Simulation Evidence on Theory‐based and Statistical Identification under Volatility Breaks

Abstract

Beside a priori theoretical assumptions on instantaneous or long‐run effects of structural shocks, sign restrictions have become a prominent means for structural vector autoregressive (SVAR) analysis. Moreover, changes in second order moments of systems of time series can be fruitfully exploited for identification purposes in SVARs. By means of Monte Carlo studies, we examine to what degree theory‐based and statistical identification approaches offer an accurate quantification of the true structural relations in a standard model for monetary policy analysis. Subsequently, we discuss how identifying information from theory‐based and statistical approaches can be combined on the basis of a low‐dimensional empirical model of US monetary policy.

Statistical Identification in Svars - Monte Carlo Experiments and a Comparative Assessment of the Role of Economic Uncertainties for the US Business Cycle

“…afterwards. The magnitude of the variance shift is empirically plausible and in line with comparable simulation studies (e.g., Cavaliere et al (2010) and Herwartz and Plödt (2016a)).…”

“…3.1.1 Autoregressive dynamics VAR processes y t are generated by means of an economically reasonable simulation framework as described in Herwartz and Plödt (2016a). The employed DGP resembles a log-linearized version of a stylized three-equation DSGE model comprising the output gap (x t ), inflation (π t ) and nominal interest rates (r t ) (Gertler et al, 1999;Carlstrom et al, 2009;Castelnuovo, 2013Castelnuovo, , 2012Castelnuovo, , 2016.…”

“…In this respect, the issue of so-called 'shock labelling' is an important modelling step in the context of statistical identification. Herwartz and Lütkepohl (2014) and Herwartz and Plödt (2016a) show how statistical identification approaches can be combined with external information garnered from economic theory. Secondly, yet having a viable variety of alternative statistical approaches to the modelling oflastly -latent structural relationships at hand, the uniqueness of shocks is somehow traded against the multitude of identification schemes.…”

“…Lütkepohl and Netšunajev (2017a) review alternative heteroskedasticity-based models by pointing out their advantages and drawbacks, while Lütkepohl and Schlaak (2018) provide guidelines on how to choose between these models conditional on distinct forms of underlying covariance changes. Herwartz and Plödt (2016a) compare stylized sign restrictions and identification by means of covariance shifts. Yet, however, little is known about how independence-based methods perform under covariance changes compared with sign restrictions and heteroskedasticity-based identification schemes.…”

“…In the first part of this study and similar to Herwartz and Plödt (2016a), we assess the relative merits of alternative identification schemes (theory-based sign restrictions and statistical techniques) using the log linearized counterpart of a stylized dynamic stochastic general equilibrium (DSGE) model for data generation. Our simulation set-up highlights numerous aspects of performance characteristics conditional on the underlying data generating process (DGP).…”