| Summary: | 碩士 === 臺灣大學 === 國際企業學研究所 === 98 === The paper implements a comparative study of calendar anomalies in international stock markets. The paper has three purposes. First is to test the existence of calendar anomalies. Second is to shift test standard from zero to the average return rate, and to test whether the calendar anomalies exist. Third is to take dividends into consideration and recalculate the total return index. We could apply this index to re-test the calendar anomalies in Taiwan stock market.
The following phenomena are observed:
1.The day of the week effect exists but the return rate varies with stock markets and also decreases over time.
2.Among these various holiday effects, the most common one appears before the holiday of 3 days or more. In addition, there is a positive correlation between pre-holiday return rate and the number of holidays in all stock markets excluding Japan, U.S. and U.K.
3.The turn of the month effect in Japan stock market occurs during the period of trading days -5 to +2. In Malaysia, it occurs during the period of trading days -9 to -7. In the remaining stock markets, it occurs during the period of trading days -1 to +4.
4.The monthly effect exists in all stock markets excluding Korea, Shanghai and U.S.. The January effect exists only in Taiwan and U.K. stock markets.
5.When test standard is changed into the average return rate , the distribution of calendar anomalies varies dramatically.
6.The payment of dividends does not affect the test result of calendar anomalies.
|
|---|
