期刊名称：Euro Area Balance of Payments and International Investment Position Statistics
印刷版ISSN：1830-3420
电子版ISSN：1830-3439
出版年度：2018
出版社：European Central Bank
摘要：We propose a class of prior distributions that discipline the long-run behaviorof Vector Autoregressions (VARs). These priors can be naturally elicited using economictheory, which provides guidance on the joint dynamics of macroeconomic time seriesin the long run. Our priors for the long run are conjugate, and can thus be easilyimplemented using dummy observations and combined with other popular priors. InVARs with standard macroeconomic variables, a prior based on the long-run predictionsof a wide class of theoretical models yields substantial improvements in the forecastingperformance.
关键词：Bayesian vector autoregression; forecasting; over tting; initial conditions;hierarchical model.
