The week of the month effect of stock returns: Empirical evidence from the Colombo Stock Exchange

Abstract

Introduction: The purpose of this study is to examine the existence of the week-of-the-month effect in the Colombo Stock Exchange (CSE) from 2014 to 2024 while considering the effects of the significant economic event in Sri Lanka that has any impact on the week-of-the-month effect.

Methodology: Weekly closing prices of the All Share Price Index (ASPI) were collected from CSE for a sample period of 10 years, from 2014 to 2024. The weekly returns of the ASPI were calculated using the logarithm rerun calculation formula. The study used the Ordinary Least Square Method (OLS), GARCH model and EGARCH model to examine the effect. To explore the best-fitted models, GARCH and EGARCH models were compared using AIC and SIC.

Findings: The results of the study revealed that there is a third-week effect in period 01 and a fifth-week effect in period 02 at a 5% significant level. In the full period, there is a negative third-week effect at a 10% significant level and a positive fifth-week return at a 5% significant level.

Conclusion: The findings of the study indicate that there is a week-of-the-month effect exists in Period 1, Period 2, and the Full period in CSE. Also, it is highlighted that the Colombo Stock Exchange is not a weak form efficient market since the investors can earn abnormal returns using trading strategies constructed using the historical information of stock prices.