CMStatistics 2021: Start Registration
View Submission - CFE
A1661
Title: REMIS (Retrieval from Mixed Sampling Frequency) for VAR(MA)s Authors:  Philipp Gersing - Vienna University of Technology (Austria) [presenting]
Manfred Deistler - Vienna University of Technology (Austria)
Abstract: A novel estimation technique for a general approach to mixed frequency data is proposed which we label REtrieval from MIxed Sampling Frequency data (REMIS): Given data observed at mixed sampling frequency, we retrieve the parameters of an underlying high frequency VAR(MA) system. A canonical state space representation is derived for the blocked process which second moments correspond to all those second moments observable under mixed frequency. Based on this representation, we derive a consistent estimator for the interpolation of the missing second moments. The high-frequency parameters are then obtained by Yule Walker estimation with the full set of second moments. We demonstrate the effectiveness of our method compared to other techniques in a simulation study and on an empirical application for GDP imputation and forecasting.