On which deals this year will reputations will be made or lost? A few chief executives, most notably Warren Buffett, are prepared to wait years to be proved right. His purchase of Burlington Northern Santa Fe for Berkshire Hathaway will be seen as a masterstroke if a sluggish decade leaves the utility-like returns of a train company as the best available.

For others, it is already possible to spot dangerous groupthink without knowing exactly which merger will prove least productive. Several technology companies have spent billions in the past 12 months buying into areas beyond their technical expertise. Software designer Oracle is trying to buy server maker Sun Microsystems, while Xerox and Dell have both made big bets that adding IT services to their hardware businesses will produce a seamlessly integrated, faster-growing whole. Rising competition as companies search for growth means not all these deals can end well.

Had miner Rio Tinto sold part of itself to China's state-owned Chinalco in June, it would have been a clear frontrunner for worst transaction. By default, raising $15bn via a rights issue instead ranks as one of the best decisions of the year. In comparison Ferrovial, the Spanish construction group, was forced to sell Gatwick airport for a loss, another sign it was foolish to spend ￡10bn to buy airport group BAA back in 2006. US taxpayers might also point out their stake in General Motors will never be worth the $67bn needed to break even, but that takeover always owed more to political than financial logic.