學術產出-學位論文
文章檢視/開啟
書目匯出
-
題名 考慮ESG因子之AI投資策略
AI investment strategies considering ESG factors作者 吳柏賢
Wu, Po-Hsien貢獻者 楊曉文<br>黃泓智
Yang, Sheau-Wen<br>Huang, Hong-Chih
吳柏賢
Wu, Po-Hsien關鍵詞 特徵篩選
機器學習
投資組合
ESG
集成學習
Feature Selection
Machine Learning
Investment Portfolio
ESG
Ensemble Learning日期 2024 上傳時間 2-一月-2025 11:40:24 (UTC+8) 摘要 最近幾年，ESG 在⾦融市場是個熱⾨的話題，許多⾦融機構想著如何在投資的過程中納⼊永續相關的概念，以促進公司的發展以及獲得好的報酬。因此，本研究針對市場上現有的 ESG 相關因⼦結合公司財務⽐率，進⾏特徵篩選，選出對於股價有解釋能⼒的變數。本⽂選⽤的⽅法為主成份分析、套索回歸和基因演算法這三種模型，對三組變數分別篩選。再來是運⽤機器學習模型，為隨機森林、⾧短期記憶模型、⽀撐向量回歸、集成學習法和極限學習機這五種，分別進⾏預測下⼀期股票價格。本⽂會根據預測報酬率進⾏排名，選出排前幾名的股票納⼊投資組合內。最後，觀察其績效表現。研究結果表明，集成學習法在所有模型中表現最為出⾊，具有精準的預測能⼒，相⽐於另外四個機器學習模型產⽣之投資組合是能有效產⽣⾼報酬和低⾵險。除此之外，發現當投資組合考慮 ESG 相關因⼦時，能產⽣⾼報酬和低⾵險，以及展現良好的抗跌能⼒。本⽂認為有助於幫助投資⼈做決策，未來需關注的不只是企業獲利相關能⼒，還需考慮到企業在永續⽅⾯的作為。
In recent years, ESG (Environmental, Social, and Governance) has become a hot topic in the financial market, with many financial institutions considering how to incorporate sustainability-related concepts into the investment process to promote corporate development and achieve good returns. Therefore, this study focuses on the existing ESG-related factors in the market combined with company financial ratios, conducting feature selection to identify variables that can explain stock prices. The methods used in this paper include Principal Component Analysis (PCA), Lasso Regression, and Genetic Algorithms to select features from three sets of variables. Subsequently, machine learning models, including Random Forest, Long Short-Term Memory (LSTM), Support Vector Regression (SVR), Ensemble Learning, and Extreme Learning Machine (ELM), are employed to predict stock prices in the next period. This study ranks stocks based on predicted returns and selects the top-ranking stocks to form an investment portfolio. Finally, the performance of the portfolio is observed. The results indicate that the Ensemble Learning method outperforms all other models, demonstrating precise predictive capabilities and effectively generating high returns and low risk compared to portfolios generated by the other four machine learning models. Moreover, it is found that when the investment portfolio considers ESG-related factors, it can generate high returns, low risk, and exhibit good resistance to downturns. This paper suggests that in the future investors should not only focus on a company's profitability but also consider its sustainability efforts.參考文獻 朱民芮, 蔡維哲, 楊曉文, & 鄞齊. (2021). 以基本面分析強化社會責任投資績效. https://doi.org/10.6529/RSFM.202109_33(3).0001 廖奕潔. (2023). ESG 股票最適資產配置:基因演算法及機器學習模型運用. 國立政治大學碩士論文. 錢慧娟. (2022). 訊號分解對於集成學習預測股價準確率之影響—以台灣加權股價指數為例. 國立政治大學碩士論文. https://doi.org/10.6814/NCCU202200961 Ahmad, I., Basheri, M., Iqbal, M. J., & Rahim, A. (2018). Performance Comparison of Support Vector Machine, Random Forest, and Extreme Learning Machine for Intrusion Detection. IEEE Access, 6, 33789-33795. https://doi.org/10.1109/access.2018.2841987 Babatunde, O. H., Armstrong, L., Leng, J., & Diepeveen, D. (2014). A Genetic Algorithm-Based Feature Selection.pdf. Breiman, L. (2001). Random forests. Machine learning, 45, 5-32. Chen, S., & Zhou, C. (2021). Stock Prediction Based on Genetic Algorithm Feature Selection and Long Short-Term Memory Neural Network. IEEE Access, 9, 9066-9072. https://doi.org/10.1109/access.2020.3047109 De