We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Economics > Econometrics

Title: Hierarchical Regularizers for Reverse Unrestricted Mixed Data Sampling Regressions

Abstract: Reverse Unrestricted MIxed DAta Sampling (RU-MIDAS) regressions are used to model high-frequency responses by means of low-frequency variables. However, due to the periodic structure of RU-MIDAS regressions, the dimensionality grows quickly if the frequency mismatch between the high- and low-frequency variables is large. Additionally the number of high-frequency observations available for estimation decreases. We propose to counteract this reduction in sample size by pooling the high-frequency coefficients and further reduce the dimensionality through a sparsity-inducing convex regularizer that accounts for the temporal ordering among the different lags. To this end, the regularizer prioritizes the inclusion of lagged coefficients according to the recency of the information they contain. We demonstrate the proposed method on an empirical application for daily realized volatility forecasting where we explore whether modeling high-frequency volatility data in terms of low-frequency macroeconomic data pays off.
Subjects: Econometrics (econ.EM); Methodology (stat.ME)
Cite as: arXiv:2301.10592 [econ.EM]
  (or arXiv:2301.10592v1 [econ.EM] for this version)

Submission history

From: Ines Wilms [view email]
[v1] Wed, 25 Jan 2023 13:53:16 GMT (129kb,D)

Link back to: arXiv, form interface, contact.