Mutual fund performance benchmarking using a quadratic directional distance function approach

The purpose of this paper is to apply a quadratic directional distance function in order to validate a relative performance indicator for portfolio benchmarking analysis. Based on a daily data set generated from a sample of 43 equity mutual funds the present study compares the performance results given by a directional and two standard distance function models as well as with a traditional fund performance index for periods ranging from 6 months to 4 years. A significant difference was observed in their rankings overall time horizons. This finding implies that ignoring the diversification effect of co-variances in portfolio risk reduction and the potential improvements in returns will yield a biased estimate of mutual fund performance.