Yapay Zeka Yöntemleri İle Hisse Senedi Fiyat Öngörüleri

Abstract

Finansal varlık fiyatlarının geleceğinin tahmin edilmesi literatür ve uygulamada uzun zamandır ilgi çeken bir konudur. Son yıllarda, borsaya kote şirketlerin hisse senetlerinin fiyat hareketleri öngörme ve geleceğe dönük değerlerini tahmin etme hedefi için yapay zeka algoritmalarının başarılı yöntemler sundukları farklı akademik çalışmalarca ortaya konulmuştur. Belirtilen akademik çalışmaların büyük çoğunluğu yurt dışında bulunan piyasalarda yapılmıştır. Bu durumun geçerliliğini BIST 30 endeksi hisselerinde test etmek için bu çalışmada yedi farklı yapay zeka algoritması programlanmış, 30 hissenin 2014-2016 yılları günlük kapanış fiyatı verileri ile algoritmalar eğitilmiş ve bir firma için üretilen kapanış değerleri tahminleri gerçekleşen değerlerle kıyaslanmıştır. Veri seti için 02/01/2014 ve 30/12/2016 tarihleri arasında işlem yapılan 755 iş günü kullanılmıştır. Kullanılan öğrenme sürelerinin performans üzerindeki etkilerini görmek için öğrenme/tahmin oranları %80/20, %90/10, %99/1 olarak belirlenen üç farklı deney yapılmıştır. Çalışmanın sonucunda doğrusal regresyon temelli algoritmaların BIST30 hisse senedi fiyat hareket yönünü tahmin etmede, nöral ağ ve Poisson regresyonu yöntemlerinin ise kapanış fiyatı değerini tahmin etmede etkili oldukları görülmüştür.

Keywords

• Finansal Zaman Serileri
• Yapay Zeka
• Makine Ogrenmesi
• Borsa Analizi
• Regresyon

Stock Price Predictions Using Artificial Intelligence Methods

Abstract

Forecasting the future of stock market prices has been an interesting topic for researchers and professionals for a long time. Lately, numerous academic papers have shown that artificial intelligence algorithms generate some successful forecasts for stock prices. Most of these referenced research papers are conducted in markets outside Turkey. To test this hypothesis in BIST 30 index companies, seven different artificial intelligence algorithms have been programmed and trained with a dataset of daily closing prices between 2014 and 2016. The dataset consists of 755 market days starting 02/01/2014 and ending 30/12/2016. Then, forecasted numbers have been compared to actual prices for one particular stock. To see the effects of amount of learning days used to performance, 3 experiments have been conducted with learning to prediction ratios of %80/20, %90/10 and %99/1. In conclusion, it is seen that linear regression-based algorithms perform well to predict the price movements while neural network and Poisson regression algorithms perform well to predict closing price values for BIST 30 stocks.

Keywords

• Financial Time Series
• Artificial Intelligence
• Machine Learning
• Market Analysis
• Regression

References

1. Adams, R. (2004). Intelligent advertising. AI & Society, 18, 59-81. https://doi.org/10.1007/s00146-003-0259-9
2. Alpaydin, E. (2010). Introduction to machine learning. Cambridge, Massachusetts London: The MIT Press.
3. Altay, E. and Satman, H. (2005). Market forecasting: Artificial neural network and linear regression comparison in an emerging market. Journal of Financial Management and Analysis, 18(2), 18-33. Retrieved from https://papers.ssrn.com/
4. Arda, E. (2020). Yapay zeka yöntemleri ile finansal zaman serisi öngörüleri (Yayımlanmamış doktora tezi). Başkent Üniversitesi Sosyal Bilimler Enstitüsü, Ankara.
5. Bachelier, L. (1900). Theory of speculation (Yayımlanmamış doktora tezi). Paris: University of Paris.
6. Baciu, O. (2012). Ranking capital markets efficiency: The case of twenty European stock markets. Journal of Applied Quantitative Methods, 9(3), 24-33. Retrieved from http://www.jaqm.ro/
7. Bayraktar, A. (2012). Etkin piyasalar hipotezi. Aksaray Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi, 4(1), 37-47. Erişim adresi: http://aksarayiibd.aksaray.edu.tr/
8. Benston, G. and Hartgraves, A. (2002). Enron: What happened and what we can learn from it. Journal of Accounting and Public Policy, 21(2), 105-127. https://doi.org/10.1016/S0278-4254(02)00042-X

