Linear models, smooth transition autoregressions and neural networks for forecasting macroeconomic time series: A reexamination
In this paper we examine the forecast accuracy of linear autoregressive, smooth transition autoregressive (STAR), and neural network (NN) time series models for 47 monthly macroeconomic variables of the G7 economies. Unlike previous studies that typically consider multiple but fixed model specifications, we use a single but dynamic specification for each model class. The point forecast results indicate that the STAR model generally outperforms linear autoregressive models. It also improves upon several fixed STAR models, demonstrating that careful specification of nonlinear time series models is of crucial importance. The results for neural network models are mixed in the sense that at long forecast horizons, an NN model obtained using Bayesian regularization produces more accurate forecasts than a corresponding model specified using the specific-to-general approach. Reasons for this outcome are discussed.
Year of publication: |
2004-07
|
---|---|
Authors: | Teräsvirta, Timo ; Dijk, Dick van ; Medeiros, Marcelo Cunha |
Institutions: | Departamento de Economia, Pontifícia Universidade Católica do Rio de Janeiro |
Subject: | forecast combination | forecast evaluation | neural network model | nonlinear modelling | nonlinear forecasting |
Saved in:
freely available
Extent: | application/pdf |
---|---|
Series: | |
Type of publication: | Book / Working Paper |
Language: | English |
Notes: | Published in International Journal of Forecasting, v.21, 2005 Number 485 39 pages |
Classification: | C22 - Time-Series Models ; C53 - Forecasting and Other Model Applications |
Source: |
Persistent link: https://www.econbiz.de/10005744713