VAR Analysis and the Great Moderation

摘要

Most analyses of the US Great Moderation have been based on structural VAR methods, and have consistently pointed toward good luck as the main explanation for the greater macro-economic stability of recent years. Based on an estimated New-Keynesian model in which the only sources of change are the move from passive to active monetary policy, and the presence of sunspots under indeterminacy, we show that VAR users may misinterpret good policy for good luck. In particular, the estimated DGP exhibits decreases in population in both variances and innovation variances for all series. Policy counterfactuals based on the theoretical structural VAR representations of the model under the two regimes fail to capture the truth, whereas impulse-response functions to a monetary policy shock exhibit little change across regimes. Since these results are in line with those found in the structural VAR-based literature on the Great Moderation, our analysis suggests that existing VAR evidence is compatible with the "good policy" explanation of the Great Moderation.