( 2019) used this approach and analyzed the bipartite network of banks and firms using the Bernoulli and the two-star model. Another approach is considering the bipartite network of trade exchanges. ( 2019a, 2019b) study trades taking into account firms and their trade relations. One may study the trade network from the firm-level viewpoint for example Chakraborty et al. Trade relation networks can be constructed and analyzed from different perspectives. We aim to model the global trade network from a social network perspective and incorporate different features available on countries and their relations in order to find the factors that affect the structure formation of this network. The main advantage of the ERGM method is its capability to incorporate different structural and non-structural features and even other networks into the analyses. In this study, we use exponential random graph models (ERGMs), a family of models suitable to model the formation of dyads in relational data like network datasets. So, to do an all-embracing analysis of the global trade network, one can add the features available on the countries and the network itself and take into account all the different aspects of this network. An example of this would be including the data available on countries’ official languages, landlockedness, and country distances. Observing the global trade network from different perspectives, various features and information other than the economic and political ones can be considered for analysis, which makes the study of this network an interdisciplinary area. Features of the global trade network that pertain to the network’s structure, such as the triangular structures and other characteristics that originate from the network of the exchanges between countries, are pretty informative about the network processes and this can justify our endeavor to incorporate the structure of the network into our analysis. However, with the advances in social network analysis, more importance can be put on the network’s structure when studying such data. The network structure of the global trade network can be glossed over and the result would be just data containing information about the involving countries and the corresponding amount of export among them. Do factors other than political and economic ones influence the formation of the trade network?Īll these questions and more may be answered by studying the trade network structure and investigating the latent features extracted from it.