Loading...
|
Please use this identifier to cite or link to this item:
https://nccur.lib.nccu.edu.tw/handle/140.119/141061
|
| Title: | 機器學習演算法產生之投資人觀點結合Black-Litterman資產配置模型-以台灣上市ETF為例
Investor`s Views Derived by Machine Learning Algorithms Combined with Black-Litterman Model-The Case of Taiwan-Listed ETFs |
| Authors: | 李皇毅
Li, Huang-Yi |
| Contributors: | 廖四郎
Liao, Szu-Lang
李皇毅
Li, Huang-Yi |
| Keywords: | 機器學習
隨機森林
XGBoost
Black-Litterman模型
投資組合理論
資產配置
台灣市場ETF
籌碼面資料
Machine Learning
Random Forest
XGBoost
Black-Litterman Model
Portfolio Theory
Asset Allocation
Taiwan market ETFs
Institutional Investors Factor |
| Date: | 2022 |
| Issue Date: | 2022-08-01 17:29:02 (UTC+8) |
| Abstract: | 本研究嘗試使用隨機森林與XGBoost兩種機器學習分類模型於預測資產價格走勢，作為量化投資人觀點之依據，並結合Black-Litterman模型建構投資組合。本研究採用之基礎資產為台灣上市ETF，特徵因子選取價量相關的技術指標與台灣特有的籌碼面資料，來預測資產價格漲跌的方向及幅度，後將預測結果轉換為Black-Litterman模型的投資人觀點進行資產配置，並比較兩種機器學習方法在不同目標函數、不同限制條件與不同風險趨避係數下，所建立相應的投資組合其績效表現之優劣。實證結果顯示：（1）兩種機器學習投資組合在測試期間內，以各績效指標衡量，絕大多數優於本研究之基準投資組合；（2）以XGBoost建構之投資組合，其績效表現皆優於以隨機森林建構之投資組合；（3）以極大化效用函數形成之投資組合，其績效表現皆優於極大化Sharpe Ratio投資組合；（4）風險趨避係數（λ）大致上與報酬呈現反向關係，而與風險指標如波動度與MDD則呈現正向關係。其中，使用XGBoost並以極大化效用函數所得之投資組合，為本研究績效最佳的投資組合。
We attempt to use two machine learning classification models, random forest and XGBoost, to capture the trend of asset prices, as a basis for quantifying investors` views, and combine with the Black-Litterman model to construct portfolios. The underlying assets used in our study are Taiwan-listed ETFs, selected features in machine learnings are price-volume-related technical indicators and Taiwan-unique institutional investors Factor to predict the trends and fluctuations of asset prices, and then convert the predicted results into investor`s views of Black-Litterman model to conduct the asset allocation process. Next, we analyze and compare the performance of the corresponding portfolios established by two machine learning algorithms under different objective functions, different constraints and different risk aversion coefficients. During the test period, We find that：(1) measured by various performance evaluation indicators, portfolios formed by two machine learning algorithms outperform the benchmark portfolios in our study, (2) performance of the portfolios constructed by XGBoost outperform the portfolios constructed by random forest, (3) performance of the portfolios formed by maximizing utility function outperform the maximized Sharpe Ratio portfolios, (4) The risk aversion coefficient（λ）is approximately inversely related to returns, while it is positively related to risk indicators such as volatility and MDD. Lastly, the portfolio generated from XGBoost by maximizing the utility function gains the best performance among all portfolios in our study. |
| Reference: | [1] Basak, S., Kar, S., Saha, S., Khaidem, L., & Dey, S. R. (2019). “Predicting the direction of stock market prices using tree-based classifiers.” The North American Journal of Economics and Finance, 47, 552-567.
[2] Best, M. J., & Grauer, R. R. (1991). “On the sensitivity of mean-variance-efficient portfolios to changes in asset means: some analytical and computational results.” The Review of Financial Studies, 4(2), 315-342.
[3] Black, F., & Litterman, R. B. (1991). “Asset Allocation: Combining Investor Views with Market Equilibrium.” The Journal of Fixed Income, 1(2), 7-18.
[4] Black, F., & Litterman, R. (1992). “Global portfolio optimization.” Financial Analysts Journal, 48(5), 28-43.
[5] Breiman, L. (1996). “Bagging predictors.” Machine Learning, 24(2), 123-140.
[6] Breiman, L. (2001). “Random forests.” Machine Learning, 45(1), 5-32.
[7] Breiman, L., Friedman, J. H., Olshen, R., & Stone, C. J. (1984). Classification and Regression Trees. Wadsworth.
[8] Chen, T., & Guestrin, C. (2016, August). “Xgboost: A scalable tree boosting system.” In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining (pp. 785-794).