9. Çelik, M., Kurtaran, A. ve Kurtaran, A. (2018). Zayıf formda piyasa etkinliğinin Türkiye hisse senedi piyasasında test edilmesi [Özel Sayı]. Uluslararası İktisadi ve İdari İncelemeler Dergisi, 457-474. doi:10.18092/ulikidince.456639
10. Çelik, T. T. ve Taş, O. (2007). Etkin piyasa hipotezi ve gelişmekte olan hisse senedi piyasaları. İstanbul Teknik Üniversitesi Dergisi, 4(2), 11-22. Erişim adresi: http://itudergi.itu.edu.tr/
11. Dondurmacı, G. ve Çınar, A. (2014). Finans sektöründe veri madenciliği uygulaması. Akademik Sosyal Araştırmalar Dergisi, 2(1), 258-271. doi:10.16992/ASOS.138
12. Draper, N. R. and Smith, H. (1998). Applied regression analysis (3. ed.). New Jersey: Wiley Publishing.
13. Enke, D. and Thawornwong, S. (2005). The use of data mining and neural networks for forecasting stock market returns. Expert Systems with Applications, 29(4), 927–940. http://doi.org/https://doi.org/10.1016/j.eswa.2005.06.024
14. Fama, E. F. (1970). Efficient capital markets: A review of theory and empirical work. The Journal of Finance, 25(2), 383-417. doi:10.2307/2325486
15. Faria, S. and Gonçalves, F. (2013). Financial data modeling by Poisson mixture regression. Journal of Applied Statistics, 40(10), 2150-2162, https://doi.org/10.1080/02664763.2013.807332
16. Goia, A., May, C. and Fusai, G. (2009). Functional clustering and linear regression for peak load forecasting. International Journal of Forecasting, 26(4), 700-711. doi:10.1016/j.ijforecast.2009.05.015
17. Guerrien, B. and Gun, O. (2011). Efficient market hypothesis: What are we talking about? Real World Echonomics Review, 56, 19-30. Retrieved from http://rwer.wordpress.com/
18. Guresen, E., Kayakutlu, G. and Daim, T. (2011). Using artificial neural network models in stock market index prediction. Expert Systems with Applications, 38(8), 10389-10397. doi:10.1016/j.eswa.2011.02.068
19. Güdelek, M. (2019). Zaman serisi analizi ve tahmini: Derin öğrenme yaklaşımı (Yayımlanmamış yüksek lisans tezi). TOBB Üniversitesi Fen Bilimleri Enstitüsü, Ankara.
20. Hao, L. and Naiman, D. (2007). Quantile regression (1. ed.). California: Sage Publications.
21. Hastie, T., Tibshirani, R. and Friedman, J. (2013). The elements of statistical learning: Data mining, inference and prediction (2. ed.). New York: Springer Publishing.
22. Heinen, A. (2008). Modelling time series count data: An autoregressive conditional Poisson model (SSRN Working Paper). Retrieved from: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1117187
23. Heshmaty, B. and Kandel, A. (1985). Fuzzy linear regression and its applications to forecasting in uncertain environment. Fuzzy Sets and Systems, 15(2), 159-191. https://doi.org/10.1016/0165-0114(85)90044-2
24. Hromkovic, J. (2005). Design and analysis of randomized algorithms (1. ed.). New York: Springer Publishing.
25. Johnson, R. and Zhang, T. (2014). Learning nonlinear functions using regularized greedy forest. IEEE Transactions on Pattern Analysis and Machine Intelligence, 36(5), 942-954. doi:10.1109/TPAMI.2013.159
26. Jordan, M., Ghahramani, Z. and Saul, L. (1997, December). Hidden Markov decision trees. In M. I. Jordan and T. Petsche (Eds.), NIPS'96 (pp. 501-507). Paper presented at the Proceedings of the 9th International Conference on Neural Information Processing Systems, Denver, Colorado. Retrieved from https://dl.acm.org/doi/abs/10.5555/2998981.2999052
27. Kolmogorov, A. N. and Uspensky, V. A. (1987). Algorithms and randomness. Theory of Probability and Its Applications, 32(3), 389-412. Retrieved from www.siam.org
