參考從2001到2005在台灣股市的股息對股票的影響的研究，在根據event-study and SAR method這兩種方式之實證研究可以得到本次的台灣股市裡面除權的效應。 / Referring to our research into the ex-dividend effects on stocks traded on TSE from 2001 to 2005, we come to the following conclusions under the event-study and SAR method. There should be negative ARs on the three days prior to the ex-date and a positive AR on the ex-dividend day, no matter in a bear or bull market. Stocks with cash-only dividends present lower ARs on the three days prior to the ex-date and the ex-dividend day while stock-only dividend ones suffer from higher ARs. The performances of stocks with balanced dividends are just in-between. The most significant ex-dividend effects turn up when it comes to stocks which go ex-dividend in season, with a positive CAR to the seventh day after the ex-date. On the other hand, the effects on the early-ex-dividend stocks exhibit insignificance generally. Later-ex-dividend stocks demonstrate the lowest fluctuation of ARs. However, the simultaneous decline of the index in the ex-dividend season is likely to result in higher-significant ARs. In terms of industry, the ex-dividend effects on electronic companies are more significant than on non-electronic companies. Given a bear market, there used to be negative CARs on electronic companies after shareholders’ meeting; on the contrary, in a bull market, there were positive CARs.