Evaluation of BIST100 Index Prediction Performance of Deep and Machine Learning Algorithms

Yıl 2024, Cilt: 20 Sayı: 2, 394 - 408, 30.12.2024

Yunus Emre Gür

Öz

This study investigates the possibility of forecasting the Borsa Istanbul BIST 100 index using machine learning and deep learning techniques. The study uses the BIST 100 index as the dependent variable. In addition, gram gold price, daily dollar exchange rate (in TL), daily euro exchange rate (in TL), BIST trading volume, daily Brent oil prices, BIST trading volume, BIST overnight repo rates, and BIST Industrial Index (XUSIN) data are used as independent variables. The Central Bank of the Republic of Turkey provides daily statistics on these variables. The performance of several deep learning recurrent neural networks (RNN) and machine learning network structures—including Random Forest, K-Nearest Neighbors, Multilayer Perceptron, Radial Basis Function, and Support Vector Machine—for predicting the BIST 100 index is tested and compared in this study. The results indicate that the CNN model outperforms the other models in terms of prediction accuracy, with the lowest RMSE and MSE values, and the highest R² value. This suggests that CNN is a robust model for financial forecasting. The relevant literature is summarized in this context in the first portion of the study, after which the methods and results are described. Then the obtained comparative prediction values are presented. Finally, the study is concluded by presenting the interpretations of the results and recommendations.

Anahtar Kelimeler

Machine Learning, Deep Learning, Financial Forecasting, BIST100 Index, CNN Algorithm

Derin ve Makine Öğrenme Algoritmalarının BIST100 Endeksi Tahmin Performanslarının Değerlendirilmesi

Öz

Bu çalışma, Borsa İstanbul BIST 100 endeksinin makine öğrenmesi ve derin öğrenme teknikleri kullanılarak tahmin edilme olasılığını araştırmaktadır. Çalışmada bağımlı değişken olarak BIST 100 endeksi kullanılmıştır. Ayrıca gram altın fiyatı, günlük dolar kuru (TL cinsinden), günlük euro kuru (TL cinsinden), BIST işlem hacmi, günlük Brent petrol fiyatları, BIST işlem hacmi, BIST gecelik repo faizleri ve BIST Sanayi Endeksi (XUSIN) verileri bağımsız değişkenler olarak kullanılmıştır. Türkiye Cumhuriyet Merkez Bankası bu değişkenlere ilişkin günlük istatistikleri sağlamaktadır. Bu çalışmada, Rastgele Orman, K-En Yakın Komşular, Çok Katmanlı Algılayıcı, Radyal Taban Fonksiyonu ve Destek Vektör Makinesi dahil olmak üzere çeşitli derin öğrenme tekrarlayan sinir ağları (RNN) ve makine öğrenimi ağ yapılarının BIST 100 endeksini tahmin etme performansı test edilmiş ve karşılaştırılmıştır. Sonuçlar, CNN modelinin en düşük RMSE ve MSE değerleri ve en yüksek R² değeri ile tahmin doğruluğu açısından diğer modellerden daha iyi performans gösterdiğini ortaya koymaktadır. Bu da CNN'in finansal tahmin için sağlam bir model olduğunu göstermektedir. Çalışmanın ilk bölümünde bu bağlamda ilgili literatür özetlenmiş, ardından yöntem ve sonuçlar açıklanmıştır. Daha sonra elde edilen karşılaştırmalı tahmin değerleri sunulmuştur. Son olarak, sonuçlara ilişkin yorumlar ve öneriler sunularak çalışma sonlandırılmıştır.

Anahtar Kelimeler

Makine Öğrenimi, Derin Öğrenme, Finansal Tahmin, BIST100 Endeksi, CNN Algoritması

Etik Beyan

Çalışma İçin Etik Kurul izni almaya ihtiyaç duyulmamıştır.

