Doğuş Üniversitesi Dergisi 1308-6979 Dogus University 10.31671/doujournal.1534375 Finance Finans MAKİNE ÖĞRENMESİ İLE YATAY KESİT HİSSE SENEDİ GETİRİLERİNİN TAHMİN EDİLEBİLİRLİĞİ: KÜRESEL PİYASALARDA AMPİRİK BİR ANALİZ MACHINE LEARNING FOR CROSS-SECTIONAL RETURN PREDICTABILITY: EVIDENCE FROM GLOBAL STOCK MARKETS https://orcid.org/0000-0002-1223-1629 Kurucan Ahmet Salih İSTANBUL ÜNİVERSİTESİ, SOSYAL BİLİMLER ENSTİTÜSÜ, FİNANS (DR) https://orcid.org/0000-0002-3379-7090 Hepşen Ali İSTANBUL ÜNİVERSİTESİ, SOSYAL BİLİMLER ENSTİTÜSÜ, FİNANS (DR) 01 24 2025 26 1 315 338 08 16 2024 09 06 2024 Copyright © 2000, Dogus University Journal 2000 Dogus University Journal

Bu çalışma, küresel hisse senedi piyasa verilerini kullanarak makine öğrenimi modelleriyle incelemektedir. 63 Firma karakteristiği hesaplayarak, makine öğrenme yöntemlerini uyguladığımızda modelimizin ekonomik kazanım ve performans açısından daha iyi sonuçlar göstermiştir. Derin öğrenme modelleriyle karşılaştırıldığında gradient-boosted regresyon ağaçları, derin öğrenme modellerine kıyasla, büyük olasılıkla düşük sinyal-gürültü oranı, derin modellerinin hiper-parametrelere karşı yüksek hassasiyet göstermesi daha tutarlı ve güvenilir sonuçlar vermektedir. Sonuçlar makine öğrenmesi yöntemlerinin başarılı portföyler oluşturmak için de kullanılabileceğini göstermektedir. Ayrıca, model karmaşıklığını artırmanın Sharpe oranlarında iyileşmeler gibi ekonomik faydalar sağladığı gösterilmektedir. Sonuçlar, makine öğrenimi modellerinin tutarlılığını ve genelleme yeteneğini vurgulayarak, modern finansal sistemde makine öğreniminin önemini ortaya koymaktadır. Bu çalışma, sadelik ilkesine dair geleneksel anlayışları sorgulamakta ve belirli bir karmaşıklık derecesine dayalı olarak stratejik ekonomik kazançlar gösterdiğini ortaya koymaktadır.

This work examines cross-sectional stock returns with machine learning models using global stock market data. By calculating 63 firm level characteristics, we find that our model outperforms linear models in terms of both economic and statistical performance. Shallow models, such as gradient boosted decision trees, provides more consistent and reliable performance compared to deeper ones in the context of asset pricing, likely due to a low signal-to-noise ratio and sensitivity to parameters. The results revealed that machine learning models can be developed into effective portfolios, complexity is welcomed when it enhances performance such as Sharpe ratios. Taken together, these results demonstrate the relative importance of machine learning for a modern financial system, and specifically, the ability to synthesize information from various characteristics that impact stock returns. This study challenges traditional notions of a preference for parsimony and, based on certain degrees of complexity, demonstrates strategic economic gains.

