PREDICTION OF BIST PRICE INDICES: A COMPARATIVE STUDY BETWEEN TRADITIONAL AND DEEP LEARNING METHODS

Abstract

Financial time series prediction is a challenging task due to the noisy, non-stationary and chaotic nature series. Traditional methods, especially autoregressive integrated moving average (ARIMA) has a wide range of application. With the rapid development of information technologies in the last two decades, various deep learning methods which are inspired by human brain that consists of inter-connected neurons have been proposed in order to improve the prediction performance of time series. As the data amount increases, these methods have been seen as an alternative for traditional ones having some important limitations. The main purpose of this study is to determine whether the deep learning methods outperform than traditional ARIMA method in predicting the BIST 30, BIST 50 and BIST 100 price indices. The prediction performance of ARIMA is compared against the prediction performances of Long Short-Term Memory and Gated-Recurrent Unit for each BIST price index. According to the root mean square evaluation metric, it is found that ARIMA models have better performance in predicting BIST 30, BIST 50 and BIST 100 indices than deep learning architectures.

Keywords

• Deep learning
• ARIMA
• long short-term memory
• gated-recurrent unit.

References

1. [1] Cortez, P., and Donate, J. P. (2014) Global and decomposition evolutionary support vector machine approaches for time series forecasting, Neural Computing and Applications 25(5), 1053-1062.
2. [2] Jothimani, D., and Yadav, S. S. (2019) Stock trading decisions using ensemble-based forecasting models: a study of the Indian stock market, Journal of Banking and Financial Technology 3(2), 113-129.
3. [3] Winters, P. R. (1960) Forecasting sales by exponentially weighted moving averages, Management Science 6(3), 324-342.
4. [4] Box, G.E.P, and Jenkins, G. (1976) Time series analysis: forecasting and control. Holden Day, San Francisco, USA.
5. [5] Siami-Namini, S., and Namin, A. S. (2018) Forecasting economics and financial time series: ARIMA vs. LSTM. arXiv preprint arXiv:1803.06386.
6. [6] Meyler, A., Kenny, G., and Quinn, T. (1998) Forecasting Irish inflation using ARIMA models, MPRA Paper No 11359.
7. [7] Ariyo, A. A., Adewumi, A. O., and Ayo, C. K. (2014, March). Stock price prediction using the ARIMA model. In 2014 UKSim-AMSS 16th International Conference on Computer Modelling and Simulation (pp. 106-112). IEEE.
8. [8] Junior, P. R., Salomon, F. L. R., and de Oliveira Pamplona, E. (2014) ARIMA: An applied time series forecasting model for the Bovespa stock index, Applied Mathematics 5(21), 3383-3391.

