eThesis - Kauppakorkeakoulun sähköiset gradut

The purpose of this study is to examine the existence of the turn-of-the-month effect in equity returns for 43 different countries' stock market indices during 1990-2012. The turn-of-the-month is defined to consist of the last and three first trading days of the calendar month rest of the trading days defined as the other days of the calendar month. The phenomenon has previously been proven to be still alive after decades of its discovery and exists internationally and thus, it's is highly interesting to study whether the effect can still be found to exist, after the recent financial crisis and with fresh and more extensive data than before.

The study is conducted by solely using MSCI indices. In order for a country to be included in the analysis, index data has to be available for a period of at least 20 years. A total of 43 MSCI price indices fulfill this criterion and thus, the data consists of daily historical closing prices of the indices during 1990-2012, collected from DataStream database.

The findings of the study are highly interesting yet mixed: during the whole 22-year period studied, a strong evidence of the turn-of-the-month effect is found from 37 of 43 countries studied, indicating that stocks appear to earn positive average returns only around turn of a calendar month, specifically during the last and three first trading days of calendar months. However, when splitting the data into three unequal sub periods of 1990-2006, 2007-2008 and 2009-2012 and testing the effect with nine indices, the results change substantially and thus, suggesting that the turn-of-the-month effect has disappeared at least in Finland, Norway, Sweden, Switzerland, US, Korea and New Zealand during/after 1990-2006 and on contrary to the initial results, is not alive anymore. Only exceptions are Brazil and China where the effect is found also after the financial crisis. For rest the of 28 indices not studied with sub periods, and yielding evidence of turn-of-the-month effect in 1990-2012, the real existence of the effect is therefore inconclusive and the initial results should be determined with cautious.