Global Business and Economics Review · 2008 · 27 citations · 28 references
ComputingEconomicsFinanceEconometricsForecastingTime Series AnalysisForecast EvaluationNeural NetworksTime Series EconometricsTime Series ForecastingTime Series ModelingTime Series PredictionAutoregressive Moving Average ModelsConditional Heteroskedasticity ModelsCurrency ForecastingGeneralized Autoregressive Conditional Heteroskedasticity ModelsNeural ForecastingNeural Time Series ForecastingNonlinear Forecasting ModelsStatistical Time Series Models
In this paper, the exchange rate forecasting performance of neural network models are evaluated against the random walk, autoregressive moving average and generalised autoregressive conditional heteroskedasticity models. There are no guidelines available that can be used to choose the parameters of neural network models and therefore, the parameters are chosen according to what the researcher considers to be the best. Such an approach, however, implies that the risk of making bad decisions is extremely high, which could explain why in many studies, neural network models do not consistently perform better than their time series counterparts. In this paper, through extensive experimentation, the level of subjectivity in building neural network models is considerably reduced and therefore giving them a better chance of performing well. The results show that in general, neural network models perform better than the traditionally used time series models in forecasting exchange rates.
28
Generalized autoregressive conditional heteroskedasticity
Tim Bollerslev · Journal of Econometrics · 1986 · 22.3K citations · Full text
Business, Finance, Econometrics +11
Multilayer feedforward networks are universal approximators
HornikK., StinchcombeM., WhiteH. · Neural Networks · 1989 · 9.4K citations