Nowcasting GDP: an application to Portugal

João B. Assunção, Pedro Afonso Fernandes*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

23 Downloads

Abstract

Forecasting the state of an economy is important for policy makers and business leaders. When this is conducted in real-time, it is called nowcasting. In this paper, we present a method that shows how forecasting errors decline as additional contemporaneous information unfolds and becomes available. When the economic environment changes fast, as has happened often in the last decades across most developed economies, it is important to use forecasting methods that are both flexible and robust. This can be achieved with bridge equations and non-parametric estimates of the trend growth using only publicly available information. The method presented in this paper achieves, by the end of a quarter, an accuracy that is equivalent to the methods used by official entities.
Original languageEnglish
Pages (from-to)717-731
Number of pages15
JournalForecasting
Volume4
Issue number3
DOIs
Publication statusPublished - 15 Aug 2022

Keywords

  • Combining forecasts
  • Error correction models
  • Macroeconomic forecasting
  • Nowcasting
  • Time series

Fingerprint

Dive into the research topics of 'Nowcasting GDP: an application to Portugal'. Together they form a unique fingerprint.

Cite this