Please use this identifier to cite or link to this item:
Medeiros, Marcelo Cunha
Veiga, Álvaro
Pedreira, Carlos Eduardo
Year of Publication: 
Series/Report no.: 
Texto para discussão 432
The goal of this paper is to test for and model nonlinearities in several monthly exchange rates time series. We apply two different nonlinear alternatives, namely: the artificial neural network time series model estimated with Bayesian regularization and a flexible smooth transition specifica-tion, called the neuro-coefficient smooth transition autoregression. The linearity test rejects the null hypothesis of linearity in ten out of fourteen series. We compare, using different measures, the fore-casting performance of the nonlinear specifications with the linear autoregression and the random walk models.
Document Type: 
Working Paper
Social Media Mentions:

Files in This Item:
209.75 kB

Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.