A Three-Step Heckit Estimator for Panel Data with Fixed Effects and an Application to Bidders’ Advisor Choice and Returns in Acquisition Sequences

The study extents the literature of the advising bank's positive influence on the performance in mergers and acquisitions by modeling the double endogeneity arising from the banks' selection by their expertise and advisory relationships along acquisition sequences. The double endogeneity arising in the selection of banks as advisors, which is observed in advised deals only, by their industry expertise and advisory relationship strength is solved with a new Three-step Heckit estimator. The Three-step Heckit estimator is an extension of the (Heckman, 1976, 1979) Two-step Heckit estimator for panel data developed by (Wooldridge, 1995). The analysis shows that the matching of the most experienced banks with the strongest advisory relationships and highest past performance with serial acquirers in the most complex transactions is efficient in terms of higher returns