In our forthcoming Journal of Finance paper, Are Overconfident CEOs Better Innovators?, we find that over the 1993 to 2003 period, CEO overconfidence is associated with riskier projects, greater investment in innovation, and greater innovation as measured by the number of patent applications and patent citations even after controlling for the amount of R&D expenditures. In other words, the R&D investments of overconfident CEOs are more productive in generating innovation. However, greater innovative output of overconfident managers is achieved only in innovative industries. We also find evidence that overconfident CEOs are more effective at exploiting growth opportunities and translating them into firm value, especially within innovative industries. We find that overconfidence remains a strong and significant predictor of innovation even when we remove managers with short tenures at their firms, which suggests that the endogenous hiring of overconfident managers by innovative firms is not the main driver of our findings.
The results of this study have a bearing on the usual presumption that overconfidence is undesirable. Business commentators often point to examples of headstrong, overconfident CEOs who made disastrous decisions. However, the chance of a big defeat may be a corollary to the chance of great victory, so the lesson to draw from examples is unclear. A more serious charge is provided by the evidence of Malmendier and Tate (2008) that the market reacts more negatively to acquisitions made by overconfident CEOs. This dark side to CEO overconfidence might seem to suggest that the CEO selection process should be designed to filter out oversized egos, or that compensation and governance should be designed to severely constrain such CEOs.
On the other hand, some authors have suggested positive roles for overconfidence, such as improving decision implementation (Russo and Schoemaker (1992)), encouraging agents to take sufficient risk on behalf of principals (Goel and Thakor (2008), Gervais, Odean, and Heaton (2011)), or countering information externality problems by stimulating entrepreneurship and experimentation (Bernardo and Welch (2001)). Vinod Khosla, co-founder of Sun Microsystems Inc. and venture capitalist, describes “unbridled confidence and arrogance” as key characteristics for successful business visionaries.  Our tests identify empirically a positive side to CEO overconfidence. Why would overconfidence adversely affect acquisition decisions yet favorably affect innovation decisions? This may be a natural consequence of the bright sides and dark sides of CEO overconfidence. We have argued, consistent with Goel and Thakor (2008) and Gervais, Heaton, and Odean (2011), that if managers are risk averse, the willingness to take audacious risks can be valuable to the firm. However, overconfidence may have more adverse effects for external acquisition than for internally driven innovation. Self-aggrandizing CEOs may engage in excessive empire-building through acquisition because they overestimate their ability to exert effective control over a wider domain.
Our findings therefore suggest a possible solution to the puzzle of why so many firms hire overconfident CEOs and CFOs (Malmendier and Tate (2005a, 2008), Ben-David, Graham, and Harvey (2010)), and why such CEOs seem to be placed at the helm of growth firms (Graham, Harvey, and Puri (2010))—exactly the kind of firm in which overconfidence will have the greatest effect. The hiring of such managers, and their matching with growth firms, may be efficient if overconfident CEOs are better innovators.
The full paper is available for download here.
 “Who Will Be the ‘Next Steve Jobs’?,” Wall Street Journal, 10/8/11, http://online.wsj.com/article/SB10001424052970203476804576617432977807982.html.