DÜNYA BORSALARINDA İKİLİ İŞLEM STRATEJİLERİNİN KARŞILAŞTIRMASI: MAKİNE ÖĞRENMESİ İLE Z-SKOR TAHMİNİ VE DİNAMİK EŞİK DEĞERİ BELİRLEME MODELLERİNİN İSTATİSTİKSEL ARBİTRAJ PERFORMANS ANALİZİ

Yıl 2025, Cilt: 9 Sayı: 1, 111 - 135, 30.03.2025

Egemen Kahraman

https://doi.org/10.47525/ulasbid.1620865

Öz

İstatistiksel arbitraja dayalı ikili işlem stratejilerinin modern makine öğrenmesi teknikleriyle birleştirilmesi, finansal piyasalarda işlem performansını artırmak için yenilikçi bir yaklaşım sunmaktadır. Çalışmada, altı önemli küresel piyasa endeksinden (FCHI, GDAXI, GSPC, HSI, IXIC ve N225) oluşan endeks verileri kullanılmıştır. İkili işlemlere uygun endeks çiftlerinin seçimi için öncelikle Genişletilmiş Dickey-Fuller (ADF) testi kullanılarak zaman serilerinin durağanlık analizi gerçekleştirilmiş, ardından Engle-Granger eşbütünleşme testi ile arasında uzun dönemli ilişkinin varlığı tespit edilen endeks çiftleri belirlenmiştir.
Araştırmada, statik z-puanı eşiklerine dayalı temel stratejinin yanı sıra, Lojistik Regresyon, Rastgele Orman, XGBoost ve SVM gibi denetimli makine öğrenmesi tekniklerinin kullanıldığı hibrit bir yaklaşım geliştirilmiştir. Ayrıca, sürekli optimize edilen eşik değerlerine dayanan dinamik bir model de oluşturulmuştur. Stratejilerin performansı, Sharpe Oranı, Sortino Oranı, maksimum düşüş ve F1 skoru gibi metriklerle değerlendirilmiştir. Sonuçlar, dinamik eşik değeri ve makine öğrenmesi destekli modellerin geleneksel yaklaşımlara kıyasla üstün performans gösterdiğini ortaya koymuştur

Anahtar Kelimeler

İstatistiksel Arbitraj, Makine Öğrenmesi, Eşbütünleşme Analizi, Dinamik Eşikleme, Piyasa Endeksleri.

Etik Beyan

Makalenin yazarı, bu çalışma ile ilgili taraf olabilecek herhangi bir kişi, kurum veya kuruluşun finansal ilişkileri bulunmadığını dolayısıyla herhangi bir çıkar çatışmasının olmadığını beyan eder.

Destekleyen Kurum

Çalışmada herhangi bir kurum ya da kuruluştan destek alınmamıştır.

COMPARATIVE ANALYSIS OF PAIRS TRADING STRATEGIES IN GLOBAL MARKETS: EVALUATING STATISTICAL ARBITRAGE USING MACHINE LEARNING-BASED Z-SCORE PREDICTION AND DYNAMIC THRESHOLDS

Öz

This study explores the integration of machine learning methodologies with traditional statistical arbitrage techniques to enhance performance in financial markets. Using data from six major global indices—CAC 40 (FCHI), DAX (GDAXI), S&P 500 (GSPC), Hang Seng (HSI), NASDAQ Composite (IXIC), and Nikkei 225 (N225)—the research applies the Augmented Dickey-Fuller (ADF) test to assess stationarity and the Engle-Granger cointegration test to identify statistically significant index pair relationships. The study evaluates three strategic approaches. First, a baseline strategy employs static z-score thresholds as a reference model. Second, a dynamic approach continuously adjusts threshold values in response to market fluctuations. Finally, a hybrid methodology incorporates machine learning algorithms—including Logistic Regression, Random Forest, XGBoost, and SVM—to enhance predictive capabilities. Performance is assessed using key financial metrics such as the Sharpe Ratio, Sortino Ratio, maximum drawdown, and F1 score. Results indicate that both dynamic thresholding and machine learning-augmented models outperform traditional approaches, offering improved risk-adjusted returns.

Anahtar Kelimeler

Pairs Trade, Machine Learning, Cointegration Analysis, Dynamic Thresholding, Market İndices

