The paper proposes a novel methodology that, by exploiting information available at a higher frequency (e.g. financial variables) allows us to estimate a causal link with the variables sampled at lower frequencies (e.g. GDP, inflation and so on). The results obtained by daily identification (and by computing their monthly or quarterly average) are compared with those obtainable from the common identification based on data at a monthly frequency.
An empirical application of the methodology on the macroeconomic effects of uncertainty in the United States shows that the results change significantly with the frequency at which the identification stage is performed. Unexpected fluctuations in uncertainty shocks in the equity market do not have macroeconomic effects if the analysis is carried out monthly. By contrast, when the same identification strategy is applied to daily data, uncertainty generates a recessionary effect on macroeconomic variables that is more in line with economic theory
No. 1274 - Bridge Proxy-SVAR: estimating the macroeconomic effects of shocks identified at high frequency
Full text
- No. 1274 - Bridge Proxy-SVAR: estimating the macroeconomic effects of shocks identified at high frequency pdf 3.8 MB Data pubblicazione: 27 April 2020