Kaynakça

• Alaca, M., and Güran, A. (2022). Trend Forecasting of BIST 100 Index Using Sentiment Scores and Technical Indicators During the COVID-19 Pandemic. Journal of Information Technologies, 15(4), 379-388.
• Alkhatib, K., Najadat, H., Hmeidi, I., and Shatnawi, M,K,A. (2013). Stock price prediction using k-nearest neighbor (kNN) algorithm. International Journal of Business, Humanities and Technology, 3(3), 32-44.
• Alper, D., and Kara, E. (2017). Macroeconomic Factors Affecting Stock Returns in Borsa Istanbul: A Research on BIST Industrial Index. Journal of Süleyman Demirel University Faculty of Economics and Administrative Sciences, 22(3), 713-730.
• Alshaikhdeeb, A. J., and Cheah, Y. N. (2023). Utilizing Word Index Approach with LSTM Architecture for Extracting Adverse Drug Reaction from Medical Reviews. Journal of Advances in Information Technology, 14(3).
• Aydin, A. D., and Cavdar, S. C. (2015). Comparison Of Prediction Performances Of Artificial Neural Network (ANN) And Vector Autoregressive (VAR) Models By Using The Macroeconomic Variables Of Gold Prices, Borsa Istanbul (BIST) 100 Index And US Dollar-Turkish Lira (USD/TRY) Exchange Rates. Procedia Economics and Finance, 30, 3-14.
• Bulut, E. (2024). Market Volatility and Models for Forecasting Volatility. In Business Continuity Management and Resilience: Theories, Models, and Processes (pp. 220-248). IGI Global.
• Cho K., Van Merrienboer B., Gulcehre C. et al., (2014). Learning Phrase Representations Using RNN Encoder-Decoder For Statistical Machine Translation, arXiv preprint arXiv:1406.1078, 2014.
• Fenghua, W. E. N., Jihong, X. I. A. O., Zhifang, H. E., and Xu, G. O. N. G. (2014). Stock price prediction based on SSA and SVM. Procedia Computer Science, 31, 625-631.
• Gao, Y., Wang, R. and Zhou, E. (2021). Stock Prediction Based on Optimized LSTM and GRU Models. Scientific Programming, 2021. https://doi.org/10.1155/2021/4055281
• Heo, J., and Yang, J. Y. (2016). Stock Price Prediction Based On Financial Statements Using SVM. International Journal of Hybrid Information Technology, 9(2), 57-66.
• Hsieh, C. H., Li, Y. S., Hwang, B. J., and Hsiao, C. H. (2020). Detection of atrial fibrillation using 1D convolutional neural network. Sensors, 20(7), 2136.
• Ince, H., and Trafalis, T. B. (2008). Short Term Forecasting With Support Vector Machines And Application To Stock Price Prediction. International Journal of General Systems, 37(6), 677-687.
• Kantar, L. (2020). Forecasting The Bıst 100 Index With Artificıial Neural Networks And Arma Model. Journal of Accounting and Finance Review, 3(2), 121-131.
• Karcıoğlu, R., and Özer, A. (2014). Determination of Factors Affecting Stock Returns in BIST: Static and Dynamic Panel Data Analysis. Journal of Uludag University Faculty of Economics and Administrative Sciences, 33(1), 43-70.
• Kemalbay, G., and Alkış, B. N. (2020). Prediction of Stock Market Index Movement Direction with Multiple Logistic Regression and K-Nearest Neighbor Algorithm. Pamukkale University Journal of Engineering Sciences, 27(4), 556-569.
• Koyuncu, T. (2018). The Relationship of BIST-100 Index with Macroeconomic Variables: An Empirical Study. Journal of Finance, Economics and Social Research, 3(3), 615-624.
• Kurani, A., Doshi, P., Vakharia, A., and Shah, M. (2023). A Comprehensive Comparative Study Of Artificial Neural Network (ANN) And Support Vector Machines (SVM) On Stock Forecasting. Annals of Data Science, 10(1), 183-208.
