Immunization of systemic risk in trade–investment networks

Abstract

By constructing trade–investment networks and modeling the contagion of systemic risk in trade and investment channels within different countries, we design three immune strategies, namely, Random Immune, Target Immune, and Neighbor Immune, and compare the inhibitory effect of these strategies on the shock contagion of multilayer networks. Using bilateral trade and portfolio investment data between China and OECD countries from 2003 to 2019, results show that first, independent of the source of contagion, the network system with a self-healing capability always reaches the steady state after a certain period of the risk transmission process. The investment network is more resilient to systemic risk than the trade network. Second, Random Immune is inferior in intervening on important countries and currently provides weaker protection for trade–investment​ networks than two other types of immunization strategies because of the high concentration of trade and investment events in the networks. Third, Target Immune aiming at controlling the most important country, the US, is not optimized for improving risk resistance of countries cooperating scarcely with the US, but its optimization efficiency continues to increase as the significance of the US in the system keeps rising. Fourth, among the four Neighbor Immune approaches, Asset-Neighbor Immune provides the superior and consistently enhanced protection for the trade–investment networks, and Liability-Neighbor Immune has the dominant position in terms of the enhancement for risk resistance of each country’s trade and investment channels. Investment Asset-Neighbor Immune or Investment Liability-Neighbor Immune, which effectively reduces the negative impact of systemic risk on the whole system and all countries, is proposed to be implemented in trade–investment networks.