Abstract
This paper proposes LASSO estimation specific for panel vector autoregressive (PVAR) models. The penalty term allows for shrinkage for different lags, for shrinkage towards homogeneous coefficients across panel units, for penalization of lags of variables belonging to another cross-sectional unit, and for varying penalization across equations. The penalty parameters therefore build on time series and cross-sectional properties that are commonly found in PVAR models. Simulation results point towards advantages of using the proposed LASSO for PVAR models over ordinary least squares in terms of forecast accuracy. An empirical forecasting application including 20 countries supports these findings.
| Original language | English |
|---|---|
| Pages (from-to) | 1185-1204 |
| Number of pages | 20 |
| Journal | International Journal of Forecasting |
| Volume | 39 |
| Issue number | 3 |
| DOIs | |
| Publication status | Published - 1 Jul 2023 |
Bibliographical note
Funding Information:I thank the Editor, two anonymous referees, Helmut Lütkepohl, Efrem Castelnuovo, Richard Paap, Andreas Pick, Wendun Wang and Tomasz Woźniak for insightful comments. Earlier versions of the paper were presented at SMYE 2017, Halle, Barcelona GSE Summer Forum 2017, Barcelona, EEA-ESEM 2017, Lisbon, annual meeting of the Verein für Socialpolitik 2017, Vienna, University of Sydney, University of Melbourne, Workshop Empirical Macroeconomics at Freie Universität Berlin, University of Konstanz, University of Vienna, Erasmus University Rotterdam, University of Groningen, University of Mannheim, Stockholm School of Economics, NESG 2019, Amsterdam, and at internal seminars at DIW Berlin. Comments by the participants are gratefully acknowledged. This work was carried out on the Dutch national e-infrastructure with the support of SURF Cooperative.
Publisher Copyright:
© 2022 The Author(s)