This dissertation consists of two essays on mutual funds. In the first essay, I show that investors misallocate a substantial amount of capital in the active mutual fund industry. To this end, I develop a novel structural identification strategy to estimate returns to scale in active management and time-varying fund skill. A median fund is over-allocated by $29 million, so the majority of funds are expected to underperform. The industry can host more capital, but additional capital should go to a small fraction of funds. In particular, funds with the highest skills are severely under-allocated and account for most of the missing capital. In the time-series, under-allocated funds can outperform their benchmarks for three years. My findings suggest the active mutual fund industry deviates from a frictionless rational expectations equilibrium. The disequilibrium implies the existence of profit opportunities for informed investors and thus rationalizes the popularity of active management.
In the second essay, we investigate the performance of active sector funds whose potential outperformance has not been exhausted entirely by decreasing returns to scale. We document that, despite good track records, most sector funds are relatively smaller than their equilibrium fund sizes — at which they are expected to generate zero net alphas. In particular, from 1998 to 2016, a passive indexation strategy of actively managed sector funds earns an annual benchmark-adjusted return of 5.70% and a monthly alpha of 27 basis points. Moreover, the strategy’s outperformance is present in market downturns (i.e., resilient to tail risk) and robust to change of rebalancing frequency and inclusion of expenses. Efficient diversification and under-appreciated skill – illustrated by an alpha arithmetic to guide similar strategies – explains the strategy’s success.
| Identifer | oai:union.ndltd.org:bu.edu/oai:open.bu.edu:2144/40355 |
| Date | 24 April 2020 |
| Creators | Chen, Huangyu |
| Contributors | Schwenkler, Gustavo, Hackbarth, Dirk |
| Source Sets | Boston University |
| Language | en_US |
| Detected Language | English |
| Type | Thesis/Dissertation |
Page generated in 0.0023 seconds