UK macroeconomic forecasting with many predictors: which models forecast best and when do they do so?

Gary Koop, Dimitris Korobilis

Research output: Contribution to journalArticle

23 Citations (Scopus)

Abstract

Block factor methods offer an attractive approach to forecasting with many predictors. These extract the information in these predictors into factors reflecting different blocks of variables (e.g. a price block, a housing block, a financial block, etc.). However, a forecasting model which simply includes all blocks as predictors risks being over-parameterized. Thus, it is desirable to use a methodology which allows for different parsimonious forecasting models to hold at different points in time. In this paper, we use dynamic model averaging and dynamic model selection to achieve this goal. These methods automatically alter the weights attached to different forecasting models as evidence comes in about which has forecast well in the recent past. In an empirical study involving forecasting output growth and inflation using 139 UK monthly time series variables, we find that the set of predictors changes substantially over time. Furthermore, our results show that dynamic model averaging and model selection can greatly improve forecast performance relative to traditional forecasting methods.
LanguageEnglish
Pages2307-2318
Number of pages12
JournalEconomic Modelling
Volume28
Issue number5
DOIs
Publication statusPublished - Sep 2011

Fingerprint

Macroeconomic forecasting
Predictors
Factors
Model averaging
Model selection
Empirical study
Forecasting method
Output growth
Forecast performance
Methodology
Inflation

Keywords

  • Bayesian
  • state space model
  • factor models
  • dynamic model averaging

Cite this

@article{682160b5b9de441eb0a674ecbc24e4eb,
title = "UK macroeconomic forecasting with many predictors: which models forecast best and when do they do so?",
abstract = "Block factor methods offer an attractive approach to forecasting with many predictors. These extract the information in these predictors into factors reflecting different blocks of variables (e.g. a price block, a housing block, a financial block, etc.). However, a forecasting model which simply includes all blocks as predictors risks being over-parameterized. Thus, it is desirable to use a methodology which allows for different parsimonious forecasting models to hold at different points in time. In this paper, we use dynamic model averaging and dynamic model selection to achieve this goal. These methods automatically alter the weights attached to different forecasting models as evidence comes in about which has forecast well in the recent past. In an empirical study involving forecasting output growth and inflation using 139 UK monthly time series variables, we find that the set of predictors changes substantially over time. Furthermore, our results show that dynamic model averaging and model selection can greatly improve forecast performance relative to traditional forecasting methods.",
keywords = "Bayesian, state space model, factor models, dynamic model averaging",
author = "Gary Koop and Dimitris Korobilis",
year = "2011",
month = "9",
doi = "10.1016/j.econmod.2011.04.008",
language = "English",
volume = "28",
pages = "2307--2318",
journal = "Economic Modelling",
issn = "0264-9993",
number = "5",

}

UK macroeconomic forecasting with many predictors : which models forecast best and when do they do so? / Koop, Gary; Korobilis, Dimitris.

In: Economic Modelling, Vol. 28, No. 5, 09.2011, p. 2307-2318.

Research output: Contribution to journalArticle

TY - JOUR

T1 - UK macroeconomic forecasting with many predictors

T2 - Economic Modelling

AU - Koop, Gary

AU - Korobilis, Dimitris

PY - 2011/9

Y1 - 2011/9

N2 - Block factor methods offer an attractive approach to forecasting with many predictors. These extract the information in these predictors into factors reflecting different blocks of variables (e.g. a price block, a housing block, a financial block, etc.). However, a forecasting model which simply includes all blocks as predictors risks being over-parameterized. Thus, it is desirable to use a methodology which allows for different parsimonious forecasting models to hold at different points in time. In this paper, we use dynamic model averaging and dynamic model selection to achieve this goal. These methods automatically alter the weights attached to different forecasting models as evidence comes in about which has forecast well in the recent past. In an empirical study involving forecasting output growth and inflation using 139 UK monthly time series variables, we find that the set of predictors changes substantially over time. Furthermore, our results show that dynamic model averaging and model selection can greatly improve forecast performance relative to traditional forecasting methods.

AB - Block factor methods offer an attractive approach to forecasting with many predictors. These extract the information in these predictors into factors reflecting different blocks of variables (e.g. a price block, a housing block, a financial block, etc.). However, a forecasting model which simply includes all blocks as predictors risks being over-parameterized. Thus, it is desirable to use a methodology which allows for different parsimonious forecasting models to hold at different points in time. In this paper, we use dynamic model averaging and dynamic model selection to achieve this goal. These methods automatically alter the weights attached to different forecasting models as evidence comes in about which has forecast well in the recent past. In an empirical study involving forecasting output growth and inflation using 139 UK monthly time series variables, we find that the set of predictors changes substantially over time. Furthermore, our results show that dynamic model averaging and model selection can greatly improve forecast performance relative to traditional forecasting methods.

KW - Bayesian

KW - state space model

KW - factor models

KW - dynamic model averaging

UR - http://www.scopus.com/inward/record.url?scp=79960929746&partnerID=8YFLogxK

U2 - 10.1016/j.econmod.2011.04.008

DO - 10.1016/j.econmod.2011.04.008

M3 - Article

VL - 28

SP - 2307

EP - 2318

JO - Economic Modelling

JF - Economic Modelling

SN - 0264-9993

IS - 5

ER -