Test for calendar anomalies in six emerging Asian markets – results from the GARCH model

This paper focuses on the identification of calendar anomalies in six emerging Asian countries. The countries under investigation are China, India, Indonesia, Malaysia, Philippines and Taiwan. They are being investigated to determine the existence of ‘day of the week’ effect, ‘month specific’ effect and ‘holiday period’ effect. In order to obtain the daily, monthly and returns around the holiday periods Ordinary Least Squares (OLS) regression is run with dummy variables representing the different days, months and holiday periods. Since OLS can provide biased and incorrect standard error and variance measures, the GARCH (1,1) model was used to identify the variance associated with the respective returns for the different calendar anomalies. Results have shown that no systematic pattern exists when testing for ‘month specific’ effect or ‘holiday period’ effect in any of the six markets. Due to the high degree of cultural variation and different types of holidays in different countries, identical holidays were not found to be significant in each six countries. However, when testing for ‘day of the week’ effect, a possible systematic pattern was found for Indonesia and Philippines. Further investigation may be required before potential investors can fully capitalize on the existence of such anomalies