Although numerous researchers investigated weekday effects, their methodologies and results do not appear entirely satisfactory. This research analyzes weekday effects of Taiwan Stock market on bull- and bear-type markets and gives one possible explanation of weekday effects “wandering”. Using Turning Point Detection Method and Naive Moving Average Method, I determine bull- and bear-type markets and use ARMA-GARCH regression analysis to discover weekday effects on the mean and volatility equations. The empirical results support the existence of strong bull Monday effect and its mean return is significant positive and the highest among all weekday effects. Moreover, the result does not depend on the time periods and methodologies used to determine market states. Negative and the lowest average return is found in strong bear Monday that is not depending on the time periods and methodologies, though it is significant only by using Turning Point Detection methodology. Other weekday effects are observed, but they are not consistent. The present study shows market states are useful factors of studying weekday effects and demonstrates the fragility of previous studies.
