Synergy frontier of multi-factor stock selection model
學年 111
學期 1
出版(發表)日期 2022-12-19
作品名稱 Synergy frontier of multi-factor stock selection model
作品名稱(其他語言)
著者 Yeh, I-chen
單位
出版者
著錄名稱、卷期、頁數 OPSEARCH 60, p.445–480
摘要 The classical "efficiency frontier" emphasizes the combination of negatively correlated or low-correlated portfolios to reduce the diversifiable risk of the investment portfolio. While the "synergy frontier" focuses on combining stock selection factors or models with "synergy" to strengthen the ability to increase the return rate of the stock selection model. Therefore, to raise the return, the focus of the multi-factor model is to discover the synergy effects of stock-picking factors. To systematically discover the synergy of stock-picking factors, two profitability factors, ROE and ROC, and two value factors, P/B and P/S were chosen. Then stock picking models that express various styles were systematically generated by means of weighted scoring approach and mixture design. The polynomial regression analysis was employed to build the return and risk models. Then a set of optimal portfolios that offer the highest expected return for a set of various levels of risk can be generated through solving an optimization model. We used the S&P 500 constituent stocks as the stock selection pool. The results showed that (1) There are strong synergy effects of return between the two profitability factors, ROE and ROC, and the value factor, P/B. (2) The relations between factor weights and risk of portfolios are rather linear, which shows that there are no synergy effects of risk between profitability factors and value factors. (3) There is synergy rotation in stock market, and the momentum strategy can overcome the rotation phenomenon and significantly improve the investment performance.
關鍵字 Synergy;Stock selection;Multi-factor model;Weighted scoring;Mixture design;Optimization
語言 zh_TW
ISSN 0975-0320
期刊性質 國內
收錄於 ESCI
產學合作
通訊作者
審稿制度
國別 TWN
公開徵稿
出版型式 ,電子版
相關連結

機構典藏連結 ( http://tkuir.lib.tku.edu.tw:8080/dspace/handle/987654321/124301 )