A maximum entropy network reconstruction of macroeconomic models

TL;DR

This paper introduces a maximum entropy approach to reconstruct macroeconomic networks from partial data, focusing on network topology and money flows, and compares models against empirical data.

Contribution

It develops a novel maximum entropy method for macroeconomic network reconstruction that incorporates empirical constraints and compares different models.

Findings

Reconstructed networks match empirical properties.

Maximum entropy models effectively capture network sparsity.

Comparison of models highlights key structural differences.

Abstract

In this article the problem of reconstructing the pattern of connection between agents from partial empirical data in a macro-economic model is addressed, given a set of behavioral equations. This systemic point of view puts the focus on distributional and network effects, rather than time-dependence. Using the theory of complex networks we compare several models to reconstruct both the topology and the flows of money of the different types of monetary transactions, while imposing a series of constraints related to national accounts, and to empirical network sparsity. Some properties of reconstructed networks are compared with their empirical counterpart.