| Abstract |
ABSTRACT
Since Wachtel (1942) pointed out that seasonal anomaly returns existed in stock market, many scholars started to study the phenomena of anomalous returns in stock markets, such as the January effect of Rozeff and Kinney (1976) and Keim and Stambaugh (1984), celebration effect in October of Jack Yang (1992), weekly effect of Cross (1973), French (1980), Richard Rogalski (1984), J. Y. Huang (1985), Y. C. Wang (1997) and C.C. Wang (1999), they all figured out that there were systematic dynamic phenomena existing in stock returns.
However, there were quite a few researches, which would deeply analyze the seasonal anomaly returns during the periods of six days work week, twice a month two days weekend holiday and five days work week, either in the academic or practical community in Taiwan. We hope to use descriptive statistics, OLS, ARCH and GARCH model in combination with Taiwan historical time-series data to clarify whether the seasonal anomaly returns exist in Taiwanˇ¦s stock market in each sub-period.
We respectively used descriptive statistics, OLS, ARCH and GARCH models with daily returns of weighted stocksˇ¦ index of Taiwan stock market from1995 to 2002 to examine the weekly effect on Taiwanˇ¦s stock market.
The results showed that:
1. Total periods (1995-2002): significantly negative effect on Mondays and Tuesdays but positive effect on Saturdays.
2. Six days work week (1995-1997): significantly negative effect on Tuesday and positive effect on Saturday.
3. Twice a month two days weekend holiday (1998-2000): significantly negative effect on Monday, Tuesday effect disappeared.
4. Five days work week (2001-2002): significantly negative effect on Mondays, the highest return on Wednesdays, and weekend effect disappeared. |
|---|
