Abstract
The motivation for this chapter is to investigate the use of alternative novel neural network architectures when applied to the task of forecasting and trading the ASE 20 Greek Index using only autoregressive terms as inputs. This is done by benchmarking the forecasting performance of six different neural network designs representing aHigher Order Neural Network (HONN), aRecurrent Network (RNN), a classicMultilayer Perceptron (MLP), a Hybrid Higher Order Neural Network, a Hybrid Recurrent Neural Network and a Hybrid Multilayer Perceptron Neural Network with some traditional techniques, either statistical such as an autoregressive moving average model (ARMA) or technical such as a moving average convergence/divergence model (MACD), plus a naïve trading strategy. More specifically, the trading performance of all models is investigated in a forecast and trading simulation on ASE 20 fixing time series over the period 2001–2008 using the last one and a half year for out-of-sample testing. We use the ASE 20 daily fixing as many financial institutions are ready to trade at this level and it is therefore possible to leave orders with a bank for business to be transacted on that basis. As it turns out, the hybrid-HONNs do remarkably well and outperform all other models in a simple trading simulation exercise. However, when more sophisticatedtrading strategies usingconfirmation filters andleverage are applied, the hybrid-HONN network produces better results and outperforms all other neural network and traditional statistical models in terms of annualised return.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
We examine the ASE 20 since its first trading day on 21 January 2001, and until 31 December 2008, using the continuous data available from datastream.
- 2.
Confirmation of its stationary property is obtained at the 1% significance level by both the Augmented Dickey Fuller (ADF) and Phillips-Perron (PP) test statistics.
- 3.
A “long” ASE 20 position means buying the index at the current price, while a “short” position means selling the index at the current price.
- 4.
- 5.
Backpropagation networks are the most common multilayer networks and are the most commonly used type in financial time series forecasting [20].
- 6.
Associative recall is the act of associating two seemingly unrelated entities, such as smell and colour. For more information see [22].
- 7.
Since most of the models have a volatility of about 20%, we have chosen this level as our basis. The leverage factors retained are given in Table 4.8.
- 8.
The interest costs are calculated by considering a 4% interest rate p.a. divided by 252 trading days. In reality, leverage costs also apply during non-trading days so that we should calculate the interest costs using 360 days per year. But for the sake of simplicity, we use the approximation of 252 trading days to spread the leverage costs of non-trading days equally over the trading days. This approximation prevents us from keeping track of how many non-trading days we hold a position.
References
O. Adam, L. Zarader, M. Milgram, Identification and prediction of non-linear models with recurrent neural networks, in New Trends in Neural Computation, ed. by J. Mira, J. Cabestany, A. Prieto. Lecture Notes in Computer Science, vol. 686 (Springer, Berlin, 1993), pp. 531–535
C. Bishop, Mixture density networks. Technical report NCRG/4288, Neural Computing Research Group, Aston University (1994)
G. Box, G. Jenkins, G. Gregory, Time Series Analysis: Forecasting and Control (Prentice-Hall, New Jersey, 1994)
G. Brown, J. Wyatt, R. Harris, X. Yao, Diversity creation methods: A survey and categorization. Inf. Fusion 6, 5–20 (2005)
R. Clemen, Combining forecasts: A review and annotated bibliography. Int. J. Forecast. 5, 559–583 (1989)
J. Connor, L. Atlas, Recurrent neural networks and time series prediction, in Proceedings of the International Joint Conference on Neural Networks (1993), pp. 301–306
C. Dunis, Y. Chen, Alternative volatility models for risk management and trading: Application to the EUR/USD and USD/JPY rates. Derivatives Use, Trading and Regulation 11(2), 126–156 (2005)
C. Dunis, X. Huang, Forecasting and trading currency volatility: An application of recurrent neural regression and model combination. J. Forecast. 21(5), 317–354 (2002)
C. Dunis, J. Laws, B. Evans, Modelling and trading the gasoline crack spread: A non-linear story. Derivatives Use, Trading and Regulation 12, 126–145 (2006)
C. Dunis, J. Laws, B. Evans, Trading futures spreads: An application of correlation and threshold filters. Appl. Financ. Econ. 16, 1–12 (2006)
C. Dunis, J. Laws, G. Sermpinis, Modelling and trading the EUR/USD exchange rate at the ECB fixing. Eur. J. Finance 16(6), 541–560 (2010)
C. Dunis, J. Laws, G. Sermpinis, Higher order and recurrent neural architectures for trading the EUR/USD exchange rate. Quant. Finance 11(4), 615–629 (2011)
J.L. Elman, Finding structure in time. Cognit. Sci. 14, 179–211 (1990)
