Designing A Trading Strategy to Buy and sell the stock of Companies Listed on The New York Stock Exchange Based on Classification Learning Algorithms

Advances in Mathematical Finance and Applications

مقالات آماده انتشار، پذیرفته شده، انتشار آنلاین از تاریخ 10 تیر 1402

نوع مقاله: Research Paper

شناسه دیجیتال (DOI): 10.22034/amfa.2022.1967149.1796

نویسندگان

Nasser Heydari¹؛ Majid Zanjirdar^{* 1}؛ Ali Lalbar²

¹Department of Financial Management, Arak Branch, Islamic Azad University, Arak, Iran

²Department of Accounting, Arak Branch, Islamic Azad University, Arak, Iran

چکیده

This study aimed to design a stock trading strategy for companies listed on the New York Stock Exchange as one of the largest stock markets worldwide. The data were collected from available libraries and the yahoo finance information base. Indicators and oscillators of technical analysis were considered as the input components of the model, stock trading strategies were designed based on classification machine learning algorithms, and the optimal model was introduced based on statistical indicators. Accuracy, recall, and F-measure indicators were used to evaluate the classification algorithms. In addition, advanced statistical methods and Python, Spyder, SPSS, and Excel software were utilized, and Kruskal-Wallis Test was applied to investigate the difference between the designed strategies. A total of 41 active companies listed on the New York Stock Exchange in financial services, health care, technology, communications services, consumer cycles, consumer support, and energy with a market capitalization greater than $200 trillion and average trading volume greater than 1 million were selected by Filter writing method on 06/28/2021. The model evaluation indicators showed that the designed random forest trading strategy model fits the data and significantly differs from other strategies based on classification learning algorithms regarding statistical indicators

کلیدواژه‌ها

Trading Strategy؛ Machine Learning؛ Classification Algorithms