基于支持向量机模型的多因子量化选股策略

作     者：咪纳 赵予宁 何秉坤 

作者机构：贵州大学经济学院贵州 贵阳 

出 版 物：《电子商务评论》 (E-Commerce Letters)

年 卷 期：2024年第13卷第3期

页      面：8619-8628页

学科分类：12[管理学] 02[经济学] 0202[经济学-应用经济学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 020204[经济学-金融学（含∶保险学）] 

主　　题：多因子 支持向量机 量化选股 

摘      要：随着金融计算机领域的迅猛发展,量化投资正在扮演越来越重要的角色。多因子选股模型作为量化投资领域的重要组成部分,是量化投资策略选取优质股票组合的有效工具。本文以沪深300成分股为研究对象,综合考虑了成长类、技术类、价值类、情绪类以及动量类对股价有影响的因子,运用因子暴露分析、相关系数以及因子IC确定能够显著影响股票收益率的因子,再建立具有最优参数的核函数支持向量机模型,选取2010年1月1日到2023年1月1日作为时间区间进行训练与回测。得出结论:沪深300成分股的有效因子为市值、股本、换手率、ROE和PE;由高斯核函数支持向量机模型选股策略构建的投资组合策略能够保持投资组合的多样性,具有较高的风险回报能力,且收益较为稳定,能够跑赢大盘,证明了模型的有效性。With the rapid development of the financial computer field, quantitative investment is playing an increasingly important role. The multi factor stock selection model, as an important component of quantitative investment, is an effective tool for selecting high-quality stock portfolios in quantitative investment strategies. This article takes the 300 constituent stocks of Shanghai and Shenzhen as the research object, comprehensively considers the factors that affect stock prices, including growth, technology, value, emotion, and momentum. Factor exposure analysis, correlation coefficients, and factor IC are used to determine the factors that can significantly affect stock returns. Then, a kernel function support vector machine model with the optimal parameters is established, selecting January 1, 2010 to January 1, 2023 as the time interval for training and backtesting. Conclusion: The effective factors of the 300 constituent stocks in Shanghai and Shenzhen are market value, share capital, turnover rate, ROE, and PE;The investment portfolio strategy constructed by Gaussian kernel support vector machine model stock selection strategy can maintain the diversity of the investment portfolio, have high risk return ability, and stable returns, which can outperform the market, proving the effectiveness of the model.