S. Fatima, G. Hussain, Statistical models of KSE100 index using hybrid financial systems. Neurocomputing 7, 2742–2746 (2008)
J. Fulcher, M. Zhang, S. Xu, Application of higher-order neural networks to financial time series, in Artificial Neural Networks in Finance and Manufacturing, ed. by J. Kamruzzaman, R. Begg, R. Sarker (Idea Group, Hershey, 2006), pp. 80–108
L. Giles, T. Maxwell, Learning, invariance and generalization in higher order neural networks. Appl. Optic. 26, 4972–4978 (1987)
J. Hansen, R. Nelson, Time-series analysis with neural networks and ARIMA-neural network hybrids. J. Exp. Theor. Artif. Intell. 15(3), 315–330 (2003)
H. Hibbert, C. Pedreira, R. Souza, Combining neural networks and arima models for hourly temperature forecast, in IEEE International Joint Conference on Neural Networks (IJCNN’00), vol. 4 (2000), pp. 414–419
M. Hibon, T. Evgeniou, To combine or not to combine: Selecting among forecasts and their combinations. Int. J. Forecast. 22, 15–24 (2005)
I. Kaastra, M. Boyd, Designing a neural network for forecasting financial and economic time series. Neurocomputing 10, 215–236 (1996)
K. Kamijo, T. Tanigawa, Stock price pattern recognition: A recurrent neural network approach, in Proceedings of the International Joint Conference on Neural Networks (1990), pp. 1215–1221
N. Karayiannis, A. Venetsanopoulos, On the training and performance of high-order neural networks. Math. Biosci. 129, 143–168 (1994)
A. Knowles, A. Hussein, W. Deredy, P. Lisboa, C.L. Dunis, Higher-order neural networks with Bayesian confidence measure for prediction of EUR/USD exchange rate, in Artificial Higher Order Neural networks for Economics and Business, ed. by M. Zhang (Idea Group, Hershey, 2009), pp. 48–59
E. Kosmatopoulos, M. Polycarpou, M. Christodoulou, P. Ioannou, High-order neural network structures for identification of dynamical systems. IEEE Trans. Neural Network 6, 422–431 (1995)
A. Lindemann, C. Dunis, P. Lisboa, Level estimation, classification and probability distribution architectures for trading the EUR/USD exchange rate. Neural Network Comput. Appl. 14(3), 256–271 (2004)
S. Makridakis, Why combining works? Int. J. Forecast. 5, 601–603 (1989)
S. Makridakis, A. Anderson, R. Carbone, R. Fildes, M. Hibdon, R. Lewandowski, J. Newton, E. Parzen, R. Winkler, The accuracy of extrapolation (time series) methods: Results of a forecasting competition. J. Forecast. 1, 111–153 (1982)
P. Newbold, C.W.J. Granger, Experience with forecasting univariate time series and the combination of forecasts (with discussion). J. Stat. 137, 131–164 (1974)
F.C. Palm, A. Zellner, To combine or not to combine? issues of combining forecasts. J. Forecast. 11, 687–701 (1992)
R. Pindyck, D. Rubinfeld, Econometric Models and Economic Forecasts, 4th edn. (McGraw-Hill, New York, 1988)
D. Psaltis, C. Park, J. Hong, Higher order associative memories and their optical implementations. Neural Network 1, 149–163 (1988)
N. Redding, A. Kowalczyk, T. Downs, Constructive higher-order network algorithm that is polynomial time. Neural Network 6, 997–1010 (1993)
A.F. Shapiro, A hitchhikers guide to the techniques of adaptive nonlinear models. Insur. Math. Econ. 26, 119–132 (2000)
P. Tenti, Forecasting foreign exchange rates using recurrent neural networks. Appl. Artif. Intell. 10, 567–581 (1996)
N. Terui, H. van Dijk, Combined forecasts from linear and nonlinear time series models. Int. J. Forecast. 18, 421–438 (2002)
H. Theil, Applied Economic Forecasting (North-Holland, Amsterdam, 1996)
P. Tino, C. Schittenkopt, G. Dorner, Financial volatility trading using recurrent neural networks. IEEE Trans. Neural Network 12(4), 865–874 (2001)
F.M. Tseng, H.C. Yu, G.H. Tzeng, Combining neural network model with seasonal time series ARIMA model. Technol. Forecast. Soc. Change 69, 71–87 (2002)
Y.-H. Wang, Nonlinear neural network forecasting model for stock index option price: Hybrid GJR-GARCH approach. Expert Syst. Appl. 36(1), 564–570 (2009)
G.P. Zhang, Time series forecasting using a hybrid ARIMA and neural network model. Neurocomputing 50, 159–175 (2003)
R.L. Winkler, Combining forecasts: A philosophical basis and some current issues. Int. J. Forecast. 5, 605–609 (1989)
G.P. Zhang, M. Qi, Neural network forecasting for seasonal and trend time series. Eur. J. Oper. Res. 160(2), 501–514 (2005)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2012 Springer Science+Business Media New York
About this chapter
Cite this chapter
Dunis, C.L., Laws, J., Karathanasopoulos, A. (2012). Modelling and Trading the Greek Stock Market with Hybrid ARMA-Neural Network Models. In: Doumpos, M., Zopounidis, C., Pardalos, P. (eds) Financial Decision Making Using Computational Intelligence. Springer Optimization and Its Applications, vol 70. Springer, Boston, MA. https://doi.org/10.1007/978-1-4614-3773-4_4
Download citation
DOI: https://doi.org/10.1007/978-1-4614-3773-4_4
Published:
Publisher Name: Springer, Boston, MA
Print ISBN: 978-1-4614-3772-7
Online ISBN: 978-1-4614-3773-4
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)