Self-selection Bias and the Listing Status of Target Firms: Value Effects in the Spanish Market

Abstract

As corporate announcement decisions are non–random events, standard OLS estimations must be corrected for the self–selection bias. In the M&A field several studies suggest that previous evidence on univariate analysis of abnormal returns is not fully reliable. We examine whether using the standard Heckman two–step estimation procedure to correct for endogeneity significantly changes the previous evidence with respect to the decision to acquire a listed versus an unlisted firm. Our results show that this correction does not change the conclusions drawn from unconditional abnormal returns. Therefore, we emphasize that the existence of self–selection bias should not mean a general invalidation of the previous evidence.