A common cognitive bias is the base rate fallacy - in judging probabilities, we focus too much on individual cases, and not enough upon background probabilities.
Much discussion of Barclays possible merger with ABN Amro seems to be committing just this error. People are focusing on the merits (or not) of the deal, and are not asking the question: what's the general success rate of cross-border banking mergers?
This paper by Dirk Schmautzer fills in the gap. It suggests Barclays is onto a loser. He says:
Transactions are found to be value-creating, albeit only on a net basis with target gains more than compensating for bidder losses.
In other words, the gains of the merger have already been reaped - by ABN Amro's shareholders.
There's worse. Schmautzer says:
Bidders create most value when they take over targets that lack stand-alone growth prospects but are already relatively cost efficient ex-ante.
But ABN Amro is thought to have too high costs.
Worse still, Barclays has bad form here. Its takeover of the Woolwich is regarded as a failure.
So why is it repeating the trick? Maybe it's not just the base rate fallacy at work. 20 years ago, Richard Roll suggested that takeovers were motivated partly by hubris (unfriendly pdf); bosses over-rate their abilities. Maybe little has changed in 20 years.