Combining Forecasts under Structural Breaks Using Graphical LASSO

Tae-Hwy Lee () and Ekaterina Seregina ()
Additional contact information
Tae-Hwy Lee: Department of Economics, University of California Riverside
Ekaterina Seregina: Colby College

No 202413, Working Papers from University of California at Riverside, Department of Economics

Abstract: In this paper we develop a novel method of combining many forecasts based on Graphical LASSO. We represent forecast errors from different forecasters as a network of interacting entities and generalize network inference in the presence of common factor structure and structural breaks. First, we note that forecasters often use common information and hence make common errors, which makes the forecast errors exhibit common factor structures. We separate common forecast errors from the idiosyncratic errors and exploit sparsity of the precision matrix of the latter. Second, since the network of experts changes over time as a response to unstable environments, we propose Regime-Dependent Factor Graphical LASSO (RD-FGL) that allows factor loadings and idiosyncratic precision matrix to be regime-dependent. The empirical applications to forecasting macroeconomic series using the data of the European Central Bank’s Survey of Professional Forecasters and Federal Reserve Economic Data monthly database demonstrate superior performance of a combined forecast using RD-FGL.Â

Keywords: Common Forecast Errors; Regime Dependent Forecast Combination; Sparse Precision Matrix of Idiosyncratic Errors; Structural Breaks (search for similar items in EconPapers)
JEL-codes: C13 C38 C55 (search for similar items in EconPapers)
Pages: 57 Pages
Date: 2024-12
New Economics Papers: this item is included in nep-ets, nep-for and nep-net
References: Add references at CitEc
Citations:

Downloads: (external link)
https://economics.ucr.edu/repec/ucr/wpaper/202413.pdf First version, 2024 (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:ucr:wpaper:202413

Access Statistics for this paper

More papers in Working Papers from University of California at Riverside, Department of Economics Contact information at EDIRC.
Bibliographic data for series maintained by Kelvin Mac ().