In an influential recent paper, Beaudry and Portier (2006) propose a sequential approach for identifying technological news shocks. Thereby, the correlation coefficient between news shocks of a short-run identification scheme and technology shocks of a long-run identification scheme in the VAR framework measures the extent to which news incorporated into forward-looking variables could reflect future technological developments. While structural VARs can potentially provide a useful guide for modelers as well as policy-makers, the ability of such models to recuperate structural shocks in general and news shocks in particular from the data is a contentious issue in the literature. In the current paper, I find by means of Monte Carlo simulations that the sequential approach can be quite successful in recuperating technological news shocks from artificial data.
News Shocks Identification Structural Vector Autoregressive Model