Title: Hierarchical regularizers for reverse unrestricted MIDAS
Authors: Marie Ternes - Maastricht University (Netherlands) [presenting]
Alain Hecq - Maastricht University (Netherlands)
Ines Wilms - Maastricht University (Netherlands)
Abstract: Reverse unrestricted MIDAS (RU-MIDAS) regressions are used to model high-frequency variables by means of low-frequency variables. However, in practice, the dimensionality of RU-MIDAS grows quickly due to the frequency mismatch between the high- and low-frequency components and the number of explanatory variables included. We propose tackling dimensionality through sparsity-inducing convex regularizers built upon the group lasso with nested groups. The regularizer encourages hierarchical sparsity patterns by prioritizing the inclusion of coefficients according to the recency of the information they contain.