Three essays on finance

Abstract

In chapter one, we find that mutual fund managers with degrees from elite universities tend to outperform their counterparts from less elite universities. We show that the better performance of elite graduates is generated from their connections with underwriters that facilitate allocations to underpriced IPOs. Indeed, we find that the funds outperform only in months when they are connected to underwriters issuing IPOs. A strategy of buying mutual funds in months when they are connected to underwriters scheduled to issue IPOs generates significant abnormal returns, as high as 4.08 percent per annum in hot markets. In chapter two, I show that there are greater occurrences of fraud committed by elite school graduates even after taking into account the fact that there are more CEOs graduated from such schools. The firm-level analysis further reveals that there is a stronger peer effect of fraud in elite schools, suggesting that frauds are more contagious among elite school graduates. Furthermore, elite school CEOs are no more likely than non-elite school graduates to commit fraud if it were not for peer effect. Using school endowment as a proxy for the tightness of the alumni network, I find evidence suggesting that there is an elite school culture that promotes tighter school network, which, in turn, enhances the peer effect. An instrumental variable approach provides consistent evidence of causality. I further examine the possible channels, and document that the performance of peer firms positively influences the peer effect of fraud, which is consistent with the “keeping up with Joneses” argument. My paper sheds light on the costs of educational connection in the context of financial misconduct. In chapter three, we document a negative cross-sectional relationship between the ownership of smart investors (high return gap mutual funds) and future stock returns controlling for total institutional ownership. We interpret this result as an attenuation of cost of capital by smart investors, which is further corroborated by a similar relation between smart investor ownership and implied cost of capital estimated from various models. Our finding is consistent with the predictions from the noisy rational expectation models like Easley and O'hara (2004) that when there are more informed investors in a firm and/or when the private signals they receive are more precise, uninformed investors face lower information risk, which leads to lower cost of capital. These predictions are further supported by evidence from natural experiments including brokerage closure and the passage of Reg. FD.