Out-of-sample evaluation of macro announcements, linearity, long memory, heterogeneity and jumps in mini-futures markets

• Order of significance of volatility properties: long memory, linearity, jumps and heterogeneity.
• Not all categories of macroeconomic announcements are significant in mini-futures markets.
• The trade balance and producer price index announcements are the most significant announcements.
• The ARFIMA model best forecast realized volatility in announcement days.

This paper examines the significance of macroeconomic announcements, linearity, long memory, heterogeneity and jumps via the out-of-sample forecasting performance in mini-futures markets. The property of long memory is the most significant. Second in-class is linearity. Then, comes the property of jumps and finally heterogeneity. The property of the effect of macroeconomic announcements is evident only for few categories of announcements. The trade balance and producer price index are the most significant announcements across mini-futures markets and evaluation criteria.