This paper investigates the focused information criterion and plug-in average for vector autoregressive models with local-to-zero misspecication. These methods have the advantage of focusing on a quantity of interest rather than aiming at overall model t. Any (suciently regular) function of the parameters can be used as a quantity of interest. We determine the asymptotic properties and elaborate on the role of the locally misspecied parameters. In particular, we show that the inability to consistently estimate locally misspecied parameters translates into suboptimal selection and averaging. We apply this framework to impulse response analysis. A Monte Carlo simulation study supports our claims.

Focused information criteria, frequentist model averaging, impulse responses, local misspecication, model selection, model uncertainty, vector autoregressive models
Econometric and Statistical Methods: General (jel C1), Model Construction and Estimation (jel C51), Forecasting and Other Model Applications (jel C53), Quantitative Policy Modeling (jel C54),
Econometric Reviews
Department of Econometrics

Lohmeyer, J, Palm, F.C., Reuvers, J.W.N., & Urbain, J-P. (2018). Focused Information Criterion for Locally Misspecified Vector Autoregressive Models. Econometric Reviews, 38(7), 763–792. doi:10.1080/07474938.2017.1409410