Volume 2, Issue 4, July 2013, Page: 94-101
Evaluation of Artificial Neural Networks in Foreign Exchange Forecasting
Akintunde Mutairu Oyewale, Department of Statistics, University of Botswana, Botswana
Received: Jun. 17, 2013;       Published: Jul. 10, 2013
DOI: 10.11648/j.ajtas.20130204.11      View  3396      Downloads  210
This study investigates the modeling, description and forecasting of exchange rates of four countries (Great Britain Pound, Japanese Yen, Nigerian Naira and Batswana Pula) using Artificial Neural Network, the objective of this paper is to use ANN to predict the trend of these four currencies. ANN was used in training and learning processes and thereafter the forecast performance was evaluated or measured making use of various loss functions such as root mean square error (RMSE), mean absolute error (MAE), mean absolute error (MAE), mean absolute precision error (MAPE) and Theill inequality coefficient (TIC). The loss functions used are good indicator of measuring the forecast performance of these series, the series with the lowest function gave a best forecast performance. Results show that the ANN is a very effective tool for exchange rate forecasting. Classical statistical methods are unable to efficiently handle the prediction of financial time series due to non-linearity, non-stationarity and high degree of noise. Advanced intelligence techniques have been used in many financial markets to forecast future development of different capital markets. Artificial neural network is a well tested method for financial markets analysis.
Artificial Neural Networks, Foreign Exchange, Loss Functions, Training and Learning Processes
To cite this article
Akintunde Mutairu Oyewale, Evaluation of Artificial Neural Networks in Foreign Exchange Forecasting, American Journal of Theoretical and Applied Statistics. Vol. 2, No. 4, 2013, pp. 94-101. doi: 10.11648/j.ajtas.20130204.11
Azoff, E.M., (1994). Neural networks time series forecasting of financial markets. Wiley publishing company (Chichester and New York).
Erims, K., Midilli, A., Dincer, I. and Rosen, M. A. (2007). Artifical Neural Network analysis of World Green Energy Use. Energy Policy, 35, no. 3, p. 1731-1743.
Franses, P.H and Van Griensen, K. (1998). Forecasting exchange rates using neural networks for technical trading rules. Studies in non-linear dynamics and Econometrics, 4, 109-114.
Franses, P.H and Van Homelin, P. (1998). On forecasting exchange rates using neural networks. Journal of applied financial Economics, 8, 589-596.
Gately, E. (1995). Neural Networks for financial forecasting. Wiley trader’s exchange publishing company
Gencay, R.(1999). Linear non-linear and essential foreign exchange prediction with simple technical rules. Journals of international Economics, 47, 91-107.
Gencay, R. and Stengos, T. (1998). Moving average rules, volumes and predictability of security returns with feed forward neural networks. Journals of forecasting, 17, 401-414.
Haefke, C. and Helmenstein, C., (1996a,1996b). Forecasting Austrian IPOs: An Application of Linear and Neural Network Error-Correction Models
Kuan, C.M. and Liu, T. (1995).Forecasting exchange rates using feed forward and recurrent neural networks. Journal of Applied Econometrics 10(4), 347-364.
Qi, M. and Madalla, G.S. (1999). Economic factors and stock market: A new perspective. Journal of forecasting,18, 151-166.
Refenes, A. P N. (1995). "Neural Networks in the Capital Markets", Wiley & Sons, Chichester, ISBN 0-471-94364-9.
Shadboth, J. and Taylor, J. (2002). neural networks and the financial markets. Springer publishing company Limited.
Trippi, R. R. and Turban, E. (1993). Neural Networks in finance and investing: using artificial intelligence to improve real world performances. Probus publishing company (Chicago III).
Browse journals by subject