• Li, M., Zhu, Y., Shen, Y., and Angelova, M. (2023). Clustering-Enhanced Stock Price Prediction Using Deep Learning. World Wide Web, 26(1), 207-232.
• Mehtab, S., and Sen, J. (2020, November). Stock Price Prediction Using CNN And LSTM-Based Deep Learning Models. In 2020 International Conference on Decision Aid Sciences and Application (DASA) (pp. 447-453). IEEE.
• Mehtab, S., Sen, J., and Dasgupta, S. (2020). Robust Analysis Of Stock Price Time Series Using CNN And LSTM-Based Deep Learning Models. In 2020 4th International Conference on Electronics, Communication and Aerospace Technology (ICECA) (pp. 1481-1486). IEEE.
• Mei, J., He, D., Harley, R., Habetler, T., and Qu, G. (2014). A Random Forest Method For Real-Time Price Forecasting İn New York Electricity Market. In 2014 IEEE PES General Meeting| Conference & Exposition (pp. 1-5). IEEE.
• Mukherjee, S., Sadhukhan, B., Sarkar, N., Roy, D., and De, S. (2023). Stock Market Prediction Using Deep Learning Algorithms. CAAI Transactions on Intelligence Technology, 8(1), 82-94.
• Nikou, M., Mansourfar, G., and Bagherzadeh, J. (2019). Stock Price Prediction Using DEEP Learning Algorithm And İts Comparison With Machine Learning Algorithms. Intelligent Systems In Accounting, Finance and Management, 26(4), 164-174.
• Reddy, V. K. S. (2018). Stock Market Prediction Using Machine Learning. International Research Journal of Engineering and Technology (IRJET), 5(10), 1033-1035.
• Sarıkoç, M., and Çelik, M. (2022). BIST100 Index Price Prediction with Dimension Reduction Techniques and LSTM Deep Learning Network. European Journal of Science and Technology, (34), 519-524.
• Sethia, A., and Raut, P. (2019). Application of LSTM, GRU and ICA for stock price prediction. In Information and Communication Technology for Intelligent Systems: Proceedings of ICTIS 2018, Volume 2 (pp. 479-487). Springer Singapore.
• Sevinç, E. (2014). Determination of the Effects of Macroeconomic Variables on Stock Returns Traded in BÌST-30 Index Using Arbitrage Pricing Model. Istanbul University Journal of the School of Business Administration, 43(2), 271-292.
• Simsek, A. I. (2024a). Using Machine Learning and Deep Learning Methods in Predicting the Islamic Index Price. In Fintech Applications in Islamic Finance: AI, Machine Learning, and Blockchain Techniques (pp. 272-285). IGI Global.
• Simsek, A. I. (2024b). Improving the Performance of Stock Price Prediction: A Comparative Study of Random Forest, XGBoost, and Stacked Generalization Approaches. In Revolutionizing the Global Stock Market: Harnessing Blockchain for Enhanced Adaptability (pp. 83-99). IGI Global.
• Togacar, M., Ergen, B., and Sertkaya, M. E. (2019). Subclass separation of white blood cell images using convolutional neural network models. Elektronika ir Elektrotechnika, 25(5), 63-68.
• Vijh, M., Chandola, D., Tikkiwal, V. A., and Kumar, A. (2020). Stock Closing Price Prediction Using Machine Learning Techniques. Procedia Computer Science, 167, 599-606.
• Wenjie, L. U., Jiazheng, L. I., Jingyang, W. A. N. G., and Shaowen, W. U. (2022). A Novel Model For Stock Closıng Prıce Predıctıon Usıng Cnn-Attentıon-Gru-Attentıon. Economic Computation & Economic Cybernetics Studies & Research, 56(3).
• Yan, W. L. (2023). Stock Index Futures Price Prediction Using Feature Selection and Deep Learning. The North American Journal of Economics and Finance, 64, 101867.