This thesis examines the effects of managerial overconfidence on corporate financing decisions. Overconfident managers tend to overestimate the mean of future cash flow and underestimate the volatility of future cash flow. We propose a novel time-varying measure of overconfidence, which is based on computational linguistic analysis of what the managers said (i.e. Chairman s Statement). The overconfidence of CEO and CFO is also constructed based on what the managers did (i.e. how they trade their own firms shares). We conduct three empirical studies that offer new insights into the roles of managerial overconfidence in the leverage decision (i.e. debt level), pecking order behaviour (i.e. the preference for debt over equity financing) and debt maturity decision (i.e. short-term debt vs. long-term debt). Study 1 documents a negative overconfidence-leverage relationship. This new finding suggests that debt conservatism associated with managerial overconfidence might be a potential explanation for the low leverage puzzle: some firms maintain low leverage, without taking tax benefits of debt, because overconfident managers believe that firm securities are undervalued by investors and thus are too costly (Malmendier, Tate and Yan, 2011). Study 2 finds managerial overconfidence leads to reverse pecking order preference especially in small firms, which sheds light on the pecking order puzzle that smaller firms with higher information costs surprisingly exhibit weaker pecking order preference. This new evidence is consistent with Hackbarth s (2008) theory that overconfident managers who underestimate the riskiness of earnings tend to prefer equity to debt financing. Study 3 finds managerial overconfidence leads to higher debt maturity. This evidence supports our proposition that overconfidence can mitigate the underinvestment problem (which is often the major concern of long-term debt investors) (Hackbarth, 2009), which in turn allows overconfident managers to use more and cheaper long-term debt. This evidence also implies that overconfidence may mitigate the agency cost of debt. Overall, our empirical analysis suggests that managerial overconfidence has significant incremental explanatory power for corporate financing decisions.
A Doctoral Thesis. Submitted in partial fulfilment of the requirements for the award of Doctor of Philosophy of Loughborough University.
Loughborough University School of Business and Economics