本研究目的為藉由機器學習搭配投資組合建構，觀察其績效，由於近年來，ETF成為一種十分熱門的投資工具，能夠追蹤特定指數或資產組合，以低成本和高流動性的方式投資市場，標的資產的分散風險配置，也利於長期投資策略的實施，而機器學習在財務領域的應用也越來越廣，透過分析過去的歷史數據分析，利用大量的數據建立模型訓練後再進行各種方向預測，幫助投資人更準確地預測市場走勢和價格變動。在本研究中，以指數型ETF為投資標的，再結合極限梯度提升(XGBoost)的機器學習模型，並用不同的資產配置建立投資組合，包括最大化報酬率、最小化風險和最大化夏普比率等等三種不同組成方式，而實證結果顯示，投資組合的資產配置雖然不同，但績效都能夠有提升，且進場時機愈早，投資組合的績效愈佳，因此對於投資人而言，投資組合結合機器學習也是一種可行的投資方法。

The objective of this study is to construct investment portfolios through the integration of machine learning to observe their performance. In recent years, Exchange-Traded Funds (ETFs) have gained significant popularity as investment tools, offering the ability to track specific indices or asset portfolios with low costs and high liquidity. This approach facilitates risk diversification of underlying assets and supports the implementation of long-term investment strategies. Moreover, the application of machine learning in the field of finance has been increasingly widespread. By analyzing historical data, extensive models are trained using a wealth of data to make various directional predictions, assisting investors in more accurately forecasting market trends and price fluctuations.
In this study, index-based ETFs are chosen as investment targets and combined with the machine learning model of eXtreme Gradient Boosting (XGBoost). Various portfolio compositions are established through different asset allocations, including maximizing return, minimizing risk, and maximizing the Sharpe ratio. Empirical results demonstrate that despite varying asset allocations, portfolio performance can be enhanced. Additionally, entering the market earlier leads to better portfolio performance. Consequently, the integration of machine learning into investment portfolios is deemed a feasible strategy for investors.

Contents
論文審定書i
摘要ii
Abstractiii
List of Figuresvi
List of Tablesvii
Chapter1 Introduction1
1.1 Motivation1
1.2 Purpose3
1.3 Framework5
Chapter2 Literature Review7
2.1 Advantages and Applications of ETFs7
2.2 The Importance of Asset Allocation10
2.3 Machine Learning in Finance11
2.4 XGBoost Introduction and Application in Portfolio12
Chapter3 Research Data and Methodology13
3.1 Research Procedure13
3.2 Research Data13
3.3 Feature Selection27
3.4 Portfolio Construction31
Chapter4 Results41
4.1 The Best Portfolio and Weights Adjustment41
4.2 Model Accuracy and Important Feature44
4.2 Portfolio Performance Analysis46
Chapter5 Conclusions52
References55

Agrrawal, P. (2013). Using Index ETFs for Multi-Asset-Class Investing: Shifting the Efficient Frontier Up. Journal of Index Investing, 4(2), 83-94.
Assoé, K., L’Her, J.-F., & Plante, J.-F. (2006). The Relative Importance of Asset Allocation and Security Selection. The Journal of Portfolio Management, 33(1), 46–55.
Ban, G.-Y., El Karoui, N., & Lim, A. E. B. (2018). Machine Learning and Portfolio Optimization. Management Science, 64(3), 1136–1154.
Charupat, N., & Miu, P. (2013). Recent developments in exchange‐traded fund literature: Pricing efficiency, tracking ability, and effects on underlying securities. Managerial Finance, 39(5), 427–443.
Chaweewanchon, A., & Chaysiri, R. (2022). Markowitz Mean-Variance Portfolio Optimization with Predictive Stock Selection Using Machine Learning, Int. J. Financial Stud. 2022, 10(3), 64.
Chen, W., Zhang, H., Mehlawat, M. K., & Jia, L. (2021). Mean–variance portfolio optimization using machine learning-based stock price prediction. Applied Soft Computing, 100, 106943.
Drobetz, W., & Köhler, F. (2002). The contribution of asset allocation policy to portfolio performance. Financial Markets and Portfolio Management, 16, 219–233.
Elton, E. J., Gruber, M. J., & de Souza, A. (2019). Passive Mutual Funds and ETFs: Performance and Comparison. Journal of Banking & Finance, 106, 265-275.
Huang, M.-Y., & Lin, J.-B. (2011). Do ETFs provide effective international diversification? Research in International Business and Finance, 25(3), 335–344.

Ibbotson, R. G., & Kaplan, P. D. (2000). Does Asset Allocation Policy Explain 40, 90, or 100 Percent of Performance? Financial Analysts Journal, 56(1), 26–33.
KIM, H. (2021). Mean-Variance Portfolio Optimization With Stock Return Prediction Using XGBoost. Economic Computation and Economic Cybernetics Studies and Research, 55(4), 5-20.
Liebi, L. J. (2020). The effect of ETFs on financial markets: a literature review. Financial Markets and Portfolio Management, 34, 165–178.
Li, H., Cao, Y., Li, S., Zhao, J., & Sun, Y. (2020). XGBoost Model and Its Application to Personal Credit Evaluation. IEEE Intelligent Systems, 35(3), 52–61.
Lin, C.-C., Chan, S.-J., & Hsu, H. (2006). Pricing efficiency of exchange traded funds in Taiwan. Journal of Asset Management, 7(1), 60–68.
Ma, Y., Han, R., & Wang, W. (2020). Portfolio optimization with return prediction using deep learning and machine learning. Expert Systems with Applications, 165(1), 113973.
Miffre, J. (2007). Country-specific ETFs: An efficient approach to global asset allocation. Journal of Asset Management, 8(2), 112–122.
Nobre, J., & Neves, R. F. (2019). Combining Principal Component Analysis, Discrete Wavelet Transform and XGBoost to trade in the financial markets. Expert Systems with Applications, 125(1), 181–194.
Paiva, F. D., Cardoso, R. T. N., Hanaoka, G. P., & Duarte, W. M. (2018). Decision-Making for Financial Trading: A Fusion Approach of Machine Learning and Portfolio Selection. Expert Systems with Applications, 115, 635-655.
Young, M. N., Chuahay, T. T. N., & Diaz, J. F. T. (2020). Comparative Analysis of Mean-Variance and Safety-First Portfolio Utilizing Exchange Traded Funds in Asia. IEEE 7th International Conference on Industrial Engineering and Applications (ICIEA).