Evaluating GDP Forecasting Models for Korea
Electronic Access:
Free Download. Use the free Adobe Acrobat Reader to view this PDF file
Summary:
This paper develops a new forecasting framework for GDP growth in Korea to complement and further enhance existing forecasting approaches. First, a range of forecast models, including indicator- and pure time-series models, are evaluated for their forecasting performance. Based on the evaluation results, a new forecasting framework is developed for GDP projections. The framework also generates a data-driven reference band for the projections, and is therefore convenient to update. The framework is applied to the current World Economic Outlook (WEO) forecast period and the Great Recession to compare its performance to past projections. Results show that the performance of the new framework often improves the forecasts, especially at quarterly frequency, and the forecasting exercise will be better informed by cross-checking with the new data-driven framework projections.
Series:
Working Paper No. 2011/053
Subject:
Economic forecasting GDP forecasting Private consumption Time series analysis Vector autoregression
English
Publication Date:
March 1, 2011
ISBN/ISSN:
9781455220977/1018-5941
Stock No:
WPIEA2011053
Pages:
23
Please address any questions about this title to publications@imf.org