The Validity of the Halloween Effect in the Istanbul Stock Exchange

In this study, we analyze the validity of Halloween effect in Istanbul Stock Exchange (ISE) between January 1990 - December 2010 which implies stock returns are lower during the May-October period versus the November-April period. As well as the Least Squares Method, we use Huber’s M-estimator which is a robust estimator against to outliers, and conclude that there is no Halloween effect in the ISE which shows the finding of Bouman and Jacobsen (2002) is due to disregarding outliers.