Forecasting of IBOVESPA returns using feedforward evolutionary artificial neural networks
Facing the challenges of anticipating financial market uncertainties and movements, and the necessity of taking buy or sell decisions supported by rational methods, market traders found in statistics and econometrics methods, the base to support their decisions. In several scientific papers about fo...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Universidade Federal de Santa Catarina
2011-12-01
|
Series: | Revista Produção Online |
Subjects: | |
Online Access: | http://producaoonline.org.br/rpo/article/view/784 |
id |
doaj-2e3b3a97efbf4d01bd87f913f73b8493 |
---|---|
record_format |
Article |
spelling |
doaj-2e3b3a97efbf4d01bd87f913f73b84932020-11-25T02:04:17ZengUniversidade Federal de Santa CatarinaRevista Produção Online1676-19012011-12-011141114114010.14488/1676-1901.v11i4.784452Forecasting of IBOVESPA returns using feedforward evolutionary artificial neural networksEdgar Leite dos Santos Filho0Wesley Vieira da Silva1Claudimar Pereira da Veiga2Ubiratã Tortato3Pontifícia Universidade Católica do ParanáPontifícia Universidade Católica do ParanáPontifícia Universidade Católica do ParanáPontifícia Universidade Católica do ParanáFacing the challenges of anticipating financial market uncertainties and movements, and the necessity of taking buy or sell decisions supported by rational methods, market traders found in statistics and econometrics methods, the base to support their decisions. In several scientific papers about forecasting financial time series, method selection keeps as central concern. This paper compares the performance of evolutionary feedforward artificial neural network (EANN) and an AR+GARCH model, for one step ahead forecasting of IBOVESPA returns. The EANN is trained by self-adapting differential evolution algorithm and AR+GARCH model is adjusted to be used as performance reference. The root mean square error (RMSE) and U-Theil inequality coefficient were used as performance metrics. Simulation results showed EANN feedforward achieved better results, fit better and captured the nonlinear behavior of returns.http://producaoonline.org.br/rpo/article/view/784Redes Neurais Artificiais, Evolução Diferencial, Modelos GARCH, Erro Quadrático Médio, Índice de Desigualdade de Theil |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Edgar Leite dos Santos Filho Wesley Vieira da Silva Claudimar Pereira da Veiga Ubiratã Tortato |
spellingShingle |
Edgar Leite dos Santos Filho Wesley Vieira da Silva Claudimar Pereira da Veiga Ubiratã Tortato Forecasting of IBOVESPA returns using feedforward evolutionary artificial neural networks Revista Produção Online Redes Neurais Artificiais, Evolução Diferencial, Modelos GARCH, Erro Quadrático Médio, Índice de Desigualdade de Theil |
author_facet |
Edgar Leite dos Santos Filho Wesley Vieira da Silva Claudimar Pereira da Veiga Ubiratã Tortato |
author_sort |
Edgar Leite dos Santos Filho |
title |
Forecasting of IBOVESPA returns using feedforward evolutionary artificial neural networks |
title_short |
Forecasting of IBOVESPA returns using feedforward evolutionary artificial neural networks |
title_full |
Forecasting of IBOVESPA returns using feedforward evolutionary artificial neural networks |
title_fullStr |
Forecasting of IBOVESPA returns using feedforward evolutionary artificial neural networks |
title_full_unstemmed |
Forecasting of IBOVESPA returns using feedforward evolutionary artificial neural networks |
title_sort |
forecasting of ibovespa returns using feedforward evolutionary artificial neural networks |
publisher |
Universidade Federal de Santa Catarina |
series |
Revista Produção Online |
issn |
1676-1901 |
publishDate |
2011-12-01 |
description |
Facing the challenges of anticipating financial market uncertainties and movements, and the necessity of taking buy or sell decisions supported by rational methods, market traders found in statistics and econometrics methods, the base to support their decisions. In several scientific papers about forecasting financial time series, method selection keeps as central concern. This paper compares the performance of evolutionary feedforward artificial neural network (EANN) and an AR+GARCH model, for one step ahead forecasting of IBOVESPA returns. The EANN is trained by self-adapting differential evolution algorithm and AR+GARCH model is adjusted to be used as performance reference. The root mean square error (RMSE) and U-Theil inequality coefficient were used as performance metrics. Simulation results showed EANN feedforward achieved better results, fit better and captured the nonlinear behavior of returns. |
topic |
Redes Neurais Artificiais, Evolução Diferencial, Modelos GARCH, Erro Quadrático Médio, Índice de Desigualdade de Theil |
url |
http://producaoonline.org.br/rpo/article/view/784 |
work_keys_str_mv |
AT edgarleitedossantosfilho forecastingofibovespareturnsusingfeedforwardevolutionaryartificialneuralnetworks AT wesleyvieiradasilva forecastingofibovespareturnsusingfeedforwardevolutionaryartificialneuralnetworks AT claudimarpereiradaveiga forecastingofibovespareturnsusingfeedforwardevolutionaryartificialneuralnetworks AT ubiratatortato forecastingofibovespareturnsusingfeedforwardevolutionaryartificialneuralnetworks |
_version_ |
1724943328931217408 |