Abstract
This study examines whether combining previously identified explanations of post earnings-announcement drift (PEAD) may lead to a new and more insightful explanation of the drift when compared with each explanation on its own. Tests are carried out over the period from quarter one 1991 to quarter one 2011 on the 500 firms comprising the S&P 500 as at the 14th September 2011. The roles of several control variables are investigated in predicting differential drift levels where cumulative abnormal returns from days 1 to 30 relative to the announcement is the measure of the PEAD effect. The results show that the PEAD occurs mostly because of transactions costs (Bhushan, 1994), arbitrage risk (Mendenhall, 2004) and investor distraction (Hirshleifer, Lim and Teoh, 2009). In new research, further tests examine the effect of the business cycle and the type of earnings surprise (positive/negative) on PEAD. It is found that investor distraction is only significant in explaining PEAD during expansions, while transactions costs and arbitrage risk are significant regardless of the stage of the business cycle. These results support a new explanation of PEAD that combines two of the existing explanations and suggests that the cause of PEAD varies and changes over the business cycle. The results of this study are consistent with Mendenhall (2004) with the addition of investor distraction which was not included in Mendenhall (2004). None of the tests in this study find a complete explanation of PEAD, so further research could be undertaken to suggest new explanations for the cause of the drift in addition to those examined in this study.