ABSTRACT. This paper investigates the existence of Calendar Anomalies in the Colombo Stock Exchange (CSE) over a period from 1985 to 2005. The regression results suggest that stock returns in the CSE to a certain extent are not in consistent with the random walk hypothesis during the period 1995–2005. The results substantiate the fact that the previous finding of turn-of-the year anomaly in the CSE by Elyasiani et al. (1996) is merely a statistical artifact. The invalidation of their finding reinforces the finding by Abeysekara (2001) that shows ‘substantial autocorrelation’ that is present in the CSE returns. This study further reveals that the irrelevancy of incorporating any specific autocorrelation scheme to the classical regression model is seemingly not appropriate to the Sri Lankan scenario. pp. 84–105
JEL: G14, N25, E44

Keywords: Calendar Anomalies, month-of-the-year, day-of-the-week, turn-of-the year, turn-of-the-financial year, turn-of-the-month effects