Abstract
Recent studies on exchange rate forecasting showed that the cointegration concept for economic time series analysis proved to be fruitful. The Johansen procedure in particular, with its evolving multiple-equation Vector Error Correction models, is able to capture more complex, higher-order cointegrated relationships between financial market time series than single-equation, semi-reduced models following the Engle/Granger approach. Following promising results achieved by the application of Neural Networks to predict financial data, this study additionally uses Neural Networks for modeling any non-linearities within the estimated cointegration systems. To make full use of the interdependencies among economic time series and to pay attention to the increased integration of international financial markets, this paper presents research on Multi-Task Learning. The idea is that by learning different, yet related, tasks simultaneously, underlying interdependencies between the various learning outputs can be exploited. The paper presents a neural Vector Error Correction approach with multiple output units as a Multi-Task Learning methodology of practical use in finance. By focusing on forecasting the DEM/USD-exchange rate, the performance is compared and evaluated out-of-sample.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Abu-Mostafa Y. S. (1990): Learning from Hints in Neural Networks. Journal of Complexity, 6(2): 192–198.
Abu-Mostafa Y. S. (1994): Learning from Hints. Journal of Complexity,10(1):165–178.
Abu-Mostafa Y. S. (1995): Hints. Neural Computation, 7:639–671.
Bartlmae K. (1997): Multi-Task Lernen bei neuronalen Netzen zur Aktienkursprognose. Master Thesis, University of Karlsruhe, Institute of Logic,Complexity and Deduction Systems, Karlsruhe, Germany.
Bartlmae K., Gutjahr S., Nakhaeizadeh G. (1997): Incorporating prior knowledgeabout financial markets through neural multitask learning. Poster Presentation, Fifth International Conference on Computational Finance,London Business School, London, England.
Bilson J. F. O. (1987): The monetary approach to the exchange rate-some empirical evidence. IMF Staff Papers, 25:48–75.
Caruana R. (1993): Multitask Learning: A Knowledge-Based Source of InductiveBias. Proceedings of the 10th International Conference on Machine Learning,Amherst, USA, 41–48.
Caruana R. (1995): Learning Many Related Tasks at the Same Time WithBackpropagation. Advances in Neural Information Processing Systems, 7:657–664.
Caruana R. (1996): Algorithms and Applications for Multitask Learning.Proceedings of the 13th International Conference on Machine Learning, Bari, Italy, 87–95.
Caruana R. (1997): Multitask Learning. Machine Learning, 28:41–75.
Cuthbertson K., Hall S., Taylor M. P. (1992): Applied econometric techniques. Phillip Allan, London, England.
Engle R., Granger C. W. J. (1987): Co-integration and error-correction: representation, estimation and testing. Econometrica, 49:251–276.
Frenkel J. A. (1976): A monetary approach to the exchange rate: doctrinal aspectsand empirical evidence. Scandinavian Journal of Economics, 78:200–224.
Gandolfo G., Padoan P. C., Paladino G. (1990):. Exchange rate determination: single-equation or economy-wide models?. Journal of Banking and Finance,14:965–992.
Gardeazabal J., Regúlez M. (1992): The monetary model of exchange rate-estimation, testing and prediction. Springer, Berlin, Germany.
Gerhards T. (1994): Theorie und Empirie flexibler Wechselkurse. Physica, Heidelberg, Germany.
Granger C. W. J., Newbold P. (1979): Spurious regression in econometrics. Journal of Economics, 2:111–120.
Hansen H., Juselius K. (1995): CATS in RATS-cointegration analysis of timeseries. Software Manual, Estima, Evanston, USA.
Johansen S. (1988): Statistical analysis of cointegration vectors. Journal ofEconomic Dynamics and Control, 12:231–254.
Johansen S. (1991): Estimation and hypothesis testing of cointegration vectors ingaussian vector autoregressive models. Econometrica, 59:1551–1580.
Johansen S., Juselius K. (1990): Maximum likelihood estimation and inference oncointegration-with applications to the demand for money. Oxford Bulletin of Economics and Statistics, 52:169–210.
Kim B. J. C., Mo, S. (1995): Cointegration and long-run forecast of exchangerates. Economic Letters, 48:353–359.
Kugler P., Lenz C. (1993): Multivariate cointegration analysis and the long-run validity of PPP. The Review of Economics and Statistics, 75:180–184.
MacDonald R., Taylor M. P. (1993): The monetary approach to the exchange rate-rational expectations, long-run equilibrium and forecasting. IMF Staff Papers, 40:89–107.
Muscatelli A., Hurn S. (1992): Cointegration and dynamic time series models. Journal of Economic Surveys, 6:1–43.
Phillips P. C. B. (1987): Time series regression with a unit root. Econometrica,55:277–301.
Poddig T., Rehkugler H. (1996): A “World” Model of Integrated Financial Markets using Artificial Neural Networks. Neurocomputing, 10:251–273.
Rauscher F. A. (1997): Exchange Rate Forecasting: a Neural VEC Approach.Neural Network World Journal, 4/5:461–471.
Refenes A.-P. N. (1995): Neural Networks in Capital Markets. Wiley, Chichester, England.
Rehkugler H., Poddig T., Jandura D. (1996): Einsatz integrierter Modelle für die simultane Prognose von Aktienkursen, Zinsen und Währungen für mehrere Länder mit Neuronalen Netzen, in: Bol G. et al. (Eds): Finanzmarktanalyse und-prognose mit innovativen quantitativen Verfahren. Physica, Heidelberg, Germany, 207–236.
Rehkugler H., Zimmermann H. G. (1994): Neuronale Netze in der Ökonomie.Vahlen,Munich, Germany.
Said S. E., Dickey D. A. (1984): Testing for unit roots in autoregressive moving average models of unknown order. Biometrica, 59:599–607.
Steurer E. (1997): Ökonometrische Methoden und maschinelle Lernverfahren zurWechselkursprognose-theoretische Analyse und empirischer Vergleich.Physica, Heidelberg, Germany.
Weigend A. S., Huberman B. A., Rumelhart D. E. (1991): Predicting Sunspots and Exchange Rates with Connectionist Networks, in: Casdagli M., Eubank S.: Nonlinear Modeling and Forecasting. Addison-Wesley, Reading, USA, 395–432.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1998 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Rauscher, F.A. (1998). Multi-Task Learning in a Neural Vector Error Correction Approach for Exchange Rate Forecasting. In: Refenes, AP.N., Burgess, A.N., Moody, J.E. (eds) Decision Technologies for Computational Finance. Advances in Computational Management Science, vol 2. Springer, Boston, MA. https://doi.org/10.1007/978-1-4615-5625-1_13
Download citation
DOI: https://doi.org/10.1007/978-1-4615-5625-1_13
Publisher Name: Springer, Boston, MA
Print ISBN: 978-0-7923-8309-3
Online ISBN: 978-1-4615-5625-1
eBook Packages: Springer Book Archive