The overconfident portfolio manager: Bold, brash, and beneficial

Individuals are often overconfident, particularly those in positions that influence economic outcomes. We develop a principal–agent model to study the consequences of hiring an overconfident portfolio manager who overestimates the precision of his private information. In the model, managerial overconfidence increases the manager’s (agent) optimal effort and induces riskier portfolio choices. The investor (principal) can mitigate excessive managerial risk-taking by increasing incentive intensity, which raises the variance of the manager’s compensation and causes the manager to bear more performance risk. When the model imposes additional constraints on the manager’s portfolio choices, we find that the likelihood of portfolio constraints binding increases with managerial overconfidence, which in turn affects effort choice. Overall, when compensation is determined endogenously, we show that moderate managerial overconfidence can increase investor welfare by stimulating effort, whereas excessive overconfidence is detrimental because it leads to excessive risk exposure. The model delivers sharp, testable predictions regarding the interaction among overconfidence, portfolio risk, and compensation design. Using mutual fund data, we provide empirical support for several key implications of the theory, particularly with respect to portfolio constraints and incentive alignment.