The pass-through of macro variable to volatility co-movement among U.S. currency and commodity futures markets system

Abstract

This paper explores how the U.S. macro variable influences volatility co-movement in currency and commodity futures markets system. It does this by using the Dynamic Equicorrelation-Mixed Data Sampling-X model, and then calculating the daily realized volatility (RV), good volatility (GV), and bad volatility (BV) of 22 futures using 5-min high-frequency data. The Hodrick-Prescott filter method is applied to compute the raw, cycle, and trend components of the news for 17 macro variables and 4 principal components of these macro variables. There are three key study findings. First, the raw component of monetary policy uncertainty is the best fit for RV co-movement, while the raw component of trade policy uncertainty is the best fit for GV and BV co-movement. Second, almost all macro variables show that the trend component news do not affect volatility co-movement. Finally, the duration of the impact of macro variables exceeds 4 months, while the influence of raw news on BV co-movement is generally shorter. The macro variable information also helps currency and commodity futures investors to make global minimum variance portfolio optimization.