- | Thursday, December 3
- 12:15 PM
- Kaiserplatz 7-9, 4th floor, Room 4.006
Unsurprising Shocks: Measuring Responses to Monetary Announcements using High-Frequency Data
This paper proposes a new method to construct high-frequency futures-based proxies for the identification of monetary policy shocks in structural VARs. I discuss how the common use of narrow time frames tightly surrounding policy announcements to measure monetary surprises is not sufficient to guarantee their exogeneity. Raw monetary “surprises” are in fact predictable using both private central bank’s forecasts and past information – in the form of both private forecasts and lagged economic and financial data – that were available to market participants well before the announcements. This results in a violation of the assumptions in Proxy SVARs and it (i) inflates the statistics capturing the instruments relevance and (ii) induces a potentially non-trivial bias in the estimated contemporaneous transmission coefficients. Consequences for the estimation of structural IRFs can be dramatic, both qualitatively and quantitatively. Applications on both the US and the UK deliver new proxies that are in line with theoretical desiderata and independent from the model specification and the choice of endogenous variables in the VAR information set. Crucially, identification of monetary policy shocks via the orthogonal proxies is shown to correctly retrieve the contemporaneous transmission coefficients even in small, potentially informationally insufficient VARs.