| dc.description.abstract | With the evolution of asset pricing models, from single-factor to multi-factor models, there is still no perfect model. Although Taiwan′s stock market is not large, it still has its articularity. Therefore, in order to understand the source of returns in Taiwan’s stock market, this study will use Fama and French′s three-factor and fivefactor model to test. This paper refers to Fama and French (2015) and collect monthly data of listing and delisting companies from the TEJ database in Taiwan from July 1990 to December 2020 (not include financial industry). Then, construct different investment portfolios (4x4, 3x3, 2x4x4 and 2x3x3) and types (Size-bm, Size-OP, Size-Inv, Sizebm-OP, Size-bm-Inv and Size-OP-Inv) on the first trading day in July each year. The test result shows that the five-factor model can improve the anomalous especially for those portfolios with extreme poor returns from the three-factor model. The reason for the improvement is not only the value factor (HML), but also the profitability factor (RMW) and the investment factor (CMA), especially the profitability factor (RMW) which provides significant negative explanation to the poor portfolios. However, the five-factor model makes the better portfolios’ anomalous worse, that is, even if the factors provide more explanatory power, there are still more unexplained intercepts. In addition, the coefficient of the value factor (HML) has decreased in Taiwan’s stock market since the addition of the profitability factor (RMW) and the investment factor (CMA) (the positive value decreases; the negative value expands). Compared with Fama and French (2015), the biggest difference is that the investment factor (CMA) shows that companies invest more and their returns would be higher than others, which means that the investment opportunities in Taiwan may be higher than other countries. | en_US |