我們利用三種不同的模型來驗證Fama和French在1993年所提出的三因子模型。在驗證的過程中，Fama及French所採用的方式是1973年由Fama以及MacBeth所提出的兩階段的方式，我們藉由代入不同的模型使得因子的顯著性不同。 此外，在驗證過程中，對資料分組或是端視其個股的表現對於結果也會有所差異。其中以分組方式可獲得最穩定的估計但相對也使用了較少的資訊；反之，在兩個階段都使用個股資料可以利用到較多的訊息，估計值卻呈現相對不穩定的狀態。最後，我們納入兩種不同估計貝他的方式以檢驗其對結果產生的影響。 This thesis uses three different approaches to examine the three-factor model proposed by Fama and French(1993).They use traditional two-pass procedure to solve the estimation problem, and we modify the method by taking different models. Whether we do grouping or not is also an important decision in dealing with our data, we find that grouping in both the two-stage get the stable estimates but lose much information. However, when we consider individuals in both step,and we include most information but the estimates vary a lot. We also consider different kinds of estimation of betas, and the appearance of estimates differ a lot. We can get different significant factors in the three models.
