JBS Vol 07. Num 1. 2005 - AN EMPIRICAL TEST OF THE SHORT-TERM OVERREACTION HYPOTHESIS USING DATA FROM BANGLADESH CAPITAL MARKETS

This paper presents an empirical test on short-term overreaction in Bangladesh Capital Markets. By observing the
behavior of the daily average market-adjusted excess returns (ERs) and holding period average market-adjusted
excess returns (ARs) following both positive and negative events, this paper presents empirical results that are
consistent with the implication of the overreaction hypothesis (OH). This paper also presents evidence that
statistically significant holding period average market-adjusted returns (HARs) can be earned on zero-investment
portfolios formed following event days by taking short positions on firms experiencing positive events and long
positions on firms experiencing negative events. The magnitudes of the HARs on zero-investment portfolios are
quite large. Even after adjustment for transaction cost and risk, the zero investment portfolios will provide large
positive returns to the investors. A market that allows investors to earn large positive returns duly adjusted for
transaction cost and risk is not even-weak-form efficient. Therefore, this study provides evidence that Bangladesh
capital markets both overreact in the short run and that they are not even-weak-form efficient.