Business Intelligence

Consistent Estimation Of Global VAR Models

Free registration required

Executive Summary

In this paper, the author proposes an Instrumental Variable (IV) estimation procedure to estimate Global VAR (GVAR) models and shows that it leads to consistent and asymptotically normal estimates of the parameters. The author also provides computationally simple conditions that guarantee that the GVAR model is stable. Vector AutoRegressions (VAR) have become a useful part of the toolbox of empirical economists. VARs are used for both model estimation and evaluation, as well as for a-theoretical data analysis. In practical applications in macroeconomics, VAR models are often estimated using data for a particular cross-sectional unit (typically a country), ignoring any possible international linkages.

  • Format: PDF
  • Size: 221.4 KB