28. Koop, G. (2003). Bayesian econometrics (1. ed.). London: John Wiley & Sons Inc.
29. Koop, G. and Korobilis, D. (2009). Bayesian multivariate time series methods for empirical macroeconomics. Foundations and Trends in Econometrics, 3(4), 267-358. http://dx.doi.org/10.1561/0800000013
30. Krollner, B., Vanstone, B. and Finnie, G. (2010) Financial time series forecasting with machine learning techniques: A survey. Paper presented at the Proceedings of the 18th European Symposium on Artificial Neural Networks: Computational Intelligence and Machine Learning. Retrieved from https://pure.bond.edu.au/ws/files/27498056/Financial_time_series_forecasting_with_machine_learning_techniques.pdf
31. Lee, C.-M. and Ko, C.-N. (2009). Time series prediction using RBF neural networks with a nonlinear time-varying evolution PSO algorithm. Neurocomputing, 73(1), 449-460. https://doi.org/10.1016/j.neucom.2009.07.005
32. Leetaru, K. (2008). An open source study of international media coverage of the WorldCom scandal. The Journal of International Communication, 14(2), 66-86. https://doi.org/10.1080/13216597.2008.9674733
33. Marcek, D., Marcek, M. and Babel, J. (2009) Granular RBF NN approach and statistical methods applied to modelling and forecasting high frequency data. International Journal of Computational Intelligence Systems, 2(4), 353-364. doi:10.1080/18756891.2009.9727667
34. Mazzola, E. and Muliere, P. (2011). Reviewing alternative characterizations of Meixner process. Probability Surveys, 8, 127-154. doi:10.1214/11-PS177
35. Meek, C., Chickering, D. and Heckerman, D. (2002). Autoregressive tree models for time-series analysis. In R. Grossman, J. Han, V. Kumar, H. Mannila and R. Motwani (Eds.), 2002 SIAM International Conference on Data Mining (pp. 229-244). Paper presented at the Proceedings of the 2002 Society for Industrial and Applied Mathematics International Conference on Data Mining, Virginia, USA.
36. Meinshausen, N. (2006). Quantile regression forests. Journal of Machine Learning Research, 7(35), 983-999. Retrieved from https://www.jmlr.org/
37. Millionis, A. and Maschos, D. (2000). On the validity of the weak-form efficient markets hypothesis applied to the London stock exchange. Applied Economics Letters, 7(7), 419-421. doi:10.1080/135048500351087
38. Minsky, M. and Papert, S. A. (1969). Perceptrons (1. ed.). Massachusetts: The MIT Press.
39. Minsky, M. and Papert, S. A. (2017). Perceptrons: An introduction to computational geometry. Massachusetts: MIT press.
40. Moghaddam, A., Moghaddam, M. and Esfandyari, M. (2016). Stock market index prediction using artificial neural network. Journal of Economics Finance and Administrative Science, 21(41), 89-93. doi:10.1016/j.jefas.2016.07.002
41. Nannavecchia, A. (2015). The Meixner process for financial data. Megatrend Review, 12(2), 33-44. doi:10.5937/MegRev1502033N
42. Özdemir, Z., Atan, S. ve Atan, D. (2016). Hisse senedi piyasasında zayıf formda etkinlik: İMKB üzerine ampirik bir çalışma. Dokuz Eylül Üniversitesi İktisadi İdari Bilimler Fakültesi Dergisi, 24(2), 33-48. Erişim adresi: https://dergipark.org.tr/tr/pub/ije
43. Pole, A., West, M. and Harrison, J. (1994). Applied Bayesian forecasting and time series analysis. New York: Chapman & Hall.