Lucia, C., Pazienza, P., & Bartlett, M. (2020). Does Good ESG Lead to Better Financial Performances by Firms? Machine Learning and Logistic Regression Models of Public Enterprises in Europe. Sustainability, 12(13). https://doi.org/10.3390/su12135317 Gu, S., Kelly, B., & Xiu, D. (2020). Empirical asset pricing via machine learning. The Review of Financial Studies, 33(5), 2223-2273. Hair, J. F. (2009). Multivariate data analysis. Hochreiter, S., & Schmidhuber, J. (1997). Long Short-Term Memory. Neural Computation, 9(8), 1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735 Holland, J. H. (1992). Genetic algorithms. Scientific american, 267(1), 66-73. Huang, G.-B., Zhu, Q.-Y., & Siew, C.-K. (2006). Extreme learning machine: theory and applications. Neurocomputing, 70(1-3), 489-501. Li, X., Xu, N., Liu, Q., & Luo, W. (2016). Corporate Pyramids and Stock Price Crash Risk: Evidence from China. China Accounting and Finance Review, 18, 1-35. Lin, H.-Y., & Hsu, B.-W. (2023). Empirical Study of ESG Score Prediction through Machine Learning—A Case of Non-Financial Companies in Taiwan. Sustainability, 15(19). https://doi.org/10.3390/su151914106 Ma, Y., Han, R., & Wang, W. (2021). Portfolio optimization with return prediction using deep learning and machine learning. Expert Systems with Applications, 165, 113973. Markowitz, H. (1952). The utility of wealth. Journal of political Economy, 60(2), 151-158. Muthukrishnan, R., & Rohini, R. (2016). LASSO: A feature selection technique in predictive modeling for machine learning. 2016 IEEE international conference on advances in computer applications (ICACA), Orimoloye, L. O., Sung, M.-C., Ma, T., & Johnson, J. E. (2020). Comparing the effectiveness of deep feedforward neural networks and shallow architectures for predicting stock price indices. Expert Systems with Applications, 139, 112828. Oukhouya, H., & El Himdi, K. (2023). Comparing Machine Learning Methods—SVR, XGBoost, LSTM, and MLP— For Forecasting the Moroccan Stock Market Iocma 2023, Saputra, D. C. E., Sunat, K., & Ratnaningsih, T. (2023). A New Artificial Intelligence Approach Using Extreme Learning Machine as the Potentially Effective Model to Predict and Analyze the Diagnosis of Anemia. Healthcare, 11(5). https://doi.org/10.3390/healthcare11050697 Sonkavde, G., Dharrao, D. S., Bongale, A. M., Deokate, S. T., Doreswamy, D., & Bhat, S. K. (2023). Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications. International Journal of Financial Studies, 11(3). https://doi.org/10.3390/ijfs11030094 Tibshirani, R. (1996). Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society Series B: Statistical Methodology, 58(1), 267-288. Ullah, S. (2012). Relationship between corporate governance score and stock prices: evidence from KSE-30 Index companies. International Journal of Business and Social Science, 3(4). Waqar, M., Dawood, H., Guo, P., Shahnawaz, M. B., & Ghazanfar, M. A. (2017). Prediction of Stock Market by Principal Component Analysis 2017 13th International Conference on Computational Intelligence and Security (CIS), Welikala, R. A., Fraz, M. M., Dehmeshki, J., Hoppe, A., Tah, V., Mann, S., Williamson, T. H., & Barman, S. A. (2015). Genetic algorithm based feature selection combined with dual classification for the automated detection of proliferative