學術產出-Theses
Article View/Open
Publication Export
-
題名 建構低波動量化交易模型的金融資產配置優化策略
Building a Quantitative Trading Model with Low Volatility to Achieve Optimal Financial Asset Allocation Strategy作者 陳韻清
Chen, Yun-Ching貢獻者 胡毓忠
Hu,Yuh-Jong
陳韻清
Yun-Ching Chen關鍵詞 深度強化學習
金融資產配置
低波動投資
Deep Reinforcement Learning
Asset Allocation
Low-volatility investing日期 2023 上傳時間 1-Sep-2023 15:40:11 (UTC+8) 摘要 近年來，受到「地緣政治」、「貨幣緊縮」、「高通膨」、「景氣衰退」等因素影響，金融市場更加趨於波動與不穩定。為了協助一般資產客群在如此波動環境下可以穩定且安心投資，本論文提出了一種建立在低波動策略基礎上的投資組合配置模型。該模型融合了深度強化學習技術與量化交易策略，並透過獎勵函數的設計，讓模型能夠根據投資資產與市場價格近三個月波動程度的相對比例來調整獎勵。在此設計下，相對於市場波動度高的資產獲得的報酬會減少，相對於市場波 動度低的資產獲得的報酬則會增加。因此，會驅使模型分配更多的權重給低波動的資產，從而建立起低波動量化交易模型的金融資產配置優化策略。我們使用美國市場交易所交易基金(Exchange Traded Funds)與總體經濟之歷史數據進行訓練與回測。實驗結果顯示，本論文建構之低波動量化交易模型，可於波動市場環境 下，適度降低資產波動率，穩定投資組合，避免投資者因情緒波動而做出不理性的決策，並且可以獲得適度之報酬。
Recently, the financial market has experienced heightened volatility, exacerbated by geopolitical tensions, monetary tightening, surging inflation rates, and a prolonged reces- sion. This thesis proposes a portfolio allocation model based on a low-volatility strategy to meet mass affluents’ needs to invest stably and safely in such a volatile environment. This model integrates deep reinforcement learning (DRL) and quantitative trading strate- gies. Through the design of the DRL’s reward function, the model can adjust the reward according to the relativity of investment assets and market price fluctuations for the past three months. Under this design, assets with relatively higher market volatility will re- ceive less rewards, while those with lower volatility will receive more. Hence, the model will be driven to assign higher weightage to low-volatility assets and establish a financial asset allocation optimization strategy for a low-volatility quantitative trading model. In this thesis, we use historical data of Exchange Traded Funds in the US market and the overall economy for training and backtesting. The results of the experiments show that the low-volatility quantitative trading model constructed by this thesis can, in a volatile market environment, appropriately reduce asset volatility, stabilize investment portfolios, prevent investors from making irrational decisions due to emotional fluctuations, and ob-tain moderate returns.參考文獻 [1] A. Ang, R. J. Hodrick, Y. Xing, and X. Zhang. High idiosyncratic volatility and low returns: International and further us evidence. Journal of Financial Economics, 91(1):1–23, 2009.[2] K. Anthony. The low volatility anomaly. https://www.etfcentral.com/news/ the-low-volatility-anomaly, 2023.[3] M. Armstrong. Rise of the robo-advisors. https://www.statista.com/chart/ 30114/robo-abdvisor-assets-under-management-and-revenue/, 2023.[4] N. L. Baker and R. A. Haugen. Low risk stocks outperform within all observable markets of the world. Available at SSRN 2055431, 2012.[5] E. Benhamou, D. Saltiel, S. Ungari, and A. Mukhopadhyay. Bridging the gap between markowitz planning and deep reinforcement learning. arXiv preprint arXiv:2010.09108, 2020.[6] D. Blitz and P. Van Vliet. The volatility effect: Lower risk without lower return. Journal of portfolio management, pages 102–113, 2007.[7] P. Bouchey, V. Nemtchinov, A. Paulsen, and D. M. Stein. Volatility harvesting: Why does diversifying and rebalancing create portfolio growth? The Journal of Wealth Management, 15(2):26–35, 2012.[8] P. Bouchey, V. Nemtchinov, and T.-K. L. Wong. Volatility harvesting in theory and practice. The Journal of Wealth Management, 18(3):89–100, 2015.