The market reaction to stock splits: A new look

Following the evidence presented in a previous paper, this paper compares the observed market reaction to the announcement of stock splits when measured using the standard CAPM benchmark against that found when using an alternative APT benchmark. The authors find that, contrary to conclusions drawn when using the CAPM, namely that there are significant positive abnormal returns around the announcement date, when performance is assessed using an alternative Arbitrage Pricing Theory based framework the evidence is less clear. The presence of abnormal returns after the announcement is significantly reduced and there is evidence to suggest that the systematic or factor risk increases. These findings can explain the slight announcement effect observed and suggests that the abnormal returns usually reported may be the result of model misspecification in the estimation of ‘normal’ returns. Consequently the rationales provided to explain the observed behaviour may be spurious.