Monetary policy shocks for Sweden

This paper estimates monetary policy shocks for Sweden between 1996-2019. I employ the Romer and Romer (2004) (R&R) approach and use annual forecasts of output growth and inflation to estimate monetary policy shocks. I complement the analysis with shocks from a recursive VAR including output, prices, and the repo rate, as well as a set of high-frequency shocks. A comparison of the three sets of shocks shows that the R&R and VAR shocks are similar, while the high-frequency shocks are fewer and smaller in size. Local projections show expected impulse responses on most economic variables, regardless of data frequency, but responses to the recursive VAR shocks are more in line with textbook findings compared to responses to the R&R and high-frequency shocks. Overall, results are robust to alternative model specifications and lag lengths in local projections.