| Previous research documented that money flow into and out of mutual funds is a convex and asymmetric function of prior year's relative performance. Many studies have looked at fund managers' incentives to modify portfolio risk from this convex compensation structure under the annual tournament hypothesis. Specifically, these studies argue that funds that performed poorly in the first half of the year have greater incentives to increase their risk later in the year more than funds that performed well to game the option-like contract. However, according to option pricing theory, it is those funds ranked in the middle at midyear that should have greater incentives to take more risk later in the year than either top winners or bottom losers. In my first essay, I reexamine this modified tournament hypothesis. In general I find evidence in support for my hypothesis. Results are stronger for the period from 1980 to 1996. For long-term growth funds the evidence is exactly opposite to what I hypothesize. Furthermore, when measures of relative return volatility are used, previous result that midyear losers increase volatility later in the year more than midyear winners does not hold in many cases.; While previous literature found an asymmetric and convex flow-performance relationship in mutual funds, there is no reason to believe that the relationship is of the same shape for all funds. In my second essay, I examine the role of performance history in investors' learning about fund quality, and thus its effect on the degree of asymmetry and convexity in flow-performance relationship. While it takes time for investors to learn, it appears that investors learn about fund quality through their performance history. And they react differently to the same performance measure for funds of different performance history. As a result, funds perceived to be of better quality face a more asymmetric and convex flow-performance relationship, even after controlling for age. The evidence is stronger when investors condition their quality inference on longer performance history. The results have implications for studies of mutual fund risk-taking under incentives from the asymmetric and convex flow-performance relationship. |
