1950 年哈利·馬克思·馬可維茲推出了現代投資組合理論，並在 1990 年獲得了諾貝爾經濟學獎，但是此模型還是存在許多問題。其 中一個重要的問題就是參數估計的問題，此模型如果要規劃一定數量 的資產時，就必須估計很大量規模的參數，這樣的估計誤差可能在投 資組合最佳化的過程中造成最優化的權重解可能不是最佳解。1990 年費雪·布萊克和羅伯特·李特曼為了解決這個參數估計的問題，在高 盛推出了 Black-Litterman model，這個模型利用貝氏估計量的理 論，讓投資者能用專業的知識去優化這些估計參數。而在本篇論文 中，我們將透過文字探勘量化穆迪的信評報告作為投資者觀點。信評 機構如標準普爾 (Standard & Poor's Global Rating)，穆迪 (Moody's Investors Service) 和惠譽 (Fitch Group) 一直被認為是保護投資人的 一道重要大門。許多公司再被信評機構升級為投資等級之前是沒辦法 被一些養老基金所投資的。這可以說明市場普遍對於信評機構的公司 調查能力是有一定認可度的。而在本篇論文中我們將使用穆迪的信評 報告，將其信評機構的文字量化成對於資產有解釋力的系統因子，並 利用其系統因子建構出更客觀且專業的投資人觀點，最後成功利用其 觀點透過 Black-Litterman model 建構更優良的投資組合，和有統計 上顯著為正 Alpha 的指數型基金交易策略。

Harry Markowitz introduced modern portfolio theory in 1950 and won the Nobel Prize in Economics in 1990, but there are still many problems with this model. One of the important problems is the problem of parameter estimation. If this model wants to plan a certain number of assets, it must estimate large amounts of parameters. This might cause tremendous estimation errors, and cause problems through the process of optimization. In 1990, Fisher Black and Robert Littman introduced the Black-Litterman model at Goldman Sachs to solve the problem of parameter estimation. This model uses the theory of Bayesian estimator to allow investors to optimize these parameters with professional view. In this paper, we will quantify Moody's credit rating reports as investor views through text mining. Credit rating agencies such as Standard & Poor's Global Rating, Moody's Investors Service, and Fitch Group have long been considered an important gateway to investor protection. Many companies' bonds can not be invested by pension funds until they are upgraded to investment grade by credit rating agencies. This shows that the market generally has a certain degree of recognition for the capabilities of credit rating agencies. In this paper, we will quantify Moody's credit rating report into systematic factors that have explanatory power for assets' return, and use it as a systematic factor to construct a more objective and professional investor view. Finally, the author successfully used rating agency views to construct a better investment portfolio through the Black-Litterman model and an index fund trading strategy with a statistically significant positive alpha.

1. 前言.......................1
2. 文獻回顧....................2
3. 假說建立與研究方法............5
4. 實證結果...................26
5. 結論.......................32
附錄..........................35
參考文獻.......................39

1. Agarwal, S., Chen, V. Y., & Zhang, W. (2016). “The information value of credit rating action reports: A textual analysis,” Management Science, 62, 2218-2240.
2. Black, F., & Litterman, R. (1990). “Asset allocation: combining investor views with market equilibrium,” Goldman Sachs Fixed Income Research, 115.
3. Bolsen, T., & Shapiro, M. A. (2018). “The US news media, polarization on climate change, and pathways to effective communication,” Environmental Communication, 12, 149-163.
4. Calomris , C.W. , & Mamaysky ,H (2018) . “How news and its context drive risk and returns around the world,” Journal of Financial Economics, 133, 299-366
5. Engle, R. F., Giglio, S., Kelly, B., Lee, H., & Stroebel, J. (2020). “Hedging climate change news,” The Review of Financial Studies, 33, 1184-1216.
6. Fodor, I. K. (2002).“A survey of dimension reduction techniques,” Lawrence Livermore National Lab., CA (US).
7. 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. Academic Press.
8. Jia-long, L., Bo-wei, L., & Min, L. (2013, July). “Model contest and portfolio performance: Black-Litterman versus factor models,” In 2013 International Conference on Management Science and Engineering 20th Annual Conference Proceedings. IEEE.
9. Kara, M., Ulucan, A., & Atici, K. B. (2019). “A hybrid approach for generating investor views in Black–Litterman model,” Expert Systems with Applications, 128, 256-270.
10. Lai, T. L., Xing, H., & Chen, Z. (2011). “Mean–variance portfolio optimization when means and covariances are unknown,” The Annals of Applied Statistics, 5, 798-823.
11. Ledoit, O., & Wolf, M. (2002). “Some hypothesis tests for the covariance matrix when the dimension is large compared to the sample size,” The Annals of Statistics, 30, 1081-1102.
12. LEE, B. (2000). “A Practitioner's Guide to Active Portfolio Management Using Implied View,” The Journal of Risk Finance.
13. Loughran, T. and B. Mcdonald (2011), “When is a liability not a
liability? Textual analysis, dictionaries, and 10‐Ks,” Journal of
Finance, 66, 35-65.
14. Löffler, Gunter, Norden, L., & Rieber, A. (2021). “Negative news and
the stock market impact of tone in rating reports,” Journal of Banking
& Finance, 133, 106256.
15. Markowitz, H. M. (1991). “Foundations of portfolio theory,” The
journal of finance, 46, 469-477.
16. Michaud, R. O. (1989). “The Markowitz optimization enigma: Is
‘optimized’optimal ?” Financial analysts journal, 45, 31-42.
17. Meucci, A. (2010), “The Black-Litterman Approach: Original Model and Extensions,” Shorter version in The Encyclopedia of Quantitative
Finance.
18. Min, L., Dong, J., Liu, D., & Kong, X. (2021). “A Black-Litterman
portfolio selection model with investor opinions generating from machine learning algorithms,” Engineering Letters, 29, 710-721.
19. Partnoy, F. (2006). How and why credit rating agencies are not like other gatekeepers.
20. Rhee, R. J. (2015). “Why credit rating agencies exist,” Economic Notes: Review of Banking, Finance and Monetary Economics, 44, 161-176.
21. Satchell, S., & Scowcroft, A. (2007). “A demystification of the Black-Litterman model: Managing quantitative and traditional portfolio construction,” In Forecasting Expected Returns in the Financial Markets. Academic Press.