[9] J. Y. Campbell, L. M. Viceira, L. M. Viceira, et al. Strategic asset allocation: port- folio choice for long-term investors. Clarendon Lectures in Economic, 2002.[10] E. P. Chan. Quantitative trading: how to build your own algorithmic trading busi- ness. John Wiley & Sons, 2021.[11] V. DeMiguel, L. Garlappi, and R. Uppal. Optimal versus naive diversification: How inefficient is the 1/n portfolio strategy? The review of Financial studies, 22(5):1915– 1953, 2009.[12] R. Durall. Asset allocation: From markowitz to deep reinforcement learning. arXiv preprint arXiv:2208.07158, 2022.[13] T. Dutt and M. Humphery-Jenner. Stock return volatility, operating performance and stock returns: International evidence on drivers of the"low volatility`anomaly. Journal of Banking & Finance, 37(3):999–1017, 2013.[14] E.F.FamaandK.R.French.Thecross-sectionofexpectedstockreturns.theJournal of Finance, 47(2):427–465, 1992.[15] T. G. Fischer. Reinforcement learning in financial markets-a survey. Technical re- port, FAU Discussion Papers in Economics, 2018.[16] Geoffrey Kelley, CFA. Risk-averse investors vs. low-volatility strategies.https://www.manulifeim.com/retail/ca/en/viewpoints/equity/ risk-averse-investors-vs-low-volatility-strategies.[17] T. Harnpadungkij, W. Chaisangmongkon, and P. Phunchongharn. Risk-sensitive portfolio management by using distributional reinforcement learning. In 2019 IEEE 10th International Conference on Awareness Science and Technology (iCAST), pages 1–6. IEEE, 2019.[18] R. A. Haugen and A. J. Heins. On the evidence supporting the existence of risk premiums in the capital market. Available at SSRN 1783797, 1972.[19] Y.-J. Hu and S.-J. Lin. Deep reinforcement learning for optimizing finance portfo- lio management. In 2019 Amity International Conference on Artificial Intelligence (AICAI), pages 14–20. IEEE, 2019.[20] J. Kim, M.-J. Kang, K. Lee, H. Moon, and B.-K. Jeon. Deep reinforcement learning for asset allocation: Reward clipping. arXiv preprint arXiv:2301.05300, 2023.[21] X. Li, R. N. Sullivan, and L. Garcia-Feijóo. The low-volatility anomaly: Market evidence on systematic risk vs. mispricing. Financial Analysts Journal, 72(1):36– 47, 2016.[22] Z. Liang, H. Chen, J. Zhu, K. Jiang, and Y. Li. Adversarial deep reinforcement learning in portfolio management. arXiv preprint arXiv:1808.09940, 2018.[23] T. P. Lillicrap, J. J. Hunt, A. Pritzel, N. Heess, T. Erez, Y. Tassa, D. Silver, and D. Wierstra. Continuous control with deep reinforcement learning. arXiv preprint arXiv:1509.02971, 2015.[24] L. Magwa and L. Bargjo. Low volatility defies the basic finance principlesof risk and reward. https://www.robeco.com/en-int/insights/2021/11/ low-volatility-defies-the-basic-finance-principles-of-risk-and-reward, 2021.[25] H. M. Markowitz. Portfolio selection. The Journal of Finance, 1952.[26] V. Mnih, K. Kavukcuoglu, D. Silver, A. A. Rusu, J. Veness, M. G. Bellemare, A. Graves, M. Riedmiller, A. K. Fidjeland, G. Ostrovski, et al. Human-level control through deep reinforcement learning. nature, 518(7540):529–533, 2015.[27] J. Moody and L. Wu. Optimization of trading systems and portfolios. In Proceed- ings of the IEEE/IAFE 1997 computational intelligence for financial engineering (CIFEr), pages 300–307. IEEE, 1997.[28] J. Moody, L. Wu, Y. Liao, and M. Saffell. Performance functions and reinforcement learning for trading systems and portfolios. Journal of Forecasting, 17(5-6):441– 470, 1998.[29] K. I. of Private Enterprise. Why are all investments down in 2022? it’s all about real rates. https://kenaninstitute.unc.edu/commentary/ why-are-all-investments-down-in-2022-its-all-about-real-rates/ #_ftn1, 2022.