RESUMEN
To enhance the international division of labor of developing countries and scientifically respond to the risks and conflicts in the network system, it is important to examine the evolutionary characteristics of the global value-added trade network, simulate the impact of risk shocks, and propose corresponding measures. Based on UNCTAD-Eora value-added trade data, this paper measured and evaluated the evolution characteristics of the global value-added trade network from 2003 to 2018 using the social network analysis and value-added decomposition methods. Then we analyze the impact of risk shocks on the evolution of the trade network using the bootstrap percolation model, building global trade networks and proposing countermeasures. The results show that the global value-added trade network has formed a complex structure and structurally stable distribution pattern, with Germany, China, and the U.S. as the core and the most crucial supports. Among which, China's core position is mainly due to the rapid rise in its export center status. The trade benefits of the three core countries are both competitive and complementary along the "One Belt and One Road". Furthermore, simulations of bootstrap percolation model reveals that the adoption of trade protection policies (caused by poor institutional quality) by different countries will spread and diffuse non-linearly in the network, and the impacts triggered by low-centered countries are comparatively more widespread. By improving defense capabilities and changing the network structure, the "cascading" impacts of trade policy uncertainty can be reduced.
