LAUSR

Abstract The relationship between international trade and capital flow has been extensively discussed. Using bilateral data from 2001 to 2016, we construct international trade and capital flow networks, defined as a weighted graph in which nodes are countries and edges represent trade and capital flow linkages. Centrality is employed as the main variable in this study to represent the importance of a country in the networks. To gain further insight into the network characteristics, we adopt a turning-parameter centrality to combine both the original degree and strength measurements within the weighted network. Moreover, we develop a new indicator called “partner quality centrality” to identify the quality of neighbors. Within the panel cointegration framework, we argue that a positive long-run equilibrium exists between the trade and capital flow networks as constructed. In addition, we employ a panel causality test to investigate the short-run dynamics, indicating that the international capital flow network has predictive power on the trade network from the short-run perspective but not vice versa.