Previous literature has documented mixed evidence on estimating mutual fund size-performance relation. This thesis tests the diseconomies of scale hypothesis of Berk and Green (2004) by exploiting an exogenous variation in capital flow generated by investors attention on a mutual fund ranking list. Wall Street Journal publishes top ten performance mutual funds for every category in each quarter, which enables a clean RD design setting. Mutual funds that just make the list receive 2.4 percentage point additional capital flow in the next quarter compared with the ones that just miss the list. I find that a 10% unexpected increase in capital flow causes 0.95 percentage point of reduction in alpha during the following quarter. The RDD estimation is around seven times larger than OLS...[ Read more ]

Previous literature has documented mixed evidence on estimating mutual fund size-performance relation. This thesis tests the diseconomies of scale hypothesis of Berk and Green (2004) by exploiting an exogenous variation in capital flow generated by investors attention on a mutual fund ranking list. Wall Street Journal publishes top ten performance mutual funds for every category in each quarter, which enables a clean RD design setting. Mutual funds that just make the list receive 2.4 percentage point additional capital flow in the next quarter compared with the ones that just miss the list. I find that a 10% unexpected increase in capital flow causes 0.95 percentage point of reduction in alpha during the following quarter. The RDD estimation is around seven times larger than OLS estimation, which provides strong evidence in favour of the Berk and Green hypothesis.