairwise directional volatility connectedness in the short-term horizons of BK model in the full-sample period.