Estimating Macroeconomic News and Surprise Shocks
Abstract: The importance of understanding the economic effects of TFP news and surprise shocks is widely recognized in the literature. A common VAR approach is to identify responses to TFP news shocks by maximizing the variance share of TFP over a long horizon. Under suitable conditions, this approach also implies an estimate of the surprise shock. We find that these TFP max share estimators tend to be strongly biased when applied to data generated from DSGE models with shock processes that match the TFP moments in the data, both in the presence of TFP measurement error and in its absence. Incorporating a measure of TFP news into the VAR model and adapting the identification strategy substantially reduces the bias and RMSE of the impulse response estimates, even when there is sizable measurement error in the news variable. When applying this method to the data, we find that news shocks are slower to diffuse to TFP and have a smaller effect on real activity than implied by the TFP max share method.
DOI: https://doi.org/10.24149/wp2304r2
Appendix DOI: https://doi.org/10.24149/wp2304app
Read paper | See all Working Papers