Structural VAR models in the Frequency Domain✩

This paper proposes a joint methodology for the identification and inference of structural vector autoregressive models in the frequency domain. We show that identifying restrictions can be written naturally as an asymptotic least squares problem (Gourieroux et al., 1985) in which there is a continuum of nonlinear estimating equations. Following Carrasco and Florens (2000), we then propose a continuum asymptotic least squares estimator (C-ALS) that efficiently exploits the continuum of estimating equations, thereby allowing to obtain optimal consistent estimates of impulse responses and reliable confidence intervals. Moreover, the identifying restrictions can be formally tested using an appropriate J-stat, and the frequency band can be selected using a data-driven procedure. Finally, we provide some Monte Carlo simulations and an application regarding the hours-productivity debate.

Automated summary

This paper proposes a joint methodology for identifying and inferring structural vector autoregressive models in the frequency domain. The authors utilize asymptotic least squares problems to establish identifying restrictions and propose a continuum asymptotic least squares estimator (C-ALS) to efficiently estimate impulse responses and confidence intervals. They also suggest using a formal test and a data-driven procedure to test the identifying restrictions and select the frequency band. Monte Carlo simulations and an application on the hours-productivity debate are provided.