It’s no secret that firms often make mistakes. Whether it’s a simple manufacturing over-run or a catastrophic blunder like the Fukushima Daiichi nuclear disaster, mistakes can be sudden and brutal.

Companies respond to this negative reaction in very different ways. Some deal with product recalls and declining product sales by fundamentally overhauling production processes or making substantial investments in research and development, while others may make small, superficial changes.

The ultimate decision on how a company reacts is a reflection of its CEO and is indicative of whether the mistake will be repeated again.

Making the same mistakes all over again

Diligent CEOs will assess what went wrong, how the company was to blame and what they could have done better. They will accept the mistakes as valuable lessons and move on. CEOs who don’t listen to the feedback, those who refuse to believe an error was the result of something they had control over and instead see it as the consequence of external or accidental forces, are likely to keep making the same mistakes again and again to the detriment of the company.

To date, there has been very little research focusing on the characteristics of CEOs when seeking a better understanding of why firms act as they do. In our research paper – Making the Same Mistake All Over Again: CEO Overconfidence and Corporate Resistance to Corrective Feedback, we look at how a CEO’s overconfidence impacts his or her firm’s actions, specifically the extent to which his or her firm incorporates and responds to corrective feedback ensuing from prior errors. Overconfidence, one of the most widely-studied cognitive biases, refers to the extent to which individuals overestimate the veracity of their knowledge and judgments.  Overconfident people tend to be generally optimistic and to underestimate the likelihood they may have made an error.  Like most individuals who succeed at a task or have their predictions proven correct, overconfident people tend to take credit themselves. However, when they fail or are proven incorrect, they blame bad luck and unforeseeable factors. As with all cognitive biases, the extent and nature of overconfidence varies substantially.

While past documentation associates a CEO’s overconfidence with a firm’s acquisition frequency, the size of acquisitions, and general risk taking, we looked at how the overconfidence levels of top executives influenced their responsiveness to corrective feedback in improving management forecasts, using forecasting behaviours of more than 300 CEOs over a 15-year period.

Forecasting accuracy is a powerful mechanism by which CEOs and firms build positive reputations in capital markets. While forecasting errors (the difference between a management forecast and actual earnings) are inevitable in an uncertain business environment, significant misjudgments have negative consequences for CEOs and investors. These errors are easy to measure and provide important feedback to executives as they construct subsequent management forecasts.

Ignoring feedback

Overconfidence has been measured in a number of different ways. To ensure our findings were robust, we assessed it in three different ways, taking into account the CEO’s portrayal in the media, their tendency to hold on to “in-the-money” stock options, and their recent successes.

As expected, the more overconfident CEOs were, the less notice they took of previous errors when making subsequent earnings forecast. By ignoring this feedback, they made significantly smaller improvements in forecast accuracy over time compared to their less overconfident counterparts.

We also found CEOs with low levels of overconfidence were more likely to cease issuing voluntary forecasts if a significant error was made, while CEOs with high levels of overconfidence continued to make regular forecasts regardless of prior errors, which is consistent with their attributing these previous errors to random, period-specific events.

Further analysis showed that the negative relationship between CEO overconfidence and improvement in forecast accuracy was more pronounced when there had been a long period of time between the management forecast date and the fiscal year-end. This suggests that when overconfident CEOs were more able to attribute blame to external factors, they were even less likely to learn from their mistakes.

These results can be extended beyond the context of corporate earnings forecasts. Firms with highly overconfident CEOs can be expected to be much slower to react when faced with major accidents or evidence of pervasive corporate malfeasance, and may be more likely to attribute events to bad luck, external parties such as customers, suppliers or to a few “bad apples” within the company. By shifting the blame, they are less likely to investigate the root cause of errors and are therefore more likely to repeat them.

The attractiveness of an overconfident leader

Struggling firms have a tendency to “search for a saviour” when engaging a new CEO. They look for an individual who is larger than life. But be warned - charismatic leaders with visibility and track records of success are more likely to possess high levels of overconfidence.

In some situations, this may work to a firm’s advantage. In fact, there may be some situations where a high level of overconfidence is a highly desirable attribute in a firm’s senior executives. Overconfident CEOs may be more likely to persevere with a product or strategy that has been initially unsuccessful, and in doing so can help promote an innovative culture within a firm. But, although an overconfident CEO’s optimism can create an energetic and positive work environment, this may also come with a cost.  As we document in this study, overconfident CEOs tend to be significantly less responsive to corrective feedback, suggesting that their firms are more likely to make the same mistakes all over again.

Guoli Chen is an Assistant Professor of Strategy at INSEAD