eThesis - School of Business electronic theses

The effect of fund flows on the stock-picking ability of fund managers is still largely unresolved. Most of the previous studies have focused on fund level returns that are affected by investor flow driven mechanistic issue of liquidity motivated trading that incurs transaction costs, liquidity consumption, and excess cash holdings. Therefore, the effect of flows on the stock picking ability is yet to be studied separately. Concentrating in the performance of the synthetic portfolio of stocks that fund managers over-weight the most, allows me to study how the fund flows affect the behavior and success of fund managers in their stock picking work. My hypothesis is that negative fund flows pose the fund managers in greater job market risk and thus they decrease their risk taking by diversifying rather than over-weighting their best ideas or choosing less unconventional stocks. This can lead to lower stock picking success. Also, I argue that abnormal and volatile fund flows move fund managers' attention from stock picking to managing the cash inflows or paying for redemptions by liquidating assets, which distracts and decreases their stock picking success. I collect the stock market data from CRSP (The Center for Research in Security Prices), the fund holdings data from Thomson Reuters's Mutual Fund s12 data base, and the fund flow data from CRSP Survivor-Bias-Free US Mutual Fund Data Base. I limit my scope to consider US domestic equity funds that are open to retail investors in 2003-2010, that results in 2,161 funds and 2,939,901 quarterly holdings analyzed in the 2007-2010 crisis sample and ca. 2,403 funds and 3,062,516 holdings in the 2003-2006 pre-crisis sample. As proxy for the stock picking ability I use the synthetic portfolio that includes only the stocks, towards which fund managers show the greatest belief following the method introduced by Cohen et al. (2009). I run OLS regressions on the returns of the "best ideas" portfolio controlling for the market risk premium, size, value, momentum, short term reversal, idiosyncratic risk, and liquidity factors. I incorporate the preceding flow measures and flow related control variables in the regressions to analyze their effect on the returns. I provide evidence that volatile and negative fund flows cast a statistically and economically significant negative drag on the stock-picking ability that is separated from the negative effects of liquidity trading. This negative effect spans over a quarter and is rather a result of decreased ability to conduct fundamental analysis or fund managers' decreased willingness to take idiosyncratic risk, than inability to execute short term trading strategies indicated by the daily results of Rakowski (2010). I add to the literature of the widely accepted model of Berk and Green (2004) by noting that large inflows not only decrease fund level alpha but also affect the stock picking ability negatively. On the contrary, their model predicts that fund level returns would increase in outflows but I show that large outflows may radically reduce the underlying stock picking ability.