Estimating GVAR weight matrices

Abstract

This paper aims to illustrate how weight matrices that are needed to construct foreign variable vectors in Global Vector Autoregressive (GVAR) models can be estimated jointly with the GVAR's parameters. An application to real GDP and consumption expenditure price inflation as well as a controlled Monte Carlo simulation serve to highlight that 1) In the application at hand, the estimated weights differ for some countries significantly from trade-based ones that are traditionally employed in that context; 2) misspecified weights might bias the GVAR estimate and therefore distort its dynamics; 3) using estimated GVAR weights instead of trade-based ones (to the extent that they differ and the latter bias the global model estimates) shall enhance the out-of-sample forecast performance of the GVAR. Devising a method for estimating GVAR weights is particularly useful for contexts in which it is not obvious how weights could otherwise be constructed from data.

Cross‐sectional Dependence and Spillovers in Space and Time: Where Spatial Econometrics and Global Var Models Meet

“…Recent econometric-theoretical contributions using socioeconomic variables to model W are of Kelejian and Piras (2014) and Qu and Lee (2015). In the GVAR model literature, there has been an attempt to estimate the weights along with all other GVAR model parameters (Gross, 2018). The identifying assumptions for the weights to be estimable include one, stating that a significant relation between a unit and the "rest of the world" must exist, as otherwise they become nuisance parameters that may be altered without changing the likelihood of the model.…”

“…is an N × k matrix of exogenous explanatory variables, and W an N × N nonnegative matrix of known constants describing the linkages among the cross-sectional units. 4 The terms τ , δ, and η denote the response parameters of, respectively, the time-lagged dependent variable Y t−1 , the spatially lagged dependent variable WY t , and the spatially and time-lagged dependent variable WY t−1 , while β and ϑ are k × 1 vectors of response parameters of the exogenous explanatory variables. The N × 1 vector α = (α 1 , .…”

“…Due to the slope heterogeneity assumption, they would not be estimable in a pure cross-section setting. 4. The spatial Durbin model (SDM), in autoregressive distributed lag panel form, is one of the most general specifications in spatial econometrics.…”

Spatiotemporal Analysis of Regional Systems

“…As a result, a new class of spatial VAR (SpVAR) models has recently appeared in the spatial econometrics literature (see, e.g., Beenstock and Felsenstein 2007;Canova and Ciccarelli 2009;Di Giacinto 2010;Dewachter, Houssa, and Toffano 2012;Márquez, Ramajo, and Hewings 2010, 2013, 2014a. SpVARs are a general type of multivariate vector autoregressions that include spatial as well as temporal lags of the state variables.…”

Raising the bar (11)

“…The fact that direct and indirect effects estimates can be different for different units is also the topic of Gross (2018, in this issue), but from the perspective of a global vector autoregressive (GVAR) model. Unfortunately, many researchers are not aware that spatial econometric and GVAR studies have many similarities.…”