Risk-aversion versus risk-loving preferences in nonparametric frontier-based fund ratings: A buy-and-hold backtesting strategy

The eventual risk-loving nature of preferences of investors has largely been ignored in the existing frontier-based fund rating literature. This contribution develops a series of nonparametric frontier-based methods to rate mutual funds accounting for both mixed risk-loving and mixed risk-aversion preferences. These new methods are proposed by defining the corresponding shortage functions that can allow for increases in all moments, or increases in odd moments and reductions in even moments. The empirical part designs a buy-and-hold backtesting to test the out-of-sample performance of the proposed rating methods corresponding to different risk preferences on the actual MF selection. The evidence indicates that the backtesting strategies based on the output frontier-based rating models with risk-loving preferences exhibit an overwhelming dominance compared to most existing frontier-based and traditional financial ratings.