本研究為跟隨Fama and French , “The Cross-Section of Expected Stock Returns”以及Turan and Nusret Cakici , “Value at risk and expected stock return”,過去研究探討三因子模型,本研究加入另一新興的風險因子-涉險值(VaR),試著探討此四風險因子對股票報酬率的關聯性,本研究方法以Panel data模型進行分析研究,將檢定結果加以比較三因子模型與四因子模型是否有何差異性,在三因子模型中,可以看到所有的風險因子皆顯著,且具有解釋能力;在四因子模型中,即使加入涉險值,所有解釋變數仍是具有解釋能力,並且四因子模型中的判定係數較三因子模型高,因此,可得知四因子模型的配適能力較三因子模型為佳。與被解釋變數的關聯性,得知貝他係數、淨價市值比對股票報酬率呈現負向關係,與過去研究有不同的論調,然而,這可能是因為本研究所選取的樣本數,本質的不同,以及所選取的時間長度也有所差異,亦或是使用的研究方法也不同,因此,會產生不同的影響。在規模效果上,發現與股票報酬率呈現正向關係,雖然存在規模效應,卻與過去學者所做之研究結果,恰為截然不同方向,本文研判其原因應與貝他係數相同。在另一解釋變數-涉險值,發現與報酬率呈現正相關,因此,若提列的涉險值越高,則股票報酬率將會越高,與Turan(2004)研究發現有著相同結果。 In this study , we follow Fama and French (1992), and Turan and Nusret Cakici (2004) . In previous studies, three-factor model has been extensive tested. We take another common used risk factor-Value at Risk(VaR)into our model. We want to know the relationship among these four risk factors and the stock return. We will analyze our empirical data with Panel data model. We compare the results of “three-factor model” with “four-factor model” to see the difference. In summary, we can see all factors are significant in explaining stock returns in three-factor model. In the four-factor model, even we take VaR into our model; all risk factors are still significant in explaining stock returns. And the in four-factor model is higher than in three-factor model. In this study, we find there is negative relation between Beta(or B/M)and stock return. Our results are quite different from previous works. Besides, we also find there is positive relation between Market value and stock return. Although size effect exits, our testing result has different way from previous work. Other risk factor-VaR has also positive relations with stock return. Our result is the same with the results from Turan etc.(2004). The reasons why our results are different from previous works could be selected samples bias, or selected time period, or selected model, or firm specified features.