diabetic retinopathy. Computerized Medical Imaging and Graphics, 43, 64-77. Whelan, T., Atz, U., Holt, T. V., & Clark, C. (2021). Uncovering the Relationship by Aggregating Evidence from 1,000 Plus Studies Published between 2015 – 2020.pdf. Yao, J., Zhang, X., Luo, W., Liu, C., & Ren, L. (2022). Applications of Stacking/Blending ensemble learning approaches for evaluating flash flood susceptibility. International Journal of Applied Earth Observation and Geoinformation, 112. https://doi.org/10.1016/j.jag.2022.102932 描述 碩士
國立政治大學
金融學系
111352020資料來源 http://thesis.lib.nccu.edu.tw/record/#G0111352020 資料類型 thesis dc.contributor.advisor 楊曉文<br>黃泓智 zh_TW dc.contributor.advisor Yang, Sheau-Wen<br>Huang, Hong-Chih en_US dc.contributor.author (作者) 吳柏賢 zh_TW dc.contributor.author (作者) Wu, Po-Hsien en_US dc.creator (作者) 吳柏賢 zh_TW dc.creator (作者) Wu, Po-Hsien en_US dc.date (日期) 2024 en_US dc.date.accessioned 2-一月-2025 11:40:24 (UTC+8) - dc.date.available 2-一月-2025 11:40:24 (UTC+8) - dc.date.issued (上傳時間) 2-一月-2025 11:40:24 (UTC+8) - dc.identifier (其他 識別碼) G0111352020 en_US dc.identifier.uri (URI) https://nccur.lib.nccu.edu.tw/handle/140.119/154980 - dc.description (描述) 碩士 zh_TW dc.description (描述) 國立政治大學 zh_TW dc.description (描述) 金融學系 zh_TW dc.description (描述) 111352020 zh_TW dc.description.abstract (摘要) 最近幾年，ESG 在⾦融市場是個熱⾨的話題，許多⾦融機構想著如何在投資的過程中納⼊永續相關的概念，以促進公司的發展以及獲得好的報酬。因此，本研究針對市場上現有的 ESG 相關因⼦結合公司財務⽐率，進⾏特徵篩選，選出對於股價有解釋能⼒的變數。本⽂選⽤的⽅法為主成份分析、套索回歸和基因演算法這三種模型，對三組變數分別篩選。再來是運⽤機器學習模型，為隨機森林、⾧短期記憶模型、⽀撐向量回歸、集成學習法和極限學習機這五種，分別進⾏預測下⼀期股票價格。本⽂會根據預測報酬率進⾏排名，選出排前幾名的股票納⼊投資組合內。最後，觀察其績效表現。研究結果表明，集成學習法在所有模型中表現最為出⾊，具有精準的預測能⼒，相⽐於另外四個機器學習模型產⽣之投資組合是能有效產⽣⾼報酬和低⾵險。除此之外，發現當投資組合考慮 ESG 相關因⼦時，能產⽣⾼報酬和低⾵險，以及展現良好的抗跌能⼒。本⽂認為有助於幫助投資⼈做決策，未來需關注的不只是企業獲利相關能⼒，還需考慮到企業在永續⽅⾯的作為。 zh_TW dc.description.abstract (摘要) In recent years, ESG (Environmental, Social, and Governance) has become a hot topic in the financial market, with many financial institutions considering how to incorporate sustainability-related concepts into the investment process to promote corporate development and achieve good returns. Therefore, this study focuses on the existing ESG-related factors in the market combined with company financial ratios, conducting feature selection to identify variables that can explain stock prices. The methods used in this paper include Principal Component Analysis (PCA), Lasso Regression, and Genetic Algorithms to select features from three sets of variables. Subsequently, machine learning models, including Random Forest, Long Short-Term Memory (LSTM), Support Vector Regression (SVR), Ensemble Learning, and Extreme Learning Machine (ELM), are employed to predict stock prices in the next period. This study ranks stocks based on predicted returns and selects the top-ranking stocks to form an investment portfolio. Finally, the performance of the portfolio is observed. The results indicate that the Ensemble Learning method outperforms all other models, demonstrating precise predictive capabilities and effectively generating high returns and low risk compared to portfolios generated by the other four machine learning models. Moreover, it is found that when the investment portfolio considers ESG-related factors, it can generate high returns, low risk, and exhibit good resistance to downturns. This paper suggests