Potential Overreactions and Reversal Effects, Evidence from the Swedish Market

Sammanfattning: Ever since De Bondt and Thaler first formulated the Overreaction Hypothesis researchers have been studying different financial markets searching for evidence for an overreacting market. This paper examines if the OMXS30 stocks overreact to extreme events and if certain variables, such as the underlying causational event of the potential overreaction and furthermore the characteristics of the potential overreaction itself, can explain the probability of reversals taking place. The main result of this study is that on average there are no reversals taking place. Therefore, the Overreaction Hypothesis is rejected for the OMXS30 stocks. Reversals are however found to exist in over one third of the events examined which is considered a high fraction keeping in mind the size of the companies studied. Short time reversals occurred after 34.7% of the potential overreactions and long time reversals occurred after 43.3% of the cases. The main tests of the thesis show on a statistically significant level that increased traded volume on the event day increases the probability of reversals taking place. Some of the causational events are with statistic significance found to increase or decrease the probability of reversals. The conclusion that the likelihood of a reversal is dependent on the causational event is therefore drawn.