 Machine Learning Asset Pricing Equity Risk Premium Predictive Modeling Return Predictability Financial Analysis Makine Öğrenimi Varlık Fiyatlaması Hisse Senedi Risk Primi Tahminsel Modelleme Getirilerin Tahmin Edilebilirliği Finansal Analiz Asness, C. S., Moskowitz, T. J., & Pedersen, L. H. (2013). Value and Momentum Everywhere. The Journal of Finance, 68(3), 929–985. Atsalakis, G. S., & Valavanis, K. P. (2009). Surveying stock market forecasting techniques – Part II: Soft computing methods. Expert Systems with Applications, 36(3), 5932–5941. https://doi.org/10.1016/j.eswa.2008.07.006 Bali, T. G., Goyal, A., Huang, D., Jiang, F., & Wen, Q. (2020). The Cross-Sectional Pricing of Corporate Bonds Using Big Data and Machine Learning. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.3686164 Bryzgalova, S., Pelger, M., & Zhu, J. (2020). Forest Through the Trees: Building Cross-Sections of Stock Returns. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.3493458 Diebold, F., & Mariano, R. (1995). Comparing Predictive Accuracy. Journal of Business & Economic Statistics, 13(3), 253–263. Dillschneider, Y. (2022). A Machine Learning Framework for Asset Pricing (SSRN Scholarly Paper 4097100). https://doi.org/10.2139/ssrn.4097100 Drobetz, W., & Otto, T. (2020). Empirical Asset Pricing via Machine Learning: Evidence from the European Stock Market. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.3640631 Fama, E. F., & French, K. R. (1992). The Cross-Section of Expected Stock Returns. The Journal of Finance, 47(2), 427–465. https://doi.org/10.1111/j.1540-6261.1992.tb04398.x Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics, 33(1), 3–56. https://doi.org/10.1016/0304-405X(93)90023-5 Fama, E. F., & French, K. R. (2015). A five-factor asset pricing model. Journal of Financial Economics, 116(1), 1–22. https://doi.org/10.1016/j.jfineco.2014.10.010 Fama, E. F., & French, K. R. (2017). International tests of a five-factor asset pricing model. Journal of Financial Economics, 123(3), 441–463. https://doi.org/10.1016/j.jfineco.2016.11.004 Freund, Y. (1995). Boosting a Weak Learning Algorithm by Majority. Information and Computation, 121(2), 256–285. https://doi.org/10.1006/inco.1995.1136 Giglio, S., & Xiu, D. (2021). Asset Pricing with Omitted Factors. Journal of Political Economy, 129(7), 1947–1990. https://doi.org/10.1086/714090 Gu, S., Kelly, B., & Xiu, D. (2020). Empirical Asset Pricing via Machine Learning. Review of Financial Studies, 33(5). https://doi.org/10.1093/rfs/hhaa009 Harvey, C. R., & Liu, Y. (2021). Lucky factors. Journal of Financial Economics, 141(2), 413–435. https://doi.org/10.1016/j.jfineco.2021.04.014 Heaton, J. B., Polson, N. G., & Witte, J. H. (2016). Deep Learning in Finance (Version 3). arXiv. https://doi.org/10.48550/ARXIV.1602.06561 Jerome H. Friedman. (2001). Greedy function approximation: A gradient boosting machine. The Annals of Statistics, 29(5), 1189–1232. https://doi.org/10.1214/aos/1013203451 Kelly, B. T., & Malamud, S. (2021). The Virtue of Complexity in Machine Learning Portfolios. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.3984925 Kelly, B. T., & Xiu, D. (2023). Financial Machine Learning. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.4501707 Kozak, S. (2019). Kernel Trick for the Cross Section. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.3307895 Leippold, M., Wang, Q., & Zhou, W. (2022). Machine learning in the Chinese stock market. Journal of Financial Economics, 145(2). https://doi.org/10.1016/j.jfineco.2021.08.017 Moritz, B., & Zimmermann, T. (2016). Tree-Based Conditional Portfolio Sorts: The Relation between Past and Future Stock Returns. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.2740751 Rasekhschaffe, K. C., & Jones, R. C. (2019). Machine Learning for Stock Selection. Financial Analysts Journal, 75(3). https://doi.org/10.1080/0015198X.2019.1596678 Schapire, R. E. (1990). The strength of weak learnability. Machine Learning, 5(2), 197–227. https://doi.org/10.1007/BF00116037 Subrahmanyam, A. (2010). The Cross‐Section of Expected Stock Returns: What Have We Learnt from the Past Twenty‐Five Years of Research? European Financial Management, 16(1), 27–42. https://doi.org/10.1111/j.1468-036X.2009.00520.x Swade, A., Hanauer, M. X., Lohre, H., & Blitz, D. (2023). Factor Zoo. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.4605976 Tobek, O., & Hronec, M. (2021). Does it pay to follow anomalies research? Machine learning approach with international evidence. Journal of Financial Markets, 56, 100588. https://doi.org/10.1016/j.finmar.2020.100588 Tukey, J. W. (1961). Box and Jenkins. In Time Series Analysis: Forecasting and Control.