[30] W. F. Sharpe. Capital asset prices: A theory of market equilibrium under conditions of risk. The journal of finance, 19(3):425–442, 1964.[31] D. Silver, G. Lever, N. Heess, T. Degris, D. Wierstra, and M. Riedmiller. Determin- istic policy gradient algorithms. In International conference on machine learning, pages 387–395. Pmlr, 2014.[32] F. Soleymani and E. Paquet. Financial portfolio optimization with online deep rein- forcement learning and restricted stacked autoencoder—deepbreath. Expert Systems with Applications, 156:113456, 2020.[33] R. H. Tütüncü and M. Koenig. Robust asset allocation. Annals of Operations Re- search, 132:157–187, 2004.[34] G. Zdzienicki and R. Diamant. The low volatility effect—pursuing a smoother investment experience. https://www.cibc.com/content/dam/ cam-public-assets/documents/cibcam-low-volatility-en.pdf, 2021.[35] Z. Zhang, S. Zohren, and R. Stephen. Deep reinforcement learning for trading. The Journal of Financial Data Science, 2020. 描述 碩士
國立政治大學
資訊科學系碩士在職專班
110971027資料來源 http://thesis.lib.nccu.edu.tw/record/#G0110971027 資料類型 thesis dc.contributor.advisor 胡毓忠 zh_TW dc.contributor.advisor Hu,Yuh-Jong en_US dc.contributor.author (Authors) 陳韻清 zh_TW dc.contributor.author (Authors) Yun-Ching Chen en_US dc.creator (作者) 陳韻清 zh_TW dc.creator (作者) Chen, Yun-Ching en_US dc.date (日期) 2023 en_US dc.date.accessioned 1-Sep-2023 15:40:11 (UTC+8) - dc.date.available 1-Sep-2023 15:40:11 (UTC+8) - dc.date.issued (上傳時間) 1-Sep-2023 15:40:11 (UTC+8) - dc.identifier (Other Identifiers) G0110971027 en_US dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/147098 - dc.description (描述) 碩士 zh_TW dc.description (描述) 國立政治大學 zh_TW dc.description (描述) 資訊科學系碩士在職專班 zh_TW dc.description (描述) 110971027 zh_TW dc.description.abstract (摘要) 近年來，受到「地緣政治」、「貨幣緊縮」、「高通膨」、「景氣衰退」等因素影響，金融市場更加趨於波動與不穩定。為了協助一般資產客群在如此波動環境下可以穩定且安心投資，本論文提出了一種建立在低波動策略基礎上的投資組合配置模型。該模型融合了深度強化學習技術與量化交易策略，並透過獎勵函數的設計，讓模型能夠根據投資資產與市場價格近三個月波動程度的相對比例來調整獎勵。在此設計下，相對於市場波動度高的資產獲得的報酬會減少，相對於市場波 動度低的資產獲得的報酬則會增加。因此，會驅使模型分配更多的權重給低波動的資產，從而建立起低波動量化交易模型的金融資產配置優化策略。我們使用美國市場交易所交易基金(Exchange Traded Funds)與總體經濟之歷史數據進行訓練與回測。實驗結果顯示，本論文建構之低波動量化交易模型，可於波動市場環境 下，適度降低資產波動率，穩定投資組合，避免投資者因情緒波動而做出不理性的決策，並且可以獲得適度之報酬。 zh_TW dc.description.abstract (摘要) Recently, the financial market has experienced heightened volatility, exacerbated by geopolitical tensions, monetary tightening, surging inflation rates, and a prolonged reces- sion. This thesis proposes a portfolio allocation model based on a low-volatility strategy to meet mass affluents’ needs to invest stably and safely in such a volatile environment. This model integrates deep reinforcement learning (DRL) and quantitative trading strate- gies. Through the design of the DRL’s reward function, the model can adjust the reward according to the relativity of investment assets and market price fluctuations for the past three months. Under this design, assets with relatively higher market volatility will re- ceive less rewards, while those with lower volatility will receive more. Hence, the model will be driven to assign higher weightage to low-volatility assets and establish a financial asset allocation optimization strategy for a low-volatility quantitative trading model. In this thesis, we use historical data of Exchange Traded Funds in the US market and the overall economy for training and backtesting. The results of the experiments show that the low-volatility quantitative trading model constructed