SUMMARY
Winner-loser anomaly explained that the shares are initially generate extreme positive abnormalreturn (winner) or the extreme negative abnormal returns (loser) will experience a reversal,especially for the loser stocks that outperform the stock winner.The objective of this paper is to explore and analyze overreaction for stocks which include calculatedIslamic Index (Jakarta Islamic Index) for period 2005 - 2006. This paper used method ofDe Bond — Thaler (1985), Dissainake (1997), Adler Manurung and Pondra Permana (2005) toselect sample. Winner and loser portfolio was formed based their return and also used marketadjusted excess returns.The results show overreaction phenomenon occurred several times in the three-month observationperiod (see calculation of CAAR). But these reactions do not have a high level of consistencybecause overreactions lasted only one time so as not to represent other periods, eventhe pickles winner overreaction does not show any symptoms (all grades pickles winner positivethree months period). Statistical analysis showed no anomaly overreactions on three-monthsobservation period.Research using six monthly observation periods overreaction symptoms occur randomly in allthree replications. But the significance test results again showed the lack of overreaction at theJII. So it can be concluded that the hypothesis of overreaction in the Jakarta Islamic Index inthe period January 2005 to December 2006 can not be proven statistically. Although there aresymptoms of overreaction, the reactions are not consistent and not significant.Keyword: overreaction, Jakarta Islamic Index ( JII), return, winner, losser, market adjusted return, cumulative abnormal Return.