本研究主要將Fama-French(1993)所提出的三因子(市場因子、規模因子、淨值市價比因子),延伸為Carhart(1997)所提出四因子模型(four-factor model)中亦即加入動能因子,並將其改寫為緃橫平滑轉換迴歸(panel smooth transition regression, PSTR)的架構,再以研發費用(R&D)為轉換變數,評估股價報酬的非線性路徑,並檢視隨時間與公司變動的四種風險溢酬。 實證的樣本期間為2010年1月至2014年12月,研究對象為41家台灣上市(櫃)生醫科技公司。 實證結果顯示: 1.股價報酬呈現非線性的路徑,決定於轉換變數各公司各期的:研發費用。 2.傳統的Fama-French三因子模型及Carhart四因子模型係採線性模型進行估計,由估計結果得知,市場因子(Rm-Rf)、規模因子(SMB)、動能因子(MRL)對股價報酬呈顯著的正向影響,而價值因子(HML)對股價報酬則呈現不顯著負向的影響。由此可知,對生醫科技公司而言,成長型公司較價值型公司具有較高的超額報酬;小規模公司較大規模公司具有超額的報酬;高淨值市價比公司比低淨值市價比公司具有超額的報酬。 3.在PSTR模型下,四個風險溢酬均是隨個別公司在不同時期的研發費用比率而變動,而非如傳統三因子或四因子模型所主張的「固定的風險溢酬」。其中價值風險溢酬是不顯著的,其餘三者(包括:市場風險溢酬、規模風險溢酬及動能風險溢酬)均是正向且顯著的。
This thesis rewrites the four-factor model, developed by Carhart(1997), as a panel smooth transition regression(PSTR) framework, and uses R&D expenditures as the transition variable in the framework, to evaluate the nonlinear path of stock returns and four time-varying risk premiums. Sample period spans from 2010:M1 to 2014:M12. Sample objects are 41 biomedical technology companies listed on Taiwan Security Exchange Corporation. The empirical results can be summarized as follows: 1.The stock returns of biomedical technology companies display a nonlinear path, depending on the changes in the R&D of individual companies in each period. 2.Traditional Fama-French three-factor model and Carhart four-factor model are estimated by linear approaches. The results show that the market premium size premium, and momentum premium are significantly positive; however, the value premium is significantly negative. Thus, biomedical technology, holding growth stocks can lead to higher returns than the value; ones in addition, holding small size stocks can lead to excess returns than big size stocks. Holding high book to market stocks can lead to excess returns than low book to market. 3.In the PSTR model, four risk premiums vary with time and across stocks, depending R&D locates in different regimes, not constant obtained from the traditional three-factor model and four-factor model. However, Value risk premium is insignificant. The remaining three risk premiums are significantly positive.