by this thesis can, in a volatile market environment, appropriately reduce asset volatility, stabilize investment portfolios, prevent investors from making irrational decisions due to emotional fluctuations, and ob-tain moderate returns. en_US dc.description.tableofcontents 第一章 前言 11.1 研究動機 11.2 研究目的 41.3 研究架構 5第二章 文獻探討 62.1 資產配置 62.2 低波動投資 72.2.1 低波動異常現象 72.2.2 波動率收益策略 82.3 深度強化學習102.4 相關研究 132.4.1 低波動投資 132.4.2 深度強化學習 15第三章 研究方法 183.1 研究命題 183.2 實驗流程 213.3 實驗設計 27第四章 研究實作 364.1 資料蒐集 364.2 模型訓練 404.3 模型測試 454.4 成果衡量 50第五章 研究結論與未來展望 545.1 研究結論 545.2 未來展望 56參考文獻 57 zh_TW dc.format.extent 5243115 bytes - dc.format.mimetype application/pdf - dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0110971027 en_US dc.subject (關鍵詞) 深度強化學習 zh_TW dc.subject (關鍵詞) 金融資產配置 zh_TW dc.subject (關鍵詞) 低波動投資 zh_TW dc.subject (關鍵詞) Deep Reinforcement Learning en_US dc.subject (關鍵詞) Asset Allocation en_US dc.subject (關鍵詞) Low-volatility investing en_US dc.title (題名) 建構低波動量化交易模型的金融資產配置優化策略 zh_TW dc.title (題名) Building a Quantitative Trading Model with Low Volatility to Achieve Optimal Financial Asset Allocation Strategy en_US dc.type (資料類型) thesis en_US dc.relation.reference (參考文獻) [1] A. Ang, R. J. Hodrick, Y. Xing, and X. Zhang. High idiosyncratic volatility and low returns: International and further us evidence. Journal of Financial Economics, 91(1):1–23, 2009.[2] K. Anthony. The low volatility anomaly. https://www.etfcentral.com/news/ the-low-volatility-anomaly, 2023.[3] M. Armstrong. Rise of the robo-advisors. https://www.statista.com/chart/ 30114/robo-abdvisor-assets-under-management-and-revenue/, 2023.[4] N. L. Baker and R. A. Haugen. Low risk stocks outperform within all observable markets of the world. Available at SSRN 2055431, 2012.[5] E. Benhamou, D. Saltiel, S. Ungari, and A. Mukhopadhyay. Bridging the gap between markowitz planning and deep reinforcement learning. arXiv preprint arXiv:2010.09108, 2020.[6] D. Blitz and P. Van Vliet. The volatility effect: Lower risk without lower return. Journal of portfolio management, pages 102–113, 2007.[7] P. Bouchey, V. Nemtchinov, A. Paulsen, and D. M. Stein. Volatility harvesting: Why does diversifying and rebalancing create portfolio growth? The Journal of Wealth Management, 15(2):26–35, 2012.[8] P. Bouchey, V. Nemtchinov, and T.-K. L. Wong. Volatility harvesting in theory and practice. The Journal of Wealth Management, 18(3):89–100, 2015.[9] J. Y. Campbell, L. M. Viceira, L. M. Viceira, et al. Strategic asset allocation: port- folio choice for long-term investors. Clarendon Lectures in Economic, 2002.[10] E. P. Chan. Quantitative trading: how to build your own algorithmic trading busi- ness. John Wiley & Sons, 2021.[11] V. DeMiguel, L. Garlappi, and R. Uppal. Optimal versus naive diversification: How inefficient is the 1/n portfolio strategy? The review of Financial studies, 22(5):1915– 1953, 2009.[12] R. Durall. Asset allocation: From markowitz to deep reinforcement learning. arXiv preprint arXiv:2208.07158, 2022.[13] T. Dutt and M. Humphery-Jenner. Stock return volatility, operating performance and stock returns: International evidence on drivers of the"low volatility`anomaly. Journal of Banking & Finance, 37(3):999–1017, 2013.[14] E.F.FamaandK.R.French.Thecross-sectionofexpectedstockreturns.theJournal of Finance, 47(2):427–465, 1992.[15] T. G. Fischer. Reinforcement learning in financial markets-a survey. Technical re- port, FAU Discussion Papers in Economics, 2018.