Kaynakça

• Baek, S., Glambosky, M., Oh, S. H., & Lee, J. (2020). Machine learning and algorithmic pairs trading in futures markets. Sustainability, 12(7).
• Bağcı, M., & Soylu, P. K. (2024). The Optimal Threshold Selection for High-Frequency Pairs Trading via Supervised Machine Learning Algorithms. doi:10.13140/RG.2.2.26440.53769
• Bertram, W. K. (2010). Analytic solutions for optimal statistical arbitrage trading. Physica A: Statistical mechanics and its applications, 389(11), pp. 2234-2243.
• Bogomolov, T. (2011). Pairs trading in the land down under.
• Caldeira, J. F., & Moura, G. V. (2012). Selection of a portfolio of pairs based on cointegration: the Brazilian case. Federal University of Rio Grande do Sul, Federal University of Santa Catarina.
• Caldeira, J. F., & Moura, G. V. (2013). Selection of a portfolio of pairs based on cointegration: A statistical arbitrage strategy. Revista Brasileira de Financas, 11(1), 49-80.
• Chaudhuri, T. D., Ghosh, I., & Singh, P. (2017). Application of Machine Learning Tools in Predictive Modeling of Pairs Trade in Indian Stock Market. IUP Journal of Applied Finance, 23(1).
• Chen, C. W., Chen, M., & Chen, S. Y. (2014). Pairs trading via three-regime threshold autoregressive GARCH models. Modeling Dependence in Econometrics: Selected Papers of the Seventh International Conference of the Thailand Econometric Society,Faculty of Economics, Chiang Mai University (pp. 127-140). Thailand: Springer International Publishing.
• Chen, Z., Wang, Z., & Sun, P. (2022). A Novel Machine Learning-assisted Pairs Trading Approach for Trading Risk Reduction. 2022 IEEE 1st Global Emerging Technology Blockchain Forum: Blockchain & Beyond (iGETblockchain) (pp. 1-6). IEEE.
• Chicco, D., & Jurman, G. (2020). The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation. . BMC genomics, 1-13.
• Chiu, M. C., & Wong, H. Y. (2015). Dynamic cointegrated pairs trading: Mean–variance time-consistent strategies. Journal of Computational and Applied Mathematics, 290, 516-534.
• Cummins, M., & Bucca, A. (2012). Quantitative spread trading on crude oil and refined products markets. Quantitative Finance, 12(12), pp. 1857-1875.
• Dickey, D. A., & Fuller, W. A. (1979). Distribution of the estimators for autoregressive time series with a unit root. Journal of the American statistical association, 74(366a), pp. 427-431
• Do, B., Faff, R., & Hamza, K. (2006). A new approach to modeling and estimation for pairs trading. Proceedings of 2006 financial management association European conference, 1, pp. 87-99.
• Elliott, R. J., Van Der Hoek, J., & Malcolm, W. P. (2005). Pairs trading. Quantitative Finance, 5(3), pp. 271–276.
• Engle, R. F., & Granger, C. W. (1987). Co-integration and error correction: representation, estimation, and testing. Econometrica: journal of the Econometric Society, 251-276.
• Fu, N., Kang, M., Hong, J., & Kim, S. (2024). Enhanced Genetic-Algorithm-Driven Triple Barrier Labeling Method and Machine Learning Approach for Pair Trading Strategy in Cryptocurrency Markets. Mathematics, 12(5), 780.
• Gatev, E., Goetzmann, W. N., & Rouwenhorst, K. G. (2006). Pairs trading: Performance of a relative-value arbitrage rule. The Review of Financial Studies, 19(3), 797-827.
• Hong, G., & Susmel, R. (2003). Pairs-trading in the Asian ADR market. University of Houston, Unpublished Manuscript.
• Huck, N., & Afawubo, K. (2015). Pairs trading and selection methods: is cointegration superior? Applied Economics, 47(6), 599-613.
• Kim, T., & Kim, H. Y. (2019). ptimizing the Pairs‐Trading Strategy Using Deep Reinforcement Learning with Trading and Stop‐Loss Boundaries. Complexity, 1.
• Krauss, C. (2017). Statistical arbitrage pairs trading strategies: Review and outlook. Journal of Economic Surveys, 31(2), 513-545.
• Liew, R. Q., & Wu, Y. (2013). Pairs trading: A copula approach. Journal of Derivatives & Hedge Funds, 19, pp. 12-30.
• Lin, Y. X., McCrae, M., & Gulati, C. (2006). Loss protection in pairs trading through minimum profit bounds: A cointegration approach. Advances in Decision Sciences.
• Nath, P. (2003). High frequency pairs trading with US treasury securities: Risks and rewards for hedge funds. vailable at SSRN 565441.
• Rad, H., Low, R. K., & Faff, R. (2016). The profitability of pairs trading strategies: distance, cointegration and copula methods. Quantitative Finance, 16(10), pp. 1541-1558.
• Sarmento, S. M., & Horta, N. (2020). Enhancing a pairs trading strategy with the application of machine learning. Expert Systems with Applications, 158.
• Van der Have, R., van Dijk, D. J., Kole, H. G., Eisma, I. H., & Diepen, G. (2017). Pairs trading using machine learning: An empirical study. Erasmus School of Economics, 12.
• Vidyamurthy, G. (2004). Pairs trading: Quantitative methods and analysis (Vol. 217). John Wiley & Sons.
• Wu, J. (2015). A pairs trading strategy for GOOG/GOOGL using machine learning.
• Zeng, Z., & Lee, C. G. (2014). Pairs trading: optimal thresholds and profitability. Quantitative Finance,, 14(11), 1881-1893.
• Zhu, J. (2022). The Performances of Different Strategies Based on Time Series Analysis for Chinese. 2022 International Conference on mathematical statistics and economic analysis (MSEA 2022) (pp. 470-476). Atlantis Press.