9. [9] Gay Jr, R. D. (2016) Effect of macroeconomic variables on stock market returns for four emerging economies: Brazil, Russia, India, and China, International Business & Economics Research Journal (IBER) 15(3), 119-126.
10. [10] Zhang, G. P. (2003) Time series forecasting using a hybrid ARIMA and neural network model, Neurocomputing 50, 159-175.
11. [11] Hochreiter, S., and Schmidhuber, J. (1997) Long short-term memory, Neural Computation 9(8), 1735-1780.
12. [12] Pascanu, R., Mikolov, T., and Bengio, Y. (2013, February). On the difficulty of training recurrent neural networks. In International conference on machine learning (pp. 1310-1318).
13. [13] Sezer, O. B., Gudelek, M. U., and Ozbayoglu, A. M. (2020) Financial time series forecasting with deep learning: A systematic literature review: 2005–2019, Applied Soft Computing 90, 106181.
14. [14] Di Persio, L., and Honchar, O. (2016) Artificial neural networks architectures for stock price prediction: Comparisons and applications, International journal of circuits, systems and signal processing 10(2016), 403-413.
15. [15] Lee, S. I., and Yoo, S. J. (2018) Threshold-based portfolio: the role of the threshold and its applications, The Journal of Supercomputing 2018, 1-18.
16. [16] Bircan, H., and Karagöz, Y. (2003) Box-Jenkıns Modelleri ile Aylık Döviz Kuru Tahmini Üzerine Bir Uygulama, Kocaeli Üniversitesi Sosyal Bilimler Enstitüsü Dergisi (6) 2003/2, 49-62.
17. [17] Yenice, S., and Tekindal, M. A. (2015) Forecasting the stock indexes of fragile five countries through Box-Jenkins methods, International Journal of Business and Social Science 6(8), 180-191.
18. [18] Sekreter, Α., and Gursoy, A. (2014) Combining forecasting method vs. individual forecasting methods: Evidence from Istanbul Stock Exchange National 100 Index, The Empirical Economics Letters 13(7), 735-743.
19. [19] Sezer, Ö. B. (2018) Zaman serisi verilerinin derin yapay sinir ağları ile analizi ve eniyilemesi: Finansal tahmin algoritmaları. PhD Thesis, Graduate School of Engineering and Science, TOBB Ekonomi ve Teknoloji Üniversitesi, Ankara, Turkey.
20. [20] Fischer, T., and Krauss, C. (2018) Deep learning with long short-term memory networks for financial market predictions, European Journal of Operational Research 270(2), 654-669.
21. [21] Yamak, P. T., Yujian, L., and Gadosey, P. K. (2019, December). A Comparison between ARIMA, LSTM, and GRU for Time Series Forecasting. In Proceedings of the 2019 2nd International Conference on Algorithms, Computing and Artificial Intelligence (pp. 49-55).
22. [22] McNally, S., Roche, J., and Caton, S. (2018, March). Predicting the price of bitcoin using machine learning. In 2018 26th Euromicro International Conference on Parallel, Distributed and Network-based Processing (PDP) (pp. 339-343). IEEE.
23. [23] Selvin, S., Vinayakumar, R., Gopalakrishnan, E. A., Menon, V. K., and Soman, K. P. (2017, September). Stock price prediction using LSTM, RNN and CNN-sliding window model. In 2017 international conference on advances in computing, communications and informatics (icacci) (pp. 1643-1647). IEEE.
24. [24] Çetin, K., Aksoy, S., and İşeri, İ. (2019, September). Steel Price Forcasting Using Long Short-Term Memory Network Model. In 2019 4th International Conference on Computer Science and Engineering (UBMK) (pp. 612-617). IEEE.
25. [25] Di Persio, L., and Honchar, O. (2017). Recurrent neural networks approach to the financial forecast of Google assets. International journal of Mathematics and Computers in simulation, 11.
26. [26] Brownlee, J. (2017). Long Short-term Memory Networks with Python: Develop Sequence Prediction Models with Deep Learning. Machine Learning Mastery.
27. [27] Bsir, B., and Zrigui, M. (2018, September). Bidirectional LSTM for author gender identification. In International Conference on Computational Collective Intelligence (pp. 393-402). Springer, Cham.
28. [28] Yu, L., Chen, J., Ding, G., Tu, Y., Yang, J., and Sun, J. (2018). Spectrum prediction based on Taguchi method in deep learning with long short-term memory. IEEE Access, 6, 45923-45933.
29. [29] Soutner, D., and Müller, L. (2013, September). Application of LSTM neural networks in language modelling. In International Conference on Text, Speech and Dialogue (pp. 105-112). Springer, Berlin, Heidelberg.
30. [30] Cho, K., Van Merriënboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., and Bengio, Y. (2014). Learning phrase representations using RNN encoder-decoder for statistical machine translation. arXiv preprint arXiv:1406.1078.
31. [31] R. B. D’Agostino, A. J. Belanger and R. B. D’Agostino Jr., “A suggestion for using powerful and informative tests of normality”, American Statistician 44, pp. 316-321, 1990.
32. [32] Anscombe, F. J., and Glynn, W. J. (1983) Distribution of the kurtosis statistic b 2 for normal samples, Biometrika 70(1), 227-234.
33. [33] Saigal, S., and Mehrotra, D. (2012) Performance comparison of time series data using predictive data mining techniques, Advances in Information Mining 4(1), 57-66.
34. [34] Jain, S., Shukla, S., and Wadhvani, R. (2018) Dynamic selection of normalization techniques using data complexity measures, Expert Systems with Applications 106, 252-262.
35. [35] Han, J., Pei, J., & Kamber, M. (2011). Data mining: concepts and techniques. Elsevier.
36. [36] Abadi, Martín, Paul Barham, Jianmin Chen, Zhifeng Chen, Andy Davis, Jeffrey Dean, Matthieu Devin et al. "TensorFlow: A System for LargeScale Machine Learning." In OSDI, vol. 16, pp. 265-283. 2016.
37. [37] Kingma, D. P., and Ba, J. (2014). Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980.
38. [38] Luo, H., and Wang, S. (2016) Based on the PCA-ARIMA-BP hybrid model of stock price prediction research, ANZIAM Journal 58, 162-178.
39. [39] Makridakis, S., Spiliotis, E., and Assimakopoulos, V. (2018), Statistical and Machine Learning forecasting methods: Concerns and ways forward. PloS one, 13(3).
40. [40] Jiang, L., Rollins, K. M., Ludlow, M., and Sadler, B. (2020). Demand Forecasting for Alcoholic Beverage Distribution. SMU Data Science Review, 3(1), 5.