Empirically-transformed linear opinion pools

Anthony Garratt; Timo Henckel; Shaun P. Vahey

doi:10.1016/j.ijforecast.2022.02.003

Empirically-transformed linear opinion pools

Anthony Garratt^*
, Timo Henckel
, Shaun P. Vahey

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

4 Citations (Scopus)

Abstract

The linear opinion pool (LOP) produces potentially non-Gaussian combination forecast densities. In this paper, we propose a computationally convenient transformation for the LOP to mirror the non-Gaussianity exhibited by the target variable. Our methodology involves a Smirnov transform to reshape the LOP combination forecasts using the empirical cumulative distribution function. We illustrate our empirically transformed opinion pool (EtLOP) approach with an application examining quarterly real-time forecasts for U.S. inflation evaluated on a sample from 1990:1 to 2020:2. EtLOP improves performance by approximately 10% to 30% in terms of the continuous ranked probability score across forecasting horizons.

Original language	English
Pages (from-to)	736-753
Number of pages	18
Journal	International Journal of Forecasting
Volume	39
Issue number	2
DOIs	https://doi.org/10.1016/j.ijforecast.2022.02.003
Publication status	Published - 1 Apr 2023

Access to Document

10.1016/j.ijforecast.2022.02.003

Cite this

@article{32f6c6851b484a668c757ba94ec6e4ce,

title = "Empirically-transformed linear opinion pools",

abstract = "The linear opinion pool (LOP) produces potentially non-Gaussian combination forecast densities. In this paper, we propose a computationally convenient transformation for the LOP to mirror the non-Gaussianity exhibited by the target variable. Our methodology involves a Smirnov transform to reshape the LOP combination forecasts using the empirical cumulative distribution function. We illustrate our empirically transformed opinion pool (EtLOP) approach with an application examining quarterly real-time forecasts for U.S. inflation evaluated on a sample from 1990:1 to 2020:2. EtLOP improves performance by approximately 10\% to 30\% in terms of the continuous ranked probability score across forecasting horizons.",

keywords = "Density forecast combination, Inflation, Linear opinion pool, Smirnov transform",

author = "Anthony Garratt and Timo Henckel and Vahey, \{Shaun P.\}",

note = "Publisher Copyright: {\textcopyright} 2022 International Institute of Forecasters",

year = "2023",

month = apr,

day = "1",

doi = "10.1016/j.ijforecast.2022.02.003",

language = "English",

volume = "39",

pages = "736--753",

journal = "International Journal of Forecasting",

issn = "0169-2070",

publisher = "Elsevier B.V.",

number = "2",

}

TY - JOUR

T1 - Empirically-transformed linear opinion pools

AU - Garratt, Anthony

AU - Henckel, Timo

AU - Vahey, Shaun P.

PY - 2023/4/1

Y1 - 2023/4/1

N2 - The linear opinion pool (LOP) produces potentially non-Gaussian combination forecast densities. In this paper, we propose a computationally convenient transformation for the LOP to mirror the non-Gaussianity exhibited by the target variable. Our methodology involves a Smirnov transform to reshape the LOP combination forecasts using the empirical cumulative distribution function. We illustrate our empirically transformed opinion pool (EtLOP) approach with an application examining quarterly real-time forecasts for U.S. inflation evaluated on a sample from 1990:1 to 2020:2. EtLOP improves performance by approximately 10% to 30% in terms of the continuous ranked probability score across forecasting horizons.

AB - The linear opinion pool (LOP) produces potentially non-Gaussian combination forecast densities. In this paper, we propose a computationally convenient transformation for the LOP to mirror the non-Gaussianity exhibited by the target variable. Our methodology involves a Smirnov transform to reshape the LOP combination forecasts using the empirical cumulative distribution function. We illustrate our empirically transformed opinion pool (EtLOP) approach with an application examining quarterly real-time forecasts for U.S. inflation evaluated on a sample from 1990:1 to 2020:2. EtLOP improves performance by approximately 10% to 30% in terms of the continuous ranked probability score across forecasting horizons.

KW - Density forecast combination

KW - Inflation

KW - Linear opinion pool

KW - Smirnov transform

UR - https://www.scopus.com/pages/publications/85126888037

U2 - 10.1016/j.ijforecast.2022.02.003

DO - 10.1016/j.ijforecast.2022.02.003

M3 - Article

SN - 0169-2070

VL - 39

SP - 736

EP - 753

JO - International Journal of Forecasting

JF - International Journal of Forecasting

IS - 2

ER -

Empirically-transformed linear opinion pools

Abstract

Access to Document

Other files and links

Fingerprint

Cite this