On The Managerial Skills Of Mutual Fund Managers Manipulation Proof Performance Measure PDF Download

This thesis investigates how manipulation of fund performances affects fund flow in the US open-ended mutual fund industry. The flaws of conventional performance measures (CPMs) enable fund managers to artificially augment fund performance so as to attract more money. By comparing CPMs with manipulation proof performance measures (MPPM) introduced by Goezmann, Ingersoll, Spiegel, and Welch (2007), we verify that manipulation exists in the mutual fund industry. Using U.S. open-ended mutual fund data from 1991 to 2007, we classify the sample into manipulated and un-manipulated funds, and further demonstrate that individual investors, rather than institutional investors, are more prone to being deceived by manipulation behaviors and thus provide more money to manipulated funds. We start in Chapter 2 with the question of the best model for predicting fund flow. Using a US mutual fund sample as empirical evidence, we compare multiple models of predicting expected flow, and find that models considering a variety of regressors (e.g. past performance, fund size, age) outperform the models that only include lagged flow as the explanatory variable. We then generate the expected flow from the best predicting model, which together with total flow would be used when assessing the investors' reactions to manipulation. Chapter 3 examines whether fund performance measures are manipulated. We show that MPPM can help avoid manipulation, and there is a significant performance discrepancy between MPPM and CPMs when compared to the market. Hence we verify that there are performance manipulations in the mutual fund industry. In addition, we find that the manipulated funds are mainly funds with excess returns below the mean, and the manipulation on retail funds and new funds are more significant. Moreover, we find that after the new Morning Star Rating, which applies a similar intuition as MPPM, was popularized in 2002, the manipulations of performance significantly decreased. Given that CPMs can be manipulated, Chapter 4 investigates whether investors are deceived and provide more money to these funds. After controlling for endogeneity between fund flow and performance, we find that the manipulated funds attract significantly more money in comparison with a group of un-manipulated funds. Specifically, we show that in the retail sample, manipulations have a significantly positive effect on flows, whereas the effect is insignificant in the wholesale subsample. -- provided by Candidate.