that in the future investors should not only focus on a company's profitability but also consider its sustainability efforts. en_US dc.description.tableofcontents 第一章 緒論 1 第一節 研究背景與動機 1 第二節 研究目的與貢獻 2 第三節 研究架構 4 第二章 文獻回顧 5 第一節 特徵篩選研究 5 第二節 機器學習預測股票市場的研究 7 第三節 結合ESG之投資組合 9 第三章 研究方法 10 第一節 特徵篩選模型 (Feature Selection) 10 一、主成份分析(Principal Component Analysis, PCA) 10 二、套索回歸(Lasso Regression) 12 三、基因演算法(Genetic Algorithm, GA) 12 第二節 機器學習預測模型 (Machine Learning) 17 一、隨機森林(Random Forest , RF) 17 二、長短期記憶模型(Long Short-Term Memory, LSTM) 17 三、支撐向量回歸(Support Vector Regression, SVR) 18 四、集成學習法(Ensemble Learn) 20 五、極限學習機(Extreme Learn Machines, ELM) 20 第三節 模型參數選擇與優化 21 第四節 模型評估指標 25 一、誤差(Error) 26 二、回測績效指標(Backtesting Performance Metrics) 26 第四章 特徵篩選與預測結果 28 第一節 資料描述及處理 29 第二節 模型最佳參數 32 第三節 篩選結果 33 第四節 預測表現 37 第五章 各種因子投資組合績效評估 40 第一節 投資組合之配置方法 40 一、均值方差(Mean-variance, MV) 40 二、等權重(Equal-weight, EW) 41 三、最小變異數(Minimum-variance, MinVar) 41 第二節 投資組合之績效評估 42 一、年化報酬率 42 二、年化標準差 46 三、夏普比率 49 四、最大回撤率 53 五、投組內相同公司 57 六、篩選後的ESG評分表現 59 第六章 結論與建議 61 第一節 結論 61 一、特徵篩選搭配機器學習的評估 61 二、投資組合績效的評估 62 第二節 未來展望 64 一、考慮總經因子變化 64 二、適當模型選取 64 三、模型應用範圍增加 65 參考文獻 66 附錄 68 zh_TW dc.format.extent 11841040 bytes - dc.format.mimetype application/pdf - dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0111352020 en_US dc.subject (關鍵詞) 特徵篩選 zh_TW dc.subject (關鍵詞) 機器學習 zh_TW dc.subject (關鍵詞) 投資組合 zh_TW dc.subject (關鍵詞) ESG zh_TW dc.subject (關鍵詞) 集成學習 zh_TW dc.subject (關鍵詞) Feature Selection en_US dc.subject (關鍵詞) Machine Learning en_US dc.subject (關鍵詞) Investment Portfolio en_US dc.subject (關鍵詞) ESG en_US dc.subject (關鍵詞) Ensemble Learning en_US dc.title (題名) 考慮ESG因子之AI投資策略 zh_TW dc.title (題名) AI investment strategies considering ESG factors en_US dc.type (資料類型) thesis en_US dc.relation.reference (參考文獻) 朱民芮, 蔡維哲, 楊曉文, & 鄞齊. (2021). 以基本面分析強化社會責任投資績效. https://doi.org/10.6529/RSFM.202109_33(3).0001 廖奕潔. (2023). ESG 股票最適資產配置:基因演算法及機器學習模型運用. 國立政治大學碩士論文. 錢慧娟. (2022). 訊號分解對於集成學習預測股價準確率之影響—以台灣加權股價指數為例. 國立政治大學碩士論文. https://doi.org/10.6814/NCCU202200961 Ahmad, I., Basheri, M., Iqbal, M. J., & Rahim, A. (2018). Performance Comparison of Support Vector Machine, Random Forest, and Extreme Learning Machine for Intrusion Detection. IEEE Access, 6, 33789-33795. https://doi.org/10.1109/access.2018.2841987 Babatunde, O. H., Armstrong, L., Leng, J., & Diepeveen, D. (2014). A Genetic Algorithm-Based Feature Selection.pdf. Breiman, L. (2001). Random forests. Machine learning, 45, 5-32. Chen, S., & Zhou, C. (2021). Stock Prediction Based on Genetic Algorithm Feature Selection and Long Short-Term Memory Neural Network. IEEE Access, 9, 9066-9072. https://doi.org/10.1109/access.2020.3047109 De Lucia, C., Pazienza, P., & Bartlett, M. (2020). Does Good ESG Lead to Better Financial Performances by Firms? Machine Learning and Logistic Regression Models of Public Enterprises in Europe. Sustainability, 12(13). https://doi.org/10.3390/su12135317 Gu, S., Kelly, B., & Xiu, D. (2020). Empirical asset pricing via machine learning. The Review of Financial Studies, 33(5), 2223-2273. Hair, J. F. (2009). Multivariate data analysis. Hochreiter, S., & Schmidhuber, J. (1997). Long Short-Term Memory. Neural Computation, 9(8), 1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735 Holland, J. H. (1992). Genetic algorithms. Scientific american, 267(1), 66-73. Huang, G.