Three essays on stock selection ability and agency problem of mutual funds

The first essay examines the impact of market frictions on mutual funds' trading on the accruals anomaly. On average, mutual funds are "accruals traders" who hold and trade towards stocks with low accruals. However, they avoid trading on stocks with extreme accruals, and the gains from actively using the accruals strategy are modest. We provide evidence that transaction costs constrain the use and reduce the profitability of the accruals strategy. Funds with more active accruals trading also have higher idiosyncratic return volatility. However, idiosyncratic risks do not significantly affect fund flows and therefore do not serve as a limit to arbritrage in the sense of Shleifer and Vishny (1997).; The second essay analyzes the performance of active equity mutual funds managed by insurance companies and their subsidiaries. We document that insurance funds underperform non-insurance peers by more than one percent in average annual returns. Two possible explanations are examined: conservatism in risk-taking and lack of incentive to pursue superior performance. There is no evidence that insurance funds make less risky investments; instead they have lower risk-adjusted returns. Further, insurance fund flows are less sensitive to performance when they perform poorly, and the investors they attract appear to be less sophisticated. Insurance funds more likely to suffer from incentive problems, such as those using insurers' brand names, those with close ties to the insurance parents, and those managing a substantial amount of non-mutual-fund assets, are more likely to underperform. As a conclusion, insurers' efforts to cross-produce and cross-sell mutual funds create agency problems that erode fund performance.; The third essay investigates whether funds managed by independent asset management companies (independent funds) are less affected by the agency problem between investors and fund managers, thus they outperform funds affiliated with a non-investment company parent. We find independent funds outperform other funds by more than one percent in their net return. Independent funds are more likely to continue their good performance and improve their poor performance. In addition, independent fund flow is more sensitive to performance for poor performers. This market discipline yields more pressures on independent fund managers to improve their performance. Our results support the conjecture that the agency problem is less severe in independent funds. In a separate examination, we decompose the performance of independent funds and find the outperformance is mainly from fund managers' stock selection ability.