[16] Geoffrey Kelley, CFA. Risk-averse investors vs. low-volatility strategies.https://www.manulifeim.com/retail/ca/en/viewpoints/equity/ risk-averse-investors-vs-low-volatility-strategies.[17] T. Harnpadungkij, W. Chaisangmongkon, and P. Phunchongharn. Risk-sensitive portfolio management by using distributional reinforcement learning. In 2019 IEEE 10th International Conference on Awareness Science and Technology (iCAST), pages 1–6. IEEE, 2019.[18] R. A. Haugen and A. J. Heins. On the evidence supporting the existence of risk premiums in the capital market. Available at SSRN 1783797, 1972.[19] Y.-J. Hu and S.-J. Lin. Deep reinforcement learning for optimizing finance portfo- lio management. In 2019 Amity International Conference on Artificial Intelligence (AICAI), pages 14–20. IEEE, 2019.[20] J. Kim, M.-J. Kang, K. Lee, H. Moon, and B.-K. Jeon. Deep reinforcement learning for asset allocation: Reward clipping. arXiv preprint arXiv:2301.05300, 2023.[21] X. Li, R. N. Sullivan, and L. Garcia-Feijóo. The low-volatility anomaly: Market evidence on systematic risk vs. mispricing. Financial Analysts Journal, 72(1):36– 47, 2016.[22] Z. Liang, H. Chen, J. Zhu, K. Jiang, and Y. Li. Adversarial deep reinforcement learning in portfolio management. arXiv preprint arXiv:1808.09940, 2018.[23] T. P. Lillicrap, J. J. Hunt, A. Pritzel, N. Heess, T. Erez, Y. Tassa, D. Silver, and D. Wierstra. Continuous control with deep reinforcement learning. arXiv preprint arXiv:1509.02971, 2015.[24] L. Magwa and L. Bargjo. Low volatility defies the basic finance principlesof risk and reward. https://www.robeco.com/en-int/insights/2021/11/ low-volatility-defies-the-basic-finance-principles-of-risk-and-reward, 2021.[25] H. M. Markowitz. Portfolio selection. The Journal of Finance, 1952.[26] V. Mnih, K. Kavukcuoglu, D. Silver, A. A. Rusu, J. Veness, M. G. Bellemare, A. Graves, M. Riedmiller, A. K. Fidjeland, G. Ostrovski, et al. Human-level control through deep reinforcement learning. nature, 518(7540):529–533, 2015.[27] J. Moody and L. Wu. Optimization of trading systems and portfolios. In Proceed- ings of the IEEE/IAFE 1997 computational intelligence for financial engineering (CIFEr), pages 300–307. IEEE, 1997.[28] J. Moody, L. Wu, Y. Liao, and M. Saffell. Performance functions and reinforcement learning for trading systems and portfolios. Journal of Forecasting, 17(5-6):441– 470, 1998.[29] K. I. of Private Enterprise. Why are all investments down in 2022? it’s all about real rates. https://kenaninstitute.unc.edu/commentary/ why-are-all-investments-down-in-2022-its-all-about-real-rates/ #_ftn1, 2022.[30] W. F. Sharpe. Capital asset prices: A theory of market equilibrium under conditions of risk. The journal of finance, 19(3):425–442, 1964.[31] D. Silver, G. Lever, N. Heess, T. Degris, D. Wierstra, and M. Riedmiller. Determin- istic policy gradient algorithms. In International conference on machine learning, pages 387–395. Pmlr, 2014.[32] F. Soleymani and E. Paquet. Financial portfolio optimization with online deep rein- forcement learning and restricted stacked autoencoder—deepbreath. Expert Systems with Applications, 156:113456, 2020.[33] R. H. Tütüncü and M. Koenig. Robust asset allocation. Annals of Operations Re- search, 132:157–187, 2004.[34] G. Zdzienicki and R. Diamant. The low volatility effect—pursuing a smoother investment experience. https://www.cibc.com/content/dam/ cam-public-assets/documents/cibcam-low-volatility-en.pdf, 2021.[35] Z. Zhang, S. Zohren, and R. Stephen. Deep reinforcement learning for trading. The Journal of Financial Data Science, 2020. zh_TW
