Factor Network Autoregressions

TL;DR

The paper introduces a novel factor network autoregressive model that captures complex multilayer network structures in high-dimensional time series, providing consistent estimation and useful macroeconomic insights.

Contribution

It develops a tensor-based principal component approach to reduce dimensionality and estimates network effects in high-dimensional time series data.

Findings

Accurate estimation of network factors and coefficients in simulations.

Effective modeling of cross-country GDP growth interdependence.

Good forecasting performance for macroeconomic variables.

Abstract

We propose a factor network autoregressive (FNAR) model for time series with complex network structures. The coefficients of the model reflect many different types of connections between economic agents ("multilayer network"), which are summarized into a smaller number of network matrices ("network factors") through a novel tensor-based principal component approach. We provide consistency and asymptotic normality results for the estimation of the factors, their loadings, and the coefficients of the FNAR, as the number of layers, nodes and time points diverges to infinity. Our approach combines two different dimension-reduction techniques and can be applied to high-dimensional datasets. Simulation results show the goodness of our estimators in finite samples. In an empirical application, we use the FNAR to investigate the cross-country interdependence of GDP growth rates based on a variety of international trade and financial linkages. The model provides a rich characterization of macroeconomic network effects as well as good forecasts of GDP growth rates.