Overconfidence costs us money. This is the message of new research by Su Hyun Shin of the University of Alabama and Andrew Hanks of Ohio State University. They studied thousands of Americans aged over 50 and found that those who were overconfident about their thinking and memory skills had significantly less net wealth than others, even controlling for obvious causes of such differences such as age, education, health and income. “Overconfidence has a negative impact on net worth,” says Dr Shin.
There are many reasons why overconfidence can impoverish us. If we’re overconfident about our job prospects we’ll save too little and borrow too much in anticipation of good times that never actually materialise. And savers who are overconfident about their abilities to pick good shares or fund managers will trade too often and pick too many funds thereby losing money in dealing charges and fund management fees.
It’s not just our own overconfidence that can hurt us, though. So too can other people’s. You will all have your own favourite examples of how politicians’ or central bankers' overconfidence has led to expensive mistakes. But chief executives’ overconfidence also hurts us. We’re all still paying the price of Fred Goodwin’s belief that it was a good idea for RBS to buy ABN Amro. And that decision was only the most egregious example of a tendency pointed out 30 years ago by Richard Roll – for takeovers to be often motivated by management hubris rather than a coldly rational assessment of the value of mergers. (One feature of overconfidence, of course, is a failure to learn from others' mistakes.)
Organic expansion, however, can also be driven by overconfidence, with regrettable effects. Ulrike Malmendier at the University of California at Berkeley says a “substantial share” of CEOs are overconfident and shows that these tend to invest too much when internal funds are plentiful. (They don’t over-expand with external finance because their over-estimation of their abilities leads them to believe that share and bond markets are undervaluing their companies thus making external finance unnecessarily expensive.) This fits a finding by Charles Lee and Salman Arif. They’ve shown that rises in capital spending around the world lead to earnings disappointments as the projects prove less profitable than bosses anticipated.
All this, however, runs into an obvious question: if overconfidence is so damaging, why is it so common? Why don’t market forces select against it?
One reason is simply that overconfident investors sometimes do well. The one who buys an overpriced asset at the start of a bubble will make big profits – big enough, sometimes, to withstand a considerable deflation of that bubble. Smart money might eventually drive stupid money out of the market – but eventually is a very long time.
Also, it’s because overconfidence has big benefits. If our forefathers had not been overconfident about their attractiveness they wouldn’t have had the courage to ask out their future wives so we wouldn’t be here. Equally, it is those who are overconfident about their abilities who are disproportionately likely to apply for top jobs. And they are likely to get them because, as Sebastien Brion and Cameron Anderson have shown, hirers mistake overconfidence for actual ability. In this way, decision-makers – and perhaps humans ourselves – are selected for overconfidence.
Which isn’t wholly a bad thing. As Richard Nisbett and Lee Ross wrote in Human Inference, one of the first books on cognitive biases, overconfidence emboldens people to give us new arts and scientific theories and to become entrepreneurs and innovators. Progress thus depends upon overconfidence. And this has big spillovers. William Nordhaus, who recently won the Nobel prize in economics, has estimated that companies get only a “minuscule fraction” of the benefits of innovations: the rest go to customers or workers.
Herein, perhaps, lies an underappreciated reason for the slowdown in productivity in western economies this century. Maybe companies have learned that innovation doesn’t pay as well as they thought and so have become less overconfident and so are innovating less. We might be seeing the vindication of one of Joseph Schumpeter’s predictions – that economic growth would slow down as the buccaneering spirit of entrepreneurs is supplanted by the cautious calculations of corporate bureaucrats.
Other people’s overconfidence can help us in another way, pointed out by Kent Daniel and David Hirshleifer. They show that overconfidence contributes to two of the best-attested stock market anomalies – the tendencies for defensive and momentum stocks to do better than they should.
Defensives, they believe, do well because investors avoid them in the mistaken belief that they have the skills to pick winners from volatile growth stocks. This causes the latter to be overpriced and defensives underpriced. And momentum, they say, works in part because investors’ overconfidence about their prior opinions means they do not sufficiently update their beliefs in light of new evidence. This means that stocks don’t immediately rise sufficiently on good news or fall sufficiently on bad; instead they drift up and down in the following weeks.
Perhaps, therefore, we should not decry the fact that so many people – especially in positions of authority – have inflated opinions of their abilities. Not only is this inevitable given that hirers select for overconfidence, but it is also not entirely a bad thing.
