Asymmetric Risk Spillover Networks and Risk Contagion Driver in Chinese Financial Markets: The Perspective of Economic Policy Uncertainty

Figure 3

Interinstitutional risk spillover networks (average). Consistent with the risk spillover networks constructed for each sector of Chinese financial markets, the return series of 19 financial institutions, and 5 real estate companies are all stationary time series. Based on the lag order test, we construct a second-order VAR model and calculate the proportion of variation in forecast error for the forward 10-step forecast. In a single risk spillover network, network nodes and connected edges are identified using the same color for institutions belonging to the same sector. The node size of the interinstitutional risk spillover network corresponds to the market capitalization of the institution. The larger the market capitalization, the larger the network node, and the thickness of the contiguous edges of the network nodes indicates the strength of the risk contagion level. The thicker the contiguous edges of the network, the stronger the risk contagion effect. (a) The upper triangular network of risk spillover. (b) The lower triangular network of risk spillover.

(a)

(b)