Forecasting economic time series using score-driven dynamic models with mixed-data sampling

Research output: Contribution to JournalArticleAcademicpeer-review

Abstract

We introduce a mixed-frequency score-driven dynamic model for multiple time series where the score contributions from high-frequency variables are transformed by means of a mixed-data sampling weighting scheme. The resulting dynamic model delivers a flexible and easy-to-implement framework for the forecasting of low-frequency time series variables through the use of timely information from high-frequency variables. We verify the in-sample and out-of-sample performances of the model in an empirical study on the forecasting of U.S. headline inflation and GDP growth. In particular, we forecast monthly headline inflation using daily oil prices and quarterly GDP growth using a measure of financial risk. The forecasting results and other findings are promising. Our proposed score-driven dynamic model with mixed-data sampling weighting outperforms competing models in terms of both point and density forecasts.

Original languageEnglish
JournalInternational Journal of Forecasting
DOIs
Publication statusAccepted/In press - 1 Jan 2019

Fingerprint

Economic forecasting
Sampling
Weighting
GDP growth
Inflation
Point forecasts
Multiple time series
Oil prices
Financial risk
Empirical study
Density forecasts

Keywords

  • Generalized autoregressive score models
  • Gross domestic product
  • Inflation
  • Mixed frequency time series
  • Time-varying parameters

Cite this

@article{d997894b818d4d9e9c3f32bb5c264680,
title = "Forecasting economic time series using score-driven dynamic models with mixed-data sampling",
abstract = "We introduce a mixed-frequency score-driven dynamic model for multiple time series where the score contributions from high-frequency variables are transformed by means of a mixed-data sampling weighting scheme. The resulting dynamic model delivers a flexible and easy-to-implement framework for the forecasting of low-frequency time series variables through the use of timely information from high-frequency variables. We verify the in-sample and out-of-sample performances of the model in an empirical study on the forecasting of U.S. headline inflation and GDP growth. In particular, we forecast monthly headline inflation using daily oil prices and quarterly GDP growth using a measure of financial risk. The forecasting results and other findings are promising. Our proposed score-driven dynamic model with mixed-data sampling weighting outperforms competing models in terms of both point and density forecasts.",
keywords = "Generalized autoregressive score models, Gross domestic product, Inflation, Mixed frequency time series, Time-varying parameters",
author = "Paolo Gorgi and Koopman, {Siem Jan} and Mengheng Li",
year = "2019",
month = "1",
day = "1",
doi = "10.1016/j.ijforecast.2018.11.005",
language = "English",
journal = "International Journal of Forecasting",
issn = "0169-2070",
publisher = "Elsevier",

}

Forecasting economic time series using score-driven dynamic models with mixed-data sampling. / Gorgi, Paolo; Koopman, Siem Jan; Li, Mengheng.

In: International Journal of Forecasting, 01.01.2019.

Research output: Contribution to JournalArticleAcademicpeer-review

TY - JOUR

T1 - Forecasting economic time series using score-driven dynamic models with mixed-data sampling

AU - Gorgi, Paolo

AU - Koopman, Siem Jan

AU - Li, Mengheng

PY - 2019/1/1

Y1 - 2019/1/1

N2 - We introduce a mixed-frequency score-driven dynamic model for multiple time series where the score contributions from high-frequency variables are transformed by means of a mixed-data sampling weighting scheme. The resulting dynamic model delivers a flexible and easy-to-implement framework for the forecasting of low-frequency time series variables through the use of timely information from high-frequency variables. We verify the in-sample and out-of-sample performances of the model in an empirical study on the forecasting of U.S. headline inflation and GDP growth. In particular, we forecast monthly headline inflation using daily oil prices and quarterly GDP growth using a measure of financial risk. The forecasting results and other findings are promising. Our proposed score-driven dynamic model with mixed-data sampling weighting outperforms competing models in terms of both point and density forecasts.

AB - We introduce a mixed-frequency score-driven dynamic model for multiple time series where the score contributions from high-frequency variables are transformed by means of a mixed-data sampling weighting scheme. The resulting dynamic model delivers a flexible and easy-to-implement framework for the forecasting of low-frequency time series variables through the use of timely information from high-frequency variables. We verify the in-sample and out-of-sample performances of the model in an empirical study on the forecasting of U.S. headline inflation and GDP growth. In particular, we forecast monthly headline inflation using daily oil prices and quarterly GDP growth using a measure of financial risk. The forecasting results and other findings are promising. Our proposed score-driven dynamic model with mixed-data sampling weighting outperforms competing models in terms of both point and density forecasts.

KW - Generalized autoregressive score models

KW - Gross domestic product

KW - Inflation

KW - Mixed frequency time series

KW - Time-varying parameters

UR - http://www.scopus.com/inward/record.url?scp=85059463907&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85059463907&partnerID=8YFLogxK

U2 - 10.1016/j.ijforecast.2018.11.005

DO - 10.1016/j.ijforecast.2018.11.005

M3 - Article

JO - International Journal of Forecasting

JF - International Journal of Forecasting

SN - 0169-2070

ER -