This paper examines the presence of the day-of-the-week effect in the Italian stock market index (MIB) sub-sectoral returns. The study, by using GARCH-M (1,1) models, did not find evidence of the day-of-the-week effect in mean equations, while some evidence was present in variance equations. The study also investigates the validity of random walk hypothesis for all the MIB sub-sectoral returns. The results indicate that almost all sub-sectoral returns did not follow a random walk process as required by market efficiency hypothesis.

This study investigates the existence of the day-of-the-week effect in the MIB index disaggregated at the sectoral level. The day-of-the-week effect is a regularity in the stock market that usually takes the form of significantly negative mean returns on the first day of the trading week (i.e., on Monday) and abnormally high mean returns on the last day (i.e., Friday). Many empirical studies have analyzed this issue relative to developed stock markets. For instance, Barone (1990) investigates the impact of particular dates on the MIB stock index between 1975 and 1989. The calendar anomalies observed are the start of the month, the 30^th and 31^st of the month as well as Tuesdays and Fridays. Comparing these results with those of the US stock markets, the author argues that while in the US the largest falls in stock prices occur on Mondays, in Italy, they occur both on Mondays and Tuesdays (whereas the most pronounced are on Tuesdays). Taking into account the different time zones, Italy's results suggest that calendar effect may be imported from the US. Berument and Kiymaz (2001) test the presence of the day-of-the-week effect during the period 1973-1997 by using the US S&P 500 stock index.

The results indicate that the day-of-the-week effect is present in the form of highest returns on Wednesdays, while the lowest return are on Mondays. Analyzing the day-of-the-week effect in the UK stock prices, Steeley (2001) argues that this effect seems to have disappeared in the 1990s. On the other side, by portioning the returns into negative and positive, the author finds that negative returns on Mondays and Fridays are significantly different from their mid-week counterparts. Several researchers (Keim and Stambaugh, 1984; Smirlock and Starks, 1986; and Damodaran, 1989) explore the possible factors that contribute to the day-of-the-week anomaly. They found that this phenomenon can be due to factors such as settlement procedure, thin trading, measurement errors as well as a high number of investment decisions taken at the weekends. Also, firm's size seems to be a major factor of the weekend effect. For instance, Abraham and Ikenberry (1994) show that there is high stock return differences in small- and medium-sized companies.

Applied Finance Journal, Italian Stock Market, US Stock Markets, Investment Decisions, Efficient Market Hypothesis, EMH, Europe Equity Markets, Random Walk Hypothesis, European Union, Autocorrelation Function Test, Standardized Maximum Modulus Distribution, Financial Services.