[9] Donthireddy, P. (2018). “Black-Litterman portfolios with machine learning derived views.” Research Gate. Retrieved April 22, 2022, from https://www.researchgate.net/publication/326489143_Black-Litterman_Portfolios_with_Machine_Learning_derived_Views.
[10] Friedman, J., Hastie, T., & Tibshirani, R. (2000). “Additive logistic regression: a statistical view of boosting (with discussion and a rejoinder by the authors).” The Annals of Statistics, 28(2),337-407.
[11] Frost, P. A., & Savarino, J. E. (1988). “For better performance: Constrain portfolio weights.” Journal of Portfolio Management, 15(1), 29-34.
[12] Gu, S., Kelly, B., & Xiu, D. (2020). “Empirical asset pricing via machine learning.” The Review of Financial Studies, 33(5), 2223-2273.
[13] He, G., & Litterman, R. (2002). “The intuition behind Black-Litterman model portfolios.” Available at SSRN: https://ssrn.com/abstract=334304.
[14] Idzorek, T. (2007). “A step-by-step guide to the Black-Litterman model: Incorporating user-specified confidence levels.” In Forecasting expected returns in the financial markets (pp. 17-38). Academic Press.
[15] Israel, R., Kelly, B. T., & Moskowitz, T. J. (2020). “Can Machines `Learn` Finance ?.” Journal of Investment Management, 18(2), 23-36.
[16] Ledoit, O., & Wolf, M. (2003). “Improved estimation of the covariance matrix of stock returns with an application to portfolio selection.” Journal of Empirical Finance, 10(5), 603-621.
[17] Liew, J. K. S., & Mayster, B. (2017). “Forecasting etfs with machine learning algorithms.” The Journal of Alternative Investments, 20(3), 58-78.
[18] Lintner, J. (1965). “Security prices, risk, and maximal gains from diversification.” The Journal of Finance, 20(4), 587-615.
[19] Lintner, J. (1965). “The Valuation of Risk Assets and the Selection of Risky Investments in Stock Portfolios and Capital Budgets.” The Review of Economics and Statistics, 47(1), 13–37.
[20] Markowitz, H. (1952). “Portfolio Selection.” The Journal of Finance, 7(1), 77–91.
[21] Markowitz, H. (1959). Portfolio Selection: Efficient Diversification of Investments (Vol. 16). New York: Wiley and Sons.
[22] Meucci, A. (2010). “The Black-Litterman approach: original model and extensions. In R. Cont (Ed.),” The Encyclopedia of Quantitative Finance (pp.196-199). New York, NY: Wiley.
[23] Michaud, R. O. (1989). “The Markowitz optimization enigma: Is ‘optimized’ optimal ?.” Financial Analysts Journal, 45(1), 31-42.
[24] Mossin, J. (1966). “Equilibrium in a Capital Asset Market.” Econometrica, 34(4), 768–783.
[25] Patel, J., Shah, S., Thakkar, P., & Kotecha, K. (2015). “Predicting stock and stock price index movement using trend deterministic data preparation and machine learning techniques.” Expert Systems with Applications, 42(1), 259-268.
[26] Sharpe, W. F. (1964). “Capital asset prices: A theory of market equilibrium under conditions of risk.” The Journal of Finance, 19(3), 425-442.
[27] Sharpe, W. F. (1974). “Imputing expected security returns from portfolio composition.” Journal of Financial and Quantitative Analysis, 9(3), 463-472.
[28] Tang, M. L., Wu, F. Y., & Hung, M. C. (2021). “Multi-asset allocation of exchange traded funds: Application of Black–Litterman model.” Investment Analysts Journal, 50(4), 273-293.
[29] Theil, H. (1971). Principles of Econometrics. New York: Wiley and Sons.
[30] Theil, H. (1978). Introduction to Econometrics. New Jersey: Prentice-Hall, Inc.
[31] Walters J. (2014). “The Black-Litterman model in detail.” Working Paper. Available at SSRN: https://ssrn.com/abstract=1314585.
[32] Zhu, M., Philpotts, D., Sparks, R., & Stevenson, M. J. (2011). “A hybrid approach to combining CART and logistic regression for stock ranking.” The Journal of Portfolio Management, 38(1), 100-109. |
| Description: | 碩士
國立政治大學
金融學系
109352017 |
| Source URI: | http://thesis.lib.nccu.edu.tw/record/#G0109352017 |
| Data Type: | thesis |
| DOI: | 10.6814/NCCU202200809 |
| Appears in Collections: | [金融學系] 學位論文
|
Files in This Item:
| File |
Description |
Size | Format | |
|---|
| 201701.pdf | | 2093Kb | Adobe PDF2 | 0 | View/Open |
|
All items in 政大典藏 are protected by copyright, with all rights reserved.
|