44. Poyarkov, A., Drutsa, A., Khalyavin, A., Gusev, G. and Serdyukov, P. (2016). Boosted decision tree regression adjustment for variance reduction in online controlled experiments. In B. Krishnapuram and M. Shah (Eds.), 22nd ACM SIGKDD International Conference (235-244). Paper presented at the Proceedings of the 22nd Knowledge Discovery and Data Mining International Conference.
45. Rokach, L. (2016). Decision forest: Twenty years of research. Information Fusion, 27, 111-125. doi:10.1016/j.inffus.2015.06.005
46. Romashchenko, A., Shen, A. and Vershchagin, N. (2000). Combinatorial interpretation of Kolmogorov complexity. In 15th Annual IEEE Conference on Computational Complexity (131-137). Paper presented at the Proceedings of the Annual Institute of Electrical and Electronics Engineers Conference on Computational Complexity, Florence, Italy.
47. Rosenberg, A. and McIntyre, L. (2011). Philosophy of science. Oxforshire: Routledge.
48. Schoutens, W. (2002). The Meixner Process: Theory and applications in finance (Semantic Scholar Working Paper). Retrieved from https://www.eurandom.tue.nl/reports/2002/004-abstract.pdf
49. Selvamuthu, D., Kumar, V. and Mishra, A. (2019). Indian stock market prediction using artificial neural networks on tick data. Financial Innovation, 5(1), 1-12. Retrieved from https://link.springer.com/
50. Shen, S., Jiang, H. and Zhang, T. (2012) Stock market forecasting using machine learning algorithms (Stanford University Working Paper). Retrieved from http://cs229.stanford.edu/proj2012/ShenJiangZhang-StockMarketForecastingusingMachineLearni ngAlgorithms.pdf
51. Siringnano, J. and Spiliopoulos, K. (2017). Stochastic gradient descent in continuous time. Society for Industrial and Applied Mathematic Journal on Financial Mathematics, 8(1), 933-961. doi:10.1137/17M1126825
52. Smith, C., McGuire, B., Huang, T. and Yang, G. (2006). The history of artificial intelligence (Washington University Working Paper). Retrieved from https://courses.cs.washington.edu/courses/ csep590/06au/projects/history-ai.pdf
53. Swednerg, R. (2010). The structure of confidence and the collapse of Lehman Brothers. Bingley: Emerald Group Publishing Limited.
54. Şıklar, E. (1999). Regresyon analizinde Bayesçi yaklaşım. Anadolu Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi, 15(1), 113-122. Erişim adresi: https://dergipark.org.tr/tr/pub/anadoluibfd
55. Theofilatos, K., Karathanasopoulos, A., Sermpinis, G. and Amorgianiotis, T. (2012). Modelling and trading the DJIA financial index using neural networks optimized with adaptive evolutionary algorithms. Communications in Computer and Information Science, 311, 453-462. doi:10.1007/978-3-642-32909-8_46
56. Tsai, C.-F. and Wang, S.-P. (2009). Stock price forecasting by hybrid machine learning techniques. Paper presented at the International MultiConference of Engineers and Computer Scientists. Retrieved from http://www.iaeng.org/
57. Xu, M., Watanachaturaporn, P., Varshney, P. and Arora, M. (2005). Decision tree regression for soft classification of remote sensing data. Remote Sensing of Environment, 97(3), 322-336. doi:10.1016/j.rse.2005.05.008
58. Yan, X. and Su, X. (2009). Linear regression analysis: Theory and computing. London: World Scientific Publishing Company
59. Yan, X.-B., Wang, Z., Yu, S.-H. and Li, Y.-J. (2005) Time series forecasting with RBF neural network. Paper presented at the 2005 International Conference on Machine Learning and Cybernetics. Retrieved from https://ieeexplore.ieee.org/document/1527764
60. Yu, K. and Moyeed, R. (2001). Bayesian quantile regression. Statistics & Probability Letters, 54(4), 437-447. doi:10.1016/S0167-7152(01)00124-9