Endogenous Network Effects in Global Trade: A Computational Analysis of Dependency Evolution
摘要
International trade operates as a networked system where trade interdependence between two countries is indirectly shaped by dependencies among other nations within the network. While existing studies predominantly concentrate on bilateral trade relationships, they often neglect the complex dynamics of multilateral network interactions. To address this gap, we first analyze structural features of global commodity trade networks (2000–2022) using Gephi software and then employ a Temporal Exponential Random Graph Model (TERGM) to examine the formation and evolution of the Belt and Road trade network. Results demonstrate that endogenous structural dependencies and temporal dynamics critically govern trade network formation. The interplay between structural cohesion and temporal persistence emerges as a dual mechanism that sustains network stability while allowing evolutionary adaptation. By shifting from a bilateral to a systemic network perspective, this study extends conventional trade analysis frameworks and provides actionable insights for strengthening multilateral economic cooperation. Our methodology advances computational economics research through a network-oriented paradigm for modeling global trade dynamics.