This paper introduces the notion of nowcasting causality for mixed-frequency VARs as the mixed-frequency version of instantaneous causality. We analyze the relationship between nowcasting and Granger causality in the mixed-frequency VAR setting of Ghysels (2012) and illustrate that nowcasting causality can have a crucial impact on the significance of contemporaneous or lagged high-frequency variables in standard MIDAS regression models.
|Place of Publication||Maastricht|
|Publisher||Maastricht University, Graduate School of Business and Economics|
|Publication status||Published - 1 Jan 2013|
|Series||GSBE Research Memoranda|
Götz, T. B., & Hecq, A. W. (2013). Nowcasting causality in mixed frequency vector autoregressive models. Maastricht University, Graduate School of Business and Economics. GSBE Research Memoranda, No. 050