| This dissertation consists of three essays on the relations among investors, mutual funds, and fund families.; Chapter one presents a model of new fund openings as a function of the past performance of a family's existing funds. At the fund level, we model the relations among fund performance, investment flows, and the risk-taking behavior of the fund manager. Our model predicts that families dominated either by outperforming funds or by underperforming funds are more likely to open a new fund than are families composed of average performers. We predict that an asymmetric performance-fund flow relation combined with expected intra-family flows from existing underperformers to a new fund provide an incentive for families with severely under-performing funds to open a new fund in hopes of managing a 'star'.; Chapter two presents an empirical analysis of new fund openings. We study fund performance, investment flows, and risk level and examine the relation between the distribution of performance across funds within a family and new fund openings. We find that new fund openings are positively correlated with measures of both extreme underperformance and extreme outperformance of existing funds as well as measures of the number of 'dog' funds within a family. The evidence supports our predictions in Chapter 1.; Chapter three addresses the relation between advisory firm organization and mutual fund performance and expenses. Specifically, we hypothesize three relations. First, the ownership structure of a fund family---mutualized, privately held, or publicly owned---may impact fund manager behavior and be reflected in expenses and/or performance. Second, fund families may experience some net pecuniary benefit or harm as a result of subsidiary affiliation. Finally, we examine expense and performance differences across directly advised versus subadvised funds. We find evidence that publicly owned fund families provide investors with lower style-adjusted returns and alpha at higher cost than do privately owned or mutualized families. Similarly, we find that bank and insurance affiliates underperform their peers in both returns net of expenses and alpha net of expenses, and that diversified financial services affiliates outperform in these measures. |
