by Arijit Chatterjee
, 26.06.12
Everyone has different attitudes to risk. Our individual judgments, interpretations and preferences influence the way we approach risky decisions. Sometimes, we can weigh the odds in a fairly rational, mathematical way. But when chief executive officers (CEOs) take big decisions such as acquiring another firm, it is very hard for them to know in advance how likely different outcomes are. Risk taking is not so much an economic calculus as an interpretive act.
An important factor in that interpretation is confidence. Unlike gamblers betting on the turn of a card, executives often believe that their capabilities (and those of their organizations) can improve the outcome of a risky decision. For example, they may feel that they are good at picking the right projects, or skilled in forecasting future events such as price movements, competitors’ actions or technology developments. Of course, they may be wrong, but their beliefs still affect their decisions. Acquisitions of other companies, for example, often destroy shareholder value, but CEOs still make acquisitions because they believe they have the ability to make them work.
When we do something well, we grow in confidence and enthusiasm. When things go badly, however, we feel self-doubt and hesitancy. In other words, feedback from our actions affects our confidence – not just in our principal domains of work but spills over in unrelated areas as well. Interestingly, this happens even when skill is not a factor in the outcome. For example, if we win at a game of luck, we may feel that we are ‘on a roll’.
CEOs receive feedback on the quality of their decisions all the time. Firm performance gives a strong objective cue about their capability. CEOs also receive ‘softer’ signals from media coverage, awards, colleagues’ remarks and so on. Arijit Chatterjee and Professor Donald C. Hambrick of Pennsylvania State University (hereafter, the authors) call these different types of signals ‘capability cues’, which affect CEOs’ confidence, which in turn influences their risk-taking behavior.
Not all CEOs react to capability cues in the same way. Their values, how they think, their education, and other factors play their part. For CEOs, one of the most important aspects of personality is narcissism: an inflated sense of self, coupled with a need to have that sense of self reinforced.
Narcissists are full of contradictions. They admire themselves, but depend on the admiration of others. They crave approval, but often incur disapproval with the way they act. And they often bolster themselves by undermining others. However, researchers disagree on the effect of narcissism on executives. Some think narcissistic CEOs go their own way, ignoring the outside world, while others see them as motivated by praise. So which one is it?
To solve this puzzle, the authors took an interactionist approach – looking at the way capability cues interact with CEOs’ personalities to affect their confidence – and examined two types of capability cues: recent performance (how well the firm is doing) and social praise (what people say about the CEO).
Arijit Chatterjee and Donald Hambrick’s overarching theory was that recent performance has less of an effect on narcissistic CEOs, while social praise has a significant effect on them. When it comes to measures of success, narcissistic CEOs ignore or downplay poor performance as irrelevant, while taking strong performance as proof of their strength. Either way, they don’t let objective evidence affect their decisions. In contrast, less narcissistic CEOs are made timid by poor performance and emboldened by good performance.
Social praise, however, has a significant effect on narcissists, who need the ‘fuel’ of applause and admiration. Glowing press accounts or media awards may give narcissistic CEOs a generally elevated mood, which research shows could make them more likely to take risks.
From extant theories and available evidence, the authors put forward four plausible hypotheses. First, there will be a positive relationship between the recent performance of the firm and current risk taking. Second, there will be a positive relationship between the amount of recent social praise for the firm’s CEO and current risk taking. CEOs tend to believe their own press; the more they are praised, the more likely they are to take risks. Third, CEO narcissism will dampen the effect of recent firm performance on risk taking. More narcissistic CEOs take risks regardless of performance data, while less narcissistic CEOs are more grounded in reality. Fourth, CEO narcissism will enhance the effect of recent social praise on risk taking. More narcissistic CEOs are emboldened by social praise to take risks, while less narcissistic CEOs are more immune to flattery.
The authors tested their theories with two studies. To measure risk taking, they looked at the spending of 152 CEOs from the computer industry in three areas: acquisitions, research and development, and capital expenditure. They also evaluated 131 CEOs who made major acquisitions, looking at the ‘acquisition premium’ or the extra amount paid over the target’s pre-takeover price.
The authors assessed CEO narcissism using four indicators: the relative prominence of the CEO’s photograph in their company’s annual reports, the relative prominence of the CEO in company press releases and their relative cash and non-cash pay. To measure recent performance, the authors used total shareholder returns and return on assets, both calculated net of industry average. Social praise was determined by looking at press evaluations of the CEO and the number of media awards received.
The results provided considerable support for the four hypotheses. Both types of capability cue had a general effect on risk taking and more narcissistic CEOs tended to ignore objective performance, while being far more stimulated by media praise (although not media awards). In contrast, less narcissistic CEOs paid a lot more attention to objective performance, while remaining relatively unmoved by social praise.
The authors call these different types of signals ‘capability cues’, which affect CEOs’ confidence, which in turn influences their risk-taking behavior. In sum, Arijit Chatterjee and Donald Hambrick’s study makes an important point: both context and character can be a source of confidence but instead of considering them in isolation it is a fruitful endeavour to understand how they interact to affect decisions. This is a very exciting area for future research.
