Time varying parameter estimation in macro-econometric models

This paper addresses the issue of Time-Varying Parameter (TVP) estimation, a technique recently introduced in the field of Macroeconometrics, and especially in FAVAR (Factor-Augmented Vector Auto-Regression) modeling. FAVAR is here extended to Vector Error Correction (VEC) methodology to yield a new and enriched model denominated FAVEC. Different from classic VAR/VEC models where Time-Fixed Parameter (TFP) estimation dominates over the entire sample and may be conducive to the “Lucas Critique”, TVP models produce changing parameters which may be utilized by the analyst to infer the dynamics underlying the data process, such as structural breaks, changes in covariances and in parameter significance, and so on. This advantage, however, comes at a high cost represented by burdensome program coding and by the CPU-machine time required to produce multi-draw Gibbs sampling to enable Bayesian parameter estimation. Sizable costs, among others, may also ensue from the construction of impulse responses and variance decompositions for the purpose of policy evaluation. In- and out-sample forecasting applied to competing TFP and TVP models of the US economy and monetary policy during the years 1959-2006, using quarterly observations, produces very interesting results that definitely favor the TVP-FAVEC model representation.