Investigation of the day-of-the-week effect on stock returns in an emerging market: the case of Sri Lanka

Abstract

Calendar anomalies in security market returns that have been documented extensively in financial literature are inconsistent with the Efficient Market Hypothesis. Among the different seasonal effects observed in stock markets, the day of the week effect has attracted wide interest. This paper examined the prevalence of the day of the week effect in Sri Lanka’s fledgling Colombo Stock Exchange (CSE). The CSE, with a market capitalization of only US$ 12 billion, however, is a vibrant market and is attracting a lot of interest with the conclusion of the long standing ethnic conflict that drained precious resources and investor confidence in the country. The All Share Price Index (ASPI) was evaluated over a 24 year period between 1985 and 2009. An Ordinary Least Squares (OLS) regression model was used to test whether there is any statistically significant difference among index returns on different days of the week. The results found strong support for the existence of the day-of-the-week anomaly at the CSE. Friday returns are abnormally high while Tuesday returns are negative and recorded the lowest. The period between 1994 and 2009 which signified major changes in the political landscape was sub divided into four periods. The results were similar with the full sample results for three of the four periods considered. In the period between 1999 and 2002, Monday returns and not the Tuesday returns provided the lowest return. The Milanka Price Index and three other sector indices, namely, diversified (DIV), banking, finance and insurance (BFI), and telecommunication (TELCO) were evaluated to establish whether the Tuesday effect was driven by foreign investor activity. The MPI, DIV and BFI indicated the prevalence of the dayof-the-week effect. The returns on Friday were significantly different from those on Tuesday. For the period between 1985 and 1994, where the market was growing and creating local investor interest, it was found that no differences could be established for daily returns during this period, enhancing the possible link between foreign investor activity and the prevalence of the-day-of-the-week effect.