| 摘要: | 本研究是以Fama and French的三因子模型為基礎，研究在台灣股票市場上：市場因素、規模效應、淨值市價比效應三個變數對台灣股票市場報酬的解釋力，並且再加入一風險指標-風險值，形成四因子模型，進一步探討此四因子對於台灣股市報酬的解釋能力以及影響方向，並且看看加入風險值之後，對於模型的解釋能力是否增加。
實証結果發現在三因子模型中，只有規模因素對於股票報酬有顯著的解釋能力，而市場因素與帳面市價比效應在三因子中並不顯著。而在風險值的實証發現，利用指數加權平均法(EWMA)來估計波動性所算出來的風險值，在回溯測試的檢驗下，發現在樣本期間內的穿透次數，皆在可容許的誤差範圍內，所以用此法所算出來的風險值再現實生活中是有一定的準確性。在四因子模型中，我們發現規模因素、市場因素以及風險值在四因子模型中皆具有顯著的解釋能力，而帳面市價比還是沒有顯著的解釋能力，而在加入風險值之後也發現模型的解釋能力大幅提升，代表風險值因子為在解釋股票報酬的因子裡，為一重要的解釋因子。
This research is based on Fama and French’s three factor models , studying on the stock market in Taiwan: Explanation strength of stock market remuneration in Taiwan correctly of three parameters than tf market factor , scale effect , BM ratio effect , and join one more risk index - value at risk , form four factor models , probe into these four factors to the explanation ability of the remuneration of stock market in Taiwan and influence the direction further, and see that after joining value at risk , the explanation ability to the model increased.
The empirical result shows that in three factor models, only the size factor has significant explanation ability to the stock remuneration, the effect compared with BM ratio factor and market factor are not significant in three factors. The empirical result shows that value at risk making use of method of exponentially weighted averages of the index (EWMA) to estimate the volatility in value at risk, under back testing , find in sample period, all within the range of tolerable error, so there is certain accuracy in actual life again with the value at risk calculated of this law. In four factor model, we find that size factor, market factor and value at risk all have significant explanation ability, the BM ratio does not have significant explanation ability and the explanation ability to find the model after joining risk value improves by a wide margin, represent risk value at risk factor for on explaining factor, stock of remuneration, it is an important explanation factor. |