-B., Zhu, Q.-Y., & Siew, C.-K. (2006). Extreme learning machine: theory and applications. Neurocomputing, 70(1-3), 489-501. Li, X., Xu, N., Liu, Q., & Luo, W. (2016). Corporate Pyramids and Stock Price Crash Risk: Evidence from China. China Accounting and Finance Review, 18, 1-35. Lin, H.-Y., & Hsu, B.-W. (2023). Empirical Study of ESG Score Prediction through Machine Learning—A Case of Non-Financial Companies in Taiwan. Sustainability, 15(19). https://doi.org/10.3390/su151914106 Ma, Y., Han, R., & Wang, W. (2021). Portfolio optimization with return prediction using deep learning and machine learning. Expert Systems with Applications, 165, 113973. Markowitz, H. (1952). The utility of wealth. Journal of political Economy, 60(2), 151-158. Muthukrishnan, R., & Rohini, R. (2016). LASSO: A feature selection technique in predictive modeling for machine learning. 2016 IEEE international conference on advances in computer applications (ICACA), Orimoloye, L. O., Sung, M.-C., Ma, T., & Johnson, J. E. (2020). Comparing the effectiveness of deep feedforward neural networks and shallow architectures for predicting stock price indices. Expert Systems with Applications, 139, 112828. Oukhouya, H., & El Himdi, K. (2023). Comparing Machine Learning Methods—SVR, XGBoost, LSTM, and MLP— For Forecasting the Moroccan Stock Market Iocma 2023, Saputra, D. C. E., Sunat, K., & Ratnaningsih, T. (2023). A New Artificial Intelligence Approach Using Extreme Learning Machine as the Potentially Effective Model to Predict and Analyze the Diagnosis of Anemia. Healthcare, 11(5). https://doi.org/10.3390/healthcare11050697 Sonkavde, G., Dharrao, D. S., Bongale, A. M., Deokate, S. T., Doreswamy, D., & Bhat, S. K. (2023). Forecasting Stock Market Prices Using Machine Learning and Deep Learning Models: A Systematic Review, Performance Analysis and Discussion of Implications. International Journal of Financial Studies, 11(3). https://doi.org/10.3390/ijfs11030094 Tibshirani, R. (1996). Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society Series B: Statistical Methodology, 58(1), 267-288. Ullah, S. (2012). Relationship between corporate governance score and stock prices: evidence from KSE-30 Index companies. International Journal of Business and Social Science, 3(4). Waqar, M., Dawood, H., Guo, P., Shahnawaz, M. B., & Ghazanfar, M. A. (2017). Prediction of Stock Market by Principal Component Analysis 2017 13th International Conference on Computational Intelligence and Security (CIS), Welikala, R. A., Fraz, M. M., Dehmeshki, J., Hoppe, A., Tah, V., Mann, S., Williamson, T. H., & Barman, S. A. (2015). Genetic algorithm based feature selection combined with dual classification for the automated detection of proliferative diabetic retinopathy. Computerized Medical Imaging and Graphics, 43, 64-77. Whelan, T., Atz, U., Holt, T. V., & Clark, C. (2021). Uncovering the Relationship by Aggregating Evidence from 1,000 Plus Studies Published between 2015 – 2020.pdf. Yao, J., Zhang, X., Luo, W., Liu, C., & Ren, L. (2022). Applications of Stacking/Blending ensemble learning approaches for evaluating flash flood susceptibility. International Journal of Applied Earth Observation and Geoinformation, 112. https://doi.org/10.1016/j.jag.2022.102932 zh_TW
