Abstract
The definition of a news surprise plays a crucial role in the analysis of the impact of unexpected macroeconomic news announcements. In this paper, we study the properties of the most commonly used measure of news surprise, defined as the difference between the announced and expected value of the indicator. Due to the high vulnerability of this measure to outliers, we consider alternative definitions of macroeconomic surprises. Based on the analysis of announcements of 15 American macroeconomic indicators, we show that taking into account the heterogeneity of analysts’ forecasts or the variability of the previous surprises, noticeably improves the properties of the distribution of surprise measures. An additional study performed with the use of a dynamic model proves a strong linear relationship between surprise measures and WIG20 returns in the first five minutes after news announcements.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
