本研究之貢獻在於將在股票市場中已廣為使用之因子模型架構，應用於美國投資級公司債市場，發展出可以解釋公司債市場中橫斷面超額報酬的多因子風險模型，並應用於數量化投資組管理。本研究首先使用橫斷
面迴歸估計因子報酬，並將因子報酬結果搭配動態主動權重配置方法應用於指數增值，後使用長期與短期因子報酬共變異數估計投資組合風險，短期風險估計用以滿足對於短期市場波動有較高準確度需求的主動管理者，長期風險估計則可用於穩定資產配置者的需求。

在模型建構上本研究參考了Barra (2009)所提出的風險模型標準流程，使用Size、Quality、Momentum、Value和Volatility五個風險因子和11個GICS產業因子建立多因子模型。在投資組合風險估計方面，本研究建構了短期與長期估計期之風險模型，並使用Bias Test測試三種投資組合分別是產業投組、因子投組和隨機抽樣投組的長短期風險估計準確度。最後，本研究應用多因子報酬結果來預測債券預期收益並轉化為排名，透過排名分數決定增值指數投資組合主動權重。

實證結果表明在2009/08至2020/03中，多因子模型的平均解釋力為45.79％，並且因子間並沒有產生共線性之現象。在估計風險的能力上，三種測試投資組合之短期和長期風險估計Bias statistic落於95%信賴區間比率平均為83.9%至87.1%和77.4%至81.6%。最後，透過Size、Quality、Momentum、Value和Volatility五個因子動態預估報酬並調整主動權重進行指數增值，表現可以優於標竿指數，全期間資訊比率可達0.71、追蹤誤差為1.2%。

One of the main contributions of this research is to develop, from a factor model that has been widely used in the stock market, a multi-factor risk model that can be used in the US investment-grade corporate bond market to find factors that can explain cross-sectional excess returns in the corporate bond market and to take advantage of the multi-factor model to achieve quantitative portfolio management. In this study, a cross-sectional regression is first used to estimate factor returns to decide active weighting allocations to enhance indexing dynamically. Then, the long-term and short-term factor covariance structures are used to estimate the portfolio risk. The short-term risk estimation can be catered for active managers who need higher accuracy risk estimation to reflect recent volatility. The long-term risk estimation can meet the needs of stable asset allocators.

This study establishes a multi-factor model, using five risk factors of size, quality, momentum, value and volatility, and 11 GICS industry factors based on Barra (2009). This study constructs a risk model for the short-term and long-term estimation periods and uses the bias test to examine the accuracy of risk estimation for three types of portfolios: a sector portfolio, a factor portfolio and a random sampling portfolio. Finally, this study uses the results of the multi-factor returns to predict bond returns and convert them into rankings to determine the active weights of the enhanced index portfolio through the ranking scores.

Our empirical results show that, from August 2009 to March 2020, the average explanatory power of our multi-factor model is 45.79%, and there is no phenomenon of collinearity between the factors. In terms of the risk forecasts, the bias statistics for the short-term and long-term risk estimation of the three test portfolio groups are between 83.9% and 87.1% and between 77.4% and 81.6% respectively, falling below the confidence interval ratio of 95%. Finally, the five factors of size, quality, momentum, value and volatility are used to enhance the benchmark index by adjusting the weights dynamically. The performance is better than the unenhanced benchmark index. The information ratio for the whole period is 0.71, and the tracking error is 1.2%.

論文審定書 i
摘要 ii
ABSTRACT iii
List of Figures v
List of Tables v
I. Introduction 1
1 Background Information 1
2 Our Research Purpose 2
3 Our Research Framework 3
II. Literature Review 4
1 Modern Portfolio Theories 4
2 Multi-Factor Models 5
3 Factor Models in the Corporate Bond Markets 6
4 Enhanced Index Funds 8
III. Data and Methodology 9
1 Analytical Procedures 9
2 Data Description 12
3 Factor Definitions and Measures 13
4 Multi-Factor-Model Construction 17
5 Multi-Factor-Model Application – Enhanced Index Fund 32
IV. Empirical Results 38
1 Multi-Factor Model Results 38
2 Enhanced Index Fund Results 51
3 Sensitivity Analysis 52
V. Conclusion 54

BARRA (2009), Europe Equity Model (EUE3).
Connor, Gregory. (1995). The Three Types of Factor Models: A Comparison of Their Explanatory Power. Financial Analysts Journal, 42-46.
Demir Bektić, Josef-Stefan Wenzler, Michael Wegener, Dirk Schiereck and Timo Spielmann (2019). Extending Fama–French Factors to Corporate Bond Markets. Journal of Portfolio Management, 45 (3) 141-158.
Fama, Eugene F and French, Kenneth R. (1992). The Cross-Section of Expected Stock Returns. Journal of Finance, 47(2), 427-465.
Fama, Eugene F., and James D. MacBeth (1973). Risk, Return, and Equilibrium: Empirical Tests.” Journal of Political Economy 81, no. 3 : 607–636.
Hill, J. M. and H. Naviwala (1999), Synthetic and enhanced index strategies using futures on U.S. indexes, Journal of Portfolio Management, 25 (5) 61-74.
Houweling , Patrick and Van Zundert, Jeroen (2017). Factor Investing in the Corporate Bond Market. Financial Analysts Journal, Vol. 73, No. 2.
Hottinga, Jouke., Leeuwen, Eric van and Ijserloo, Judith van. (2001). Successful Factors to Select Outperforming Corporate Bonds. Journal of Portfolio Management, 28 (1) 88-101.
Jordan Brooks, Diogo Palhares and Scott Richardson (2018). Style investing in fixed income. Journal of Portfolio Management, 44, 127-139.
Jostova, Gergana, Nikolova, Stanislava, Philipov, Alexander and Stahel, Christof W. (2013). Momentum in corporate bond returns. Review of Financial Studies, 26, 1649–1693.
Marielle De Jong and Frank J. Fabozzi (2020). The Market Risk of Corporate Bonds. Journal of Portfolio Management, 46 (2) 92-105.
Peter Mladina and Steven Germani (2019). Bond-Market Risk Factors and Manager Performance. Journal of Portfolio Management, 45 (6) 75-85.
Ronen Israel, Diogo Palhares and Scott Richardson (2018). Common Factors in Corporate Bond Returns. Journal of Investment Management, Vol. 16, No. 2.
Steven S. Crawford, Pietro Perotti, Richard A. Price III and Christopher J. Skousen (2019). Financial Statement Anomalies in the Bond Market. Financial Analysts Journal, 75:3, 105-124.
Shen, Shawn., Pathammavong, Arom., and Chen, Alex (2019). Fixed Income Value Factor. Journal of Fixed Income, 29 (1), 21-43.
Sugitomo, Seisuke and Shotaro, Minami (2018). Fundamental Factor Models Using Machine Learning. Journal of Mathematical Finance, 08(01), 111-118.
Jostova, Gergana, Nikolova, Stanislava, Philipov, Alexander and Stahel, Christof W. (2013). Momentum in corporate bond returns. Review of Financial Studies 26, 1649–1693.