It has been laborious – but we are now close to a “clean” data set for the M&A analysis. The latest work has focused on winnowing the 1,000+ transactions down to a more narrowly defined set of genuine merger transactions in which it is conceivable that senior management might have considered any one of the 10 potential brand strategies.
After a number of false starts, the decision rules for what constituted a merger - as opposed to an acquisition - were relatively easy to specify in an objective, transparent way. We have set the maximum discrepancy in revenue/value to be equal to the maximum observed discrepancy in a transaction that had adopted either a “best of both” or a “different in kind” strategy.
Using this decision rule, we have reduced the number of examples of strategy 1 or 10 from over 80% of the total sample to closer to the 65% figure observed in our earlier research. By doing so, we have hopefully eliminated a lot of “noise” from the data – all of those transactions that were clearly acquisitions (the vast difference in size between the acquirer and target made it inconceivable that it would be regarded as a merger).
As a footnote, I have learned that the technical distinction between a merger and an acquisition is simply whether the target’s legal entity survives for legal reporting purposes. If it does, then the transaction is a merger. If it does not, then the transaction is an acquisition.
