Identification of structural shocks can be subject to nonfundamentalness, as the econometrician may have an information set smaller than the economic agents´i one. How serious is that problem from a quantitative point of view? In this work we propose a simple diagnosis statistics for the quantitative importance of nonfundamentalness in structural VARs. The diagnosis is of interest as nonfundamentalness is not an either/or question, but is a quantitative issue which can be more or less severe. Using our preferred strategy for identifying news shocks, we find that nonfundamentalness is quantitatively unimportant and that news shocks continue to generate significant business cycle type fluctuations when adjust the estimating procedure to take into account the potential nonfundamentalness issue.
Non-Fundamentalness; Business Cycles; SVARs; News;
- C32: Time-Series Models • Dynamic Quantile Regressions • Dynamic Treatment Effect Models • Diffusion Processes
- E32: Business Fluctuations • Cycles
TSE Working Paper, n° 16-738, novembre 2016