Can Excess CEO Confidence Increase Risk of Corporate Failure?

A recent report by KPMG [1] on the behavior of chief executive officers (CEOs) suggests that 67 percent of UK CEOs trust their intuition over data. The impact of intuition may become problematic if it is driven by biased perception. One of the most common biases among CEOs is overconfidence, a tendency to believe that they are better than they objectively are, particularly in their judgment, ability, and knowledge.

In a recent paper, we investigate whether CEO overconfidence can help explain the probability of corporate failure. Despite extensive research exploring the consequences of managerial overconfidence for corporate policies and outcomes, the question of how CEO overconfidence affects the likelihood of corporate bankruptcy remains unaddressed.

The literature focuses largely on the dark side of CEO overconfidence, but there is also a beneficial side. For example, overconfident CEOs believe strongly in their leadership, which leads others to view them as more competent and therefore more respected and influential. As better leaders, they may enhance personal motivation and strengthen labor commitments that in general contribute to better performance [2]. However, studies exploring the dark side [3] suggest that overconfident CEOs overestimate future cash flows and engage in value-destroying investment and financing decisions. Is one view more useful than the other in predicting corporate failure? Would the bright side (effective leadership) outweigh the dark side (suboptimal decision making)?  The answers to these questions are unclear.

Anecdotal evidence suggests that the overconfidence of CEOs could trigger corporate failure. For example, Walter A. Fallon, CEO of Kodak, was blamed for the firm’s bankruptcy. Fallon’s bias could reassure him in his judgment that Kodak’s position on the market was safe, but it led to delays in innovation and the failure to recognize the significance of digital cameras. In a similar vein, the co-CEOs of BlackBerry’s mobile phones, which once led the mobile industry, were blamed for the failure, due to overconfidence, of the product.

To investigate the link between CEO overconfidence and the likelihood of failure, we employ in our paper a large sample of non-financial UK firms.  The UK offers an interesting context for studying the impact of behavioral biases on corporate decisions because CEOs of UK firms have more leeway in their corporate decision-making compared with managers in other countries and institutional investors tend to be passive.

Overall, we find that firms managed by overconfident CEOs are more likely to fail. Our analysis shows that there are two main contexts in which overconfident CEOs affect the probability of bankruptcy. First, we provide evidence that the positive relation between CEO overconfidence and corporate failures is more pronounced in firms that operate in innovative sectors or have higher R&D spending. This finding suggests that the adverse impact of overconfidence is greater in firms operating within innovative environments. Second, we show that the association between overconfidence and bankruptcy risk is more pronounced in firms with less conservative accounting practices, which implies that overconfident CEOs increase the likelihood of bankruptcy by delaying the reaction to bad news.

Are there ways to monitor the detrimental impact of the bias? Are existing corporate governance mechanisms effective in moderating the impact of CEO overconfidence on the probability of bankruptcy?  To answer this question, we consider in the analysis both internal and external monitoring mechanisms. To test the former, we incorporate three important board characteristics: size, independence, and gender diversity. We find that there is a significant association between CEO overconfidence and the probability of bankruptcy only in subsamples of firms that are managed by weaker boards, i.e. smaller, less independent, and male-only.   To examine the effectiveness of external governance, we consider the effect of monitoring by banks and institutional owners and find that the positive association between CEO overconfidence and the probability of bankruptcy is mainly driven by our subsample of financially healthy firms when CEOs are not under the scrutiny of their creditors and in firms with low institutional ownership.

Our findings suggest that it is important to incorporate personal managerial attributes and biases in corporate-failure prediction models. Our findings also suggest that stricter monitoring by creditors and board characteristics play an important role in moderating the adverse effects of CEO overconfidence. While this is a step forward in our understanding of the effectiveness of corporate governance in restraining behavioral biases, further research is needed to shed light on the relationship between the behavioral characteristics of directors and the likelihood of failure of the companies they manage.

REFERENCES

[1] KPMG (2018) 2018 UK CEO Outlook Report, available at https://home.kpmg/uk/en/home/insights/2018/05/uk-ceo-outlook-survey-2018.html.

[2] Hirshleifer,D., A. Low, and S.H. Teoh. 2012. “Are Overconfident CEOs Better Innovators?” The Journal of Finance 67 (4): 1457–1498. doi:10.1111/j.1540-6261.2012.01753.x.

[3] Malmendier,U., and G.Tate. 2015. “Behavioral CEOs: The Role of Managerial Overconfidence.” The Journal of Economic Perspectives, 29 (4): 37–60. doi:10.1257/jep.29.4.37.

[4] Kim, J.-B., Wang, Z., and L. Zhang. 2016. “CEO Overconfidence and Stock Price Crash Risk.” Contemporary Accounting Research 33 (4): 1720–1749. doi:10.1111/1911-3846.12217.

This post comes to us from professors Jingsi Leng at Loughborough University, Neslihan Ozkan at the University of Bristol, Aydin Ozkan at the University of Huddersfield, and Agnieszka Trzeciakiewicz at the University of Hull. It is based on their recent paper, “CEO overconfidence and the probability of corporate failure: evidence from the United Kingdom,” available here.

Leave a Reply

Your email address will not be published. Required fields are marked *