Andrey Duván Rincón-Torres, 
Kimberly Rojas-Silva, 
The most recent
Abstract
We study the effect of macroeconomic announcements surprises on Colombian treasury bond spot rates in the medium term. For this, we employ a two-step regression approach proposed by Altavilla, Giannone and Modugno (2017), which takes into account the high frequency response to these surprises while filtering out the noise in the estimation of its medium to long term effect. We found that the share of variation of one day Colombian treasury bond spot rates changes explained by these surprises lies below 10%. Moreover, Colombian macroeconomic announcement surprises other than the nominal exchange rate depreciation do not seem to significantly affect spot rate changes, although important ones have big (but not significant) effect. Furthermore, the explanatory power of macroeconomic news surprises increases substantially at longer horizons, i.e. monthly and quarterly changes, reaching 34% for the latter. These results arise from the fact that spot rate changes show a delayed effect to shocks. This is mostly due to the features of the shocks contained in the error and the persistence of macroeconomic news surprises effect’s. Our results are robust to the appetite for risk of international investors measure employed in the model.