Evaluating non-linear models on point and interval forecasts: an application with exchange rates

The aim of this paper is to compare the forecasting performance of SETAR and GARCH models against a linear benchmark using historical data for two bilateral dollar exchange rates, namely the Japanese Yen and the British Pound. The analysis is carried out with series sampled at weekly and daily freq...

Full description

Bibliographic Details
Main Authors: Gianna Boero, Emanuela Marrocu
Format: Article
Language:English
Published: Associazione Economia civile 2012-04-01
Series:PSL Quarterly Review
Subjects:
Online Access:https://rosa.uniroma1.it/rosa04/psl_quarterly_review/article/view/9837
_version_ 1811172411840659456
author Gianna Boero
Emanuela Marrocu
author_facet Gianna Boero
Emanuela Marrocu
author_sort Gianna Boero
collection DOAJ
description The aim of this paper is to compare the forecasting performance of SETAR and GARCH models against a linear benchmark using historical data for two bilateral dollar exchange rates, namely the Japanese Yen and the British Pound. The analysis is carried out with series sampled at weekly and daily frequencies. The relative performance of the models is evaluated on point forecasts and interval forecasts. Point forecasts evaluation tends to favour on average the linear models, though the analysis produces some evidence of forecasting gains from nonlinear models in sub-samples characterised by stronger non-linearity. Evaluation of interval forecasts clearly favours the GARCH models and shows that they are more accurate than the AR and SETAR models, especially at forecasting events in the tail regions of the distribution.     JEL Codes: F31, F37
first_indexed 2024-04-10T17:29:35Z
format Article
id doaj.art-c2daf0c4128b486e963b54936c0ee80e
institution Directory Open Access Journal
issn 2037-3635
2037-3643
language English
last_indexed 2024-04-10T17:29:35Z
publishDate 2012-04-01
publisher Associazione Economia civile
record_format Article
series PSL Quarterly Review
spelling doaj.art-c2daf0c4128b486e963b54936c0ee80e2023-02-03T16:48:39ZengAssociazione Economia civilePSL Quarterly Review2037-36352037-36432012-04-015823210.13133/2037-3643/9837Evaluating non-linear models on point and interval forecasts: an application with exchange ratesGianna BoeroEmanuela Marrocu The aim of this paper is to compare the forecasting performance of SETAR and GARCH models against a linear benchmark using historical data for two bilateral dollar exchange rates, namely the Japanese Yen and the British Pound. The analysis is carried out with series sampled at weekly and daily frequencies. The relative performance of the models is evaluated on point forecasts and interval forecasts. Point forecasts evaluation tends to favour on average the linear models, though the analysis produces some evidence of forecasting gains from nonlinear models in sub-samples characterised by stronger non-linearity. Evaluation of interval forecasts clearly favours the GARCH models and shows that they are more accurate than the AR and SETAR models, especially at forecasting events in the tail regions of the distribution.     JEL Codes: F31, F37 https://rosa.uniroma1.it/rosa04/psl_quarterly_review/article/view/9837Exchange Rates
spellingShingle Gianna Boero
Emanuela Marrocu
Evaluating non-linear models on point and interval forecasts: an application with exchange rates
PSL Quarterly Review
Exchange Rates
title Evaluating non-linear models on point and interval forecasts: an application with exchange rates
title_full Evaluating non-linear models on point and interval forecasts: an application with exchange rates
title_fullStr Evaluating non-linear models on point and interval forecasts: an application with exchange rates
title_full_unstemmed Evaluating non-linear models on point and interval forecasts: an application with exchange rates
title_short Evaluating non-linear models on point and interval forecasts: an application with exchange rates
title_sort evaluating non linear models on point and interval forecasts an application with exchange rates
topic Exchange Rates
url https://rosa.uniroma1.it/rosa04/psl_quarterly_review/article/view/9837
work_keys_str_mv AT giannaboero evaluatingnonlinearmodelsonpointandintervalforecastsanapplicationwithexchangerates
AT emanuelamarrocu evaluatingnonlinearmodelsonpointandintervalforecastsanapplicationwithexchangerates